1. Between Critique and Anthropology

Commentators have long noted the deep tension between Kant's better-known, Critical account of human freedom and the ‘developmental’ conception that recurs throughout his impure ethics (most notably, in his moral anthropology. See, for example, Frierson (Reference Frierson2003), Kain and Jacobs (Reference Kain and Jacobs2003), Louden (Reference Louden2000), Munzel (Reference Munzel1999), Wilson (Reference Wilson2006), and Wood (Reference Wood1999)). This tension is most clearly manifested in the post-Critical writings of the 1790s, in which Kant increasingly turned to consider the phenomenal dimensions of human life. This interest was by no means cursory; while Kant is best known for the Critical philosophy elaborated over the course of the 1780s, he lectured extensively on history, anthropology and physical geography throughout his entire teaching career. His post-Critical works show a serious and sustained philosophical interest in the embodied, phenomenal dimensions of moral personhood; Kant appears in his later years to have become increasingly preoccupied with lived, human agency, with the imperfect material to which the Critical account of moral duty was to apply.Footnote ² In the Metaphysics of Morals, the Religion, the Anthropology, the Contest of Faculties and a number of shorter essays, Kant sketches out a ‘developmental’ picture of human agency. Turning away from the Critical account of our moral nature and duties, these writings attend to the ways in which imperfect human agents develop an ever-increasing capacity (and propensity) to realize their ends as moral beings. Robert Louden (Reference Louden2003) describes this as the ‘second part’ of morality, addressing the conditions within which we learn and ameliorate the skills required for autonomous self-determination.

This presents us with a paradox: while the account of freedom advanced in the Critique of Pure Reason is predicated on the independence of the intelligible will, divorced from any and all empirical circumstances, moral anthropology attends to ‘the subjective conditions in human nature that hinder people or help them in fulfilling the laws of a metaphysics of morals’ (Kant Reference Kant1996: 372; MM 6: 217).Footnote ³ An obvious tension runs between Kant's incompatibilist account of the intelligible will's freedom and an anthropology addressing the empirical conditions affecting human beings’ abilities to fulfil their moral duties. As Patrick Frierson (Reference Frierson2003: 1) notes, this seeming contradiction has long been subject to criticism; in the late 1790s, Friedrich Schleiermacher argued that Kant's anthropology wholly undermined the transcendental account of freedom developed his in Critical philosophy. In an 1804 review of Kant's published lectures on pedagogy, Johann Friedrich Herbart similarly noted the apparent incompatibility of moral education and transcendental freedom (Louden Reference Louden2000: 56); how are we to ‘learn’ a freedom situated outside of time, space and experience?

Compounding the problem, moral anthropology appears inextricable from a complete metaphysics of morals. Anthropology is not merely appended to an otherwise independent system of ethics, but is rather a constitutive, necessary part of it: ‘a metaphysics of morals cannot dispense with principles of application, and we shall often have to take as our object the particular nature of human beings, which is cognized only by experience, in order to show in it what can be inferred from universal moral principles’ (Kant Reference Kant1996: 372; MM 6: 217). How are we to reconcile this apparent contradiction?

2. Reconciling Two Views of Freedom: A Few Approaches

Recent Kantians address the problem in a number of fruitful ways.Footnote ⁴ While instructive, these treatments tend to overlook Kant's moral psychology, leading to certain interpretative limitations or failures.

One way of understanding the moral value of empirical ‘aids’ is to treat them as attuning human beings to morally significant data in the world, without which moral action would be impossible. Barbara Herman most clearly develops this line of argument in her account of ‘rules of moral salience’ (1993: 73–93). While the moral value of a given act lies in the will's alignment with the right set of maxims, moral action in the world – the object of Kant's practical philosophy, within which his ethics falls – requires the ability to identify morally relevant information within it. As Herman explains: ‘To be a moral agent, one must be trained to perceive situations in terms of their morally significant features’ (1993: 83). Without the capacity to discern morally relevant data in our lived experience – a capacity that we are clearly bound to develop in anthropological contexts – we would be entirely incapable of moral action. Empirical aids accustom us to recognizing and appraising morally salient information in the world, enabling us to fulfil our moral duties; through habituation, moral education, civic training and social proprieties, we develop the judgement and discernment required to engage in moral action at all.

While Herman illustrates the necessity of learning to recognize morally pertinent data in the world, she does not clearly address the thorny relationship between mutable rules of moral salience and transcendental freedom. She persuasively demonstrates that the categorical imperative alone would be unintelligible to agents who lacked all prior knowledge of the moral features of their world, and so that anthropological considerations must – to some degree or another – enter into embodied Kantian agents’ moral calculus. But she is somewhat less clear in articulating the relationship between those empirical considerations and the transcendental foundations of freedom. Herman aims to show ‘how the RMS [rules of moral salience] can have a foundation or source in the Moral Law … [such that they are] neither arbitrary nor conventional’ (Herman Reference Herman1993: 85); and yet, she consistently portrays them as bound to, generated by and directed towards our phenomenal, embodied condition. It is difficult to see how the moral law itself could generate ‘non-conventional’ RMS, since these are variously described as contextually variable ‘vehicle[s] for moral education’ (83), as revisable in light of ‘new facts’ (88), as ‘a function of a community's particular circumstances’ (83), or as ‘constitut[ing] the structure of moral sensitivity’ (78) – all, in other words, addressing facets of our anthropological nature. While Herman asserts that RMS are ‘a product of the Moral Law’ (87), it is unclear how a transcendentally determined moral law might generate rules of moral salience that so directly respond to the particularities of our phenomenal condition. It is, in fact, precisely because of our anthropological differences that we require RMS at all; RMS are those ‘structures of moral sensitivity’ that enable us to discern morally relevant information in variable empirical contexts. Herman persuasively argues that anthropological knowledge is incorporated in the moral evaluations of embodied agents, but is less clear as to exactly how these mutable, contextually contingent RMS relate to (or are generated by) the transcendental foundations of human freedom, if at all.

Nancy Sherman's exploration of Kantian virtue delves a little deeper into the problem, arguing that ‘[t]he distinction between a metaphysics of morality and an anthropology is meant to capture this division between an a priori grounding and the specific circumstances of the human case’ (Reference Sherman1997b: 123). Like Herman, Sherman emphasizes the unavoidable interconnections between morality and anthropology. The categorical imperative itself points to our empirical nature's inescapable effects on our moral duties: imperatives are, after all, only possible (and necessary) for phenomenal beings such as ourselves. While ‘the conception of free practical agency … grounds our fundamental moral norms’, she argues, ‘the “actual execution” of moral principles, or full moral practice itself, relies on the resources of a supportive empirical nature’ (129).

Sherman convincingly argues that ‘cultivated emotions become an important expression of the dominion of practical reason’ (Reference Sherman1997b: 135); while moral agency depends on a transcendentally free will, imperfect creatures such as we are require the additional supports of certain ‘strengthening’ empirical emotions. And yet, she fails to distinguish between intelligible and phenomenal affects, conflating morally necessary and morally supportive forms of affect. While Sherman shows the value of the morally edifying emotions that we are under an imperfect obligation to foster, she ignores the deeper necessity of certain forms of intelligible affect demanded by the very structure of moral motivation – namely, the moral feeling. As I argue in section 4, the moral feeling sustains Kant's account of moral motivation – and importantly, this feeling is subject to cultivation and to the influence of empirical conditions. By disregarding the moral feeling, which is ‘produced solely by reason’ and so ‘cannot be called pathologically effected but must be called practically effected’ (Kant Reference Kant1996: 201; CpV 5: 76), Sherman fails to address the form of affect most pertinent to the relationship between transcendental freedom and moral anthropology. Her otherwise astute account of the moral relevance of phenomenal affects neglects this vital connection in Kant's moral psychology.

Robert Louden similarly aims to untangle the relationship between nature and freedom, on the presumption that Kant ‘definitely assumes here that interrelationship between the two realms is possible’ (2000: 7). Louden's account of this interrelationship focuses on moral education: the moral learning imparted by ‘[c]ulture, art, science, politics, law’, he claims, ‘helps set the stage for moral life by shaping empirical character in ways that are analogous to that required by a virtuous moral disposition’ (59). While Kant's treatments of moral education support the claim that pedagogy shapes empirical character, Louden is decidedly less clear about its influence over our intelligible character – the object of our moral concern. He asserts that ‘Kant does believe that efficacious moral education is education that somehow cuts through the surface causal network in order to effect the grounding of character. How this process works is something human beings cannot fully understand; we cannot know intelligible character’ (59). This is somewhat unconvincing for (at least) two reasons. First, Kant's most comprehensive accounts of moral pedagogy (in the second Critique and Lectures on Pedagogy) provide little evidence to sustain the contention that education ‘somehow cuts through’ to the noumenal realm; his treatments of intelligible character in fact overwhelmingly emphasize its complete and necessary independence from all sensible determination. As Patrick Frierson points out (2003: 83–4), Louden's focus on the empirical dimensions of practical freedom leads to his almost complete neglect of its transcendental foundations; his account of Kant's practical freedom all but disregards its noumenal grounding. Secondly, Louden's explanatory appeal to the unknowability of intelligible character takes aim at the wrong target, misrepresenting what it is that we are unable to know. Kant asserts that human cognition is unable to grasp the objective reality of intelligible character and transcendental freedom, not the connection between a possible noumenal realm and the empirical world. Louden thus conflates what Kant claims we cannot know (intelligible character and noumenal freedom) with the exact object under investigation (the relationship between the noumenal world/intelligible character and the phenomenal world/empirical character).

Finally, taking an altogether different tack, Thomas McCarthy argues that Kant's transcendental accounts of freedom and moral equality are, quite simply, incompatible with his teleological-anthropological view of human development. Not only are Kant's Critical and anthropological perspectives at odds, but more broadly, the transcendental foundation of shared human dignity is directly contradicted by the teleological need to develop the rational capacities in which this dignity inheres. There is, as McCarthy puts it, ‘a lack of fit between how things look from the normative point of view of morality or right and how they look from the functional point of view of human progress’ (Reference McCarthy2009: 62). This slippage enables Kant to uphold a universalistic view of humanity's shared moral equality, while also failing to treat women and non-Europeans as fully free and moral persons. In McCarthy's view, there is no reconciliation; Kant's transcendental account of human dignity is qualified by a developmental logic that puts freedom beyond the reach of the untold unfortunates destined for, in Hegel's words, the slaughter-bench of history.

3. Subjective and Objective Dimensions of Moral Agency

While this scholarship illuminates important dimensions of the relationship between Kant's Critical and anthropological philosophies, it tends to neglect important facets of his moral psychology, leading to problems of interpretation.

I would like to suggest that the tension between these two perspectives is substantially mitigated when we recognize them in precisely this light: as different perspectives, attending to different dimensions of human moral agency. The problem stems (at least in part) from failing to clearly differentiate between the objective derivation of the moral law (a principle of appraisal) and the development of a subjective receptivity to our moral obligations (a principle of volition). By distinguishing objective and subjective dimensions of moral action, we can better understand Kant's transcendental and anthropological perspectives as enacting a philosophical division of labour, each attending to different aspects of a fuller conceptualization of moral agency, rather than as in tension.Footnote ⁵ Rather than treating them as making the same kinds of claims (leading to the question: are human beings transcendentally free or not?), we need to carefully consider what these distinctive forms of argumentation aim to do and say.

How exactly does Kant conceive of the transcendental grounds of our freedom? In the Critique of Pure Reason, he distinguishes between empirical and intelligible character, which describe different forms of causality in which human beings partake. While empirical character concerns a person's relation, as a sensible being, with other objects in the phenomenal world, intelligible character refers to the capacity to initiate acts of will entirely independent of, and unconditioned by, the phenomenal world. Our empirical character, as Henry Allison puts it, ‘functions as the empirical cause of [an] action’ (1990: 5); it pertains to our motives and decisions within the causal order of the sensible world. As an intelligible being, however, the ‘subject must be considered to be free from all influence of sensibility and from all determination through appearances’ (Kant Reference Kant1953: 469; CPR A541/B569). The possibility of human freedom turns on this intelligible character; without the capability to act on purely rational, self-given motivations, we would remain bound by the causal chains of the empirical world. Freedom depends on our capacity to exclude sensible influences from self-generated action; the autonomy of the will is predicated on its determination ‘in accordance with laws that are independent of any empirical condition and thus belong to the autonomy of pure reason. … The law of this autonomy … is the moral law’ (Kant Reference Kant1996: 174; CpV 5: 43).

The moral law is, of course, at the heart of Kant's practical philosophy: as free beings living alongside one another, our duties are determined by the moral law that stands as the very condition of our freedom. What, then, exactly is the relation between autonomy and freedom? And how exactly does Kant understand freedom at all? Without here having the space to fully address this notoriously difficult question, Kant distinguishes between transcendental and practical senses of freedom.Footnote ⁶ Transcendental freedom refers to the capacity of the intelligible will to act spontaneously and to initiate events or states from itself – that is, to act as an initial cause, independent of prior causation.Footnote ⁷ As a capacity operative only within the constraints of the noumenal realm, transcendental freedom can only be the subject of speculative reason. Practical freedom, conversely, describes the particular freedom with which we are endowed as rational, finite agents, as moral actors subject to the sway of the phenomenal world. Kant's conception of practical freedom has a few distinctive characteristics. First, as Paul Guyer notes, it includes the negative dimension of ‘independence from domination by one's own inclinations and … independence from domination by others’ (2005: 118). Second, practical freedom incorporates the positive attribute of autonomy, that is, the determination of the will by reason itself.Footnote ⁸ Henry Allison captures this nicely, arguing that ‘a will with the property of autonomy is one for which there are (or can be) reasons to act that are logically independent of the agent's needs as a sensuous being’ (1990: 97). This capacity – to act for reasons independent of our sensuous being – turns on the transcendental freedom of the will, on an intelligible character capable of initiating events and actions spontaneously; as Allen Wood maintains, ‘Kant's metaphysical contention is that the will can be practically free only if it is transcendentally free’ (1999: 172). Drawing on Kant's assertion that ‘the practical concept of freedom is based on this transcendental idea’ (Kant Reference Kant1953: 465; CPR A533/B561), Guyer argues that ‘transcendental freedom is a necessary [but not sufficient] condition of practical freedom … the ability to free oneself from the domination by one's sensory impulses presupposes the ability to initiate new series of actions, independent of natural laws’ (2005: 121–2).Footnote ⁹ While practical freedom (the capacity to act independently of inclination/coercion, and in conformity with self-legislated reason) presupposes the possibility of transcendental freedom (the capacity for unconditioned and spontaneous action), it is not reducible to it; autonomy requires not just any choice, but the right kinds of choices, for the right kinds of reasons. While this leaves open a number of ambiguities in their relationship, the relevant point remains: both transcendental and practical accounts of freedom (however we might interpret their connection) appear to require the strict exclusion of phenomenal considerations.

Kant's transcendental and practical views of freedom thus leave little space for the influence of a moral anthropology. How, then, are we to respond to the ‘paradox that Pistorius and so many others have found in Kant's attempt to reconcile causal determinism at the phenomenal level with his incompatibilist conception of freedom’ (Allison Reference Allison1990: 41)?

We need to differentiate two elements in moral action: the objective determination of the moral law and the duties consequent on it (what is the moral law, and what are the obligations which it incurs?) from the subjective volition in choosing to adopt the moral law as the grounds of one's action, that is, in the determination of one's maxim (should I adopt the moral law as the principle determining my action?).Footnote ¹⁰ Christine Korsgaard's examination of Kantian motivation is instructive in this regard. Korsgaard argues that, for Kant, any action incorporates two basic components, which she describes as ‘incentive’ and ‘principle’.Footnote ¹¹ An incentive is, most simply, ‘a motivationally-loaded or evaluative representation of an object’ (Korsgaard Reference Korsgaard2009: 19); roughly speaking, the incentive provides the volitional impetus behind any action, moral or not. And yet we do not act directly from the impulsion that incentives exert over us; if we did, we would remain beholden to the desire that motivated the action, and so be unfree. As Korsgaard maintains:

This is the key to what Allison describes as the Incorporation Thesis: for a free being, ‘an inclination or desire does not of itself constitute a reason for acting. It can become one only with reference to a rule or principle of action, which dictates that we ought to pursue the satisfaction of that inclination or desire’ (1990: 40). Human beings are perpetually assailed by competing desires and motivations to action; practical freedom lies in our capacity to adopt certain principles (or, to follow Kant, certain maxims) – the right principles – as the grounds for selecting the incentives on which we choose to act.Footnote ¹²

Broadly speaking, the first and second Critiques articulate what those principles are (or rather, what that principle is), addressing the first of the two elements determining the will described above. The Critical account delineates an objective, transcendental grounding for our freedom (self-determined action in conformity with the moral law) and an equally unequivocal description of the incentive structure grounding moral action (immediate respect for the moral law).Footnote ¹³ And yet, it pays significantly less attention to the second, volitional dimension of moral action implicit in Kant's practical philosophy; it does little to illuminate how we, as fallible, human agents, learn to recognize, internalize and adopt this proper incentive structure consistently as a life-long orientation. While the Critiques largely treat the determination of the moral law and the good will, Kant's ‘empirical’ works address the conditions under which we develop the subjective propensity to obey the moral law, how we learn to choose to act as morality objectively compels us to.Footnote ¹⁴ We now turn to more carefully examine this distinction.

4. Elateres Motiva: Moral Feeling and the Subjective Grounds of Choice

The Groundwork and Critiques address the formal qualities possessed by rational beings generally, and by human beings, subject to both empirical and intelligible motivations, more specifically. While our rational nature establishes the moral law's dominion over us, our phenomenal inclinations lead us to require the categorical imperative in determining the duties descended from it. Kant draws a contrast between humans and holy beings such as angels in describing our particular moral nature: as rational creatures, both are bound by the moral law but human beings alone are subject to the categorical imperative, as angels are incapable of acting otherwise than in conformity with the moral law. While the categorical imperative responds to our imperfections, the moral law which it clarifies is determined transcendentally, based on the nature of rational, end-setting beings more generally. The moral law and our duties are, then, formal attributes of what we are, as finite, rational beings; they are determined a priori, based on the nature of rational agency in conjunction with the fact of human finitude.

Yet human beings are complicated by their unique capacity – in fact, by their natural predilection – to ignore the moral law. Moral action requires not only a consciousness of the objectively given moral law, but also, a given subjective disposition to incorporate it within one's incentive structure. Under the heading ‘Of the Supreme Principle of Morality’ in the Lectures on Ethics, Kant outlines an important distinction:

As we have seen, the Critiques focus on the objective judgement involved in the appraisal of the moral law (‘What is morally good or not?’); as Kant here asserts, this is the first of two elements constituting moral action. The second, volitional dimension lies in the ‘subjective ground’ of moral choice (‘What moves me to live according to this law?’), which Kant describes as the moral feeling, the principle of impulsion particular to human beings. So what exactly is the moral feeling, this principle of volition, and how does it fit within Kant's broader account of freedom and moral action?

In a close examination of the incentive structure of the morally good will in the second Critique, Kant asserts that

This is a serious problem for moral action: as sensible beings, we are naturally inclined to prioritize phenomenal inclinations over the rational compulsion to act on moral incentives. Anticipating the Religion's account of humanity's radical evil, Kant describes this ‘propensity to make oneself as having subjective determining grounds of choice into the objective determining ground of the will’ (Kant Reference Kant1996: 200; CpV 5: 74) as self-love in general, and as self-conceit when the will adopts these subjective incentives as law-giving. The problem is volitional: we are naturally compelled by incentives of pleasure or inclination, while the sanctity of moral action lies in moving the will by what Kant acknowledges is the weaker motivational impulse of respect for the moral law. How, then, are we to choose the right incentives on the right grounds if we are naturally pre-disposed to pursue our inclinations?

Kant outlines both the problem and its solution in his account of the moral feeling:

How exactly does the moral feeling resolve the problem of morality's weaker elateres? Kant argues that the confrontation between self-love/self-conceit and the authority of the moral law creates both negative and positive feelings. By limiting self-love and striking down self-conceit, the immediate determination of the will by the moral law both humiliates self-conceit, producing a negative feeling, and generates respect for the law, a positive feeling (Kant Reference Kant1996: 200; CpV 5: 74).Footnote ¹⁶ As rationally determined affects, grounded in pure respect for the moral law, these moral feelings resolve the basic volitional quandary by generating the motivational impulses to adopt the moral law over the incentives of inclination. The moral feeling, comprised of both negative and positive affects, enables moral action by redressing the motivational deficit of the understanding before the stronger volitional force of inclinations.Footnote ¹⁷ As Patrick Frierson (Reference Frierson2005) notes, Kant distinguishes between intelligible and sensible pleasures; the affective pull – both positive and negative – of the moral feeling counters the influence of sensible inclinations without itself being rooted in the phenomenal world. The moral feeling, Kant asserts, ‘is therefore produced solely by reason. It does not serve for appraising actions and certainly not for grounding the objective moral law itself, but only as an incentive to make this law its maxim’ (Kant Reference Kant1996: 201; CpV 5: 76).

As Kant describes it, the ‘moral feeling is a capacity for being affected by a moral judgment. When I judge by understanding that the action is morally good, I am still very far from doing this action of which I have so judged’ (Kant 1997a: 71; LE 27: 1428). It thus belongs to the practical dimension of Kant's ethics, connecting moral judgement and action. The moral feeling is the bridge between what I ought to do as a matter of moral obligation and how I come to recognize this obligation as incumbent upon me at all; it develops ‘the will's receptivity to finding itself subject to the law as unconditional necessitation’ (Kant Reference Kant1996: 285: TP 8: 283).

And importantly, the moral feeling is a matter of cultivation, subject to the influence of empirical circumstances. Like Locke, Kant recognizes the absurdity in trying to compel the moral feeling; like faith, the moral feeling is an affect that resists rational argument. Yet we can inculcate a receptivity to it:

As an inborn capacity to sense the pull of intelligible obligations, the moral feeling can be neither compelled nor commanded by duty – but it can be cultivated. We thus have an imperfect obligation to develop it, as a ‘subjective condition of receptiveness to the concept of duty’ (Kant Reference Kant1996: 528; MM 6: 399). While we cannot produce the moral feeling itself, we can certainly form agents that are more or less susceptible to it. As Kant asserts (in a distinctively un-Kantian sounding account of moral development): ‘The subject must first be habituated to morality; before coming primed with rewards and punishments, the indoles erecta must first be excited, the moral feeling first made active, so that the subject can be actuated by moral motives’ (Kant 1997: 80; LE 27: 287).

Turning back to our original conundrum – the tension between Kant's transcendental and empirical accounts of freedom – we can now see what we might best understand as a philosophical division of labour, between (respectively) principles of moral appraisal and moral volition. The ‘formal’, Critical account addresses the grounds of our freedom: freedom inheres in the capacity to act from an immediate respect for the moral law and to set our own unconditioned ends, independently of empirical circumstances. And yet, as imperfect, embodied beings subject to sensible inclinations, we nevertheless need to develop the moral feeling pushing us to both recognize and internalize the authority of the moral law. While the Groundwork and Critiques largely address the objective, transcendental determination of the moral law (the principle of appraisal), the Anthropology, Religion and certain sections of the second Critique attend to the development of our subjective receptivity to it (the principle of volition).

Seen in this light, the tension between Kant's transcendental and anthropological treatments of human freedom is significantly mitigated. While the former addresses the grounds of our freedom as rational beings, the latter speaks to the development of the moral volition that is inescapably incorporated in the freedom of human beings. The Critical, transcendental account of freedom revolves around the Groundwork's good will, the figuration of the only unqualified good imaginable. And yet, the good will is not a human will; it is an idealization that models the good of free, rational agents. If we are to better understand Kant's view of the human good, we must turn from the good will to the good human being.

5. The Good Will and the Good Human Being

As I have argued, we better understand Kant's transcendental and anthropological viewpoints as complementary, rather than as being in tension; they attend to the different dimensions of a specifically human moral agency (the angels, as we saw, require no account of moral motivation). While the former addresses the grounds of our practical freedom, as agents that partake in rational self-determination, the latter attends to the wholly unsystematizable aspects of our phenomenal, embodied, sensible natures. Kant's declaration in the Groundwork that ‘the moral law in its purity and genuineness (and in the practical this is what matters most) is to be sought nowhere else than in a pure philosophy’ (Kant Reference Kant1996: 46; G 4: 390) makes perfect sense; the very purpose of the tract is, in his words, ‘nothing more than the search for and establishment of the supreme principle of morality’ (Kant Reference Kant1996: 47; G 4: 392). The Groundwork necessarily disregards the empirical world that can have no part in determining the ends of a free, rational being.

But, as Kant himself recognizes, the moral law's determination does not encompass the totality of human moral agency. If it is to speak to the constitutively imperfect creatures we are, a complete metaphysics of morals must reckon with the development of the capacities required to realize our ends as free beings. ‘[M]orality’, Kant argues, ‘cannot exist without anthropology, for one must first know of the agent whether he is also in a position to accomplish what is required from him that he should do. One can, indeed, certainly consider practical philosophy even without anthropology, or without knowledge of the agent, only then it is merely speculative’ (Kant 1997a: 47; LE 27: 244). Kant describes a full ethics as treating both objective, systematic rules of moral obligation and their subjective, variable application to human beings; and, as he puts it, ‘this application belongs to the complete presentation of the system’ (Kant Reference Kant1996: 584; MM 6: 469). This fuller picture, incorporating both transcendental and anthropological dimensions of human agency, is elaborated in his conceptualization of the good human being.

Kant sketches out the good human being in Religion within the Boundaries of Mere Reason. Broadly speaking, Religion explores the relationship between the various (provisional) ‘historical faiths’ adopted by an imperfect humanity and ‘the pure faith of religion [that] will rule over all’ (Kant Reference Kant1998: 127; R 6: 121) – the religion of pure reason – towards which we progress. It thus falls within Kant's wide-ranging treatments of the historical-teleological conditions moving us to our moral perfection. The Groundwork's good will, ‘the will of a rational being, in which … the highest and unconditional good alone can be found’ (Kant Reference Kant1996: 56; G 4: 401) is, of course, the moral beacon that we are bound to approximate. In the Religion, however, Kant considers ‘[h]ow it is possible that a naturally evil human being should make himself into a good human being’ (Kant Reference Kant1998: 66; R 6: 45). He turns his attention from the transcendental grounding of the good will to the exploration of how imperfect human agents develop a consciousness of, and receptivity to, their moral obligations. The good human being thus draws out the subjective grounds of moral personhood, the ways in which ‘radically evil’ human beings come to internalize the imperatives of the moral law. Kant describes radical evil as our propensity to prioritize sensible impulses over intelligible ones in contexts in which we ought to do the opposite (‘evil’), despite an awareness of our moral duty (‘radical evil’). Radical evil describes the misalignment of incentives endemic in our natural constitution: human beings are naturally disposed to act on their sensible impulses, even in the face of moral obligations of which they are conscious.Footnote ¹⁸ Radical evil thus constitutes a problem to which the good will is unresponsive; while the good will delineates the ends that we ought to pursue as free, rational beings, it fails to explain how creatures of our kind come to recognize and choose to seek them at all.

Kant's conceptualization of the good human being addresses precisely this subjective, volitional dimension of human moral agency. The good human being, he argues, requires an internal revolution, an epiphany that fundamentally reorientates the motivational structure of the will. Without this revolution, this deep-seated turn in the maxims governing our incentives, we remain incapable of bridging the chasm between moral and merely legal action:

Without changing not just our actions, but the grounds upon which we choose to base them, we fail to undergo the motivational transformation reorientating us from evil to moral goodness. The ‘revolution’ which Kant describes initiates a profound shift in the maxims governing the will, turning us away from a predilection to act on sensible incentives and toward a moral disposition, orientated by the ends given by the good will. This moment of moral epiphany ‘is like a kind of rebirth … which makes the resolution and the moment when this transformation took place unforgettable to him, like the beginning of a new epoch’ (Kant Reference Kant2006: 194; AP 7: 294).Footnote ¹⁹ The good human being is born from this tectonic shift, this subjective redirection enabling the agent to transcend a condition of unprincipled volatility and develop an orientation towards moral virtue.

And yet, this internal revolution describes only the initial transformation in the incentive structure of the good human being. While the transition to moral goodness depends on this moment of revelation, a properly human good consists in the subjective, life-long commitments to principled action, rational self-determination and moral improvement that follow from it. The moral epiphany realigning the will's incentive structure is no merely temporary change, but rather instils an ongoing orientation towards, and dedication to, the principled pursuit of moral action. It is worth quoting Kant at length on this point:

The good human being's transition from radical evil to moral rectitude incorporates two related dimensions: (1) the moral revolution, and (2) the long-term commitment to moral improvement – the ‘incessant laboring and becoming’ – which follows from it. This ‘ever-continuing striving for the better’ describes the particularly human moral task of approximating, but never reaching, the good will and perfect moral virtue. It illuminates the moral imperative not to be virtuous, but to develop the capacities enabling us to approach virtue.

While Kant only fleshes out this subjective, developmental dimension of moral agency in his post-Critical texts (in particular, in the Metaphysics of Morals, the Anthropology and, as we here see, the Religion), it stretches back to (at least) the second Critique's treatment of the ‘moral disposition’. As with the moral feeling, the moral disposition is not amenable to compulsion; nothing can ‘command us to have this disposition in dutiful actions but only to strive for it’ (Kant Reference Kant1996: 207; CpV 5: 83). In other words, we do not just become moral agents through the moral revolution that turns us, once and for all, from adopting a maxim of self-love to one of moral rectitude; this is only just the beginning of our moral transformation. As finite, imperfect beings constitutionally bound to our phenomenal inclinations, moral agency is a process, and not an end. This dimension of Kant's ethics is all too often overlooked; moral action and the conditions for our freedom are not only fixed by the dictates of a stringent and Archimedean moral law, but are also goals that we need to work towards. Our moral task is to develop our subjective predilection – the moral disposition – to always work towards our moral improvement.

While the good will models the moral ideal that all rational beings are obligated to pursue, the good human being elaborates the subjective, volitional and empirical dimensions of moral personhood and practical freedom. It draws out a more robust conception of human agency that tends to be obscured by the preponderant focus on Kant's Critical works. The figuration of the good human being shows that, while practical freedom stems from our rational, intelligible character (i.e. from our capacity to initiate action and set ends in the world through our own rational faculties), the freedom that we are capable of actualizing – as human, and not merely rational beings – remains subject to the sway of the phenomenal world.Footnote ²⁰ We remain constitutively imperfect, empirical creatures that are bound to develop our rational, freedom-realizing capacities within the phenomenal realm.

A moral anthropology, then, helps us to understand who we are, and so, how we orientate ourselves towards continuous moral self-improvement. It tells us about the conditions within which we undertake this life-long task of developing our moral faculties and orientations.Footnote ²¹ It points us towards the lived, embodied dimensions of human morality and freedom that Kant does address with the good human being. While practical reason enjoins us to pursue the ideal of holiness (the will's perfect conformity with the moral law), this remains ‘a perfection of which no rational being of the sensible world is capable … it can only be found in an endless progress toward that complete conformity’ (Kant Reference Kant1996: 238; CpV 5: 122). If the perfect virtue underlying the good will can be achieved by ‘no rational being of the sensible world’, we are bound to recognize the specifically human task of pursuing an ‘endless progress’ towards it – this is the moral agency of which we are capable. The good human being's ‘incessant laboring and becoming’ is, as Kant puts it, ‘the moral vocation of our nature’ (Kant Reference Kant1996: 238; CpV 5: 122): to endlessly seek to develop our capacities as free and rational beings. It is to foster the moral disposition orientating us towards our moral ends.

Of course, this is far from resolving a number of intractable ambiguities in Kant's broader account(s) of freedom. What I hope to have shown is that while practical freedom, the particular freedom that finite moral agents such as ourselves are able to exercise, depends on the presumption of noumenal freedom, the reality of our empirical nature conditions this (practical) freedom and our moral agency in important ways. While Kant's Critical account of moral personhood focuses on the transcendental derivation of the moral law, his anthropology addresses how, as imperfect beings who are in fact disposed not to obey the moral law, we come to internalize and incorporate it within ourselves. Anne Margaret Baxley's distinction between autonomy and autocracy, as distinctive properties of the free will, is helpful in clarifying these two facets of moral personhood. ‘Autonomy’, Baxley argues, ‘designates the property of the will to give particular sorts of laws to itself. … Autocracy, on the other hand, describes the executive power of the will to enforce principles that have been given. … autocracy represents the strength that a rational autonomous being must strive to acquire so that she is master over her inclinations’ (Baxley Reference Baxley2003: 11–12). Anthropology speaks to the development of autocracy, to the conditions under which imperfect, free beings cultivate the strength of will to act as the moral law commands them to. And yet, instructive as Baxley's distinction is, it does not entirely capture the dispositional dimension of practical freedom that Kant's account of the good human being draws out. Moral personhood, Kant shows, requires not just an autocratic strength of will, but rather, the development of a disposition to always act morally; it is to integrate the moral law within our purposes consistently, as a foundational orientation. And this disposition, this orientation towards constant moral improvement, is subject to precisely those influences that a moral anthropology addresses.