Grounding the Highest Good: the omnipotent, omniscient, and omnibenevolent God of Kant’s moral argument

Introduction^[1]

Kant famously claimed that he “had to deny knowledge in order to make room for faith” (B XXX). The distinctive epistemic status of faith has attracted significant scholarly attention.^[2] Kantian faith is justified not by evidence, but rather by the role of a putative assent in promoting our necessary ends, including the highest good. But this leaves open the question of why certain faith-contents, but not others, play a role in achieving some given end. In this paper, I foreground the question of why Kant thinks faith has the very content that it does, focusing on the postulate of God. Why does Kant hold achieving the highest good requires faith in God, rather than faith in, say, a demi-God?

Kant intended to answer this question. He was aware of Hume’s challenge in the Enquiry (1748) to physico-theological arguments for the existence of God, on the grounds that such arguments could not secure knowledge in a being with traditional ‘omni’-properties such as omniscience, omnipotence, and omnibenevolence. At best, such arguments could secure assent to a being who is very knowledgeable, very powerful, and very good. Kant considered his moral arguments for faith in God’s existence (hereafter: ‘moral arguments’) to have the advantage of justifying assent to the existence of a being with the traditional omni-properties, rather than merely large finite properties. Kant made this point in presentations of the moral argument throughout the Critical period.^[3] However, it is unclear how Kant’s moral arguments secure this result. Scholars working on Kant’s moral arguments commonly deny he had a satisfactory account of faith’s content.^[4]

In this paper I support Kant’s position by providing a new reconstruction of Kant’s moral arguments, which I call the ‘grounding interpretation’, that explains how Kant secures faith in a being with all the traditional omni-properties.^[5] This interpretation holds that Kantian moral faith requires assent to a being that is all-sufficient, or the ground of all possibility, and that such a being has all the omni-properties. Scholars have examined at length both Kant’s pre-critical argument for a ground of possibility, as well as a putative Critical argument for theoretically grounded assent to a ground of possibility.^[6] In this paper, I argue that Kant’s Critical moral arguments offer practical grounds for assent to this demanding conception of God. Moreover, the metaphysically demanding conception of God as all-sufficient is given a practical interpretation, as entailed by our conception of a completely independent agent, or the “will of a self-sufficient being” (AA 28: 1060; see Kain 2021). This will be used to make sense of Kant’s position that his moral arguments offer practical grounds for assent to the full traditional ‘omni’-God.

In the first section of this paper, I introduce the problem of explaining how the moral arguments secure faith in a being with the omni-properties (1). In the second section, I present the ‘grounding interpretation,’ which is an interpretation of Kant’s moral arguments capable of securing that faith-content (2). In the third section, I clarify the sense in which the grounding interpretation provides a practical rather than theoretical argument for faith in God (3).

1 Hume and the “shrunken God” objection

Kant was sensitive to the question of whether a putative argument for God’s existence justified assent to a being with the traditional divine properties (the ‘omni’-properties), or merely a being with large finite properties. In this section I first establish that Kant was aware of this question and consistently held that his Critical-era moral arguments answered it. I then argue that it is not clear how Kant’s answer works, and that contemporary treatments of Kant’s views on this point are revisionary.

Kant first became aware of the challenge of accounting for God’s properties via Hume. In the Enquiry (1748), Hume raised a well-known objection to physico-theological arguments for the existence of God.^[7] Physico-theological arguments appeal to a need to explain the order, beauty, and/or purposiveness in the world. Hume holds that these arguments merely secure a being with large finite properties:

Allowing, therefore, the gods to be the authors of the existence or order of the universe; it follows, that they possess that precise degree of power, intelligence, and benevolence, which appears in their workmanship; but nothing farther can ever be proved, except we call in the assistance of exaggeration and flattery to supply the defects of argument and reasoning (Hume, Enquiry, § XI, 94).

To the extent that one can only observe a finite degree of power (etc.), one is only entitled (so Hume argues) to posit a cause with that precise degree of power (etc.). In the Beweisgrund (1763), Kant endorsed Hume’s charge against the physico-theological arguments (BDG, AA 02: 160).^[8] I will call the question of whether a putative argument for God’s existence genuinely secures assent to a being with the traditional omni-properties, “Hume’s challenge.”

Throughout the Critical period, Kant held that his moral arguments for the existence of God could meet Hume’s challenge. Indeed, he repeatedly emphasized that this was a major advantage of his moral arguments:

And then, in its unavoidable problem, namely that of the necessary direction of the will to the highest good, there is shown not only the necessity of assuming such an original being in relation to the possibility of this good in the world but – what is most remarkable – something that was quite lacking in the progress of reason on the path of nature, a precisely determined concept of this original being. (KpV, AA 05: 139; compare KrV, A 814/B 842; KU, AA 05: 444, V-Phil-Th/Pölitz, AA 28: 1071.)

But how did Kant’s solution work? Kant addresses this question in each of the three Critiques, the Religion, and his 1783/4 lectures on rational theology, offering variations on the same response:

This being he [i. e. the moral agent] will have to think of as the most perfect, for otherwise his morality would not obtain reality through it. It must be omniscient if it is to know the smallest stirrings of his innermost heart and all the motives and intentions of his actions. And for this merely much knowledge will not suffice, but only omniscience. – It must be omnipotent, so that it can arrange the whole of nature to accord with the way I act regarding my morality. It must even be holy and just; for otherwise I would have no hope that the fulfilment of my duties would be well-pleasing to it. (V-Phil-Th/Pölitz, AA 28: 1012, compare AA 28: 1073; KrV, A 815/B 843; KpV, AA 05: 140, KU, AA 05: 444; RGV, AA 06: 139.)

God’s role in relation to morality is that of upholding the possibility of the highest good. His role in that capacity is to distribute happiness in proportion to virtue, and (so the argument goes) only a being with all the traditional divine properties could carry out that task. Thus, the key contrast between physico-theological arguments and Kant’s moral arguments is that only the latter identifies a task that can only be performed by a being with all the traditional divine omni-properties.

Kant’s response situates him within a distinctive strand of Pietist and Lutheran philosophical theology that stressed the practical significance of achieving the correct account of the divine properties. Kant’s argument is prefigured by that of the Lutheran theologian Johann Gustav Reinbeck (1740), who cited Psalm 139 to draw a connection between God’s omniscience, God’s capacity to know the smallest mental phenomena, and emotional intimacy between God and the believer (1740, 135–137). Kant was aware of this work.^[9] Kant’s library also contained a manual for schoolteachers written by Johann Bernhard Basedow (Warda 1922), holding that an analysis of the divine properties can reinforce students’ belief in the existence of God (1781a, 57), as well as Basedow’s Christian songbook with paeans to divine omniscience, omnipotence, and omnibenevolence (1781b). Eberhard’s Vorbereitung (1781), which Kant used as a set text for his rational theology lectures, holds that errors in representing the divine properties are the source of polytheism and superstition (§ 43, 18: 580). For Kant, as for his theological sources, overcoming Hume’s objection was not just a philosophers’ game. Rather, Kant’s moral arguments are part of a broader tradition of Protestant theology that emphasizes the practical stakes of properly representing the divine properties.

However, it is hard to see how Kant’s own way of appealing to practical considerations secures faith in a being with traditional divine properties. Consider omniscience. Kant claims that, in order to distribute happiness in proportion to virtue, the distributor must know each rational agent’s “motives and intentions” in acting. But this falls short of omniscience. For instance, it fails to include mathematical knowledge. Or consider omnipotence. Kant claims that the distributor-being of the moral argument is omnipotent insofar as it makes it the case that each moral agent receives the amount of happiness that they deserve on the basis of their virtue. However, the power to bring about happiness in proportion to virtue is not omnipotence. Such a being need not have the power to create a world ex nihilo, thereby having the status of a “creator” rather than that of an “architect” who rearranges pre-existing material (V-Phil-Th/Pölitz, AA 28: 1093 f; see also Watkins 2013, 228).

It seems, then, that Kant’s moral arguments don’t succeed in addressing Hume’s challenge. Contemporary interpreters concur on this. Gordon Michalson holds that Kant’s moral arguments lead to “theistic shrinkage” by failing to supply the traditional divine properties (1999 ch. 2, 34–35, 42). Allen Wood (citing Michalson) links this point to the finitude of God’s properties, worrying that Kant’s moral arguments supply us only with a “shrunken God” (2020, 28–29). Other interpreters have noted variants of this problem, albeit largely in passing, and suggest that Kant should have lowered his expectations for his moral arguments. I consider two such treatments.

One strategy, which I will call the ‘supplemental strategy’, has been proposed by Eric Watkins (2013, 229–230) and Desmond Hogan (2014, 48). I shall focus on Watkins’ more detailed development of this proposal. Watkins suggests that Kant’s moral arguments establish only the moral attributes of the ground of the highest good (holiness, benevolence, justice), and not any of its non-moral attributes (229). Watkins argues that we should “supplement” the conclusion of Kant’s moral arguments (i. e. faith in a being with moral attributes) with the first Critique’s independent theoretical argument for the necessity of assuming an ens realissimum (i. e. faith in a being with attributes such as omnipotence and omniscience) to generate a “more differentiated case for God’s existence” (229).

Watkins offers support for this strategy by appealing to Kant’s categorization of the divine properties into transcendental, natural, and moral (230). Kant, drawing on Eberhard’s Vorbereitung (1781, § 71, 18: 597 f), identifies three moral properties or “moral perfections” of God as holiness, benevolence, and justice (AA 28: 1073; compare KpV, AA 05: 131fn, KU, AA 05: 444; RGV, AA 06: 140–141, MpVT, AA 08: 257–258). These three properties make up the concept of God that Kant associates with the Trinity in the General remark to Part III of the Religion (RGV, AA 06: 137–142).^[10] Kant identifies the intellect and will as “natural properties” (KU, AA 05: 481) or “psychological predicates” of God, which he also associates with omniscience and omnipotence (V-Phil-Th/Pölitz, AA 28: 1020–1021; see also KrV, A 631/B 659, KpV, AA 05: 137, KU, AA 05: 481). The “transcendental” properties include eternity and omnipresence (KU, AA 05: 444; see also KrV, A 631/B 659, AA 28: 1020). Kant correlates this threefold categorization of the divine properties with the division of theology into transcendental, natural, and moral (AA 28: 999–1000; cf KrV, A 631/B 659). This division reflects whether one represents God with “transcendental concepts alone,” with “physical concepts,” or with “concepts taken from morality” (AA 28: 999–1000). Watkins appeals to this threefold division to support the claim that Kant’s moral argument secures only the moral attributes of God (230).

This strategy is philosophically appealing. However, Kant’s division among the branches of theology does not offer it textual support. Rather, Kant uses this division to reiterate his claim that the moral arguments supply all of God’s moral, natural, and transcendental properties. Kant states that “moral teleology” secures the “natural properties” of omniscience, omnipotence, and omnipresence (KU, AA 05: 481). Kant likewise holds that the moral argument’s appeal to the highest good justifies attributing to God the “transcendental properties” of eternity and omnipresence (KU, AA 05: 444). In the section of the rational theology lecture notes discussing moral theology, Kant argues that the “determinate concept” of God based on “morality alone” secures “every perfection,” including omniscience, omnipotence, eternity, and atemporality (V-Phil-Th/Pölitz, AA 28: 1073; compare 1012). So while Kant does indeed identify a proper subset of the divine properties – holiness, benevolence, and justice – as making up an “entire moral concept of God” (AA 28: 1073), he holds that the moral argument secures all the transcendental and natural properties, as well.

Andrew Chignell (2022, 68; 2023, 55–56) has proposed a different strategy for responding to the shrunken God objection. I call this the “pluralistic strategy.” Chignell puts forward this proposal as part of his “moral-psychological” reading of Kant’s moral argument, which Chignell identifies within the third Critique (2022, 66–67; 2023, 52–57). On this view, faith in the achievability of the highest good is justified insofar as its non-achievability would be demoralizing, or have an “enervating effect on my resolve” (2023, 54). Chignell proposes that this gives the moral-psychological reading an advantage for responding to Michalson’s shrunken God objection: for some people, it may be the case that their resolve requires that the world-order be “entirely stable and not in principle vulnerable to disruption through mistake or intervention of a stronger or smarter power” (2023, 55; compare 2022, 68). The need for this guarantee, however, is a matter of empirical fact, and different people may psychologically require different kinds of being. Thus, Chignell proposes that some particularly resilient agents might be satisfied with the view that “nature itself” or “karmic law” underwrites the highest good, rather than an omni-God (2022, 69).

By Chignell’s own admission, such pluralism is revisionary (2022, 66). For Kant claims that highest good is possible “only on the presupposition of an author of the world possessed of the highest perfection” (KpV, AA 05: 140; former emphasis mine; compare KrV, A 815/B 843 and A 828/B 856). Likewise, the third Critique holds that we “must conceive of it [the original being]” as a “legislative sovereign,” before introducing the omni-properties as part of a list of what “must … be thought in such a being” (KU, AA 05: 444; emphasis added). Now, there is a scholarly tradition of reading the Religion, and Kant’s philosophy of religion as a whole, in a pluralistic vein, according to which Kant provides an account of rational religion that different faith traditions can in principle satisfy.^[11] But being pluralistic about whether rational faith requires an omni-God (versus a demi-God or karma) differs from being pluralistic about which religions offer an account of such an omni-God. Kant’s receptivity to attributing faith in God to non-Christian religions such as Zoroastrianism and Hinduism (e. g. RGV, AA 06: 140 fn.) does not provide evidence for pluralism about whether an omni-God is required by rational faith. Chignell puts forward his pluralistic strategy as a philosophically friendly addendum to Kant’s views. But this leaves us with the question of how Kant understood his own moral argument for an omni-God.

Kant took his moral arguments’ ability to address Hume’s challenge as one of their major advantages, yet his own reasoning on this point seems highly disappointing. Both the supplemental strategy and the pluralistic strategy offer revisionary fixes to Kant’s putative failure to address Hume’s challenge. But I would suggest instead that this shows we do not have an adequate account of Kant’s moral arguments. In the next section (§ 2), I supply such an account.

2 The grounding interpretation

In this section I present a new interpretation of Kant’s moral arguments for faith in God’s existence, which I call the ‘grounding interpretation.’ This interpretation meets Hume’s challenge. Its strategy is to examine closely what is involved in God’s grounding a proportional distribution between happiness and virtue. The core idea is that, by focusing on precisely why, how, and in what way such a proportional connection is grounded, we will see why God has all the traditional divine properties.

This interpretation begins ‘halfway through’ the moral argument, with the claim that morality requires faith in the achievability of the highest good. Scholars have paid close attention to the initial steps of the moral argument, which establish this point. They make explicit in many steps what Kant expresses in a few words,^[12] and look throughout the Kantian corpus to supply suppressed premises.^[13] My interpretation shows that the latter stage of the moral argument benefits from equivalent treatment. Two major overall approaches in the literature for understanding this first step are the “rational coherence” reading which links the achievability of the highest good to the rational structure of morality itself (e. g. Wood 1970, 2020), and the “moral-psychological” reading which links the achievability of the highest good to moral-motivational considerations (e. g. Ebels-Duggan 2016, Chignell 2022, 2023).^[14] The grounding interpretation stresses the importance that happiness and virtue be nomologically connected as part of the highest good, which is more naturally motivated by a rational coherence reading. In invoking a ‘nomological connection,’ I mean that the connection is law-governed: the virtuous receive their happiness because of a law that relates happiness to virtue. One can understand the grounding interpretation as showing how rational coherence readings have the resources to make sense of Kant’s requirement that moral faith concerns an omni-God. However, this reading can be compatible with a broader variety of explanations for our faith in the highest good, so long as such explanations secure the point that the connection between happiness and virtue is nomological or law governed.^[15]

The grounding interpretation appeals to a metaphysically substantive conception of God as a ground of possibility. This might seem surprising, if one expects that Kant’s appeal to practical considerations in moral theology was supposed to replace any appeal to metaphysical considerations. However, as I noted in the previous section, Kant’s approach can be read as part of a broader tradition of Pietist and Lutheran theology that attempted to synthesize rational-metaphysical and practical considerations. Kant’s rejection of theology as a science is compatible with a more nuanced reading of Kant’s views on the role of metaphysical principles within moral theology. I now turn to the reconstruction, which proceeds in four steps.

Step 1: [I believe that]^[16] happiness and virtue ought to be necessarily proportionate

This step specifies the content of what we must be able to believe, as part of believing in the achievability of the highest good. The highest good has two components: an individual’s complete virtue, and happiness proportioned to virtue. We are responsible for the first component, but not the second. That is, we bring about complete virtue, but not happiness proportioned to virtue, insofar as the former, but not the latter, can be brought about by our agency.^[17] The latter is the focus of the moral arguments for faith in God’s existence. This much is standardly accepted.

What my reconstruction in particular stresses is that, in the highest good (which ought to be achieved), happiness and virtue are necessarily proportionate: “virtue and happiness are thought as necessarily combined” (KpV, AA 05: 113, emphasis added; see also 115, and KrV, A 815/B 843). Kant insists on the necessity of the connection because the “combination [of happiness and virtue] is cognized a priori” (KpV, AA 05: 113), and that which is cognized a priori is necessary: “Reason in abstraction from all experience can cognize everything only a priori and necessarily, or not at all” (KrV, A 775/B 803; see Bader 2015, 187 f). The proportional connection is not merely universal but also necessary.

One might wonder if it is artificial to stress necessity in this way. Some interpreters take Kant’s moral arguments to center around some precisification of a point about the existential inadequacy of a world in which virtuous action is accompanied by misery (e. g. Ebels-Duggan 2016, Chignell 2023). On such views, it might seem unimportant whether the connection between happiness and virtue is understood as obtaining necessarily, or merely universally.

However, in order to achieve an interpretation of the moral argument capable of making sense of Kant’s requirement that this argument reaches to an infinite being, to God, we must lay emphasis on a further point: that the achievability of the highest good is posited or demanded by reason (e. g. KpV, AA 05: 107 f; 144 f; KU, AA 05: 455 f). As we have seen, reason is a faculty that cognizes only a priori and what is necessary. This point, in turn, reflects how reason is a faculty that grasps, or comprehends, through grounds. This point is reflected throughout Kant’s system; for instance, in the claim that reason cannot cognize or even think something that is unconditionally necessary, because this requires cognizing and thinking through conditions, or grounds (KrV, A 593/B 621; R 4007, Refl, AA 17: 383; see discussion of the latter in Hogan 2021 ft. 95, 435 f). Kant applies this more general account of reason’s operation to his characterization of the highest good, claiming that because the necessity of achieving the highest good is cognized a priori, “the condition of its [the highest good’s] possibility must therefore rest solely on a priori grounds of cognition” (KpV, AA 05: 113, emphasis mine). Since the connection between happiness and virtue is demanded by reason, it has to be graspable by reason itself, which means graspable via its grounds.^[18]

The necessary connection that reason posits between happiness and virtue is synthetic and nomological. Claims of a necessary connection are either analytic or synthetic. As Kant establishes in the second Critique’s Dialectic, the necessary connection between happiness and virtue is not analytic (KpV, AA 05: 122 f). For if that connection were analytic, happiness and virtue would be logically related to one another. However, that is false, and because it is a logical falsehood, not even God can make it true. This means that the necessary connection between happiness and virtue is synthetic a priori. A synthetic a priori connection requires a further principle or “third thing” in virtue of which that connection obtains (KrV, A 155/B 194, A 157/B 196, GMS, AA 04: 447). In the context of the moral argument, the “third thing” is a law: “it must be thought synthetically and, indeed, as the connection of cause and effect” (KpV, AA 05: 113). Moreover, this causal connection is not itself brute. For this connection is posited by reason, and as we have seen, reason is a faculty that grasps in virtue of grounds. Since the nomologically necessary connection between happiness and virtue is demanded by reason, it has to be graspable by reason itself, which means graspable via its grounds. The next step specifies the nature of those grounds.

Step 2: The ground of the necessary proportionality of happiness and virtue is the actualizer of the essences of things that enter into that proportional connection

This step holds that the ground of a nomological connection between happiness and virtue is the actualizer of the essences entering into that connection. This step establishes that for Kant, talk of ‘laws’ and talk of ‘essences’ are not separate; rather, laws are grounded in essences, and hence establishing laws requires (at a minimum) determining which essences are actualized.

Kant states that nature’s “complete purposive unity” is to be found “in the essence of things which constitute the whole object of experience […] hence in universal and necessary laws of nature” (KrV, A 694/B 722; emphasis added). Kant moves from talk of essences to talk of laws of nature. This reflects his more general views on causal laws: that causal laws obtain in virtue of the essences of the actual substances, which determine one another through unchanging grounds that are part of their essential natures (Watkins 2005, 244). Kant scholars have developed variants of this view on which the grounds of laws of nature lie in the essences of things or the natural kinds to which they belong (Stang 2016 ch. 8, Massimi 2017, Messina 2017). This close link between laws and essences explains why, in the Religion, Kant says that God can be “considered the ultimate source of all natural laws only because he is the creator of natural things” (RGV, AA 06: 142). For Kant, establishing the laws of nature is a matter of determining which things are actual. Kant then makes use of this point in the moral argument for God’s existence: establishing a causal law relating happiness and virtue requires that certain essences be actualized, with the result that this necessary proportionality is obtained. Faith in the achievability of the highest good requires faith in something capable of actualizing essences in that way.

So far, then, we have a justification for faith in an ‘actualizer’ of essences – in particular, an actualizer of some essences, or those essences that are actual. But it isn’t yet clear that this is a genuinely infinite being, rather than a shrunken God. The next two steps establish this.

Step 3: The ground of the necessary proportionality of happiness and virtue is the ground of the essences of all things

This step involves strengthening the conclusion of the previous step, which established that a necessary connection between happiness and virtue obtains in virtue of an actualizer of some essences. In this step, what I show is that Kant’s moral argument requires something stronger: a ground of all essences. Kant’s characterization of God as the ground of essences is central to his moral argument as a whole, and to his response to Hume’s challenge. I will first provide textual support for this point, then explore its philosophical rationale.

Kant identifies the God of moral theology as “the highest good himself through the nature of things in general” and as “self-sufficiently the source of the possibility of things” (R 6099, Refl, AA 18: 451–453, est. 1783–84; emphasis mine). In a further note, Kant states that while the “atheist” can succeed in explaining a “contingent” agreement of ends, a “necessary agreement” of ends cannot be explained if “I take the essences of things as given independently for themselves,” but rather requires as its ground the “necessary unity of the being from which all possibility is derived, in which possibility and reality is not distinct” (R 6292, Refl, AA 18: 560 f, est. 1785–88; emphasis mine). So, Kant explicitly characterizes the God of moral theology as the source of all possibility. This position is echoed in the first Critique, where Kant appeals to how the moral argument establishes God as the ground or source of possibility in order to distinguish it from the physico-theological argument, which “could at most establish a highest architect of the world, who would always be limited by the suitability of the material on which he works” (KrV, A 627/B 655; compare AA 28: 1093; 1104). Thus, in multiple instances, Kant stresses the conception of God as a ground of essences when he contrasts the infinite God of the moral argument with the shrunken God of the physico-theological argument.

It might seem surprising that Kant would place such a metaphysically-loaded conception of God at the heart of his moral theology. In fact, this is what we should expect. For Kant holds that moral theology provides support for believing in the concept of God supplied by transcendental theology – including that of God as a ground of possibility. In the first Critique’s Ideal, Kant claims that moral theology establishes the objective reality of the concept of God of transcendental philosophy, which includes God’s characterization as the most real ground of all possibility (KrV, A 641/B 669; see also A 588f/B 616 f). Kant reiterates this point in the second Critique, where he argues that the “transcendental ideal” of “speculative reason” is that very concept which receives “significance” as the ground of the highest good (KpV, AA 05: 133). In the Vigilantius lecture notes, Kant is recorded as saying that his pre-Critical ‘possibility proof’, in which Kant argues that God is the ground of all possibility or all essences, leads to faith in “a Godhood that must support our practical reason” (V-MS-Vigil, AA 27: 718, compare 724; V-Phil-Th/Pölitz, AA 28: 1034 f). Kant posits “a being which is the ground of everything possible” in order to explain “what in general the possibility of something consists in” (V-Phil-Th/Pölitz, AA 28: 1034). Commentators have previously noted Kant’s position that there is a moral use for the concept of God supplied by transcendental theology (Walsh 1997, 170; Longuenesse 2005, 235). But they have not explained why such a demanding notion of God, extending to the characterization of God as ground of possibility (among other attributes; see KrV, A 580/B 608), has a moral use.

Why would Kant’s moral arguments require conceiving of God as a ground of possibility? Kant is not explicit on this point. What I will now provide is a way of filling in the blanks on Kant’s behalf, but without bending the letter or the spirit of Kant’s texts. First, I will state the core philosophical thought, then I will explain why I take this move to dovetail with a kind of inference that Kant elsewhere makes at the intersection of theology, morality, and modality.

The thought works as follows: if the creator set the laws of nature by choosing from a pre-existing set of essences to actualize, the creator’s choices would be constrained by what essences there are. Whether there would be a choice that leads to a necessary proportionality of happiness and virtue would not be under the creator’s control. But then reason could not demand the realizability of the highest good unconditionally, for whether that demand is realizable would depend on possibility-space happening to be constituted in a way amenable to the realization of that demand. The ground of necessary nomological connection between happiness and virtue has to be the ground of possibility-space in order to exclude contingency from the realizability of the highest good. Put otherwise: the ‘ought’ of practical reason requires not only that some effect come about, but also that this effect comes about in a controlled or secured way.

The overall thought at work in this step is that, insofar as practical reason makes an unconditional demand, it cannot be contingent whether that demand is realizable. Controlled or secure world-making is world-making by a being that grounds essences, and hence is not constrained by the configuration of possibility-space, but rather has agency with respect to possibility-space itself. Kant places stress on the unconditional character of the moral ‘ought’ to explain how the moral arguments meet Hume’s challenge. For instance, in the passage from the Canon cited above, Kant claims that because “moral order” is “grounded in the essence of freedom” and not “contingently founded through external commands,” it must be grounded in something that is “inseparably connected a priori to the inner possibility of things” (KrV, A 816/B 844).

But how can Kant make the move from the unconditional or necessary character of a moral command to the kind of necessity that Kant associates with a ground of possibility (R 6292, Refl, AA 18: 560 f; BDG, AA 02: 82)? For Kant’s talk of the “unconditional” character of moral demands concerns how the moral law commands the performance of action independently of sensible conditions (GMS, AA 04: 462 f). It might seem that Kant is equivocating between a moral and a metaphysical sense of necessity. But rather than ascribe such an error to Kant, I think we can instead supply him with a thought linking the two notions: that if an unconditional moral command requires happiness being proportionate to virtue, then it cannot be contingent whether that proportionality is realized, at pain of undercutting the unconditionality of the original command. Moral unconditionality (qua independence of sensible conditions) means something different from metaphysical unconditionality (qua ground of possibility). But when it comes to the demand that happiness is proportioned to virtue, the former presupposes the latter. This is because practical reason requires that effects come about in a controlled or non-contingent way. This general requirement of practical reason is applied to the task of world-making. Controlled world-making is world-making by a being that grounds possibilities (and hence has agency with respect to them) rather than choosing among pre-given possibilities (and hence being constrained by them).

Kant elsewhere links moral unconditionality to metaphysical unconditionality or necessity. It lies beyond the scope of this paper to provide a complete defense of all such moves. But I wish to use a discussion of two parallel cases to suggest that the kind of modal move I have ascribed to Kant is consistent with his thinking at the intersection of theology and morality. First, consider Kant’s account in the third Critique of the final end [Endzweck] of creation. Kant cites the “good will” as “that alone by means of which [the human being’s] existence can have an absolute value” (KU, AA 05: 442 f), and uses this point to support the claim that the “existence of rational beings under moral laws” is that which “can alone be conceived of as the final end of the existence of the world” (KU, AA 05: 449–450). Kant takes himself to have thereby identified the “final end of creation … such as the highest reason would require for its creation,” and answered the question: “Why must human beings exist?” (KU, AA 05: 436n; compare R 6082, Refl, AA 18: 444, est. 1783–4; R 6092, Refl, AA 18: 448 f, est. 1783–84). Here Kant moves between two accounts of the unconditioned: that whose value does not require any further condition obtaining (see e. g. GMS, AA 04: 393), and that whose existence does not require any further condition obtaining. The two are not equivalent: one might think that Kant here makes the mistake of conflating value-theoretic unconditionality with something stronger, in this case, a sufficient reason to bring something into being, which is unconditional insofar as it cannot be outweighed by other considerations. But the point here is better understood as suggesting that, for Kant, value-theoretic unconditionality implies something stronger; moreover, that the inference from a value-theoretic sense of the unconditioned to a modally stronger account of the unconditioned is precisely the move that takes us from moral philosophy to moral theology.

Second, Kant rehearses a parallel move in the first Critique’s Ideal, holding that the necessity of moral laws explains why reason construes God’s own existence as necessary. At the end of the Ideal, Kant appeals to how “there are practical laws that are absolutely necessary (the moral laws)” in order to argue that, “if these necessarily presuppose any existence as the condition of the possibility of their binding force, this existence has to be postulated” (KrV, A 633f/B 661 f). Here Kant makes a prima facie surprising move from a value-theoretic to a metaphysical notion of necessity. This is prima facie surprising because the fact that an assent is based on an “absolutely necessary” law might be read as implying nothing about the modal status of any of the objects that figure in that judgment (see e. g. KpV, AA 05: 143). But Kant goes further in this passage from the Ideal, holding that assents based on theoretical considerations (“what exists”) are only “relatively necessary,” and “therefore” cannot service as the basis for cognition of the “absolute necessity of a thing” (KrV, A 634/B 662). Kant presents this as a contrast with assents based on “absolutely necessary” moral laws, which suggests that he aligns whether a law commands with absolute necessity, with the absolute necessity of the being to whose existence one assents. So Kant here presents a move from the modal status of a command to the modal status of its object.

These two examples show how, for Kant, the move from moral philosophy to moral theology can involve a move from the practically unconditioned to a metaphysically stronger notion of the unconditioned. Ascribing a version of this move to Kant in his moral arguments does not require rehabilitating the notion of absolute necessity, as that whose existence is contained in its concept, which Kant calls an “abyss” for human reason (KrV, A 613/B 641).^[19] For Kant explicitly differentiates this definition of necessity from the necessity that consists in being the ground of all possibility, which Kant associates with moral theology (R 6292, Refl, AA 18: 560 f; compare BDG, AA 02: 82). The overall point here is that moving from the unconditional necessity of a moral command (such as the requirement that we realize the highest good), to a stronger form of necessity (in this case, a ground of possibility-space) is a kind of move that is characteristic of Kant’s practically-grounded theological metaphysics. Moreover, there is a plausible thought at the core of this move: since the proportionality of happiness and virtue is required for the highest good to be achievable, it should not be a contingent matter whether that proportionality obtains, and world-making excludes such contingency when it is undertaken by a being that grounds possibility-space rather than being confronted with pre-given possibilities.

This completes the third step of the grounding interpretation: we can explain Kant’s view that his moral arguments require a ground of possibility by noting that only a ground of possibility-space, and not a mere actualiser of essences in a ‘pre-existing’ or independent possibility-space, is a secure world-maker who excludes contingency in the realizability of the highest good. The fourth and final step shows how this move enables us to explain Kant’s ascription of the traditional omni-properties to God.

Step 4: A ground of all things is omni-ɸ with respect to ɸ for all ɸ (where ɸ indicates any real property)

This final step establishes that a ground of all possibility-space has all the traditional omni-properties, and hence is God. It consists in getting clear on what, exactly, is involved in having an omni-property. Large finite properties are not omni-properties. But Kant also holds that omni-properties are not mathematically infinite, and explicitly rejects the use of “mathematical infinity” in theology (V-Phil-Th/Pölitz, AA 28: 1018).

Kant proposes that “all-sufficiency [Allgenugsamkeit] (omnisufficientia),” which “expresses the relationship of a ground to its consequences,” is a more apt term for providing a positive account of the divine properties (V-Phil-Th/Pölitz, AA 28: 1019).^[20] Kant echoes this usage in the third Critique, where he holds that moral theology’s “determinate concept” of the original being, compared with the non-determinate physico-theological concept (KU, AA 05: 480), is the concept of an “all-sufficient [allgenugsam] (and for that very reason unique, properly so called, highest) being” (KU, AA 05: 477). Kant contrasts a demi-god figure having a “merely great understanding, might, etc” with a being that has “omniscience, omnipotence, in a word, [that we conceive of] as containing the sufficient ground for such properties in all possible things” (KU, AA 05: 462). Kant spells out what it means to have an omni-property with reference to grounding that property in all possible things. This suggests the following schema for omni-properties: for any given ɸ, to be omni-ɸ is to be the ground of all possible ɸ-ness. For instance, Kant explains divine goodness as follows: “[God] is benevolent and infinite [gütig und zwar unendlich], because out of him all possible goodness flows” (R 6099, Refl, AA 18: 451–453, est. 1783–84).^[21]

Kant’s account of all-sufficiency links the ground of possibility and the omni-properties. The previous step established that I must have faith in a being that is the ground of all possibility. All possibility is all possible properties, so this being grounds all possible properties. On Kant’s view, this is what it means to be a being with all the omni-properties. Thus, if I have faith that the highest good is achievable, I have faith in a being that has all the omni-properties.^[22]

This completes the grounding interpretation, which provides a reading of Kant’s moral arguments that explains how Kant’s moral arguments overcome Hume’s challenge.

3 The practical status of the grounding interpretation

The grounding interpretation explains how the moral arguments secure faith in a being with the traditional omni-properties. It does so by supplying a practical rationale, in relation to the highest good, for faith in a being that is the ground of all possibility. However, one might worry that the grounding interpretation is objectionably ‘theoretical’ rather than genuinely ‘practical.’ After all, Kant presents his moral arguments as an alternative to the proofs of traditional speculative metaphysics (KrV, A 640–642/B 668–670). Responding to this objection requires getting clear on what it means for an argument to be ‘practical.’ In this section, I present two senses in which an argument can be practical: with respect to justification, and with respect to the content of its representation of the unconditioned. I argue that the grounding interpretation is practical by both criteria.

Kant’s own rationale for categorizing his moral arguments as practical appeals to the justificatory role played by practical reason’s ends. In the third Critique, Kant argues that faith is distinctive because we are justified in assuming the objects of practical faith (i. e. God and immortality) not due to evidence, but rather due to their relation to an “end” that is “commanded by pure practical reason”, namely, the achievement of the highest good (KU, AA 05: 469; see also KrV, A 828/B 856, KpV, AA 05: 132–136). The grounding interpretation meets this criterion: we are justified in having faith in God as a condition on the possibility of the achievability of the highest good. To be sure, explaining why the achievability of that end requires faith in God, in turn appeals to premises whose content is not explicitly practical in nature. But so does any reconstruction of Kant’s moral arguments. For instance, the moral argument of the second Critique includes the premise that a necessary connection is either analytic or synthetic (KpV, AA 05: 112–113), and this is a theoretical judgment.

However, not all theoretical judgments can be premises of a practical argument. The premises of practical arguments exclude theoretical judgments justified by the ends of theoretical reason. Kant introduces “doctrinal belief” as a kind of non-evidential assent justified by the explanatory ends of theoretical reason (KrV, A 825–827/B 853–855). He also holds that doctrinal belief is “unstable,” in contrast with moral belief (KrV, A 827–828/B 855–856). An argument for moral belief should not include doctrinal belief as a premise, lest such instability be transmitted to moral belief. Furthermore, since Kant holds that our faith in God is practical because its justification appeals to the end of pure practical reason, our argument for faith in God should not appeal to the ends of both theoretical and pure practical reason, on pain of undermining the basis for considering that argument practical.

One might worry that the grounding interpretation’s appeal to a ground of possibility violates this restriction by including doctrinal belief as a premise. Some scholars have argued that in the Critical period, Kant continues to endorse a version of his pre-critical ‘possibility proof’ as an article of doctrinal belief or regulative principle, justified with reference to the explanatory aims of theoretical reason.^[23] For instance, one might assent to the ground of possibility in order to explain “what in general the possibility of something consists in” (V-Phil-Th/Pölitz, AA 28: 1034). But the grounding interpretation does not justify assent to a ground of possibility by appealing to theoretical reason’s ends. Rather, it shows that a ground of possibility is implicit in pure practical reason’s demand for realizability of the highest good. In particular, to have faith in a ground of possibility is to have faith that the highest good is realizable in a secure or controlled way. In performing such an explication, one spells out what pure practical reason’s demand is a demand for.^[24] But such explication is not justified as a means to a theoretical end.^[25] Insofar as being ‘practical’ is understood as a matter of justification, the grounding interpretation is practical: it begins with a premise justified by an end of pure practical reason (the realizability of the highest good), and excludes any premise justified by an end of theoretical reason.

A different way of questioning whether the grounding interpretation is ‘practical’ or rather objectionably ‘theoretical’ concerns the content of its representation of the unconditioned. For the grounding interpretation argues that a ground of possibility is implicit in practical reason’s demand for the realizability of the highest good. However, this might seem to conflate the theoretically unconditioned and the practically unconditioned. In the first Critique, Kant characterizes the ens realissimum as the ground of all possibility, or “the supreme condition of the possibility of everything that can be thought (the being of all beings)” (KrV, A 334/B 391). Although Kant himself states that God of the moral arguments is “all-sufficient” (e. g. KU, AA 05: 477) and hence a ground of possibility (e. g. KU, AA 05: 462), this creates a puzzle. For it might seem that all-sufficiency is a representation of the unconditioned that only theoretical reason forms, and hence inappropriate to a practical argument. Why would practical reason have or form an idea of all-sufficiency?

The key point is that, for Kant, all-sufficiency is not only a theoretical idea. All-sufficiency [Allgenugsamkeit] also means the “will of a self-sufficient being” (V-Phil-Th/Pölitz, 28: 1060). All-sufficiency is a concept of unconditioned agency, or agency not subject to conditions, where ‘conditions’ introduce dependency. This is more demanding than an account of a will that is autonomous or practically unconditioned insofar as it is not determined by inclination (GMS, AA 04: 463, KpV, AA 05: 33). Patrick Kain (2021, 310–312) reconstructs this account of all-sufficiency as part of Kant’s theory of divine freedom. There is a close link between being all-sufficient in this practical sense (for this I will stipulatively use the term ‘practical all-sufficiency’), and a ground of possibility, which explains Kant’s use of the same term (allgenugsam/Allgenugsamkeit) for both. Since a practically all-sufficient being is absolutely independent, it neither desires anything other than itself, nor represents anything other than itself in acting (V-Phil-Th/Pölitz, AA 28: 1060–1062). Rather, its creation of a world can only occur through self-cognition. This point, in turn, means that a practically all-sufficient creator is a ground of possibility, so that “in cognizing himself, [God] cognizes everything possible which is contained in him as its ground” (V-Phil-Th/Pölitz, AA 28: 1061). Creating the world in a practically all-sufficient or independent manner requires that God not cognize possibilities external to himself, but rather be a ground of possibility.

All-sufficiency is a practical idea, insofar as it is an idea of unconditioned agency, and an all-sufficient world-maker is a ground of possibility. This point clarifies the link between the moral arguments and the ens realissimum as a ground of possibility. God, as the “ground of the practically necessary connection” between virtue and happiness (KrV, A 810/B 838; cf KpV, AA 05: 125, KU, AA 05: 457), establishes a nomological connection by making a world actual. As Step 3 of the grounding interpretation argued, we represent this ground in a way that excludes contingency from its capacity to bring about that connection. For if God’s role is to secure “the accession of nature to a condition that does not stand within our own power” (KU, AA 05: 455), then if the conditions of world-making failed to stand within God’s own power, the original problem motivating the introduction of God would recur within the account of divine creation. World-making is practically conditioned or dependent when it operates on pre-given material or possibilities, and practically unconditioned or independent when it is undertaken by a being that grounds those possibilities. Both an “architect” of the world who arranges pre-given materials, and the creator who chooses from a set of pre-given essences, have conditions on their world-making. A practically all-sufficient agent, as a ground of possibility, does not.

This last point explains why Kant links the moral argument’s God to the first Critique’s ground of possibility. In the Critical Resolution to the Antinomy of Practical Reason, Kant claims that the highest good “furnishes significance [Bedeutung]” to the first Critique’s “transcendental ideal, the theological concept of an original being,” via the account of God as the “supreme principle of the highest good in an intelligible world” (KpV, AA 05: 133). Moral faith, whose content falls short of a ground of possibility would not “furnish significance” to the very same idea of theoretical reason. But practical reason’s postulate nonetheless has its source in a practical representation of the unconditioned. Pure practical reason has an idea of unconditioned agency that entails grounding possibility when applied to the task of world-making that is part of the highest good.

This concludes this section’s defense of the grounding interpretation as a practical rather than theoretical argument for faith in God’s existence. With respect to justification, the grounding interpretation appeals to pure practical reason’s end of realizing the highest good and does not appeal to theoretical reason’s interest in explanation. With respect to content, the grounding interpretation does not supplant practical reason’s representation of the unconditioned with a theoretical representation of the unconditioned but rather shows how the latter is entailed by the idea of the world as the product of unconditioned agency.^[26]

Conclusion

Kant held that his moral arguments for God’s existence justified faith in a being with the traditional omni-properties. Scholars have struggled to make sense of how Kant’s moral arguments secure that faith-content. Both the ‘supplemental strategy’ and the ‘pluralistic strategy’ are revisionary solutions proposed on Kant’s behalf, which weaken the content secured by the moral argument. This paper has proposed a non-revisionary reading of Kant’s moral arguments that explains how Kant’s moral arguments justify faith in a being with the traditional omni-properties. The grounding interpretation starts with pure practical reason’s demand for the non-contingent realizability of the highest good, which introduces the idea of the world as the product of a practically unconditioned agent with complete control over its world-making, which is a ground of possibility. Theoretical reason’s idea of a ground of possibility acquires a practical basis as part of the idea of an unconditioned agency that creates the world. In substantiating Kant’s claims that his moral arguments yield the traditional omni-properties, we do not pull him back into a mode of metaphysical reasoning that he disavowed. Instead, we see how what we had previously understood as a purely theoretical notion, a ground of possibility, gains its meaning and justification as part of a practical notion of unconditioned agency. Kant’s moral arguments show that within the traditional conception of God in theological metaphysics, there is a genuinely practical core: the idea of absolute power.^[27]