LAWRENCE PASTERNACK | Guidebook to Kant on Religion within the Boundaries of Mere Reason | Routledge 2014


By Allen Wood

Pasternack’s Guidebook is a section-by-section commentary on Kant’s Religion within the Boundaries of Mere Reason. It sheds light on many dark corners of one of Kant’s more difficult texts. Its strength is that it provides for each part of it a thoughtful and informed discussion with which a reader of Kant’s Religion can compare his or her own readings and reactions. Pasternack avoids many (if not all) of the errors common to readers of the Religion; he offers a refreshing new approach to many topics. His reading of the Religion attempts to see Kant within the context of Christian theology—more specifically, the orthodox Lutheranism of Kant’s own time and place.

Pasternack’s fundamental aim in the book is to present Kant’s position in such a way that it can be seen to have successful replies to certain criticisms that have recently been levelled against it: for instance, Gordon Michalson’s charge that Kant “wobbles” inconsistently between traditional Christianity and Enlightenment rationalism, and John Hare’s declaration that without a genuine faith in Christ’s saving work, Kant’s position is a “failure” because it leaves a “moral gap” between guilt and justification (Pasternack, pp. 2–3). Pasternack is not afraid to point out where he thinks Kant’s position differs from Christian orthodoxy, but at some points—especially concerning Kant’s doctrine of divine grace and its relation to Pelagianism—I believe he misses some of the subtleties of Kant’s position, and fails to appreciate the ways Kant carefully steers a course between, on the one hand, mere acquiescence in Christian orthodoxy and, on the other hand, doctrines so radical his Lutheran readers could not accept them. It was Kant’s aim to appeal to traditionalist Christians, urging them to rethink Lutheran orthodoxy in rationalist terms, but never putting them in a position where they would have no option but to reject either Christian orthodoxy or else Kant’s teachings in the Religion. Many readers of the Religion believe he failed in this aim, but a sympathetic reader ought to reply to them by showing fuller awareness of what he was trying to do. I believe that was Pasternack’s intent, but I don’t think he is always as successful in carrying it out as he might have been. In my comments I will try to suggest some ways he might have done better.

These comments will address only certain parts of Pasternack’s book, namely, topics in Parts One and Two of Kant’s Religion. These limits are simply for reasons of space, since my paper will be quite long enough as it is. I divide the comments into three parts, and one appendix (which will be posted separately as Part B):


Pasternack denies that Kant’s position is either Deism or Agnosticism (p. 3). I think he is in the end correct, but neither issue is simple or unproblematic.

Deism. This is certainly correct if ‘deism’ is taken in Kant’s own sense of the word, where it refers to a conception of God that admits only ‘ontological’ predicates and refuses all predicates drawn from finite things, especially the personal predicates of intellect and will (A631/B659; AA 28:1001). For Kant, religion is the recognition of our moral duties as divine commands, which presupposes the concept of a God who can will, command, cognize our failings and react to them as a person. So Kant rejects ‘deism’ in that technical sense. But that is a very non-standard meaning of ‘deism’ as the term is used in the modern period. (I think it is adopted by Kant mainly so that he can deny that the term applies to him.) Instead, ‘deism’ usually refers to a position that recognizes a natural or rational religion that is not dependent on empirical revelation (e.g., through scripture, inspiration or ecclesiastical tradition). The very title of Kant’s Religion within the Boundaries of Mere Reason refers to precisely the position more conventionally called ‘deism’.

It nevertheless remains problematic to apply the term ‘deist’ to Kant’s own position in the Religion. This is because it is arguable that Kant holds that the natural religion of reason may, at least historically, require some kind of revealed religion—certainly for its origin, and perhaps even also for the fulfilment of its own aims. According to Kant, the title of Religion within the Boundaries of Mere Reason refers only to the inner of two concentric circles, allowing for a revealed religion as the outer circle (AA 6:12; Pasternack, pp. 3, 77). It is difficult to say precisely how or in what ways Kant thinks the outer circle is needed by the inner circle. Much would depend on Kant’s treatment of particular theological issues and doctrines, and on how a position like Kant’s might be creatively extended. Kant may think that what belongs to the outer circle is optional or indifferent, not required for saving faith, but still rationally acceptable. He does seem to think at times that it is destined to be gradually removed or cast aside (AA 6:115–24). But Kant may also regard this progressive process as indefinitely long, never actually to be completed. Further, he may also think that parts of a revealed religious tradition may be permanently acceptable as part of rational religion, if they are given the right (rational or moral) interpretation. Further, Kant seems (as we will see presently) to want to integrate parts of the revealed tradition itself (e.g., the Christian doctrine of original sin) into rational religion, by giving it a rational-moral interpretation (as the thesis that human nature contains a radical propensity to evil). So there may be something about the religion of the outer circle that will be necessary indefinitely: without it, natural or rational religion (‘deistic’ religion, in the commonly accepted sense of the word) may be incomplete even in relation to its own aims.

Agnosticism. Pasternack also denies that Kant’s position is Agnosticism, even though Kant thinks (1) God’s existence is theoretically unknowable either way (A639–40/B667–8) and the concept of God, since it is an idea of reason, entails that God can never be an object of theoretical cognition by us; and further (2), Kant also holds that the “minimum of theology” necessary for religion consists only in assent to the proposition that God is possible—so that we may say that if there is a God, then our duties are commanded by him (AA 6:154n; AA 28:998). In that sense, Kant holds that even religion, as he understands it, is compatible with certain kinds of agnosticism. Kant does think, however, there are practical grounds for assent to God’s existence. The (agnostic) “minimum of theology”, while still religiously acceptable, is not for Kant the optimal moral or religious position. Kant’s theoretical position in rational theology is therefore agnostic in a straightforward sense, but Kant nevertheless does not regard agnosticism as the most defensible position for a rational and moral human being to adopt.

Kant’s “pure rationalism” in regard to religion. In Part Four, Kant distinguishes “pure rationalism”—the position that allows and accepts divine revelation, but denies that it is necessary for religion—from both “naturalism”, which denies such revelation and “supernaturalism”, which holds that faith in revelation is necessary (AA 6:154–5). I have criticized the “pure rationalist” position, so described, as implausible on the ground that it holds there are divine commands but that we have no obligation to carry them out.[1] If taken in this sense, pure rationalism does seem a rather pointless position; but Pasternack observes that pure rationalism can be understood more sympathetically: as allowing the possibility of a divine revelation whose acceptance may not be necessary for saving faith (seligmachender Glaube) but may be religiously permissible and even religiously helpful for some if the revealed doctrines are interpreted in a way that makes them serviceable to a pure moral religion (pp. 13–14, 219–23). On this more sympathetic interpretation, I agree with Pasternack that Kant is a pure rationalist. In my comments below I will, in effect, be defending Kant’s position as a pure rationalist one in just this sense. Thus my earlier deprecation of pure rationalism must stand corrected if the term is understood in this more sympathetic way, as I now think it should be. The rhetoric of Kant’s definitions at 6:154–5 even seems designed to present pure rationalism as the best position.

According to Pasternack, the religious motivation for pure rationalism is that a saving faith must be “equally available to all”, “convincingly communicated to everyone” (p. 4; AA 6:103). I believe this criterion for “saving faith” is not correctly understood if it is taken to draw a distinction between revealed doctrines that are available only to those who have read certain scriptures or heard certain preachings, and rational or philosophical doctrines that are supposedly accessible equally to everyone, whatever they may have read or heard. Understood this way, Kant is open to the objection that critics of deism (e.g., Bishop Butler, in The Analogy of Religion) have often mounted against it—namely, that the doctrines of the rationalist philosophers are also available to people only through their empirical dissemination, and this too is limited to those who have been taught by rationalist philosophers. If we compare the two regarding their dissemination (how far the “tidings” have been made known), so the anti-deist argument goes, then perhaps it is the Christian Gospel that has had the wider availability.

This point seems well-taken if “equally available to all” is understood as referring only to the extent of a teaching’s empirical dissemination. Kant’s criterion, however, is more defensible if we emphasize the words “convincingly communicated”. Rationalists such as Kant rightly insist that empirical knowledge, its transcendental foundations, and rational morality are such that any human being whose intellect has been sufficiently developed can comprehend and accept them. By contrast, revealed teachings (whether of the Old Testament, the New Testament, the Holy Koran, or any other book represented to us as inspired prophecy) must rest on the authority of its prophetic human author or authors, and usually also the authority of its learned or ecclesiastically empowered interpreters. These matters, even according to the doctrines of revealed theology itself, can never be made credible to all, for they do not even recognize universal human reason as the proper criterion for their acceptance. They appeal instead to a faith given by divine grace, or to the authority of tradition and church; even according to them, these exceed all rational grounds on which their credibility might rest. According to Kant, moreover, such supernatural revelation cannot be self-validating, and no rational or empirical grounds could ever suffice for it to be credible. God is an idea of reason, to which no experience could provide us with any cognition. “If God could really speak to a human being, the latter could never know that it was God speaking” (AA 7:63). If supposedly revealed doctrines can be made universally credible, that could be only through the way they might serve as vehicles for a pure rational religion whose foundation is universal human reason, and not revelation, authority or inspiration.

Following Kant’s own words, Pasternack describes the method of the Religion as the conducting of two “experiments” (pp. 2–4; AA 6:12). The first experiment is to develop a conception of pure rational religion or religion, within the boundaries of mere reason (in other words, a deistic religion in the commonly accepted meaning rather than Kant’s technical meaning). The second experiment is to compare the contents of this deistic religion with those of a revealed faith (specifically, that of Christianity), and see how far the pure rational religion may be helpfully supplemented by a revealed faith, or even how, as a matter of historical necessity, pure rational religion may require revealed religion or “ecclesiastical faith” based on a revealed scripture, as its vehicle (AA 6:101–2). More specifically, then, Kant’s aim is to explore the way in which Christianity may be viewed, and also the way it must be interpreted, if it is to serve as a vehicle of rational religion.

Publication of the ‘Religion’. This aim is closely related to Kant’s legal difficulties in getting the Religion published, and the subsequent reproaches aimed at him by King Friedrich Wilhelm II through his minister J. C. Wöllner (AA 7:6–7). Kant argued in reply that his subject in the book is philosophy and not theology, and that it cannot have “disparaged” Christianity, because it made no “appraisal” of it (AA 7:8).

Here Pasternack seems sceptical, and to side rather with Kant’s critics, whom he has told us he intends to refute. He suggests that Kant may not have been entirely innocent of “guile” or “prevarication” (p. 12). I think he is wrong. In Conflict of the Faculties, Kant carefully distinguishes in terms of their functions within the political state, the faculties of philosophy, theology, law and medicine, and he locates the aims and subject matter of the Religion within the province of the philosophy faculty, not the theology faculty. It is true that in this work Kant does offer an interpretation of Christian doctrines. But he insists that this hermeneutic of scripture is not intended as an authoritative ecclesiastical exegesis. Instead, it is a purely philosophical exercise, aiming only at the possibility of reconciling Christianity as a revealed religion with the pure religion of reason. In order to view Kant as having appraised and disparaged Christianity, you would not only have to reject Kant’s own rational exegesis, declaring it to be a falsification of Christianity, but you would also have to reject even his aim of reconciling Christianity with a religion of reason, and the criteria of acceptability associated with this aim. These imply that an exegesis of Christian doctrine might become acceptable if it is the only (or the best) way of achieving such a reconciliation. Moreover, to find guile or prevarication in Kant’s procedure, you would have to see these aims themselves as not being what Kant himself says they are, but something else. (And what then are you claiming that to be?)

In my opinion, Kant’s aims are clearly and honestly stated. You may think Kant’s attempt to reconcile Christian doctrines with rational religion does not succeed—or perhaps you think he could have done a better job of reconciling Christianity with reason, or conclude that Christianity cannot be reconciled with reason at all. But Kant’s project presents itself honestly for what it is, and the interpretation of Christianity it offers on these terms is not meant to appraise Christianity negatively or disparage it, but on the contrary to defend Christianity—conditionally, from the rational or philosophical point of view adopted in the Religion. I would sooner indict Kant’s accusers of guile or deception, for refusing accurately to represent his aims.

In his Introduction, Pasternack rightly devotes attention to the conditions under which the Religion was published, since these shed light on its contents. The Religion was originally intended to appear as four long articles in Biester’s Berlinische Monatschrift. But Kant’s tussle with the censors forced him to publish it as a single book, based on approval by the philosophical faculty at Jena (which made it entirely legal, at least on Kant’s theory of the matter). Pasternack seems to me correct in thinking that this forced change was salutary, since it required Kant to provide a much more systematic and consistent treatment than he might otherwise have done. Pasternack helpfully offers a chiastic account of the structure of the Religion. The first two parts, he suggests, deal with the individual, the second two, with society. But the two middle parts deal with the redemption offered by religion, while the two outer parts deal with the corruption which makes redemption necessary (p. 9).

Part One of the Religion deals with individual corruption, sin or the “radical evil in human nature”, while Part Two deals with the way the individual may be redeemed. Part Three argues that because the origin of sin or radical evil is social, the struggle against it must also be social, and it provides an account of the ethical community required for this struggle, as well as an account of the history of ecclesiastical faith (especially the Christian faith) regarded as an empirical, historical vehicle through which the aims of an ethical community may be realized. Part Four deals with the corruption of this ideal in past and present ecclesiastical faith, involving “fetishism”, “superstition”, “priestcraft”, and the “counterfeit service of God”. Pasternack discusses all four parts, and even the subsections within each. Each of Chapters 3–6 of Pasternack’s commentary are devoted to one of the four parts. My comments, as I have said, will concern only certain aspects of Pasternack’s treatment of Parts One and Two, which are found in Chapters 3 and 4 of his Guidebook.

Before getting into his actual commentary on the Religion itself, Pasternack devotes a long Chapter 1 to Kant’s doctrine of the highest good. I won’t comment on this discussion, except to say that I think it is flawed—both too brief and too ambitious. I think he would have been wiser to have begun his commentary on the Religion with a discussion of the Prefaces, which he does in Chapter 2.

Part One: The radical propensity to evil in human nature

Readers of the Religion have spent more time on this part of the Religion than any other. Perhaps this is due to the deplorable tendency of the readers of any book to concentrate on what comes early in it rather than what comes later. This has resulted in many books being badly misunderstood. In my view, however, readers of the Religion have also been especially creative in misunderstanding this First Part itself.[2] Pasternack gets right four important points on which many other writers go conspicuously wrong.

(1) Pasternack realizes that for Kant, the term ‘radical evil’ does not refer to some especially egregious species of evil, but rather merely to the common root (radix) of human evil, whatever its degree (p. 125). (Deliberately ignoring Kant’s meaning altogether, writers on the subject of evil sometimes attack his doctrines for being mistaken when ‘radical evil’ is understood in their sense rather than in the sense Kant meant. Or sometimes even when they realize Kant’s meaning is not their favoured one, some writers on the topic of evil attack Kant simply for not meaning by the term what they want to mean by it.)

(2) Pasternack realizes that the thesis that there is in human nature a common, radical, universal, innate and inextirpable propensity to evil is for Kant an empirical thesis. It is not something that could ever be demonstrated by a priori proof; still less does Kant himself propose to offer such a proof (pp. 99–101). It is remarkable that so many writers go wrong here, because Kant is just about as explicit on this point as he could possibly be:

However, that by the ‘human being’ of whom we say that he is good or evil by nature we are entitled to understand not individuals (for otherwise one human being could be assumed to be by nature good and another evil), but the whole species—this can be proven (bewiesen) only later on (weiterhin) if it is shown in anthropological research that the grounds that justify us in attributing one of these two characters to a human being as innate are of such a nature that there is no cause for exempting anyone from it, and that the character therefore applies to the species. (AA 6:25–6)

This passage states with crystal clarity that the thesis of radical evil is empirical and anthropological: any “proof” (even any “formal proof”) of it would involve not transcendental argument but empirical evidence (“anthropological research”). It also states quite explicitly (with the word weiterhin) that no such proof yet exists.[3]

(3) Pasternack also sees that the radical evil in human nature, as Kant understands it, is connected to our unsociable sociability, and to our predisposition for humanity as it unfolds in the social condition. Kant states this too with unmistakable clarity (pp. 94–5, 109–10).

(4) Pasternack wisely distances himself from the extremely common (but absurd) tale of extravagant transcendent metaphysics according to which Kant dogmatically asserts that the human will is free only in a supernatural intelligible world hovering somewhere above the clouds (pp. 20–6). He recognizes the noumenal in Kant as only a limiting concept, and more generally, transcendental idealism not as a set of extravagant metaphysical doctrines but a device for constraining our ambitious reason within the bounds of what can be cognized. Kant never offers us a positive “theory of freedom”, but only the critical thought that we can conceive of freedom only by transgressing the boundaries of our cognition. This should be taken as a declaration that although freedom can never be proven or disproven, it is incomprehensible and the problem of its possibility is forever insoluble by us (A555–8/B583–6; AA 4:459). It may be understandable that those who find this a frustrating and unwelcome message would be motivated to punish the messenger by ascribing to him a ridiculous metaphysical theory of freedom. It is harder to see why those sympathetic to Kant would saddle him with such nonsense.

Does Kant himself accept the thesis of radical evil? This is a question worth raising for two reasons:

First, the thesis of radical evil in human nature that Kant puts forward, explicitly admitting that it is both empirical and as yet unproven, is extremely strong. It claims not only that evil is possible for us humans, but that there is in all of us a positive propensity, an “enticement” or “concupiscence”, to choose it, which is culpable and chosen by each of us. He also claims that this propensity is

  1. universal: present in every human being without exception,
  2. inextirpable or ineradicable: not removable by us through any deed we can perform,
  3. innate: not in the sense that it is literally inborn, or of biological origin, but in the sense that there is no time in our lives after our birth when it is not already present in us (AA 6:31).

The universality, innateness and inextirpability of the propensity are such that Kant claims that although it is the work of our freedom, it also belongs to human nature itself (AA 6:32). Why would Kant accept such a strong thesis, given that he admits it is as yet unproven?

Second, when we turn to the very end of the Anthropology (published three years after the Religion), Kant is more cautious concerning the moral character of our species. If we ask, he says, how our species compares with other possible rational species we can imagine, then

[the evidence of experience] presents the human species not as evil, but as a species of rational beings that strives among obstacles to rise out of evil in constant progress toward the good. In this its volition is generally good, but achievement is difficult because one cannot expect to reach the goal by the free agreement of individuals, but only by a progressive organization of citizens of the earth into and toward the species as a system that is cosmopolitically united. (AA 7:333)

This is at least minimally consistent with the Religion. For it says we begin from evil, strive towards good, and require some pertinent form of social organization if this striving is to be successful. But it falls far short of ascribing to human beings the universal, innate and inextirpable propensity to evil asserted in the Religion. This is quite appropriate if we understand Kant to think that the thesis of radical evil is not yet proven.

In the Religion itself, however, Kant does appear to assert and presuppose throughout the full thesis of radical evil. His important doctrines in Parts One and Two regarding the Change of Heart and the redemption of humanity presuppose it. He claims we are justified in accepting it, despite the absence of sufficient “formal proof”, based on our awareness of our own evil actions and what may be inferred from them about the deepest maxim that underlies them (AA 6:20), as well as the “multitude of woeful examples” of human evil in both the uncivilized and civilized conditions (AA 6:32–4). Kant never offers convincing evidence of it, however—still less the “transcendental proof” absurdly and gratuitously attributed to him by many writers on the Religion.

So why does he accept it? In this work, I suggest, Kant is addressing chiefly an audience of Lutheran Christians, who already believe in the doctrine of original sin, and do not need to be convinced of it. His aim is therefore not to convince them of what they already believe, but rather to accept the proposition that rational moral philosophy can give an adequate account of the contents of their belief. If we today are more sceptical about the thesis of radical evil, we should keep firmly in mind that Kant was not writing for us.

I think even Pasternack may not fully realize how ambitious Kant’s claim of radical evil really is. Those writers who think they have some kind of a priori proof for it (or that Kant has one, which they are kind enough to spell out for us on his behalf) clearly underestimate how extreme it is. No claim that human beings have a propensity to evil that is universal, innate and inextirpable could possibly be proven a priori—if it could ever be proven at all. To say that we have a propensity to evil is clearly a far stronger claim than saying we are not holy beings. The latter says only that evil choices are possible for us, while the former posits in us some subjective ground (albeit one we have brought on ourselves by choice) to commit further evil acts, and a positive concupiscence (Concupiscenz, Anreiz) or temptation to commit evil acts (AA 6:29, 213). Pasternack (inconsistently with his main insights) seems at times to treat the universality of evil as something having an a priori status, or as the result of one of several “a priori steps” (pp. 103–4). At one point he appears to think that we can infer a priori that if a human being commits moral evil, then he must have a propensity to do so—as though the only alternative to this would be to ascribe to the human being a species of mere imperfection that would not amount to true evil (p. 107). This inference seems plainly invalid, unless it drastically underestimates the strength of the claim that human nature contains a radical propensity to evil. It is one thing not to have a holy will—that is, for it to be possible for you not to obey the moral law—and a much stronger thing to have a propensity not to obey it.

Pasternack does not provide much discussion of the inextirpability of evil—the thesis that it cannot be finally done away with by any deed of ours in time (AA 6:37, 51). In one footnote, he even mistranslates unvertilgbar as “inextricable”, and he seems to confuse the inextirpability of the evil propensity with the idea that “even after the Change of Heart, evil may come to dominate our lives” (p. 129). This would be a quite different point. Even if we could (and even if we did) extirpate the propensity to evil, we could still do evil, and we could even subsequently let evil dominate our lives, as long as we were free beings who lack a holy will. (Pasternack’s thought here also presupposes a serious misunderstanding of Kant’s conception of the Change of Heart. He thinks it must be a datable event. This is a topic to which I will return in discussing Part Two.)

Pasternack seems also at one point to be confused about Kant’s claim that the propensity to evil is “innate”—taking this to mean that for Kant it is “in one sense an innate fact beyond our choosing” (p. 126). Kant is about as clear on this as anyone could be: by the “innateness” of the propensity, he does not mean that it is present in us at or prior to birth, still less that it is something not chosen by us. He means only that there is no identifiable point in our lives when it was acquired (AA 6:38). This could be true even if, as Kant repeatedly asserts (and Pasternack also affirms), the propensity to evil must be the work of free choice (pp. 108–10).

Pasternack seems at points to make curious concessions to Allison, Palmquist, Morgan and others who believe in the mythical “a priori (transcendental) proof” of the thesis of radical evil. He says that there are “several a priori steps in Kant’s moral anthropology” (p. 109). I would agree with him if all he means is that Kant makes two conceptual (analytical) claims that form part of his discussion. These are:

  1. that evil must result from an act of freedom, a choice of some other incentive ahead of the moral incentive, and not from the presence of mere sensuous inclinations (AA 6:34–5); and
  2. Kant’s “rigorism”, that a choice not to obey the moral law must involve the preference of some non-moral incentive ahead of the incentive of the law (AA 6:24).

But these are never a priori “steps” in Kant’s anthropology, or in any reasoning he might have used to support the thesis of radical evil. They are only conceptual points that clarify the concept of evil, and therefore help define the question whether a human will might be called good or evil.

Diabolical evil. Kant has regularly been criticized for denying the possibility of a diabolical will, an “absolutely evil will” or “evil reason”—a reason, as Kant says, for which “resistance to the law [is] raised to an incentive” and which, Kant infers, would therefore have to be “exonerated from the moral law” (AA 6:35). Pasternack seems to me to be aware of the essential point, namely, that the conception of such a will would be incoherent, based on the conceptual point (pp. 120–3, 130). Namely, a will with an original incentive to violate the law would thereby be exonerated from it and hence cannot be called evil (or therefore diabolical either, if being diabolical is a way of being evil). Inconsistently, however, Pasternack appears to agree with the critics. I find this both puzzling and disappointing.

Kant does not deny that there are “diabolical vices”—or inclinations to do what one knows is contrary to duty (AA 6:27, 385–8, 458–61). These include the desire to make the unhappiness of another itself an end, contrary to our duty always to include the happiness of others among our ends. This desire shows itself in envy, ingratitude and malice, which clearly involve the awareness that making others unhappy is evil, and the inclination to do this in part precisely because it is evil. In this sense, Kant holds that human wills do evil for its own sake.

On a couple of occasions where I have been present, Rae Langton has offered the related criticism that Kant has no place for vices that take the form of degrading or humiliating others—that is, aiming not at making them unhappy but at showing them disrespect for its own sake. I think she is mistaken.[4] Kant focuses on hatred (and making the unhappiness of others an end) rather than disrespect (and making the degradation of others an end), only because he treats the happiness of others as the encompassing duty of virtue (the morally required end) we have regarding them, and he therefore understands diabolical vices as directly opposed to this end. Kant regards duties of respect as limiting our conduct rather than as setting an end. But I think the human desire to humiliate or degrade others is something whose existence he clearly acknowledges, and also includes under diabolical vices. It seems clear that the motivation Kant sees lurking behind envy and ingratitude is our resentment of the way that the other’s superiority or beneficence puts us in an inferior position to them, and a desire to do away with it by harming or lowering the other. Thus the desire to degrade those we envy or hate is for Kant even the motivating force behind our desire to make them unhappy. For Kant, degrading others lies at the core of all the diabolical vices.

Related to the criticism that Kant does not allow for diabolical evil is the charge that he has too narrow a notion of what evil is, treating it only as self-interest. Pasternack himself even puts it this way (p. 119). But this is based on a superficial misunderstanding. When Kant speaks of subordinating morality to “self-love”, the latter must be understood only as a placeholder for whatever non-moral incentives are given preference to those of reason in a particular evil choice. It is obviously not true that Kant regards all evil as self-interested, or self-interest as the deepest enemy of morality. Kant often suggests that it is not self-interest but rather the ‘diabolical’ vices of hatred: envy, ingratitude, malice, vengeance, that are the worst that human nature has to offer (AA 6:27, 458–61). As evidence for the evil in human nature, he conspicuously cites not self-interested behaviour but, on the contrary, “never-ending cruelty […] from which no human being derives the least benefit” (AA 6:33).

Pasternack cites Milton’s Lucifer (an august authority, regularly appealed to by all sides in these disputes) who exclaims: “Evil be thou my Good!” (p. 121). Like many who appeal to Lucifer’s authority, Pasternack presents Lucifer’s choice as a possible counterexample to Kant’s claim that diabolical evil is impossible. To me this makes no sense. When Lucifer says: “Evil be thou my Good!” he is making a choice: He is choosing henceforth to treat Evil in the same way the incentives of reason would lead him (or any creature who understands the concepts of Good and Evil) to treat Good. Lucifer, as a volitional agent, necessarily understands Good as an incentive of choice, even the rationally supreme incentive. He apparently has a competing (non-moral, counter-rational) incentive (arising from satanic pride or whatever it is), which he thereby chooses systematically over the original rational incentive to Good. That choice is precisely what constitutes Evil. That agrees completely with Kant. Such a choice is the only thing that “choosing Evil to be your Good” or “doing evil for evil’s sake” could possibly consist in.

Perhaps there are people who think of those they hate (Satan, Nazis, Communists, Terrorists, Liberals, Repuglicans) as being “so evil” (as these haters would put it) that these hateful beings have only an incentive to do what is wrong or bad, and no competing incentive at all to do what is right or good. In that case, however, the haters have deprived themselves of any way of explaining coherently why they consider those they hate to be evil. For they themselves admit that these beings have no reason or incentive to behave differently from the way they do behave. So where do we get off blaming them or calling them evil? All that could be left of the claim that these hated creatures are Evil is that they have different tastes from ours—we who consider ourselves good people (but it is no longer clear why we could be justified in this). If it comes down only to a matter of differing tastes or preferences, then by our own admission Satan or the Nazis, etc. could with precisely equal justification call us Evil. (And our hatred of them might lend further justification to that.) What I fear happens here is that hatred so blinds the intellects of these haters (the haters of Satan, Nazis, Terrorists, etc.), that they have formed a conception of evil so extreme that it has become incoherent and self-undermining. When they read Kant’s account of what evil has to consist in, they criticize him for failing to ratify the correctness of their own incoherently malicious thoughts. They themselves seem to have become clear instances of those very diabolical vices of hatred whose existence Kant clearly affirms.

Pasternack, at his better moments, seems aware of all this, but he still insists that Kant allows for “the logical possibility” of a diabolical will, denying only that the human will could be diabolical (pp. 122, 130). He never cites any passage in which Kant asserts the logical possibility of something Kant clearly argues to be logically impossible. Kant does not “wobble”, but on this topic, as on some others, Pasternack sometimes seems to wobble between his own best understanding of Kant and common interpretations or criticisms of Kant, which are inconsistent with his best understanding, and which he sets out to correct.

Part Two: The Change of Heart and divine forgiveness

Part Two of the Religion, which deals with the individual response to evil and the means through which God might provide us with a way of justifying ourselves in spite of it, is the most difficult part of the book. For this reason, perhaps, it is not true that it has been overworked in the literature. Pasternack is to be commended for tackling a set of difficult problems. I think there is much in his discussion that is worth pondering, and even accepting. But as with his discussion of radical evil, I find some parts of it unclear, and other parts highly questionable.

In broad outlines, the picture Pasternack draws seems to me this, and also seems to me accurate as an account of Kant’s doctrines:

  • The correct response to radical evil on the part of the individual is an equally radical conversion or Change of Heart, which puts the moral agent on the path from bad to better. This fundamentally consists in adopting a good “disposition” (Gesinnung) (pp. 131–3).
  • Despite this conversion, the human being still apparently lies under a kind of claim, analogous to a legal claim, as a result of the fact that we all began from evil. In order to hope for participation in the highest good, the human being must somehow be saved from or absolved of this quasi-legal judgement (pp. 141–62).
  • Some role in this salvation is played by the personified ideal of the good principle, the ideal of humanity well-pleasing to God. This corresponds to the Christian idea of Jesus Christ and his vicarious atonement for our sins (pp. 133–41).
  • Some role is also played by the moral agent’s belief in divine aid, either aid that makes possible the Change of Heart, or aid in absolving the moral agent through a divine act of atonement, resulting in a divine verdict releasing the moral agent from the quasi-legal claims of the evil principle (pp. 142–60, 232–4, 248–54).

It is important to have gotten these general outlines correct, because Kant’s text does not always present them clearly or perspicuously. It is even harder, however, to get the details right. Here is where I wish Pasternack had done better. That would involve getting right Kant’s “second experiment”—seeing clearly how far Christianity as an ecclesiastical faith can not only be reconciled with pure rational religious faith but also how far it might serve as a vehicle for it, and how, interpreted symbolically and rationally, it might even satisfy some needs rational religion may not be able to satisfy all by itself without the larger concentric circle of faith.

Kant was raised as a Pietist Lutheran. The Pietists, or at least many of them, believed that we sinners are utterly powerless to do anything to save ourselves or to make ourselves worthy of salvation. Everything done towards our salvation must be done by God’s grace alone. All we can do is wait. When divine grace comes, they believed, we can feel it operating on us. This happens in a conversion (a “born again”) event that changes the sinner’s life. According to many Pietists, there can be no salvation without the experience of this saving event. They also thought that our belief in Jesus Christ’s vicarious atonement for our sins through his crucifixion and resurrection was essential to our salvation. These doctrines are the way they understood the biblical declarations that we are saved by faith and not by works.

I think the best way of understanding Part Two of the Religion is to understand what Kant is saying as a reaction to this set of beliefs. Kant finds some of the Pietist beliefs acceptable, if they are given a rational interpretation—one consistent with our capacities for knowledge, the conditions of moral responsibility, and with attitudes conducive to our moral betterment. But Kant rejects significant parts of the Pietist story, or accepts them only in a highly qualified way and on an interpretation that many Pietists might not have found acceptable. Kant also regards parts of the story as permissible to believe, but morally indifferent, and therefore insists that they cannot belong to a saving faith. He does his best to frame his account so as to enable traditional Christians to accept this optional part of the story, as long as they recognize that it is optional. They must not regard intolerantly those who do not accept it, and they must not treat the merely optional parts as essential to their own salvation. If not properly understood, Kant’s careful attempt to make these distinctions can appear to “wobble” between traditional Christianity and Enlightenment rationalism. But that is only if you have ceased to care what makes rational moral sense and what does not. To Kant, that basic distinction is what determines which interpretations of Christian doctrine are acceptable and which not, and which are permissible but optional to believe. I think Pasternack would agree with all this. But he sometimes fails to see how Kant succeeds in accommodating the optional parts of the Christian story.

The Change of Heart. Pasternack writes as if the conversion or Change of Heart is seen by Kant as a determinate event occurring in time during the life of the sinner. That, as I have said, was the view of many of the Pietist co-religionists of Kant’s childhood and youth. Above we saw that Pasternack misunderstands the inextirpability (or “inextricability”) of radical evil as the claim that evil may still come to dominate our life even after the Change of Heart, implying that it is a datable event in someone’s life (p. 129).

Kant no doubt wanted the Religion to be at least minimally acceptable to people who thought this way, but if we look carefully, we can see that Kant himself did not in fact understand the Change of Heart in this way at all. Kant distinguishes the revolution in way of thinking (Denkungsart)—which is thought of as a “single and unalterable decision”—from the “gradual reformation” in conduct which is the way this revolution must show itself.

Kant was all too familiar, I speculate, with Pietists who declared on Sunday—through tears of joy and repentance—that the hand of grace had made new people of them. Their hearts had been changed, their sins forgiven, their souls saved. Then on Monday he saw these same people promptly going back to their bad old ways just as if nothing had happened. The only change in them was the acquisition of a new layer of self-righteousness in the form of the complacent conviction that they are now saved. Kant wanted no part of this gushy, pretentious tartuffery. For him, the only sign of the Change of Heart that matters is that the human being should “find himself upon the good (though narrow) path of progress from bad to better” (AA 6:47–8). Whether this progress began with a dramatic “conversion event” might sometimes be a contingent fact about this or that person, but for Kant this is an adiaphoron: it has no moral or religious significance whatever. Further, Kant is convinced on Critical epistemic principles that the divine “hand of grace” could never be empirically perceptible by anyone. All declarations that it has occurred are necessarily self-deceptive Schwärmerei, and such illusions could have no positive moral or religious significance. All that could possibly matter is whether the person’s future course of life actually displays a gradual moral reformation. That is Kant’s only conception of the Change of Heart.

The three difficulties. Central to Kant’s discussion in Part Two is his solution to three difficulties about salvation he raises at AA 6:66–71 (Pasternack, pp. 142–8). The first difficulty is that, having begun from evil, our good disposition remains always imperfect. The second difficulty concerns the question how confident we are entitled to be that we have undergone a Change of Heart and are on the path to salvation. The third difficulty is that, since we began from evil, there is a burden of guilt (represented as a quasi-legal claim that the evil principle has over us) that must be lifted or atoned for if we are to be saved.

The first difficulty. Kant’s answer to the first difficulty is that God, viewing our progress from his timeless perspective, sees our moral striving as a perfected whole no matter at what time it might be cut short (AA 6:66–7). This is true as long as our progress is genuine. For our moral progress represents positive merit, and God can always reckon this to our credit. The first is the easiest difficulty to solve, in Kant’s view. The other two presume its solution, but concern further problems about our salvation that arise even assuming the Change of Heart and the merit that goes with moral progress.

The second difficulty. Pasternack seems to me to confuse Kant’s solution to both the second and third difficulties with his solution to the first. He diverts this discussion of the second difficulty into a digressive defence of the idea that Kant continued to believe in immortality as a condition of endless moral progress (pp. 150–3). He completely misses the direct answer Kant in fact gives to the second difficulty. Kant thinks, namely, that we are entitled to be confident only to the degree that we actually see improvement in our day-to-day conduct. Those who see steady improvement in their course of life may “reasonably hope that in this life he will no longer forsake his present course but will rather press on in it with ever greater courage.” On the other hand, someone who sees no such improvement in himself “can reasonably entertain no hope of improving”, and therefore can glimpse only a future of “boundless misery” (AA 6:68–9). Kant repeats these thoughts, applying them to a human being’s self-examination at the end of life, as part of his argument for the moral benefits of the solution offered to the third difficulty (AA 6:76–8).

This is also the way Kant reconciles the Christian conception of the Change of Heart with his ethics of virtue. Pasternack gets Kant dead wrong when he sees a conflict between Kant’s virtue ethics, as presented in the Doctrine of Virtue, and his Christian-inspired conception of the Change of Heart (pp. 132–3). For Kant, only the gradual path of moral struggle and increasing virtue exhibits a genuine Change of Heart. Kant does of course reject the notion that virtue is mechanical habituation (assuetudo) (AA 6:407). He contrasts this conception unfavourably with what he considers the Christian conception of a revolution in the way of thinking. Speaking of the former, Kant says that “a human being considers himself virtuous whenever he feels himself stable in his maxims of observance to duty” (AA 6:47; my emphasis). The point not to miss, however, is that Kant denies that mechanical habit is true virtue; virtue is a rational habit (AA 6:407). True virtue, in other words, the gradual path of steady moral improvement, acquired by rational habituation to dutiful conduct, is the only reliable sign of a Change of Heart. The extravagant emotions of the repentant sinner who declares himself to be washed in the blood of the Lamb, assuring us that he feels the hand of God’s grace upon him—these declarations should always be greeted with scepticism (since they can correspond to no true cognition). Unless they issue in better life-conduct, they should be regarded with that special disgust we ought to feel towards pious hypocrisy.[5]

Divine aid and the Change of Heart. Lutheran doctrine holds that we are saved by faith alone, not by works, and that any motion of our souls towards God is due entirely to his grace, not at all to our deeds or our merits. A modern rational human being with decent sensibilities has a hard time accepting these doctrines. Sanity tells us that we cannot be morally credited for what only another can do and we ourselves cannot do. These Lutheran doctrines surely exaggerate, and are an example of what Kant calls “hypocritical high praise” through which grovelling believers wish to ingratiate themselves with a Deity they demean by regarding him as a childish tyrant who loves to be flattered—even flattered with praise so extravagant as to be nonsensical.

But Kant does not want flatly to reject these doctrines, since he hopes he can make the religion of reason appealing to people who accept them. Kant thinks, however, that it is unknowable what, if anything, God has done to make the Change of Heart possible. The essentials of Kant’s position are these: We cannot know whether the Change of Heart is possible for us apart from divine aid, but if divine aid is needed, we are entitled to believe it is available to us if we strive to make ourselves worthy of it. It cannot be required as part of saving faith to believe in any particular account of the way it has been made available—e.g., through a “born again experience”, or even through the death and resurrection of Jesus Christ (AA 6:50–2, 66–7).

Pasternack’s way of putting this point, however, is distorted and in need of correction. He says:

If the Change of Heart is genuinely morally required of us, then Divine aid is not necessary. It may or may not be available, but if we are to hold consistently to the demand that ought implies can, then if this inner transformation is something we ought to accomplish, then it is something we can accomplish. (pp. 145–6)

Pasternack confirms explicitly that this is how he interprets Kant: “Kant opposes the view that we need Divine aid in order to undergo a Change of Heart” (p. 146). This can’t possibly be right. It tells us that for Kant the import of ought implies can is that divine aid is not necessary for the Change of Heart. That would commit Kant to a dogmatic assertion of the transcendent proposition that divine aid could not be necessary for the Change of Heart. This is more than Kant thinks we could ever know; it also brings Kant into gratuitous conflict with the Lutheran position he is doing his utmost to accommodate as far as he can. ‘Ought implies can’ does not require the ‘can’ to be unconditional; it only requires that the condition be satisfied if we do as we ought. And that is precisely what Kant holds.

What Kant says on this point is very clear: “Everyone must do as much as it is in his powers to do; and only then […] can he hope that what does not lie in his power will be made good by co-operation from above” (AA 6:52).[6] I would therefore revise what Pasternack said above, as follows: Ought implies can. We cannot know whether divine aid is necessary for us to do what we ought. Therefore, if the Change of Heart is morally required of us (if it is something we ought to accomplish), then in case divine aid is necessary for us to accomplish it, we can justifiably believe this divine aid is available to us, if we do what lies in our power to make ourselves worthy of it. This remains consistent, first, with Kant’s actual views about ‘ought implies can’, and second, with what he thinks we can and cannot know about divine activity. It further allows, third, a traditional Lutheran to continue to hold a (morally and religiously optional) belief in the necessity of divine aid—as long as they also do everything they can to make better human beings of themselves. Pasternack’s misreading would deprive Kant of all three advantages.

The third difficulty: the need for divine forgiveness. The greatest of the three difficulties, Kant says, is the third. Even if the human being has undergone the Change of Heart and put himself on the path of goodness, it remains true that “he nevertheless started from evil, and this is a debt which it is impossible for him to wipe out” (AA 6:72). But salvation requires that it be wiped out. The religion of reason, as well as Christian theology, must believe this is possible, and also give some account of how it is possible.

Christian doctrine tells us, in broad terms, that it is God who steps in and makes good what we cannot. But how is this to be understood? There are two models here that are commonly entertained by Christian theology. Both models involve the idea of a debt that can be forgiven. This is understood in the sense that, when I have borrowed money from you but cannot repay, you might forgive the debt. As Pasternack convincingly shows, Kant regards both models as morally and rationally unacceptable (pp. 153–62). The first model involves the idea that the debt does need to be repaid (by someone), but since it cannot be repaid by me, it must be repaid by someone else. The crucifixion of Christ is then seen as the repayment by God himself (or by God’s only begotten Son) of a debt we cannot repay. Our sin is so great that only the sufferings of God himself, or of an entirely innocent being sent down to earth by God, could suffice to make the payment. Applying the analogy with the case where I have borrowed money from you, it is as if some beneficent third party came along and got me off the hook by paying you what I owe. This is the meaning of Christ’s death on the cross. The second model involves the idea that since it is to God that we owe the debt, it might not be necessary for us to repay it at all, as long as God—a creditor who has discretion over whether to collect what is owed him—mercifully decides that we need not repay. Christ’s life and self-sacrifice might then be seen as God’s way of communicating to us this wondrous act of mercy and generosity on his part. This would, in the most direct sense, be forgiveness of the debt. Applying this to the example where I have borrowed money from you, you would need to have the option of, generously or mercifully, not requiring me to repay the debt at all.

Regarding the guilt we have incurred through the fact that we have begun from an innate propensity to evil, Kant rejects the entire analogy with the repayment or forgiveness of a debt. Violating the moral law is not like borrowing money from God; it is not even like borrowing money you then find you can’t repay. To put it in legal terms, as Kant does, our guilt for past misdeeds is not a transmissible liability: it is not a liability that might be made good by someone else, or that might be made simply to go away by the person to whom it is owed at that person’s discretion (AA 6:72; Pasternack, p. 154).

Where we are speaking of moral guilt and the liability to punishment, no one else may rightfully assume that. If punishment is due to me, but is inflicted on some other (innocent) person, this could not be a satisfaction of justice; it could only be a new and even more outrageous injustice. If an innocent person wanted to be punished for someone else’s sins, that would not be meritorious; it would be sick and repulsive. For these reasons, Kant rejects the whole idea of “vicarious atonement” if, on the first model, it is understood by analogy with the forgiveness of a monetary debt. Likewise, Kant rejects the idea that, on the second model, it would be possible for a just judge simply to waive the punishment that is due to me. That would not be to satisfy justice at all; it would be only unjustly to condone evil. My guilt would remain, and it would even be increased by my arrogant assumption that it had been done away with. God would be made complicit in it if he were to offer, and I to accept, such an immoral way for me to escape the punishment that is my due.

Kant’s solution involves the need to meet two requirements: First, a satisfaction of justice, a just verdict, which, second, is at the same time also a decree of grace; and these two requirements must be jointly met in a way that is self-consistent. Grace must not violate justice, and justice must require the grace rather than rendering it superfluous. Grace must also be accepted in accordance with justice. The problem is this: justice requires satisfaction, a punishment rendered to divine justice (AA 6:71–2, 73–4); but even when justice has been satisfied, there remains on account of our past guilt also the need for a “surplus over the merit of works”, for which even the satisfaction of divine justice is insufficient (AA 6:72, 75). Let us take in order the two issues of satisfaction and of the surplus.

Satisfaction. First, satisfaction can occur neither before the Change of Heart, since that simply increases while the guilt is being incurred, nor after the Change of Heart, since then it would be punishment inflicted on a person who is no longer guilty. Therefore, the satisfaction of justice must occur in the Change of Heart itself. Here Kant adopts the Christian imagery of the Old Man (Adam) and the New Man (Jesus Christ). He sees this imagery as symbolizing the moral transformation that goes on in the human being who has undergone the Change of Heart. The New Man—the changed disposition, representing the ideal of humanity well-pleasing to God—accepts the inevitable sufferings of life that are due to the Old Man; in this acceptance, justice is satisfied. Those Christians who still believe that it is Jesus Christ who atones for their sins can even adopt the morally and religiously optional doctrine (which, as we have seen, on Pasternack’s too reductive account Kant would have to deny) that their changed disposition is due to divine aid rather than their own merits. For these more conservative believers, their changed moral disposition would be due to “not I, but Christ in me” (Galatians 2:20). The talk of the New Man would not be merely symbolic, as it would be for the more rationalistically inclined. (Kant thinks of religious progress as a gradual process in which the crude, literal thinking of the tradition will be slowly replaced by symbolic thinking that both has greater aesthetic power and can be better supported intellectually.) What really matters for Kant is that the changed course of life should be real—that there should be real moral progress in their own conduct.

Nor, once again, does the Change of Heart need to be seen as a datable temporal event—a “born again” experience—for this account to apply. In fact, it applies better if we do not see it that way, but instead as the reversal of principles (or of the fundamental maxim) which is manifested during the entire process of gradual moral progress (or virtuous habituation). At each moment, the improving moral agent accepts the suffering due to the Old Man, but through his moral progress he exemplifies the disposition of the New Man. By distinguishing the principles on which we act as imperfect moral agents from those we adopt as improving moral agents, we can view this as a kind of vicarious atonement that is morally acceptable. For the New Man and the Old Man are really just the same moral agent, regarded in two ways. This point becomes even clearer if we do not see the Change of Heart as a distinct temporal event, but treat it as a relation, at any given time, between our morally imperfect self and our morally improving self (AA 6:73–5).

The surplus. There is, however, still the second point regarding the third difficulty. Even after the Change of Heart, the changed disposition itself can never do more than what is morally required, and therefore there is an outstanding guilt incurred in the past (AA 6:72). The merit of the New Man’s works can never yield a surplus on which the Old Man may draw. This is the point at which Kant thinks only God’s grace can solve the difficulty. The crucial passage explaining this is the following:

Here, then, is that surplus over the merit from works for which we felt the need earlier, one that is imputed to us by grace. For what in our earthly life (and perhaps in all future times and future worlds) is always only in mere becoming (namely, our becoming a human being well-pleasing to God) is imputed to us as if we already possessed it here in full. And to this we indeed have no rightful claim (according to the empirical cognition we have of ourselves), so far as we know ourselves (estimate our disposition not directly but only according to our deeds), so that the accuser in us would still be more likely to render a verdict of guilty. It is therefore always still a decree of grace when we are relieved (entschlagen) of responsibility for the sake of this good in which we believe, though fully in accord with eternal justice (because based on a satisfaction that for us consists only in the idea of an improved disposition, of which, however, God alone has cognition). (AA 6:75–6)

First I will consider Pasternack’s rendering of this passage. Then I will suggest an alternative interpretation that seems to me more faithful to it, and to offer Kant a more defensible doctrine.

This part of Kant’s account is plainly intended to take account of the Christian conception of vicarious atonement, but to do so in a way that is morally acceptable and does not involve the error of treating the guilt attaching to moral evil as a transmissible liability. Kant even emphasizes this point (AA 6:74–5), but I saw no discussion of it in Pasternack. Pasternack instead asserts bluntly, unqualifiedly (and repeatedly) that Kant simply rejects the doctrine of vicarious atonement (pp. 75–6, 141–2, 154, 158, 167, 208). In that assertion he is offering traditional Christians, on Kant’s behalf, a ground for rejecting outright the doctrines of the Religion that Kant himself is careful to avoid giving them. Pasternack is thereby selling Kant short, and also selling short his own project of defending Kant against the charges of Michalson, Hare and others.

Pasternack thinks that Kant’s account in the passage quoted above turns entirely on the difference between God’s omniscient perspective and our limited one. We see only our deeds, which though perhaps constantly progressing morally, never achieve perfection. God, however, sees the inner disposition expressing the Change of Heart. “From the Divine Perspective, God is able to see our inner worth, the moral status of our Gesinnung, and so is able to assess our worth in a way that we cannot” (p. 156). According to Pasternack, there is no role at all here for divine action or even divine mercy: God’s only role is cognitive; it is not in the least volitional. God merely sees better than we do what we deserve, and then gives us only what justice strictly requires him to give. Our perception that there is a “surplus” required over and above what we earn through the Change of Heart, a surplus due to God’s grace, is simply an illusion due to our merely temporal perspective on our lives.

In effect, Pasternack simply substitutes Kant’s solution to the first difficulty (AA 6:66–7) for his solution to the third. For it was only in relation to the first difficulty that Kant made reference to God’s timeless perspective on our always imperfect striving. The third difficulty, however, has nothing to do with the temporal imperfection of our perceptions. It concerns instead the quasi-legal claim that the evil principle apparently still has over us based on the fact that we started from evil. This is an irretrievably temporal matter. God’s timeless perspective on our progress cannot address that problem at all. Pasternack, however, tries to force it to do so. On Pasternack’s interpretation, we may think we have no rightful claim to God’s decree of acquittal, but that is just our error, as temporal beings. On Pasternack’s reading, God’s timelessness gets rid of the quasi-legal claim the evil principle has on us as well as of the problem of our temporal imperfection. It is as if Pasternack, blind to the difference between the first difficulty and the third, was capable of understanding only Kant’s solution to the first difficulty, which he then tries to make do for the third as well.

Pasternack correctly emphasizes that the term entschlagen, in legal contexts, refers to the dismissal of a charge, or perhaps the striking down of a legal statute (p. 157). Clearly it is the former use, not the latter, that applies, because what we need is not the repeal of God’s law, but rather the dismissal of a charge against us based on the quasi-legal claim the evil principle has on us. The only role “grace” plays in the account, according to Pasternack, is that the divine perspective is that of the timeless Leibnizian “realm of grace” as opposed to the optical illusions occasioned by the distorting temporal “realm of nature” (p. 147).[7] According to Pasternack’s interpretation, God lets us off scot free, because, when things are seen from his timeless perspective, the dismissal of the charge against us is simply what strict justice requires of him. As Pasternack reads Kant, God never had any choice in the matter.

Pasternack’s interpretation entirely dispenses with the idea of divine grace, as well as divine forgiveness and divine mercy. Kant’s account, he bluntly asserts, “rejects not only the idea of Vicarious Atonement but even the idea of Divine Forgiveness in general” (p. 142). Pasternack appears to be fully aware of this consequence. According to him, Kant had no qualms whatever about eliminating, in effect, the entire Christian doctrine of atonement: “Kant’s use of ‘grace’ here, as in ‘decree of grace’ (6:76) pertains merely to the divine perspective” (p. 157). In effect, Pasternack’s Kant declares entschlagen—dismissed or struck down—the entire issue of sin, guilt, atonement, forgiveness, since, from the true, timeless (divine) perspective, there never was any such issue at all. Pasternack’s representation of Kant’s attitude at this point even makes Kant sound stubborn, arrogant, and lightheartedly dismissive of the central idea of Christianity: “The debt of sin is not somehow repaid in the suffering that goes along with the Change of Heart […] It is not forgiven through a “Divine Supplement” […] More conservative readers may not like what Kant has to say, but an answer is given that satisfies Kant” (p. 158). Pasternack’s Kant seems even to enjoy thumbing his nose at them. He is merely a secular moralist, willing or even eager to eliminate essential parts of the Christian message because his shallow, Enlightenment rationalism is blind to it. This looks to me like precisely what hostile traditionalist Christians see in the Religion. Pasternack seems to be siding with these critics against Kant. I think a reader of the Religion both can and must do a lot better.

As we have seen, Kant does reject the model (or rather both models) of a forgiveness of debt, because they treat moral guilt as a transmissible liability. But I think the passage quoted above does permit us to see Kant’s account as one involving ‘forgiveness’ in another sense, a sense much more familiar to us in the context of modern moral sensibilities. Let me briefly sketch the moral structure of this more familiar moral concept of forgiveness, and then try to show how Kant’s account fits an interpretation appealing to it more satisfyingly than it fits Pasternack’s reductionist interpretation.

Suppose I have done you some wrong. You are morally entitled to blame me. Perhaps you are also entitled to impose some sort of penalty on me. Suppose also that I too come to regard this situation as morally unsatisfactory, even intolerable. I accept that I have done wrong, but do not want to be the continuing object of your blame. I want the deserved penalty to be lightened. But because I also have due concern for the morality of the situation, I do not want you merely to condone the wrong I have done. Instead, I want to make things right. I want to be reconciled with you on morally acceptable terms. I therefore do something in an attempt to make things right—at a minimum, I apologize to you, I repent of the wrong I have done. Perhaps I also offer you a reparation of some kind in compensation for the harm I have caused, not so much because I think it can undo my misdeed, but as a sign to you that I am sorry for what I have done and am resolved to do better in future. Because I have wronged you, you are in a position either to accept or not accept my attempt to make things right. Borrowing another legal term, we might say that you have ‘standing’ to blame me, and also standing either to accept my offer of reconciliation or not accept it. Which you do is up to you. If you do accept my attempt to make things right, then you are relinquishing or dismissing the attitude of blame towards me (here the quasi-legal German term entschlagen would also be appropriate). There is no obligation for you to reconcile with me, but it would be morally acceptable of you to do so. I have, as Kant puts it, a “(moral) receptivity” to the benefit you would voluntarily offer me, on account of my having done what I can to make things right (AA 6:75n). Accepting my repentant deed is an act of gracious generosity on your part towards me, for which my gratitude towards you would be appropriate.

I submit that we all recognize this sort of scenario. We think it is a way that a person who has done another wrong may be reconciled with the other on morally acceptable terms. The term we would now use for your acceptance, reconciliation and relinquishment of blame, is your forgiveness of me. This is a use of the word quite distinct from the forgiveness of a debt, involving a transmissible liability or a discretionary dismissal of the liability. It counts as ‘forgiveness’ in a different sense, but one that is familiar to us and accords with modern, rational moral sensibilities.[8]

As Kant has presented it in the passage quoted above, the relation between the human moral agent and God fits this concept of forgiveness. Religion understands human moral duties as divine commands, and God as legislator of the moral law, who is entitled to offer sanctions for its disobedience (AA 6:153–4). This makes the human being’s evil choices into sins, wrongs committed against God, and gives God standing to blame the human being, yet also standing to forgive the sins the human being has committed. So the human being and God do stand in a relation like the relation between me and you in the scenario just described. Further, by undergoing the Change of Heart, and accepting as the New Man the penalty of human suffering due to the Old Man, the human being has done something analogous to “making things right” in the above scenario. God may accept or not accept it, and may or may not relinquish or dismiss both his blame towards the human agent, and any penalties the human agent has incurred. Both Christianity and the religion of reason believe, however, that God has accepted the Change of Heart, and issued a decree of grace, freeing the sinner from blame. Rational religion accepts it for the same reason it accepts God as a condition for the possibility of the highest good. For only if we believe in God’s grace, in this sense, is it possible for us to be saved by making better people of ourselves.[9] The human being has no rightful claim to this decree, but it also involves no violation of justice. It is an act of gracious generosity on God’s part, for which the human being should feel gratitude towards God.

This is my proposed alternative interpretation of Kant’s resolution of the “third difficulty” (AA 6:75–6). Let’s call it the “forgiveness” interpretation. I think it has a number of advantages over Pasternack’s interpretation, not only from the standpoint of traditional Christianity, and not only from our present-day moral standpoint, but also from Kant’s own standpoint.

Pasternack’s interpretation ascribes to God no volitional act in issuing the “decree of grace”, but only the cognitive act of seeing the Change of Heart, and the human being’s new disposition, from a timeless perspective. By contrast, on the forgiveness interpretation, God’s role in our salvation is not merely cognitive; it involves God’s voluntary choice to accept the Change of Heart and forgive the human sinner. The forgiveness interpretation leaves God free either to issue this decree of grace or not to issue it. Kant would not want to view God as required by justice to issue the decree of grace; he even defines “grace” in precisely this connection as “a superior’s decision to grant a good for which the subordinate has no more than a (moral) receptivity” (AA 6:75n). Kant also thinks our disposition after the Change of Heart “takes the place” of moral perfection in God’s sight, but is not equivalent to it (AA 6:75n). This is the point on which the resolution to the third difficulty differs from the resolution to the first, and why the third difficulty is the hardest of the three to resolve.

The forgiveness interpretation explains why this is a decree of grace, and not merely a decree of justice. Pasternack must say that calling it a “decree of grace” merely reflects our distorted temporal perspective and is not true to the real moral situation. On Pasternack’s interpretation, God seems to be morally required to strike down or dismiss the verdict of guilt, as a human judge would be required to render a verdict of acquittal when the judge sees clearly that the accused is not guilty under the law; God’s so-called “decree of grace” involves no generosity (in fact, no grace at all). Kant thinks it appropriate for us to feel gratitude towards God for our salvation (AA 6:64). On Pasternack’s account, however, it is not necessary, but would even be inappropriate, for the human being to feel gratitude towards God for issuing this decree. You shouldn’t need to feel gratitude towards a judge who does no more for you than what justice strictly requires. Gratitude would seem to display a servile attitude towards a judge who let the grovelling sycophancy of the accused persuade him to do the right thing contrary to his naturally cruel and unjust disposition. If Pasternack’s account were correct, then being grateful to God for his grace would be an insult both to us and to him.

The forgiveness interpretation allows some truth to our (temporal) perception of the situation. By contrast, Pasternack’s interpretation treats our entire perspective as merely illusory. The forgiveness interpretation agrees that we have no legal claim to the decree of grace, while seeing this decree as nevertheless morally acceptable—or as Kant puts it, we have no “rightful claim” but only a “moral receptivity”. As a consequence, Pasternack can offer no reason why we should think God would issue the decree of grace. For we cannot see things from God’s atemporal perspective, from which we are allegedly judged innocent, but instead must regard ourselves as guilty and incapable of supplying the “surplus” of merit needed to satisfy divine justice. It is one thing to equate a genuinely changed disposition with a merit timelessly cognized, as happens with the first difficulty. It is quite another to claim, as Pasternack’s Kant is forced to do, that God’s timeless vision would reveal that, contrary to all appearances, we have no guilt at all based on our past misdeeds. Kant would not want it to be incomprehensible to us—a merely arbitrarily assumed metaphysical fact about the way God timelessly sees our changed disposition—how this decree satisfies the demands of justice. By contrast, the forgiveness interpretation understands the Change of Heart, and God’s gracious acceptance of it, as morally acceptable, involving an act of gracious generosity on God’s part, for which our gratitude is appropriate.

Concluding remarks. Pasternack’s principal aim in this book is admirable. He sets out to defend Kant’s religion against the charge that it is involved in a series of “wobbles” or “compromises” between Christianity and Enlightenment rationalism. He means to answer the accusation that the Religion fails to give an interpretation of key Christian doctrines that reconcile them with modern rational moral principles and sensibilities. For reasons of space, and despite the length of these comments, I have examined only about half of Pasternack’s commentary, and discussed only some of the things I found there. But I think I have shown on several points that a better job can be done of fulfilling Pasternack’s own aims. I fully agree with Pasternack—indeed, I think I agree with this to an even greater extent than he does—that Kant can be successfully defended against the charges of failure, inconsistency or “wobbling” that have commonly been brought against the Religion.

Kant’s Religion will continue to be criticized by the religious for being dry, rationalistic and moralistic, and by the irreligious for being too sympathetic to traditional religion. Moreover, both everyday morality and Christianity have continued to evolve since Kant’s day, so we cannot expect a reconciliation that worked at the end of the eighteenth century to satisfy us even as much as it might have satisfied Kant’s contemporaries (which, as we know, it already did not do for many of them). In that sense, I think we have to concede that the Religion was, has been, and still is a failure in the ambitious task Kant set himself: of showing Christians that they can adopt a rational, Enlightenment morality without abandoning the doctrines, symbols and the emotions of their faith.

All this means, however, is that Kant’s Religion is very much at odds with twenty-first century culture, just as it was already at odds with much religious culture in his own day. But who did not already know that? But we ought also to ask: Who is responsible for Kant’s failure? Even while being unable to accept for myself the terms of Kant’s reconciliation, I for one see this as a failure of our culture in the past two centuries rather than of Kant. A religion of reason, which could unite both religious people and people with rational, scientific minds, would be a great thing. It would immeasurably enrich humanity and make our sad world a far better place than it is. The road the world has not taken may be the right road. Sometimes a philosopher’s apparent failures are not really the philosopher’s failures at all, but rather failures out there in the world. Kant may be the one who is in the right, and all the rest of us—the religious and the irreligious alike—may be the ones who are in the wrong. We should have understood Kant better, and found the path towards which he was pointing. Both the secular rationalists and the traditional religious should see the failure as theirs rather than Kant’s.


[1] A. Wood, Unsettling Obligations: Essays on Reason, Reality and the Ethics of Belief (Stanford: CSLI, 2002), p. 98.

[2] For a fuller defence of some of the things I am about to say see my ‘The Evil in Human Nature’, in G. Michalson (ed.), A Critical Guide to Kant’s Religion within the Boundaries of Mere Reason (Cambridge: Cambridge UP, 2014), pp. 31–57.

[3] My guess is that many of those who offer “a priori” or “transcendental” proofs of the thesis do so with an intent which, though misguided, is charitable towards Kant. They see how important the thesis is to the Religion, and they are embarrassed to admit (as Kant does admit, though without embarrassment) that it is still unproven. Although it is an empirical thesis, the thesis does dovetail with Kant’s empirical view of human nature and his moral psychology. My (less than charitable) conjecture is that these readers have so accustomed themselves to Kantian morality that they have lost the ability to see that Kant’s empirical moral psychology as only that. Therefore, they misguidedly seek a priori arguments to confirm what Kant knows is only contingent and empirical. 

[4] Langton is certainly not mistaken about everything here. Her focus, in dealing with questions of degrading treatment of people, is on the way women have been subjected to such treatment: at the extreme end, by sexual violence, but also through pornography and also, at the milder end, through commonly accepted attitudes deprecatory to women. She correctly perceives that Kant himself shared some misogynistic attitudes. I think that perception, overgeneralized, may be what lies behind her view that he did not recognize degradation as a distinctive form of evil. Kant was certainly concerned with some ways in which women are subjected to degrading treatment. But there’s no question that he did not see these issues in the way twentieth and twenty-first century feminists see them. Langton is surely mistaken, however, in generalizing to the conclusion that he did not recognize the degradation of rational beings as a distinctive and serious form of evil. Because Kantian ethics is fundamentally about respect for humanity in the person of every human being, the vices opposing all duties to others (both of love and respect) are quite recognizably vices relating to the aim of lowering others (degrading, even dehumanizing them). Even the forms of degradation in which Langton is specifically interested are clearly versions of this kind of evil—whether or not Kant thought of them that way. 

[5] It would take us too far afield here to stop to demonstrate that there is virtually no daylight separating Kant’s conception of moral virtue from Aristotle’s on these points. But see my ‘Virtue: Aristotle and Kant’, in D. Brink, S. Sauvé Meyer & C. Shields (eds), Virtue, Happiness, and Knowledge: Essays for Gail Fine and Terence Irwin (Oxford: Oxford UP, forthcoming).

[6] Kant begins one conditional sentence with “Gesetzt, zum Gut- oder Besserwerden, sei noch eine übernaturliche Mitwirkung nöthig […]” (AA 6:44). Di Giovanni translates Gesetzt as “Granted…”, while Pluhar translates it as “Supposing that…”. Pasternack repeatedly harps on the difference, apparently thinking that it is significant, and that Pluhar’s translation is superior (pp. 144, 145, 146, 251). But there is no significant difference here at all. In the German, and in both translations, the import of the antecedent clause is conditional: “If it be granted [or supposed, this makes no difference] that supernatural co-operation is necessary […].” Kant wants to leave the truth of this antecedent undecided, since he thinks we can never know whether it is true. But Kant wants to allow that some may permissibly believe it. Pasternack apparently fears that translating gesetzt as “granted” allows for the possibility that someone might actually believe such aid is necessary, whereas he thinks “Supposing that…” is consistent with not allowing this possibility; and Pasternack’s position is that Kant does not want to allow it at all. On the substantive point of interpretation, however, Pasternack is clearly wrong. Kant does want to allow this possibility, though he thinks belief in the necessity of divine assistance is optional, morally indifferent and not required for saving faith. He also thinks that saving faith requires that if it is believed that supernatural co-operation is needed, then one must also believe that it is available only to those who do what they can to make themselves worthy of it. Pasternack’s obsessive quibbling over a minor point of translation is an unfortunate symptom of his systematic misreading. 

[7] The appeal to Leibniz here is utterly hopeless. It surely misinterprets Leibniz to think that as he uses it, the term ‘grace’ refers only to the timelessness of God’s perception, and involves only divine justice, not divine mercy, and gives human beings no cause for gratitude towards God. Leibniz, after all, was at least as traditional a Christian as Kant. See Leibniz, Discourse on Metaphysics, §36; Monadology, §84, Principles of Nature and Grace, §17. 

[8] For one recent book-length presentation of roughly this concept of forgiveness, see C. Griswold, Forgiveness: A Philosophical Exploration (Cambridge: Cambridge UP, 2007). Griswold sees forgiveness as the elimination of resentment by a wronged person, but subject to certain moral norms and conditions. These conditions include expressions of regret by the guilty party, assurances of a change of heart and sometimes actions to be understood as attempts at reparation. The result of forgiveness is a morally acceptable reconciliation of the wrongdoer with the one wronged.

[9] This defence of rational faith in divine grace was presented nearly half a century ago in a book by a naïve young Kant scholar who still had an awful lot to learn about Kant’s ethics, but nevertheless somehow managed to get this point right. See A. Wood, Kant’s Moral Religion (Ithaca, NY: Cornell UP, 1970 [reissued: 2009]), Ch. 6.

© Allen W. Wood, 2015.

Allen Wood is Ruth Norman Halls Professor of Philosopher at Indiana University, Bloomington, US. Among numerous articles and book contributions, he has published the following monographs: Kant’s Moral Religion (Cornell UP 1970, reissued 2009), Kant’s Rational Theology (Cornell UP 1978, reissued 2009), Karl Marx (Routledge 1981, 2nd expanded ed. 2004), Hegel’s Ethical Thought (Cambridge UP 1990), Kant’s Ethical Thought (Cambridge UP 1999), Unsettling Obligations (CSLI Publ. 2002), Kant (Wiley-Blackwell 2005) and Kantian Ethics (Cambridge UP 2008). His latest book is The Free Development of Each: Studies on Freedom, Right and Ethics in Classical German Philosophy (Oxford UP, 2014). He is also, alongside Paul Guyer, general editor of The Cambridge Edition of the Works of Immanuel Kant, and has translated several of Kant’s works as well as Fichte’s Attempt at a Critique of All Revelation. He currently works on a book on Fichte’s Ethical Thought. His Kant’s Groundwork for the Metaphysics of Morals: A Commentary (co-authored with Dieter Schönecker), originally published in German, has recently been published with Harvard UP.