Kant’s Idealized Imputation of the Propensity to Evil in Religion Within the Boundaries of Mere Reason

Abstract

In Religion, Kant posits an innate and natural propensity to evil, which he implicitly designates as “radical evil”. However, this notion has been criticized for its apparent incompatibility with freedom and its problematic atemporality, leading scholars to dismiss it as an obscure metaphysical speculation or an uncritical residue of Christian doctrine rather than a morally necessary element. Drawing on student notes from Kant’s lectures, I firstly contend that the apparent conflict between an innate–natural propensity to evil and human freedom can be resolved through Kant’s imputation theory and empirical psychology. While evil originates from this innate ground in human nature, it can ultimately be overcome through the exercise of free will. Secondly, my argument proceeds to show that radical evil first emerges in Religion, not as a concept of agnostic metaphysical abstraction, but as a necessary postulate for Kant’s idealized conscience-based imputation. This imputation transcends the constraints of temporality, thereby idealizing the propensity to evil into radical evil. Atemporality is merely an analogy representing this imputation’s independence from temporal constraints. Thirdly, although this idealized imputation and radical evil are absent from Kant’s pre- and post-Religion works, my analysis suggests that the idealized imputation is necessary for his project of advancing the individual and collective moral progress through the mediation of religion.

1. Introduction

In 1792, Kant published the essay “On the Radical Evil in Human Nature” in Berlinische Monatsschrift, marking his systematic attempt to explore the moral evil of human beings, a crucial ethical issue his earlier works had surprisingly neglected. By establishing the concept of an innate propensity to evil (Hang zum Bösen), implicitly equated with “radical evil”, this essay was later incorporated as Part One of Religion within the Boundaries of Mere Reason (1793; hereafter Religion). Here Kant strategically bridges radical evil with Christian theology through concepts like redemption, divine grace, and ecclesiastical community, thereby not only expanding ethical discourse to encompass evil but also establishing a connection between his ethics and the philosophy of religion. Moreover, Kant characterizes the propensity to evil as an atemporal deed of intelligible freedom, a formulation that inspired Schelling’s idealist theories (Firestone and Jacobs 2008, p. 144) and anticipated existentialist thought, particularly through his aporetic claim that the ultimate ground for this intelligible act remains itself unintelligible (Cherkasova 2005; Fremstedal 2012).

Nonetheless, the concept of the propensity to evil has long been a subject of debate. Its characterization as an innate and natural evil appears to contradict Kant’s central ethical principle of freedom. While Kant defines this propensity as an innate evil rooted in human nature (Rel. 6:32),1 he simultaneously insists that moral evil does not stem from birth or natural causes but rather from the human capacity for free choice (Rel. 6:21). After a long period of “widespread silence” (Morgan 2005, pp. 63–64), late 20th-century scholarship critically examined this tension: how an ostensibly innate and natural evil could coexist with Kantian freedom (Michalson 1990, p. 46; Horn 2011, p. 65).

Equally problematic is Kant’s claim that this propensity, though manifested as the supreme ground of unlawful maxims in time, is fundamentally a timeless deed of human free choice, namely “an intelligible deed, cognizable only through reason without any time condition” (Rel. 6:31). This formulation invites criticism for falling into agnosticism. According to Kant’s epistemology, all human knowledge is constrained by empirical conditions, namely time and space. However, such an intelligible deed of freedom transcends all of these conditions. Consequently, it remains epistemically inaccessible to human beings who are constrained by empirical conditions (Michalson 1990, p. 63). As a result, it generates a counterintuitive implication that moral agents can be held responsible for a deed they cannot consciously recognize (Horn 2011, p. 66).

To make matters worse, Kant implicitly designates the propensity to evil as radical evil. Scholars debate whether these two concepts can be equated. Some argue, for instance, that the propensity to evil serves only as a transcendental condition for human evil (Palmquist 2015, p. 75), while radical evil constitutes its empirical manifestation (Palmquist 2015, p. 103). Nevertheless, Kant himself suggests that the former is, in some sense, also the latter. He states: “we can call it [the supreme ground of unlawful maxims] a natural propensity to evil, and […] we can even call it a radical innate evil in human nature (not any the less brought upon us by ourselves)” (Rel. 6:32). He further asserts that “if there is such a propensity in human nature, then there is a natural propensity to evil in humans; and this propensity itself, because it must ultimately be sought in free choice and can therefore be attributed to it, is morally evil. This evil is radical because it corrupts the basis of all maxims” (Rel. 6:37). Thus, the tension between the innate and natural character of the propensity to evil and human freedom extends to the core concept of Religion, namely, radical evil.

As a result, these critics align themselves with Goethe’s criticism (cf. Höffe 2011, p. 20)—a view later echoed by Schiller and Karl Barth (Sussman 2005, p. 154)—by contending that Kant’s doctrine of the propensity to evil is neither a necessary nor an integral part of his ethics. Rather, they assert that Kant uncritically introduced Christian doctrines into his philosophy, or, in Goethe’s words, hid them within his “philosophical mantle.” Furthermore, they argue that Kant’s concept of radical evil is an artificial construct—either a “Christian theological heritage” (O’Connor 1985, p. 299) or an unexamined metaphysical speculation that ultimately betrays its “latent resentment against the body” (Michalson 1990, p. 69).

In this study, I employ a comparative approach to analyze the evolution of Kant’s imputation theory, focusing on the aforementioned three issues: (1) the apparent contradiction between innateness-naturality of the propensity to evil and human freedom, (2) the problematic atemporality of the propensity to evil, which invites suspicions of agnosticism and metaphysical abstraction, and (3) the necessity of radical evil for Kant’s ethics. Current Kant scholarship has achieved notable progress through this approach. First, the conceptual origins of the propensity to evil are evident in the imputation theory articulated in student notes from Kant’s lectures, demonstrating Baumgarten’s direct influence (Brunner 2025). Second, the precise formulation of radical evil as an intelligible deed is absent from Kant’s pre-Religion works and diminishes in significance in his later writings (Klemme 1999).

Building on this progress, I address the tension between the innate and natural propensity to evil and freedom by drawing on Kant’s empirical psychology and imputation theory from his Lectures on Anthropology and Lectures on Ethics (Section 2). Furthermore, I demonstrate that in Religion, Kant “idealized” his conscience-based imputation; rather than being a metaphysically atemporal abstraction, radical evil—understood as a similarly “idealized” propensity to evil—functions as a necessary postulate for this imputation (Section 3). Finally, I contend that radical evil and ideal imputation are essential to Kant’s ethics, constituting more than an uncritical remnant of Christian doctrine, despite receiving less attention in his works after Religion (Section 4).

Before we proceed, it is important to clarify the central term used in this paper. The concept of “idealized imputation” is drawn from Kant’s notion of particular ideas. These ideas are concepts of reason that are independent of experience and, therefore, free from the temporal constraints that govern empirical knowledge. As part of the method of idealization in modern natural science, they function as standards for empirical cognition. They enable us to “assess the defectiveness of our cognition and tell us what we ought to approximate” (Spagnesi 2023, p. 420). For instance, although “it is hard to find pure earth, pure water, pure air” in experience, such idealized constructs remain necessary for disciplines like chemistry (KrV A646/B674). It should be noted, however, that these ideas are not transcendental ideas, such as God, the world, or the soul (Frketich 2025, p. 2). Unlike transcendental ideas, which contain no empirical elements, particular ideas are abstracted from experience by the understanding and subsequently idealized by reason. Although independent of experience, they retain a reflective relationship to it (Frketich 2025, pp. 16–20).

Accordingly, the good or evil imputed in “idealized imputation” is also an idea of this kind. It is abstracted from observed virtue or evil in experience, yet idealized to the point of being independent of variations in time and space. By contrast, in empirical imputation, virtue or propensity to evil is graded and differs from one individual to another. It should be acknowledged that Kant does not explicitly apply this idea to evil in his Religion (a possible explanation is that Kant tries to avoid technical terms in this work to keep his presentation as accessible as a children’s story or a sermon. Rel. 6:14)2; however, he does refer elsewhere to the “only one virtue in the idea (objectively)”, as opposed to the “virtue in reality (subjectively)” encountered in experience, which are plural and varies in “completeness or deficiency” (MS 6:447). My argument is partly similar to the “quasi-transcendental argument” (for the universality of the propensity to evil), which suggests that in Religion, Kant “seeks the fundamental boundary conditions that must be assumed in order to explain how religion is possible at all” (Palmquist 2008, p. 273). Like this approach, my research also attempts to establish a necessary condition of possibility—in this case, for religious imputation. However, I would not characterize the resulting concept as “transcendental”, since it does not appeal to transcendental ideas or epistemology. This conceptual clarification provides the foundation for my interpretation of Kant’s account of the propensity to evil in the Religion. The ensuing discussion will analyze the objects, judgment, and criteria involved in this form of imputation and the empirical imputation, especially in Section 3 and Section 4.

2. The Innate–Natural Propensity to Evil and Freedom

In this section, I propose a two-stage resolution to the apparent contradiction between the innate–natural character of evil propensity and human freedom in Religion. First, drawing upon Kant’s empirical psychology as developed in his Lectures on Anthropology, I demonstrate how the innate propensity to evil functions as a hypothetical possibility for evil inclination—a subjective possibility whose realization creates the necessary conceptual space for freedom to either restrain or to foster this latent evil. Second, through an analysis of Kant’s imputation theory as articulated in his Lectures on Ethics, with particular reference to Baumgarten’s concept of imputable natural evil, I establish that moral evil constitutes a nature that free will may either suppress or foster.

The first source of the tension between the propensity to evil and freedom stems from the innate character of this propensity. Kant posits that an innate evil propensity constitutes a moral quality a priori to “the use of freedom given in experience (from the earliest youth back to birth)” (Rel. 6:22). This formulation raises a fundamental question: what justifies attributing such moral qualities to human freedom? More precisely, how is moral responsibility tenable for a propensity that is “given by birth” (angeboren)? The “transcendental freedom interpretation” offers an approach to this problem by conceptualizing innateness as a transcendental property of freedom itself—a property logically antecedent to all empirical actions (O’Connor 1985, p. 298). Thus, although the term “innate” etymologically implies birth-givenness, it operates here as a metaphorical schematization of the intelligible ground underlying evil propensity (Brunner 2025, p. 43). This distinction yields two senses of propensity: literal innateness and transcendental innateness (Formosa 2007, p. 223).

Nevertheless, this proposed resolution proves inadequate for addressing Kant’s nuanced account of the innateness-freedom relationship in Religion. As Kant specifically clarifies: although the propensity “can indeed be considered innate”, it nevertheless “need not be represented as such; it can rather be thought of (if it is good) as acquired, or (if evil) as brought by the human being upon himself” (Rel. 6:29). This distinction reveals that the authentic innate evil propensity cannot be reduced to a metaphorical schematization of transcendental freedom; rather, it is literally something given by birth. By contrast, moral good or evil propensity—constituted either acquired good or self-incurred evil—are only analogically “represented” (vorgestellt) or “thought of” (als … gedacht) as “innate” within a derivative sense.

Furthermore, this interpretation conflicts with the classification implied in the following passage in Religion, which operates on a purely empirical–psychological level without transcendental grounding.

“By propensity (propensio) I understand the subjective ground for the possibility of an inclination (habitual desire, concupiscentia), insofar as it is accidental to humanity in general. It differs from a predisposition in that it can indeed be considered innate, but need not be represented as such; it can rather be thought of (if it is good) as acquired, or (if evil) as brought by the human being upon himself. However, we are only talking here about a propensity to the genuine, i.e., to moral evil, which, since it is only possible as a determination of free will […]”.

(Rel. 6:28–29)

I argue that a threefold classification can be constructed from this passage. The phrase “the subjective ground for the possibility of an inclination” refers to the foundational level, where propensity is distinguished from mere inclinations. “It differs from a predisposition” marks the second level, which differentiates predispositions from propensities: while predispositions to good (such as the animal instinct for sexuality) function independently of experience with objects, propensities form the subjective ground for possible inclinations (i.e., habitual desires) and depend on object experience to develop. Kant illustrates this with the example of “savage peoples” who possess a physical propensity for intoxicants—a propensity that only solidifies into a concrete inclination after initial exposure. Finally, “to the genuine, i.e., to moral evil” introduces a third level of distinction, separating physical propensities from moral ones.

Consequently, I propose an alternative solution—termed the “psychological propensity interpretation”—that conceptualizes the innateness of evil propensity not as a transcendental quality, but as a psychological attribute homologous to instinct and physical propensity. It consists of two parts. First, it seeks to substantiate my construction above and further elucidate the moral significance of Kant’s threefold classification. For instance, the moral relevance of the comparative treatment of physical and moral propensities in Religion remains unclear (Morgan 2005, p. 67). Second, it aims to resolve the central tension between the innateness of the propensity to evil and human freedom within this threefold framework.

(1) Kant’s threefold classification gains conceptual clarity through his empirical psychology, as systematically developed in his anthropology lectures. His anthropology comprises divisions: the didactic and the characteristic. The didactic section originates substantially from Baumgarten’s empirical psychology (Brandt 2003, p. 89), where propensity is conceptualized alternatively as either inclination (Baumgarten 2019, p. 296) or as the predominant passion influencing human choice (Baumgarten 2011, pp. 392–93)—a view reflecting eighteenth-century mainstream psychology. Kant, however, revises this conventional understanding by redefining propensity not as an actualized inclination but as the subjective possibility of inclination to evil (cf. Brunner 2025, p. 28), establishing the crucial propensity-inclination distinction that forms the first tier of his three-level classification.

This first-level psychological distinction lies in the fact that instinct and propensity, as innate desires, fundamentally differ from inclination as acquired desire. In Kant’s empirical psychology, propensity is defined as the “hypothetical possibility” of an inclination. Inclination, by contrast, emerges as a real desire acquired through experiential interaction with objects and habituation. Instinct constitutes a real desire existing before empirical object encounter (Parow 25:408–411). This innate-acquired dichotomy sharply separates inclinations (dependent on object experience) from instincts and propensities (Friedländer 25:580).

The second-level distinction establishes a differentiation between instinct and propensity. This distinction rests not merely on the premise that propensity requires object experience to develop into an inclination, but also on the additional qualification that such inclination must be deliberately acquired by the moral agent (Michalson 1990, p. 42). Kant’s psychology explicitly characterizes such inclination formation as “artificial” (gekünstelt) and even “counter-natural” (widernatürlich), as exemplified by the case of opium use for battlefield courage enhancement (Friedländer 25:619).

The third-level distinction addresses the conceptual demarcation between physical and moral propensities. Significantly, physical propensity cannot be reduced to natural necessity; both types of propensity are, in fact, subject to artificial suppression or fostering through acting agency (cf. Brunner 2025, p. 30). The central challenge resides in establishing their differentiation. An illustration of this distinction can be found in intoxicant addiction: while it may manifest as a physical propensity among the aforementioned “savage peoples”, it can equally constitute what Kant identifies as a form of moral depravity, namely the bestial vice (MS 6:427). The fundamental question thus arises: what criterion can substantiate this differentiation?

This criterion is grounded in the moral significance of the causal factors that shape the development of human propensities. Certain artificial influences—such as education and cultural conditioning—are morally neutral. Children remain morally innocent insofar as their latent propensities have not yet been realized. However, through habitual reinforcement, these latent propensities may develop into concrete evil inclinations (Parow 25:409). This developmental trajectory enables parents to identify their children’s propensities and intervene to prevent the formation of harmful inclinations (Menschenkunde 25:1112). Yet such external influences do not themselves constitute genuinely moral determinants. Even a criminal led to the gallows could act with honor and generosity, if he were educated and cultured to channel his inclinations (Anthro. Collins 25:209). Similarly, so-called “savage peoples” might avoid intoxicant behaviors if their culture were differently constituted.

By contrast, the authentic moral cause ultimately resides in the power of freedom itself—that is, the capacity to resist or to yield to evil propensities. For instance, Socrates demonstrates the exercise of rational self-mastery in overcoming natural malicious inclinations (Friedländer 25:654). Conversely, a morally corrupt individual cannot be absolved of accountability by appealing to external factors such as poor education, harmful associations, the malice of nature, and even his previous way of life (KrV A554/B 582). Ultimately, the individual always retains the freedom to reject these influences and maintain moral integrity. This principle applies equally to cases of addiction: an alcoholic remains responsible for failing to exercise rational self-restraint.

(2) The three-level clarification outlined above resolves the apparent tension between innate evil propensity and human freedom. Unlike innate instincts, which manifest as real desires, a propensity is merely a subjectively possible ground for inclinations that only become real through habitual development. This permits either the suppression (erstiken) or guidance (lenken) of propensities toward specific inclinations (Parow 25:410). This subjective possibility establishes the necessary conceptual space for the exercise of the power of freedom, either to create favorable conditions for developing inclinations from propensities or to prevent their development (Anthro. Collins 25:209). While the subjective necessity of inclinations may diminish practical reason’s efficacy and consequently constrain the objective necessity of freedom, these inclinations themselves do not constitute moral evil. Properly speaking, it is the propensity that qualifies as evil, insofar as it provides the logical precondition for freedom to overcome it (Anthro. Collins 25:209–210; Menschenkunde 25:1112). This relationship is analogically represented as self-incurred evil by freedom, as stated above.

The second source of tension between evil propensity and freedom stems from the natural origin of this propensity. While Kant explicitly states that the propensity to evil “lies in human nature” (Rel. 6:37), this raises the question of how such a natural quality can be reconciled with human freedom (Michalson 1990, p. 46). Conventional readings maintain that Kant only systematically addressed this issue in his Religion (cf. Paton 1946, p. 275; Beck 1984, p. 203). However, two passages in the first two Critiques concerning the imputation of evil to freedom (KrV A554/B 582; KpV 5:99) demonstrate a remarkable continuity with Kant’s mature imputation theory as developed in Religion (Rel. 6:38). These texts strongly suggest that Kant had already formulated a coherent solution to the apparent nature-freedom conflict concerning natural evil well before the publication of Religion.

This coherent solution is located in Kant’s broader theory of imputation, as presented in his Lectures on Ethics. While drawing on Baumgarten’s imputation theory, Kant adapts it to fit his conception of freedom. Baumgarten identifies two types of evil rooted in human nature: weakness and fragility. Weakness refers to the natural inability to realize a certain degree of correctness in its free determinations. Fragility, on the other hand, describes the natural propensity to moral evil—a fundamental vulnerability to be tempted to moral evil. Crucially, both forms of natural evil can exist in either unconditional or conditional forms. The unconditional variety falls outside moral responsibility, whereas the conditional form remains subject to imputation when it results from, or might have been prevented by, free action (Baumgarten 2019, pp. 297–98).

Kant adopts Baumgarten’s dichotomy between weakness and fragility. Weakness means that human nature “lacks the degree of moral goodness that is necessary to make action adequate to the moral law”; fragility means that human nature contains “not only a lack of moral goodness, but even serves as a great principle and driving force for evil actions” (Kant 2004, p. 96). Meanwhile, he rejects Baumgarten’s notion of morally non-imputable natural evil, maintaining that both forms must necessarily be attributed to freedom (Vigilantius 27:571). It is suggested that Kant transformed Baumgarten’s category of unconditional evil into a conception of merely potential evil (Brunner 2025, pp. 106–7). Yet this interpretation, stands in tension with Kant’s purpose in introducing the doctrine of propensity to evil: namely, to preclude appeals to human nature from becoming what he called the “opium of conscience”, that is, using one’s nature as an excuse to evade culpability (cf. Klemme 1999, pp. 136–37; cf. Kant 2004, p. 96).

By contrast, Kant endorses Baumgarten’s concept of morally imputable natural evil, thereby resolving the apparent tension between the naturality of the propensity to evil and freedom. He emphasizes that all natural weaknesses and fragilities can be overcome through the exercise of free will (Vigilantius 27:572). The concept of freedom signifies that human beings are not predetermined by nature to act according to their propensity to evil (Powalski 27:159). On the other hand, when free will fails to adequately exercise its capacity to suppress evil propensity, this may lead to the growth of corresponding inclinations, which in turn impose constraints upon the actualization of freedom. This account of self-incurred limitation of human freedom is marked by a deficiency of moral resolve and the courage required to resist inclinations (cf. Klemme 1999, p. 146). Moreover, if free will ceases resisting fragility, it actively nurtures (nähren) those very inclinations toward evil, thereby activating latent evil propensities (Vigilantius 27:571–572). Furthermore, should evil propensity be considered non-imputable due to its natural origins, the moral law would prove inadequate as an incentive for morality. This scenario would necessitate alternative incentives, thereby undermining the essential purity of the moral law (Kant 2004, pp. 96–97).

This solution proposed in my second step can be termed the “imputation interpretation”, which is systematically coherent with the earlier “psychological propensity interpretation” proposed in the first step. The “human nature” in question corresponds to the innate yet subjective possibility for evil inclination. Within Kant’s empirical psychology, the propensity itself is not a real inclination or passion, contra Baumgarten’s formulation. Rather, it constitutes merely a hypothetical foundation for inclinations that may be either constrained or developed through the exercise of freedom. Crucially, this very possibility constitutes the imputable nature itself, determined by whether an individual exercises freedom rightly or wrongly, thus either restraining or developing the propensity to evil.

Finally, it is necessary to address a suspicion that threatens to undermine the proposed resolution to the tension between nature and freedom—namely, the concern that the concept of nature employed here retains uncritical elements of Christian doctrine. It is suggested that, despite rejecting determinism in this conventional conception of nature (Michalson 1990, p. 62), Kant may have strategically adopted the term “evil by nature” as a concession to his Lutheran audience (Wood 2014, p. 57). Maybe Kant himself was uncomfortable with using “nature” to describe the propensity to evil and may not have used this term literally (Louden 2000, p. 137). At minimum, his concept of “nature” remains ambiguous (Dalferth 2014, p. 68). Nevertheless, Kant’s Groundwork for the Metaphysics of Morals demonstrates a genuine effort to incorporate the idea of an evil nature into his ethical system. A close analysis reveals that the natural dialectic in this work closely resembles the propensity to evil discussed in Religion (GMS 4:405; Rel. 6:42; cf. Allison 1990, p. 152; Baxley 2003, p. 20; Muchnik 2010, pp. 122–23; Sussman 2005, p. 159). Moreover, the two primary stages of this propensity—namely, impurity and fragility—are already present here (GMS 4:406; Timmons 1994, p. 127).

To summarize, my two-step solution shows that the tension between the innate–natural propensity to evil and moral freedom in Religion is formally addressed by Kant in his Lectures on Anthropology and Lectures on Ethics. In Religion, with the development of Kant’s critical philosophy, the concept of freedom advances to that of transcendental freedom, which yields the idealized notion of conscience-based imputation, thereby introducing a crucial concept: radical evil. Section 3 will address this issue.

3. The Idealized Imputation and Radical Evil

In this section, I will discuss the issue of agnosticism and metaphysical speculation regarding Kant’s conception of radical evil caused by its atemporality in Religion. This analysis is conducted in three stages. First, I propose an idealized conscience-based imputation as a result of Kant’s endeavor to integrate his imputation with transcendental freedom in Religion. Second, I will demonstrate that this idealized imputation transforms the propensity to evil into radical evil, which represents the idealized form of this propensity. Finally, I argue that radical evil is a postulate for this idealized imputation, which is neither unknowable nor speculative. Atemporality is merely an analogy representing this imputation’s independence from temporal constraints.

The first stage of my analysis is the close examination of Kant’s conceptualization of an idealized imputation in Religion. I begin with an outline of his general theory of imputation and its conscience-based branch. While Religion’s predominant emphasis on radical evil—characterized as “the subjective ground of the possibility of the deviation of the maxims from the moral law” (Rel. 6:29)—might suggest that Kant equates evil exclusively with this concept, his more rigorous philosophical definition is in fact developed in the Critique of Practical Reason, where evil is systematically posited as “a necessary object of […] detestation […] according to a principle of reason” (KpV 5:58).

In distinction to this definition, the propensity to evil, as previously established, falls under imputation judgments discussed in Kant’s Lectures on Ethics. The concept of imputation goes beyond simple moral evaluation and entails a structured series of judgments. Specifically, at least three distinct judgments are required. First, the act must be determined as a free deed—an action attributable to an autonomous subject (Kant 2004, pp. 87–88). Second, this deed must be classified as either an external or internal action (Kant 2004, pp. 94–95). Third, when considering reward or punishment as consequences of the deed, a legally effective judgment must be incorporated. This ultimately requires an authoritative court or judge to determine the appropriate consequences (Kant 2004, p. 100). The propensity to evil emerges as the object of a distinct form of imputation—specifically, what I will term “conscience-based imputation” (i.e., imputation within conscience). This propensity manifests as an internal deed encompassing both the agent’s subjective intention (maxim) and its fundamental incentive (Kant 2004, p. 99; cf. Brunner 2025, pp. 115–16). Within this particular imputation, conscience operates as the adjudicating judge, a role that Kant articulates through the metaphorical notion of a “divine court” (Kant 2004, p. 101).

In Religion, Kant’s conscience-based imputation faces the task of integrating transcendental freedom, a key theme in his late-1780s ethical works. This integration involves three dimensions:

(1)

As noted earlier, the first judgment in imputation requires establishing that a deed derives from human freedom, rather than natural causes. Since the Second Critique, Kant began to insist that this freedom is transcendental, existing independently of temporal and spatial conditions. This perspective significantly shapes Religion (cf. Allison 1990, p. 140). Consequently, Kant extends the concept of imputable evil from empirical psychology to an atemporal realm and its atemporal nature (“antecedent to every use of freedom given in experience”, Rel. 6:22; “not acquired in time”, Rel. 6:25).

(2)

The second judgment must determine that this propensity to evil functions as an internal act. In Religion, Kant’s commitment to transcendental freedom requires understanding evil as an internal deed transcending temporal constraints. This requirement necessitates distinguishing two levels of deeds: particular internal deeds (concrete maxims) become cognizable through temporally conditioned self-examination within time. At the intelligible level, the propensity to evil is understood as “an intelligible deed, cognizable only through reason without any time condition” (Rel. 6:31). Fundamentally, this intelligible deed operates as a supreme maxim that grounds all empirical deeds.

(3)

The third judgment establishes conscience as the proper judge for internal imputation. In Kant’s 1780s works, he systematically grapples with the enduring opacity of the human heart, concluding that internal imputation can never attain absolute epistemic certainty. As finite beings, our self-reflection is necessarily constrained by spatiotemporal conditions. We can only examine the intentions and incentives accessible to our empirical consciousness. Even the most rigorous self-examination cannot uncover all motivations, as “secret incentives” may remain buried deep within the human heart (GMS 4:407). Because this limitation is caused by the transcendental freedom that lies beyond the empirical world, “the true morality of actions (their merit or guilt) remains entirely hidden from us, even in our own conduct” (KrV A551/B579). In his Lectures on Ethics, Kant offers a graduated scheme of empirical imputation to address this dilemma (See Section 4). Accordingly, weakness is classified as a negative evil, while fragility constitutes positive evil (Kant 2004, p. 98). By contrast, in Religion, Kant adopts a distinct approach through the transforming of conscience into a “scrutinizer of the heart” (Herzenskündiger) who transcends the constraints of experience. Here, conscience is personified as a God who is capable of assessing internal deeds through intellectual intuition that transcends sensory and temporal limitations (Rel. 6:67, 72).

To conclude my first stage of my analysis, I propose an idealized notion of a conscience-based imputation as a result of Kant’s aforementioned attempts to reconcile imputation with transcendental freedom. It has been demonstrated that in Religion, Kant introduces an intelligible perspective grounded in pure reason’s judgment (Forschner 2011, p. 80). Building upon this foundation, my analysis conceptualizes this judgment as an idealized conscience-based imputation. I use the term “idealization” to differentiate this imputation from the aforementioned empirical imputation. It means that such imputation transcends the temporal conditions of experience (though it is not transcendental in the Kantian sense, since it constitutes neither a metaphysical nor an epistemological concept, but rather a strictly ethical one, as will be further discussed below). Although no empirical sample of material fully conforms to the ideal of pure elements, chemistry requires us to consider them as such for analytical purposes. Analogously, even though empirical self-examination cannot reveal the total corruption of the heart, we are compelled to think the imputation of an idealized evil by a divine scrutinizer who apprehends the supreme ground of all maxims.

Consequently, the objects, judge, and criterion in this imputation are all idealized. This imputation comprises several idealized components: an intelligible deed anchored in transcendental freedom, a deed that transcends temporal constraints, and God understood as a “scrutinizer of the heart” who can penetrate the limitations of experience and perceive this deed. Additionally, the principle referenced in this imputation is also, in this sense, idealized. Conscience is idealized in its operation. Thus, it not only examines the propensity to evil as it is manifested in the maxims of actions we discover in self-reflection but also judges this propensity as a single, intelligible deed. It is further personalized as God. God institutes what Kant terms a “divine court of justice” that adjudicates deed “by the scales of pure reason”—standing in stark contrast to human courts that evaluate multiple virtues with “empirical standards” (Rel. 6:24). Most significantly, this principle from pure reason is rigorous: the deed must be either wholly good or wholly evil, admitting no gradations of evil.

Based on the aforementioned idealized imputation, the second stage of my analysis focuses on the central concept in the first part of Religion: the notion of radical evil as an idealized propensity to evil. It is consistently emphasized that Kant uses “radical” in its etymological sense meaning, denoting “root” (from Latin radix)—a usage distinct from Arendt’s interpretation of radical evil as something that surpasses any comprehension due to its extreme magnitude (Allison 2020, p. 475; Louden 2000, p. 135; Fremstedal 2012, p. 204; Formosa 2007, p. 225). Given that Kant speaks of “radical evil in human nature” (Rel. 6:37), it is suggested that “radical” signifies that this evil is rooted in humanity itself (Louden 2000, p. 135; Formosa 2007, p. 225). However, Kant also explicitly states that this evil is radical precisely because it “corrupts the ground of all maxims” (Rel. 6:37). This implies that this “root” might instead refer to an individual’s fundamental maxim that corrupts all other subordinate maxims (Allison 1990, p. 147; Fremstedal 2012, p. 204; Muchnik 2010, p. 120). The notion of idealized imputation strongly supports the second interpretation. In Religion, Kant typically uses “radical” concerning the problem of imputing an innate and natural evil propensity to human freedom, thereby establishing its culpability (Rel. 6:32, 37, 72). Consequently, radical evil is best understood as an imputable intelligible act, namely, an evil that grounds all unlawful maxims.

Nevertheless, radical evil does not invalidate Kant’s original conception of the propensity to evil in his lectures, which is empirically graduated. In Kant’s Lectures on Ethics, he retains Baumgarten’s dichotomy of weakness and fragility in human nature, while introducing an intermediate category: impurity (Kant 2004, p. 97). This additional form of natural evil stems from Kant’s reflection on the “opium of conscience”: Since human beings are weak by nature, the moral incentive alone proves insufficient to motivate them to fulfill their duties. As a result, non-moral incentives inevitably come into play (Powalski 27:159). Subsequently, in Religion, Kant explicitly expands the dichotomy into a three-stage propensity to evil: weakness, impurity, and depravity. While his terminology undergoes notable modifications (for example, he designates the first stage as “fragility” and characterizes the third stage alternatively as “depravity” or “perversity of the human heart”, Rel. 6:30), the fundamental content of the original conception of the propensity remains intact. Moreover, the aforementioned gradation of this propensity remains consistent: weakness and impurity arise unintentionally, whereas depravity entails deliberate guilt (Rel. 6:38).

Moreover, radical evil represents the idealized version of the propensity to evil. Similarly to the idealized conscience-based imputation, the concept of “idealized evil” means that the empirically gradated propensity to evil is idealized into an intelligible deed that transcends empirical conditions. While Kant describes the propensity to evil as “radical evil”, he refrains from explicitly clarifying the nuanced distinction between these concepts that arises from this idealization. Specifically, radical evil may be understood as the propensity to evil precisely insofar as it represents the latter’s idealized form. Within empirical imputation, individuals are assessed according to empirical standards. This process reveals gradations in the propensity to evil, ranging from unintentional guilt (such as weakness or impurity) to deliberate crime (i.e., depravity). However, through idealized imputation, individuals are measured against a strict standard of pure reason, whereby no degree of evil escapes detection, with all empirical degrees idealized as radical evil (Rel. 6:24). Although no empirical sample of water can ever be fully adequate to the idea of pure water, it is necessary to regard it as if it were such an instantiation for the purposes of chemical analysis. Analogously, any empirically observed evil propensity (including those in the first two stages) assumes the status of radical evil when subjected to this idealized lens of “radical perversity”. Consequently, each stage of the propensity expresses or manifests itself as radical evil (Allison 1990, p. 159; Formosa 2007, p. 225; Forschner 2011, p. 79).

Grounded in this understanding of the idealized propensity to evil, the third stage of my analysis addresses the charges of agnosticism and metaphysical abstraction arising from the atemporality issue. Kant posits radical evil as an intelligible deed, which serves as the supreme yet unknowable maxim governing all concrete maxims within time. Furthermore, this intelligible deed constitutes the ultimate ground of our maxims, making any more fundamental ground of radical evil unintelligible (Rel. 6:43). This epistemic limitation gives rise to accusations of agnosticism (Michalson 1990, pp. 64–69). The idea that individuals can be morally responsible for something beyond their conscious awareness seems absurd (Horn 2011, p. 66). Furthermore, this agnosticism permits individual to maintain the belief in their noumenal goodness notwithstanding their manifest empirical imperfections (Chignell 2014, p. 110).

A further unresolved issue caused by the atemporality problem concerns the relationship between the timeless deed and its role in determining concrete maxims within time (O’Connor 1985, p. 291). While Kant articulates this distinction, he fails to provide sufficient theoretical elaboration (Michalson 1990, p. 61). This theoretical gap renders the concept of radical evil somewhat problematic, representing a metaphysical speculation incorporated into Kant’s ethics. Criticism has highlighted several problematic aspects, including Kant’s “latent resentment against the body” (Michalson 1990, p. 69), his deep-rooted “Christian theological heritage” (O’Connor 1985, p. 299), and his assumption of a noumenal self that exists outside of time (O’Connor 1985, p. 295).

The prevailing defense against this criticism argues that radical evil should be understood through an additional epistemological lens, specifically as a postulate (Sussman 2005, p. 172; Forschner 2011, p. 82; Brunner 2025, p. 84). Understood this way, this postulate operates merely as an analogy from an intelligible standpoint, rather than as a reality in an atemporal world. In this context, the term “deed” is employed metaphorically to denote the acquisition of an evil propensity (Allison 1990, p. 153). Similarly, “timeless” does not signify objective reality but rather a “process of practical rationality” (Formosa 2007, p. 233; cf. Allison 1990, p. 154). Moreover, the deed is not to be interpreted as the timeless cause of individual maxims, but instead as either a regulative idea or postulate (Allison 1990, pp. 136–45; 2002, p. 340; Brunner 2025, pp. 51–52). Radical evil, therefore, should not be understood as a metaphysically substantive intelligible deed in the noumenal realm that causally determines empirical actions in the phenomenal world. Consequently, the atemporality must be interpreted minimally, that is, strictly as “a claim about the limits of rationalization” (Timmons 1994, p. 119).

My study supports this defense against the criticism of “latent resentment against the body”, but also shows that Kant’s motivation for introducing the intelligible perspective is primarily ethical rather than epistemological or metaphysical. The atemporality here functions as an analogy, highlighting the ideal character of conscience-based imputation and radical evil, in contrast to empirically graduated imputation, not to the empirical world in a metaphysical sense. As demonstrated in his Lectures on Anthropology and Lectures on Ethics, Kant defines the propensity to evil as an experience emerging from internal self-reflection of conscience. In Religion, Kant argues that conscience must attribute internal deeds to transcendental freedom, a causality that necessarily transcends all temporal conditions. This requirement compels the adoption of an idealized framework for conscience-based imputation. The possibility of this imputation, however, presupposes several key elements: the postulation of God as a “scrutinizer of the heart”; radical evil conceived as both an intelligible deed and a corrupt ground; and the notion that this evil ground undermines concrete maxims, all existing independent of temporal conditions. Consequently, the possibility of the idealized conscience-based imputation presupposes radical evil; conversely, the necessity of postulating the concept of radical evil for Kant’s ethics is justified by this idealized imputation. The atemporality of the idealized imputation and radical evil analogously signifies that this imputation is independent of the empirical objects and standards, leaving no need to refer to a metaphysical-noumenal world.

Although the necessity of radical evil can be justified with reference to the possibility of an idealized conscience-based imputation, the ethical necessity both of this idealized conscience-based imputation and consequently of radical evil within Kant’s ethical system is problematized by its marginalization in Kant’s later works after Religion. This issue will be examined in Section 4.

4. The Necessity of the Idealized Imputation and Radical Evil

In Section 4, I seek to establish the necessity of the notion of an idealized conscience-based imputation alongside radical evil. The present argument proceeds in three stages. First, I demonstrate that the absence of this notion in Kant’s pre- and post-Religion works suggests that its theoretical necessity within Kant’s ethics requires further inquiry. Second, I establish that this notion is necessary for Kant’s strategy for enhancing individual and collective morality through religious education. Third, I maintain that the diminished emphasis on this doctrine in Kant’s later works reflects a shift in his intellectual focus toward the role of education in the moral development of the human species.

Several arguments have been advanced to demonstrate the necessity of both the idealized conscience-based imputation and radical evil in Kant’s ethics. First, the supreme ground of maxims, even when evil, serves to unite the countless subordinate maxims resulting from the empirical use of freedom, thereby establishing continuity and regularity in moral agency (Michalson 1990, pp. 55–59). Second, this unified personality provides the necessary foundation for imputation (Allison 1990, p. 144). Finally, while empirical imputation might lead to an infinite regress when searching for the ultimate principles behind these maxims, this regress can be resolved by appealing to a supreme ground (O’Connor 1985, p. 291).

The absence of references to the idealized conscience-based imputation in Kant’s early works and lectures before Religion poses a challenge to these arguments: Kant initially adhered to an “empirically graduated” imputation, although transcendental freedom had already been taken into account. By “empirically graduated” imputation—as opposed to the “idealized” imputation discussed in Section 3—I refer to the evaluation of observable actions through self-reflection based on a graduated standard. This approach focuses on the propensity to evil, wherein unintentional acts are deemed less morally culpable than intentional ones (Kant 2004, pp. 93–95): weakness constitutes a negative evil, while fragility qualifies as positive evil (Kant 2004, p. 98). Moreover, in the Critique of Pure Reason, when transcendental freedom is discussed, Kant maintains that, since self-reflection is limited to our empirical self, we cannot ascertain definitively whether actions originate from “mere nature, whether from the innate defect of temperament or its fortunate constitution (merito fortunae)” or from the exercise of freedom, which would warrant praise or blame. Thus, “no one can judge with perfect justice” (KrV A551/B579).

A more pressing concern emerges in Kant’s post-Religion works, where both the notion of idealized conscience-based imputation and the doctrine of radical evil are substantially marginalized. Interpreters endeavor to trace residual elements of the propensity to evil in the Metaphysics of Morals (1797), published just four years after Religion (1793). In the “Doctrine of Right” (Part I), Kant characterizes human beings as prone to two distinct forms of evil: passive disregard for the moral law, and active rejection of its authority, the latter of which Kant dismisses as “wholly pointless” (MS 6:330; cf. Höffe 2012, p. 386). In the “Doctrine of Virtue” (Part II), evil is framed as a precondition for virtue, given that virtue is “the struggle against the influence of the evil principle in a human being “(MS. 6:440; cf. Baxley 2003, p. 20). Kant concludes that moral life commences with an “all at once” resolution to oppose evil (MS 6:477; Allison 1990, p. 171).

Nevertheless, these texts ultimately demonstrate that what persists in the Metaphysics of Morals is essentially the original conception of the propensity to evil in Kant’s lectures, not the doctrine of radical evil. In the “Doctrine of Right”, the distinction between positive and negative forms of evil corresponds to the two degrees of evil propensity identified in the Lectures on Ethics—weakness and fragility (Kant 2004, p. 98). Moreover, while Religion presents the revolution of disposition as a single, timeless act of decision that transforms the corrupt heart (Rel. 6:48), Kant’s Lectures on Anthropology reveals a temporal perspective of the “all at once” resolution in the “Doctrine of Virtue”: it occurs at age 30, with its fulfillment emerging later at age 40 (Anthro. 7:294). This resolution constitutes merely an intermediary stage of moral self-education (Mrongovius 25:1392).

Kant’s treatment of the propensity to evil in Anthropology from a Pragmatic Point of View (published in 1798, five years after Religion) equally deserves attention. Here, Kant posits that empirical observation reveals an innate evil propensity—specifically, “a propensity actively to desire what is unlawful”. Significantly, however, he no longer maintains the concept of radical evil as a corrupted intelligible character. While the intelligible character is good, it is the crudity of human nature that constitutes the evil propensity (Anthro. 7:325; cf. Klemme 1999, p. 146).

The preceding analysis, constituting the first stage of my argument, is not merely a historical examination of the evolution of Kant’s concepts of imputation and evil. Rather, it exposes the theoretical challenge of determining the precise role of both the doctrine of the idealized conscience-based imputation and radical evil: even though integrated with transcendental freedom, Kant’s ethics remains exclusively confined to empirically graduated imputation. This empirical perspective appears to be Kant’s consistent view, as demonstrated by the consistent reiteration of earlier ideas regarding graduated imputation in Metaphysics of Morals (MS 6:227–228, 405, 438). Thus, the suspicions raised in the Introduction reemerge: are all the concepts—including radical evil, imputation in a divine court, and God as “the scrutinizer of the heart”—merely residues of Christian doctrine, hidden in Kant’s “philosopher’s mantle”?

In response to this issue, the second stage of my argument aims to demonstrate that the doctrine of radical evil and idealized conscience-based imputation remain essential to Kant’s ethics, as they constitute the necessary condition for individual and collective moral progress. While Kant states in Religion that “the proposition of innate evil is of no use at all in moral dogmatics”, he nevertheless employs it as a fundamental precondition in the context of personal moral education, particularly in moral asceticism (Asketik) and moral formation (sittliche Ausbildung), since “we cannot begin from a natural innocence, but must initiate from the presupposition of an evilness of the will” (Rel. 6:50–51). It follows that radical evil acquires moral significance insofar as it serves as the indispensable starting point for moral improvement (Muchnik 2010, p. 138; Auweele 2015, p. 379).

In essence, moral progress represents the final end of “Part One” of Religion. This final end, however, is somewhat obscured by the title “General Remark” given to the final section of “Part One” in the second edition of Religion (1794). In the original version of “Part One”, titled “On the Radical Evil in Human Nature” and published in the Berlinische Monatsschrift (1792)—as well as in the first edition of Religion (1793), this section—titled “Concerning the Restoration to Its Power of the Original Predisposition to the Good”—was positioned as the fifth section, carrying equal theoretical weight and serving as the concluding section (Kant 1792, p. 369; 1793, p. 45). Therefore, Kant’s idealization of the evil propensity into radical evil through idealized conscience-based imputation ultimately serves the dual-perspective purpose of personal moral progress: the revolution against evil disposition from the intelligible perspective and the gradual reform of the evil propensity from the empirical perspective (Rel. 6:48). The second edition of Religion may also represent Kant’s intention to provide a moral education. He asserts that to understand the Religion’s essential content, “nothing more is required than common morality”. He further notes that this content is expressed in “the most popular instruction for children or in sermons, and is easily understood.” (Rel. 6:14. For a more detailed interpretation of the educational significance of Religion, see Auweele 2015, pp. 373–75).

Similarly, the idealized conscience-based imputation serves as a prerequisite for moral improvement. In his Lectures on Ethics, Kant observes that “one who binds oneself to conscience” does not reinforce objective obligation but rather enhances subjective commitment—that is, it elevates an action’s moral worth beyond mere legality (Vigilantius 27:575). Even in the “Doctrine of Virtue”, where this concept is marginalized, its idealized form persists implicitly. Kant defines self-reflection as the attempt to fathom the depths of the heart’s “abyss” (“the descent into the hell of self-knowledge”). Only by exposing radical evil—“the inner obstacles (of an evil will nest in him)”—can one eliminate it and thereby achieve moral wisdom (MS 6:441; cf. Cherkasova 2005, p. 579). When discussing conscience, Kant still portrays it in idealized terms as a “scrutinizer of the heart”, comparing it to a divine court. Yet he clarifies that this as merely an analogical representation (MS 6:439–440).

Moreover, Kant’s conception of moral progress through conscience extends beyond the personal sphere. The metaphor of a divine court, for instance, forms part of his broader project to cultivate collective morality among human beings. It has been noted that Religion envisions moral progress not merely as an individual pursuit but as a collective good (Muchnik 2010, pp. 117–18). In Religion, Kant characterizes the ethical community as the “Kingdom of God on earth” (Rel. 6:101), conceived as an ethico-civil state in which all individuals comply with the moral law as though legislated by God. Given that radical evil constitutes the primary obstacle to the collective good, the ethical community arises as an indispensable corrective: it overcomes the ethical state of nature, which is a condition of mutual corruption driven by radical evil (Klemme 1999, pp. 141–45). Significantly, conscience appears here in its role as “scrutinizer of the heart”, reinforcing the idea that idealized conscience-based imputation is foundational to the ethical community (Rel. 6:99). Thus, each person’s idealized conscience-based imputation establishes the church as the optimal context for moral improvement. This community fulfills a pedagogical function, thereby justifying “the necessity of a divine lawgiver that scans the Gesinnung of the human being” (Auweele 2015, p. 389).

Consequently, the doctrine of radical evil and the idealized imputation in conscience are necessary to both individual and collective moral progress through religious education. The religious terms merely serve as analogical representations of this doctrine, which could be termed “moral progress under a religious mantle”, contrary to Goethe’s critique. This assertion is partly supported by Kant’s 1770s Lectures on Anthropology, which demonstrate that the conscience-based imputation represents a consistent element of Kant’s ethics, aimed at enhancing collective morality, despite not being fully theorized at the time. Here, Kant presents his view on religious education: individuals cultivate their conscience through internal reflection, which compels them to act according to moral principles via self-imposed constraints. In this way, the kingdom of God on earth takes shape, and conscience emerges as the supreme judge (Friedländer 25:693). The ultimate aim of this kingdom is to realize the highest degree of human perfection (Friedländer 25:695).

The final stage of my argument addresses the central question: why, then, are the idealized conscience-based imputation and the conception of radical evil marginalized in Kant’s later work, even though they remain essential to his ethics? Building on previous analysis, I propose an education-related interpretation: Kant’s shift in focus toward moral education as the means of collective moral advancement emerges as a contributing factor, potentially even the primary one.

In the 1770s, as evidenced by Lectures on Anthropology, directly engaging with Basedow’s education reform, Kant articulated a robust Enlightenment optimism that historical progress could be driven “from bottom to top”: teachers and priests should be the first instructed in morality, after which the ruling class would be educated by them. Thus, morality would “soon ascend to the throne”, and through their influence, “the whole nation could be educated” (Friedländer 25:697; Kant 2004, p. 367).

By the 1780s, Kant had revised his original plan, arguing that the efforts of citizens (“bottom-up”) and government cooperation (“top-down”) would “meet each other” (Menschenkunde 25:1202). Finally, in 1798, Kant abandoned his earlier bottom-up vision of historical progress. He concluded that the plan to educate “the youth in home instruction and further in schools, from the lowest to the highest, in spiritual and moral culture reinforced by religious teaching” was unlikely to achieve the desired results (Streit 7:92; cf. Louden 2011, p. 143). This evolution signifies not a rejection of education’s necessity, but rather a strategic shift in focus—from formal education to the enthusiasm inspired by the French Revolution (cf. Brandt 2007, pp. 184–90). Thus, notwithstanding its continued importance in Kant’s work (particularly in the “Doctrine of Method” of the Metaphysics of Morals), moral education is no longer posited as necessary for collective moral progress.