By Susan Meld Shell
Stephen Palmquist’s commentary on Kant’s Religion performs an inestimable service for which students of both Kant and the history of philosophy and theology have reason to be deeply grateful. Palmquist’s own novel and searching interpretative approach to Kant’s religion is visible on every page; yet this commentary neither presupposes agreement with that approach nor is its usefulness intended to be limited to those who share it. Given such a wide ranging study, in which thought-provoking gems appear on almost every page, I must regrettably give short shrift to the many ways in which I have profited from reading Palmquist’s commentary, along with many points of strong agreement, including the importance of strict fidelity to the text, and due attention to the circumstances surrounding its publication. And I am entirely sympathetic with his attempt to situate Religion within a wider systematic context.
My basic point of disagreement boils down to this: For Palmquist, if I understand him aright, religion within the boundaries of bare reasons enjoys a higher rank, both systematically, and in human terms, than a religion that is merely moral. On this I am not so sure. Let me express my hesitation under four headings:
- The sense in which religion is ‘necessary’ for Kant
- The meaning of ‘religion within the boundaries of bare reason’
- Kant’s attitude toward ‘supernaturalism’
- Suggestions by way of an alternative reading
1. The Sense in Which Religion is ‘Necessary’ for Kant
Consider the following well-known but perplexing passage:
Morality, insofar as it is grounded on the concept of the human being as one who is free and even because of this also binds himself through his reason to conditional laws, needs neither the idea of another being above him in order to recognize his duty nor as an incentive anything other than the law itself in order to observe it. At least, it is his own fault if he encounters such a need in himself; but then neither can it be remedied through anything else; because what does not arise from himself and his freedom yields no substitute [Ersatz] for a lack in his morality. It thus on behalf of itself (as much objectively, concerning will, as subjectively, concerning ability) in no way requires religion, but is rather through the power [Vermögen] of pure practical reason sufficient unto itself. (RGV, AA 6:3; trans. mine. cf. RGV, AA 6:74)
I take Kant to be saying something like the following: Although a morally good will is not determined by its purpose, human willing cannot take place at all without reference to an end. But not only this: insofar as one is rational, one cannot avoid being interested in the result of one’s right conduct nor, accordingly, to what ultimate end we should direct our doings and non-doings—lest they cancel one another out. Hence our need for the idea of a highest good in the world plus that of a God who makes the former (happiness proportional to virtue) possible. Nor is such an idea “empty” (“practically considered”) since it meets a
natural need, which would otherwise be a hindrance to moral resolve [Hinderniß der moralischen Entschließung], to think for all our doings and nondoings taken as a whole some sort of ultimate end that reason can justify. (RGV, AA 6:5; trans. mine)
I would like to emphasise what Kant here implies: that the ‘expansion’ of morality to religion to which Kant immediately refers is itself based on an implicit duty of virtue, i.e., that of moral self-perfection. That suggestion is supported by Kant’s similar reference, in the Critique of Judgement, to the righteous man (Spinoza, perhaps) who believes in a moral God lest the abysmal vision of a Godless universe weaken the man’s respect for the moral law and thus impair his moral attitude (KU, 5:453). Accordingly, all of Kantian ‘religion’, including Religion within the Boundaries of Bare Reason, may indeed be more fully grounded in a morally practical standpoint than Palmquist, if I understand him, is willing to allow.
2. The Meaning of ‘Religion Within the Boundaries of Bare Reason’
It is helpful, once again, to turn to Kant himself, who (in a late passage from the Doctrine of Virtue) specifically distinguishes “religion within the boundaries of bare reason” both from “religion” generally and from what he there calls the “doctrine of religion” (Religionslehre). In “granting” that “religion is an integral part of the general doctrine of duties” Kant then adds that that problem is to determine the “boundaries” of the science to which it belongs, that is, whether it is either “part of ethics” or whether it must instead be regarded “as lying entirely beyond the boundary of a pure-philosophic morals”. Kant goes on to distinguish, by way of answer, between the “formal” element of all religion (insofar as it is defined as “the sum of all duties as [i.e. as (morally) equivalent to] (instar) divine commands” and a “material” element, containing special duties as divine commands and which would belong “only to revealed religion”. In the former (formal) case, there is no need for belief in God “as a being existing outside of our idea”, but only for the purpose of “making obligation […] intuitive for ourselves” and thereby “strengthening the moral incentive in our own lawgiving reason”. In the latter (material) case, we would indeed have to assume the existence of God as “something that could be set forth as given directly (or indirectly) in experience”. Such a religion, however “well grounded it might otherwise be” (a question that Kant here leaves open), would comprise “no part of a purely philosophic morals” (all quotations MS/RL, AA 6:487).
We are thus left, as it would seem, with a stark either/or: either a formal religion that consists in regarding moral duties as (equivalent to) divine commands and that is, as such, part of a pure philosophic morals or a material religion that is “entirely beyond the bounds” of a pure philosophic morals. Religion of the latter sort includes belief in the existence of God, not merely (bloß) “the idea of it [demselben] set forth arbitrarily” and “from a practical intention”, but that could be explained as given (whether indirectly or directly) in experience (MS/RL, AA 6:487; trans. emended) (The latter, “material” sort of belief—belief, that is to say, in a divine existence that could be explained as given in experience—is entirely excluded from anything remotely bordering on a pure philosophic morals.)
And yet—as Kant now adds—“one can indeed speak of a ‘religion within the boundaries of mere [bloße] reason’, which is not, however, derived from mere reason but is also grounded in the doctrines of history and revelation” and that only contains the “harmony of pure practical reason with the latter (that they do not conflict [nicht widerstriete] with one another)” (MS/RL, AA 6:488, trans. and boldface mine; Palmquist alludes to this passage at p. 33n.182).
Accordingly, religion in Kant seems to take the following systematic forms: first, religion as entailed by the postulates of pure practical reason and leading directly to “the recognition of all duties as divine commands” (KpV, AA 5:129); second, religion insofar as it belongs to the metaphysics of morals proper, and with a specific view to “strengthening the moral incentive of our own lawgiving reason” by “making obligation […] intuitive for ourselves” (MS/RL, AA 6:487); third, religion within the boundaries of bare reason; and fourth, religion that falls “entirely beyond” a pure philosophic morals.
The crucial question, for an interpreter of the work in question, is, of course, the status and meaning of the third: ‘religion within the boundaries of bare reason’. For Palmquist, if I understand him aright, religion of this sort, like the ‘ethico-theology’ of the Third Critique with which Religion is, on his account, roughly continuous, in some meaningful sense transcends the practical or moral standpoint, by “completing” the “abstract moral theory” to whose regard that standpoint, in his view, is limited (p. 140).
I am compelled to say that my own sense of Kant’s ‘ethico-theology’ bears no such reading. Indeed, in every passage I can find, Kant consistently emphasises that it is only from a practical (or moral) point of view that belief in a personal God and final purpose of creation has any workable meaning. And there may be other reasons (pace Palmquist’s extended argument both here and elsewhere) to resist linking Religion and the Third Critique too closely—an issue I shall return to.
3. Is Kant a ‘Supernaturalist’?
One of the most interesting and instructive aspects of Palmquist’s commentary is its detailed attention to the role of Gottlob Christian Storr, whose review, on Palmquist’s persuasive account, provoked many of the substantive changes in the second edition. And it provides a helpful corrective to the dismissively reductive treatment of Religion common in the literature until very recently. At the same time, in so doing Pallmquist may push Kant closer to what Storr calls ‘supernaturalism’ (pp. x, 38)—a position that Kant in my view gently but firmly disavowed—than can be textually justified.
A crucial passage from Part Four is here especially instructive. After distinguishing between a “natural religion”, in which “I must know something as a duty before I can recognise it as a divine command” and a “revealed religion”, in which “I must know something to be a divine command before I can recognise it as a duty”, Kant adds:
He who declares merely [bloß] natural religion to be morally necessary, i.e., for duty, can also be called a rationalist (in matters of faith). If he denies the reality [Wirklichkeit] of all supernatural divine revelation, he is called a naturalist. Should he now allow for this [revelation] but maintain that recognising [kennen] it and accepting it as real are not necessarily required for religion, he could be called a pure rationalist. But should he hold belief in this to be necessary for universal religion, he could be called the pure supernaturalist in matters of faith. (RGV, AA 6:154–5; trans. mine)
Palmquist understands Kant to here identify three kinds of ‘rationalist’; the naturalist, the pure rationalist, and the super-naturalist (p. 388). I read it, rather, to be saying that there are two types of rationalist: the ‘naturalist’ and the ‘pure rationalist’: while the third (the super-naturalist) is not a rationalist at all, but falls, as it were, ‘entirely beyond the boundaries of bare reason’.
In short: while I agree with Palmquist that Kant’s own preferred position is ‘pure rationalism’ (inasmuch as unlike ‘naturalism’ it does not ‘dogmatically’ deny the possibility of a real revelation), I do not think that Kant takes ‘supernaturalism’ (which dogmatically affirms that possibility) to be a kind of ‘rationalism’. For by insisting that ‘universal religion’ of a sort adequate for moral purposes requires belief in the reality of divine revelation is precisely what rationalism as here defined explicitly denies. And if one needed further evidence that ‘supernaturalism’ as here understood is not a form of ‘rationalism’, one need look no further than the Conflict of the Faculties where Kant calls ‘supernaturalism’ (as distinguished from his own ‘supersensualism’) excusable but also very wrong (worin sie aber sehr fehlen) (SF, AA 7:54, 59).
Thus while also agreeing with Palmquist’s ingenious hypothesis that Kant intended to lay out an alternative to both Fichtean naturalism and Storrean supernaturalism, I also think he pushes Kant too far towards the latter camp. Kant’s accommodating attitude towards theologians like Storr is, in my view, less substantive than tactical (intended to bring them over to his side, if not right away, then eventually, as with the theological faculty whose plan of study Kant hoped his own course on philosophical theology might one day cap). I do not mean to say that Kant was a full-blown Fichtean; but only that the ‘unity’ of pure moral religion and historical faith (a unity encompassed by a ‘religion within the boundaries of bare reason’) may have a different character than Palmquist here, and throughout the commentary, suggests.
4. Suggestions by Way of an Alternative Reading
I can here only hint at what such an alternative understanding of this unity might mean by turning briefly to two key passages: the first is Kant’s famous image, in the second Preface, of a circle within a circle; and the second is a less often noted passage, from the General Remark at the end of Piece/Part One, on the meaning of the so-called ‘thesis of innate evil’.
Beginning with the first: that image of two concentric circles (a wider sphere of historical faith inclosing a narrower one consisting in “the pure religion of reason”) is quickly followed up by Kant’s well-known description of a double experiment. In the first, the philosopher, who must, as such, keep within the inner sphere, “abstract[s] from all experience”. In the second experiment, by way of contrast, one “abstract[s] from the pure religion of reason insofar as it constitutes a system on its own” and starts instead from some alleged revelation. One then proceeds by holding fragments of this revelation, as a historical system, up to moral concepts, and then seeing “whether it does not lead back to the same pure rational system of religion” (RGV, AA 6:12; trans. mine). What this complex procedure actually entails, and how it bears on the structure of Religion as a whole, is, of course, a much debated question. For the present, it must suffice to note that the relation between these experiments and the proceeding image is itself none too clear, including as to where ‘religion within the boundaries of bare reason’ is itself meant to fit. It is surely not just within the inner sphere, nor yet can it constitute the entire outer sphere—which would presumably include beliefs and practices that are purely cultic (as with ancient Judaism as here depicted). According to my own translation of the passage:
From this standpoint I can now make the second experiment/attempt [Versuch], namely proceeding from some supposed revelation [dafür gehaltenen], and abstracting from rational religion (insofar as it constitutes a system subsisting for itself), hold it as an historical system to moral concepts in a merely fragmentary way and see if it does not lead back to the same pure rational system of religion, which would be self-subsistent and sufficient for a genuine [eigentliche] religion, that as an a priori rational concept (remaining over following the elimination of everything empirical) only exists [statt findet] through this relation—albeit not from a theoretical perspective (in which must also be counted the technical-practical aspect of instructional method as a doctrine of art [Kunstlehre]) but yet from a moral-practical perspective. (RGV, AA 6:12; trans. mine)
The sense of this dense and difficult passage would seem to be as follows: A rational system of religion based on wholly a priori concepts has real existence only insofar as actual human beings themselves embrace it through the application of common morality (gemeinen Moral) to the tradition handed down to them (RGV, AA 6:14).
“Genuine religion” so understood has meaning, however, only from a moral-practical perspective, that is, for those who are in a moral/practical position to themselves affect the outcome, as it were, of the experiment (either individually, in the here and now, or collectively and historically, over time). ‘Religion’ can be ‘genuine’, in other words, under the subjective assumption that the outcome of the experiment(s) itself remains undecided. Hence Kant’s subsequent suggestion that instead of constituting a true compound (‘religion within the limits of bare reason’ as tentatively described in the Doctrine of Virtue), the two systems might prove to mix only temporarily, like oil and water. On the interpretation I am suggesting, in other words, the outer circumference of religion within the boundaries of bare reason (and hence its precise relation to Kant’s ‘two circles’) is not yet determined or even determinable, hinging as it does on the success of works like Religion within the Boundaries of Bare Reason.
This brings me to my second passage: one that has slipped by most critics of whom I am aware virtually unnoticed. And yet it supplies a crucial gloss on the famous doctrine of radical evil (and related claims) from which most theologically ‘affirmative’ readings of Religion take their bearings.
The thesis of innate evil is of no use in moral dogmatics; for the precepts of the latter contain the same duties and retain the same force whether or not there be in us an innate propensity toward transgression. In moral ascetics however this thesis says more, yet no more than this: in the ethical training (Ausbildung) of the moral predisposition, created in us, to the good we cannot make a start from an innocence natural to us but must begin from the presupposition of a malice of the power of choice in adopting its maxims in opposition to the original predisposition to the good, and because the propensity thereto is ineradicable, with unceasing counteraction against it. (RGV, AA 6:50–1; trans. mine)
If one takes Kant’s statement literally (as I am wont to do), than the entire teaching concerning radical evil is itself a sub-branch of what Kant elsewhere calls “moral anthropology” (see RGV, AA 6:12): valid, or of use, only as an extension of the moral-pedagogical insight that any effort to become a better human being must perforce begin from the acknowledgement that something internal to one’s power of choice resists obeying the moral law—something for which one is oneself morally responsible. The thesis of moral evil, on this account, provides “moral ascetics” with a target of resistance against which to exercise one’s moral muscles, so to speak. But not only this: the thesis of radical evil may itself be a tool whose usefulness (on Kant’s own understanding) is of merely temporary duration: as a specific remedy for the counter-thesis of original sin (along with related doctrines of Christianity ‘inauthentically’ interpreted) that exposes men’s moral muscles to a peculiar sort of wasting (e.g. the spiritually enervating temptation to leave everything to God).
Understanding ‘radical evil’ in this way, i.e. as an example of ‘authentic’ interpretation in Kant’s specific sense—a way of making moral use of doctrines ‘handed down to us’ that might otherwise be morally confounding and/or harmful—not only reduces the interpretative pressure posed by its seeming inconsistency with Kant’s moral system as a whole (a problem with which many critics have struggled); it also limits the claims of Religion (along with its peculiar privileging of Christianity) in ways that may have special salience for non- and post-Christian present.
Such a reading would not be morally reductive in any simple sense, for it would grant positive religion an important role, both affirmatively, that is, in nudging us toward self-sufficient reliance on a pure moral religion, and negatively, that is, in weaning us away from counter-service of religious ‘priestcraft’.
At the same time, it would resist treating moral conversion (which, after all, remains a duty here and now for each of us) as somehow conditional on the prior establishment of the true church or other radical improvement of society.
To briefly conclude:
- The path from morality to religion for Kant never ceases to be contained within a grounding frame that is itself fundamentally moral.
- Palmquist’s contrary privileging of the ‘judicial’ over the ‘moral/practical’ standpoints may push Kant too far toward ‘supernaturalism’.
- Palmquist’s conflation of the systematic positions of the Critique of Judgement and that of the Religion not only gives the latter a systematic status that, in my view, is not only textually unwarranted. It may also (given its specifically Christian focus) unduly limit his appeal in circles where his thought may today be most needed.
That said, I have learned an extraordinary amount from Palmquist’s commentary on a work I had thought that I already knew well; and I am extremely grateful to him both for his devoted efforts as an astute critic and scholar to bring it forth and for the magnificent result from which many others will surely profit.
Received: 30 October 2016.
 Cf. in this regard the more positive or at least neutral Streit that characterises the Conflict of the Faculties.↩
 On moral ascetics, see also MS/RL, AA 6:484–5; see also Palmquist, p. 140, note 162.↩
 Though I sympathise with Palmquist’s use of ‘volition’ for Kant’s Willkür (instead of the more common ‘power of choice’), I also find it potentially misleading, given the etymological link between ‘volition’ and the Latin voluntas. Kant himself uses Wille to translate the Latin voluntas, and Willkür to translate the Latin word arbitrio. It is unfortunate that we have no English noun that serves as a direct equivalent of the latter term.↩
 On moral ascetics as a kind of ‘gymnastics’ see MS/RL, AA 6:485, a section that can also be usefully compared with Kant’s treatment in Religion of the so-called problem of the ‘new man’.↩
 On the meaning of ‘authentic’ interpretation, see Kant’s essay On the Miscarriage of all Philosophical Theodicies.↩
 To be sure, much turns on how one understands the term ‘dogmatics’—an unusual term for Kant, but one with a long and controversial theological history. Without entering into that debate the most reasonable interpretation of the term ‘moral dogmatics’ is to designate the ‘factum of reason’ and all that it contains by way of duties that are incumbent on us as rational moral beings. (See, in this regard, Kant’s reference to certain “universal propositions of moral faith” at RGV, AA 6:11.) Palmquist’s own gloss, namely that it refers merely to “abstract moral theory” (p. 140), strikes me as unduly restrictive.↩
© Susan Meld Shell, 2017.
Susan Shell is Chair of the Department of Political Science at Boston College. Among many other publications, she has published KANT’S OBSERVATIONS AND REMARKS: A CRITICAL GUIDE, KANT AND THE LIMITS OF AUTONOMY and THE EMBODIMENT OF REASON.