The end of all things. Kant and Cohen on religion and reason

. According to Kant’s concept of natural religion, the moral concept of God consists in the coordinated aggregate of the concepts of holiness, goodness and justice. I argue that this concept can be used to define a critical account of religion’s role within the public sphere. In order to do so, I refer to Hermann Cohen’s philosophy of religion. Cohen undertakes to explicate a concept of religion as progress toward the “dominion of the good on earth”, especially in relation to Kant’s ideas of natural religion and ethical community. It is inferred that Cohen’s difference between progress in religion and religious progress opens a path to a definition of religion as pre-institution (religion without religion). The goal is to make the concept of religion distinct in the tradition of critical philosophy and its logic. I argue that the emancipatory project of Kantian public reason presupposes a set of rules defining the translation from postulates (Kant’s rational theism) to problems and assignments. Since natural religion is a pure practical concept of reason, religions are subject to moral evaluation. The latter being guided by the pragmatic maxim of overcoming logical and moral egoism, means that any community, even many communities at once (different cultures) can occasionally represent an ethical community, if not in the sense that an ethical community is constituted.


Giving an account of religion vs making a concept of religion distinct
It has been claimed that religion, as accounted for by Kant in the second Critique and the writings of the late period, is devised to "bridge across the great gulf" (cf. O'Neill, 1996, p. 271) dividing theoretical and practical reason. Kant's concept of causality, so the argument goes, commits him to strict determinism in theoretical matters, while he resorts to the idea of a transcendental freedom, consisting of pure self-determination, in order to make rational acting possible for human reason in a world of Newtonian causation (compatibilism). The moral agent, however, is itself committed to a formalism which, more often than not, has been deemed bloodless and void. The bridge, then, stands for the hope of congruence between virtue and bliss (see KrV, B 526;Kant, 1998, p. 680). By thus considering Kant's philosophy of religion, the concept of religion is taken for granted. A clarification of the concept of natural religion is important though, because from the conjunction of coordinated markers 1 within it, an argument follows (entailment in the legal sense, not a mere logical connection or Verknüpfung): if natural religion is shown to have the force of a drive to morality, then secularisation needs to be an inner-religious dynamic, culminating in the idea of a religion without religion. One consequence of the argument is in direct contrast to the idea of postsecularism as the statement that the distinction between the secular and the religious is hopelessly blurred in our time.
Here the further claim is made that Cohen developed his philosophy of religion along the lines of a rational reconstruction of the concept of 'natural religion' and was not exclusively or mainly motivated by the wish to illustrate the historical contribution of Judaism to a "history of pure reason", as is sometimes argued. 2 2 Practical and theoretical as opposed to speculative cognition For the purpose of reconstructing Kant's concept of natural religion, a tripartite articulation of cognitions ought to be referred to. On the one hand, the distinction between theoretical and practical cognition is relevant (Log, AA 09: 86; Kant, 1992, p. 587). Notice that this opposition is reiterated within the practical realm, with the distinction of object and form of acting. The relevance to the concept of religion consists in the systematic connection of the concepts of God and soul with the concept of finite rational beings: the postulates are introduced in relation to the object (summum bonum consummatum) of the free moral act. They are, so to speak, the silent care-takers safeguarding the proportionality of virtue and bliss. When Kant's concept of religion is at stake (cf. Mackie, 1982, p. 106) this piece of reasoning within the Kantian corpus has to bear the main burden of proof. Religion after Kant, then, is equated sans phrase with an open theism. 3 On the other hand, for Kant the "practical" concept marker is defined via a double opposition: "[…] practical cognitions are, namely, either, 1. imperatives, and are to this extent opposed to theoretical cognitions; or they contain, 2. the grounds for possible imperatives and are to this extent opposed to speculative cognitions. […] By speculative cognitions we understand, namely, ones from which no rules for proceeding can be derived, or which contain no grounds for possible imperatives" (Log, AA 09: 86; Kant, 1992, p. 587).
Since religion (natural religion), for Kant, has its place within practical philosophy; and, furthermore, since theology deals mostly with speculative propositions, it follows that a proper treatment of religion within the Kantian system hinges on a concept of religion made 1 For the term Merkmal, the translation 'marker' used in Kant (1974) is preferred to 'mark', which is consistently used in the newer translation (Kant, 1992). This is because the word 'mark' corresponds to the German Markierung, also Wegmarke, while 'marker' better renders the idea of an epistemic ground: only concepts are markers in Kant's logic. Since the newer translation is widely referred to in the literature, in this contribution I will refer to it. 2 For instance see Holzhey (1997, pp. 103-104). 3 Notice that the above quoted text of Baroness O'Neill explicitly rules out deism. As for theism, O'Neill seems prone to a position close to that of Swinburne. Cf. Swinburne (1986, p. 9): "The hypothesis of theism is that the universe exists because there is a God who keeps it in being and that laws of nature operate because there is a God who brings it about that they do." On Kant's Deism see Wood (1991). Wood makes the case for natural religion and against revelation, especially concerning Religion within the Boundaries of Mere Reason. He does not, however, refer to conceptual clarifications such as the one in "On the miscarriage of all philosophical trials in theodicy" (RGV, AA 08: 257; Kant, 1996a, p. 25), where Kant clearly distinguishes the moral concept of natural religion from speculative considerations (Physikotheologie) and argues that the first derives its drive toward the moral law (duty) from its inner articulation, not from transcendental considerations. Moreover, the adjective (and adverb) bloß means actually 'naked', while the rendering 'unassisted' 'unaided' (Wood, 1991, p. 2) is a free paraphrase.
, 05002 (2023) t h   distinct. That means, in turn, understanding the concept of 'pure will' in relation to its object (summum bonum), in so far as practical principles (postulates) regarding that object are subordinated markers to the concept of 'pure will'. Furthermore, the claim is made in this paper that a clarification of 'natural religion' can be achieved only on condition that the concept entails rules of substitution to assign practical or theoretical propositions to speculative ones. A critical concept of natural religion has to legitimate a 'translation' of postulates into problems, 4 because it is the expression of a rational interest (see MS, AA 06: 212; Kant, 1996b, p. 374). For our cognition, religion in relation to quality is a matter of understanding, not of comprehension.

Natural religion and natural history of religion
In the "Lecture on Philosophia Practica universalis una cum Ethica" (1780-1782) 5 Kant expounds three arguments for the independence of natural religion from speculative theology. The argument regarding simplicity states that morality precedes religion, the first leading to the second. Thus, morality has to be applied to a coordinated marker of pure will in order to grasp the genesis of natural theology, and that is the concept of a holy legislator. Holiness originates in the concept of a qualitative whole (the moral law) and is, secondly, linked to the concept of moral completeness (the moral attribute of goodness). The argument concerning completeness states that religion is the necessary complement of morality, because it is only within the idea of sanctity, goodness and blessedness that the manifold of beliefs about what sort of person I am, becomes a set or a system under a unifying principle. Thirdly, rituals and religious cults are not, strictly speaking, religious actions, but (at best) the effects of religious acts. 6 Religion is equated with the moral attitude (Gesinnung). 7 To connect the modes of cognitions to the systematic role of what Kant calls natural religion, means devising in the concept of religion moments of a logic of transition. By contrast, the speculative type of cognition implies an option for a transformative logic or a logic of transposition. 8 The speculative kind of cognitions, therefore, defines a different research project in relation to sets of rituals, myths, supernatural beings and belief systems, as well as historical facts about such sets, insofar as they are referred to or constructed by what Kant names Humaniora (as different from Polyhistory) (Log, AA 09: 45; Kant, 1992, p. 554). 9 A second distinction becomes decisive here: preliminary judgements and prejudices. Kant deals with history as scientific inquiry (critical inquiry, in his words) in the section entitled "Logical Perfection of Cognition as to Modality" of his Logic (Log, AA 09: 65; Kant, 1992, p. 571). It is said there that history is a kind of experience, thus resulting in knowing (holdingto-be-true on objectively and subjectively sufficient grounds). Thus, history is a type of rational cognition, as opposed to beliefs and opinions. A crucial aspect of rational cognition as such is, Kant says, the reservation of judgement. Provisional judgements "are very necessary, indeed indispensable, to the use of the understanding in all meditation and investigation. For they serve to guide the understanding in its searches and to that end place at its disposal various means" (Log, AA 09: 74; Kant, 1992, p. 578). Provisional judgements, if assumed as principles, lead to prejudices. In this second type of considerations, a certain structural similarity between Hegel and Kant can be inferred. From the point of view of erudition, historical facts about religions (monotheism, religious concepts, the afterlife, immortality, eschatology), can be related to narratives, such as an unfolding pattern of rationality. If the two aspects are now brought together (the doctrine of postulates and the remarks on history), the following concept of religion can be formulated: religion is the systematic unity defined over an operational manifold (symbolic manipulations) under the concept of an ethical community. Thus, religion in the sphere of practical rationality is to be understood in functional terms; it is, in other words, not a set of institutions, but a set of functions which organise the institutional sphere in the context of what is deemed imponderable and inestimable both within the given infrastructure and beyond its boundaries. This goal is achieved by providing representations of a community's institutions and, above all, by projecting types of actions onto sets of objects (institutional orderings). Most importantly, religions organise and implement rules for inclusion-exclusion in shifting doctrinal systems. For instance, the category of quiescent or passive supernatural beings (founding fathers, heroes, animal spirits, ancestors; not Kant's example) can be understood as the result of an exclusion of certain factors from the scope of defining the scheme actions-consequences in relation to powerful interventions (be it a blessing or a retribution). This functional exclusion is an operation within a religious system. One could label this definition the osmotic concept of religion, since osmosis is defined as a spontaneous process working without energy input into a system.

Cohen on reconciliation and the recursive function of religion
The idea that a history of religions has to be pursued within the framework of historical inquiry, thus in relation to an open horizon; and, consequently, the idea that reconciliation has to be conceived as a cyclical pattern, can already be found in Kant. It is the idea that religious progress is a complex process of reiterated self-exclusion and self-inclusion of the religious syndrome into history. 10 The concept of religious progress is central to Cohen's philosophy of religion. Already in Ethics of Pure Will (first published 1904), the discovery of individual culpability in the Babylonian prophetical literature is contextualised within a genetical account of autonomy. With the fourfold articulation 11 of the latter, Cohen (1981, p. 338) identifies "Kant's mistake" 12 of self-legislation as "legislation from the self" (ibid., p. 339). The formula of autonomy as self-legislation after Cohen (1981, p. 339) is "self-legislation is not legislation from the self, but to the self". His Religion der Vernunft was published posthumously in 1919 with a Geleitwort by his wife Martha Cohen-Lewandowsky, who was to be murdered on 12 September 1942 in Auschwitz, blessed be her name. In Religion of Reason out of the Jewish Sources, Cohen's undertaking is quite the opposite of Religion within the Boundaries of Mere Reason; in Religion of Reason, freedom is the horizon, not the presupposition of a theory of agentiveness. Religion, according to this systematic approach, is a religion without religion; consequently, politics can be conceived of as the realm of conflict resolution (the problem of derivations in relation to a conflict of value where different traditions are in conflict or the strands of one tradition are juxtaposed).
One of Cohen's main contributions in his late work is the differentiation of the concepts of progress in religion and of religious progress. By the first, Cohen means the process (with pre-exile prophets) of moralisation of religion. Social morality (soziale Sittlichkeit) is the interest behind, for instance, Isaiah 2:22: "Cease ye from man, whose breath is in his nostrils: for wherein is he to be accounted of?" (Cohen, 1966, p. 209), since the "dissolution of the individual is the highest triumph of ethics" (ibid., p. 208).
Ezekiel, on the contrary, marks a progress which is to be addressed as religious progress, because he discovers the individuality of the human being by means of the idea of sin before God; surmounting the concept of kin punishment (only accountability without culpability) and pointing to reconciliation as the transcendence-individual relation, Ezekiel, in Cohen's reading, turns around one of the famous paradoxes of Luther's: what man, as man, can do, is what man, as sinner, cannot do. Is the idea of religious progress in Cohen's sense a real transition toward a cognition of structural dependencies in the field of an ethical community? If one bears in mind that the typical debates about welfare and social policies echo the theological arguments concerning atonement, the question is of a certain interest beyond the systematic point of view of tracing Kant in Cohen, and Ezekiel in both. Besides, as has been pointed out above, the idea of a religious progress is a promising one in relation to the debate on public reason. Both issues will not, regrettably, be discussed in the present paper. Cohen sees the prophetic literature up to Ezekiel as mainly engaging the problem of social ethics. In this context, says Cohen, Mensch is taken issue with only as the problem of the Mitmensch (the thou, in Buber's words). The thou is the unspoken presupposition of the I: "Yet, is the I but the bottom line of the thou?" (ibid., p. 208). Ezekiel, on the contrary, represents a stage of theological ethics, where the mere ritual correspondence between action and consequence, formalised with the means of reward and punishment, is seen in its deficiency within the context of a definite moral injury (the remnant of Israel). The religious progress in this respect is achieved by Ezekiel with the sentence: "the soul sins" (ibid., p. 222). In the retributivism of the pre-exile prophets, the medium of accountability is the trans-generational line; when the author(s) of Ezekiel state that culpability is not transmitted in the same way, they are de facto claiming that this first aspect of a progress in religion (ensuing continuity in the diaspora) does not contain, as its implication, the second of a religious progress: culpability is a strict individual category (it is not a category of the ethical community, the church).
The distinction of religious progress and progress of religion is, thus, to be understood as follows: the postulate of a righteous transcendent being is requested in order for the ethical community to emancipate its historical roots (Urzeit) from the idea of godly revenge and crude retribution (lex talionis); natural religion as the stage of simplicity (Kant) is expelled from the community's historical conscience, in order for reason (Cohen's individual) to anticipate the fulfilment of a kingdom of ends in a troubled present (to conceive of an end to all things in a moral, practical respect).

Kinship and universalism. Religious progress
Progress is the key category that allows formulation of the 'rule of replacement' to translate the postulates of pure practical reason (immortality) by substituting for them problems (practical sentences expressing a sort of practical incompleteness, characterised by the principles of reliability and maximality not holding). Progress, in the Kantian sense, is relative to the difference between logical and aesthetic perfection. Since progress is a pragmatic idea of anthropology, presupposing the manifolds of experience and historical conscience, as well as reflections based thereupon (sensus historiae); progress as the symbolic construction of a continuing turn for the better refers to aesthetic perfection ("the agreement of cognition with the subject […] based on the special sensitivity of man" -Log, AA 09: 36; Kant, 1992, p. 549). Since aesthetic perfection is compatible with logical perfection only as to essential beauty (Log, AA 09: 37; Kant, 1992, p. 550), whilst aesthetic perfection also concerns sensation and therefore the agreeable, a conflict between the two perfections arises. This conflict, which expresses itself in the sphere of culture as a rupture of rational argument from practical public issues, can be resolved if a third species of perfection is defined: religious progress. Religious progress is an intuitive representation of a duratio noumenon (EAD, AA 08: 327; Kant, 1996a, p. 221): it will be a schematic representation in relation to the practical purpose of a reconciliation between maxims; a symbolic one, with regard to the idea of an ethical community. In analogy to logical and aesthetic perfection, religious progress consists in the set of rules defining the transition from a characteristic system to a culture, or system of exhibitiones, expositions. Thus, the following can be stated: religious progress defined for the sake of applying a speculative cognition to a manifold of cases (maxims) yields a kind of generality (1); the rules that govern the institutional framework of the manifold ought to be spelled out; here, the substitution of the postulates through schemes (problems) is weakly justified in terms of an emerging subjective conflict of cognitions within a sensibility.
(2) Distinctness: The construction of a concept in intuition (Ez 18:14 "Behold, all souls are mine") cannot be substituted for the concept itself (if a scheme is taken for a concept, an idolatry or ritualism can be the consequence. This consequence is a pernicious one, because it would strike through the very rule of replacement implied by the concept of religion). That is, between (1) and (2) there is a clear hierarchy: the boundaries of religious distinctness are those of generality. In exegetical terms: allegorical interpretations of traditional maxims (for instance, "Like Fathers, Like Sons", in reference to John 5:19-30) have to be preliminarily dismissed, until a moral interpretation becomes available as a result of historical arguments.

Conclusion
Above, the idea of a religious progress has been connected to the Kantian concept of the horizon. This idea has its place in the Kantian doctrine of logical perfection as to quantity: historical horizon and rational horizon are distinguished. While a historical horizon has no boundaries (Log, AA 09: 41; Kant, 1992, p. 554), the rational horizon can and should be circumscribed as to the possible objects and the limits (indeed, this is precisely the purpose of a critique). The fact that historical cognition has an open horizon means that groups or societies are conceivable, where religion doesn't play any role in relation to the problem of symbolically ruling over imponderable and impenetrable environments. The related question of whether religion is dependent on institutions is a complex one and cannot be dealt with in this paper. Suffice it to say that religious acts carry a lot of clout in institutions. The institutionality of religion, it could be said, consists precisely in the fact that religions refer to a symbolism and have been defined as the manipulation of a set of symbolic actions (institution) for the sake of representing control. However, a reason has been given above which seems to justify that religions are pre-institutions, rather than an institution in its fullest sense: the fact that a religion can reorganise an institution, even excluding parts of a religious belief or setting supernatural beings in a passive state, as in founding myths and heroic legends. An institution, that is, that can include itself by excluding itself from a given infrastructure. If the idea of a religious progress figures in the very structure of the concept of public reason, then it can be said that any community, even many communities at once (defined by different cultures coexisting) can occasionally represent an ethical community, if not in the sense that an ethical community is constituted.