1. Kant’s Standpoint Talk in the Transcendental Aesthetic and in Groundwork III

Before discussing Groundwork III, I want to consider one further place in Kant’s writings where the notion of a standpoint makes a prominent appearance: the Transcendental Aesthetic, where Kant declares that ‘it is only from a human standpoint that we can speak of space, extended beings, etc.’ (A26/B42). He makes analogous points about temporal ‘things’ (A33–7/B49–53). Like space, ‘time is … a purely subjective condition of our (human) intuition’ (A35/B51). The human standpoint is thus introduced by reference to ‘the subjective condition under which alone we can have … intuition’ (A26/B42) of objects.

Some commentators hold that we cannot transcend this human standpoint. For Henry Allison, ‘the “human standpoint” … is the only standpoint available to us’ (Allison Reference Allison2004: 132).Footnote ¹ For Béatrice Longuenesse, the human standpoint is our one and only perspective of discursive representation, which requires the interplay of spontaneous conceptual capacities and passively received intuitions, and which contrasts with the perspectives of both animals and God (Longuenesse Reference Longuenesse2005: 2–3). However, I want to suggest that Kant’s notion of a human standpoint is significantly more specific than the notion of a discursive (non-animal, non-divine) perspective. Immediately after introducing the notion of a human standpoint, Kant invokes the possibility of ‘other thinking beings’, which might not be ‘bound by the same conditions as those which limit our intuition’ (A26/B42). For Kant the notion of thought, or of an understanding that conceptualizes sensible intuitions, entails discursivity (Log, 9: 58, 91; A51/B75, A68/B92–3, B145).Footnote ² A non-discursive, divine intellect would not ‘think’ or use concepts; it would intuit things via a non-sensible (non-receptive) mode of intuition (KU, 5: 406). Thus, the conceivability of ‘other thinking beings’ that might not be bound by the conditions of ‘our’ intuition shows that the notion of a human standpoint goes beyond the notion of a discursive perspective: there might be discursive cognizers who represent the world via sensible but non-spatiotemporal intuitions (for this logical possibility, see also B148, 150).

As Kant’s further characterization makes clear, the human standpoint combines the receptivity of spatiotemporal intuition with the spontaneity of judgement. From the human standpoint, we ‘speak of’ extended beings, ‘attribute’ the ‘predicates’ of spatiality (and temporality) to things, and make judgements such as: ‘All things are next to one another in space’ (A26/B42). Thus the human standpoint concerns a distinctive species of experience,Footnote ³ i.e. a kind of ‘empirical cognition’ (A93/B125–6) that represents things as parts of a single spatiotemporal framework, via judgements that have objective validity for all human cognizers.

After characterizing the way in which we judge things from the human standpoint, Kant suggests that things may also be ‘considered in themselves through reason, that is, without regard to the constitution of our sensibility’, when we ‘remove the condition of the possibility of all experience’ (A28/B44). We can consider things apart from the conditions of spatiotemporal experience because we possess two types of representations that involve no reference to space and time: first, the pure categories, concepts of substance or causality, which ‘are not … grounded in sensibility, like the forms of intuition, space and time’ (B305); second, ‘ideas’ that ‘are imposed by the very nature of reason itself’ (A327/B384), such as the idea of freedom. These pure concepts or ideas do not give us theoretical cognition of objects: for such cognition, we strictly depend on our human sensibility. But the conditions of theoretical cognition or experience do not delimit the bounds of all meaningful thought.Footnote ⁴

This suggests that the human standpoint is not the only perspective that is available for human beings.Footnote ⁵ We can also ‘consider’ things through representations that abstract from our human forms of sensibility because they derive from purely intellectual faculties such as the understanding or reason. These are faculties that we share with other (conceivable) thinking or rational beings that lack human forms of sense. In the Aesthetic, Kant does not give any reason why we would consider things in this way, and he does not designate this non-anthropocentric way of considering things as an independent standpoint. His notion of a ‘standpoint’ may nonetheless apply here; I discuss this issue at the end of this section.

Although the two passages are not usually discussed together, Kant’s standpoint distinction in Groundwork III is closely connected to Kant’s standpoint talk in the Aesthetic. In Groundwork III (4: 451), Kant distinguishes between two ways in which we can judge things. Insofar as our judgements are subject to the conditions of our sensibility, they give us cognition of a world of appearances that ‘may be very different according to the difference of the sensibility in various observers’. This parallels the Aesthetic’s appeal to ‘other thinking beings’ whose judgements about sensible appearances do not conform to the conditions of our human sensibility. But not all our representations depend upon our human sensibility: we can also conceive objects ‘without regard to the constitution of our sensibility’, as non-spatiotemporal things in themselves. This conception concerns an ‘intelligible world’ ‘that remains always the same’, i.e. that would also be represented by other thinking beings if they abstracted from their specific, non-human mode of sensible intuition. This parallels the Aesthetic’s appeal to a non-anthropocentric way of ‘considering’ things through reason.

In the Groundwork, Kant makes explicit that he seeks to distinguish between two stances that we can adopt not just towards ‘things’ or ‘beings’, but towards a specific class of beings, namely, ‘ourselves’ (4: 450) as ‘rational beings’:

According to this passage, each of the two standpoints affords rational beings a conception of their activity in accordance with distinctive laws. Thus, to further understand the character of the two standpoints, we need to understand the character of these two types of laws.

Clearly, the standpoint that regards rational beings as part of the sensible world invokes empirical, causal laws of nature. But which laws figure in the standpoint that regards rational beings as part of the intelligible world? The final sentence of the above passage states that these laws are ‘independent of nature’, not empirical but ‘founded in reason alone’. Kant typically contrasts empirical ‘laws of nature … according to which everything happens’ and normative ‘laws according to which everything ought to happen’ (GMS, 4: 388). This does not entail that normative laws are the only laws governing the non-natural, intelligible world. But Kant characterizes the intelligible standpoint as one from which a rational being ‘finds itself’ subject to non-empirical laws and ‘cognizes’ such laws for the exercise of its faculties. In Kant’s system, normative laws are, arguably, the only positive laws that are left for us to cognize when we consult nothing but pure reason. Thus Kant’s intelligible standpoint as characterized in Groundwork III is a normative standpoint from which rational beings consider how they should act.

How does this normative standpoint relate to the non-sensible way of ‘considering’ things in themselves that Kant introduces in the Aesthetic? Here we must first note that considering things in abstraction from human sensibility is not the same as considering things via normative judgements: we can also make descriptive judgements about non-sensible beings, such as the existential judgement that there are things in themselves (B308–9; Prol, 4: 315), or negative judgements such as ‘Things in themselves are not spatially extended’ (B149). These kinds of judgements are not yet sufficient for the normative-intelligible standpoint that Kant seeks to capture in Groundwork III, namely, a non-sensible perspective that rational beings can adopt toward their faculties and actions. Judgements such as ‘Things in themselves are not in space and time’ apply to all non-sensible beings including those that lack norm-governed rational faculties, such as the things in themselves that may underlie appearances of stones, chairs and donkeys and to which we do not attribute reason or understanding (A546/B574).

This leaves us with two interpretative options. On a first interpretation, Kant wants to posit two different intelligible standpoints: (1) a purely descriptive standpoint from which we can consider all things in themselves, including non-rational beings; (2) a more specific intelligible standpoint of normativity that we adopt if we consider specifically rational noumenal beings. On a second interpretation, (2) is the only intelligible standpoint that Kant posits, whereas (1) does not qualify as a standpoint at all. This second interpretation depends upon two assumptions: first, the very notion of adopting a standpoint entails that one can ‘consider’ the nature of things via positive, informative judgements; second, we can obtain positive characterizations of things in themselves (which go beyond the above-mentioned existential or negative judgements) only when we attend to the rational, norm-governed faculties that we can attribute only to a sub-class of things in themselves. I admit to being partial to the second interpretation, in part because Kant explicitly uses the notion of an intelligible standpoint only to designate our normative stance towards rational beings. However, I have no proof that taking up a standpoint for Kant strictly entails that we can make positive judgements about the nature of things. Moreover, one might deny that the only positive judgements we can (legitimately) make about things in themselves are normative judgements.Footnote ⁶

Given these complications, I want to leave it open whether the normative standpoint on rational beings that Kant posits in Groundwork III is the only standpoint that we can adopt towards things in themselves, or whether this normative standpoint is a specific application of a broader intelligible standpoint on the entire noumenal world. These two readings have significant overlap. On both readings, Kant’s standpoint distinction in Groundwork III is strongly indebted to his discussion in the Aesthetic: the anthropocentric human standpoint delineated in the Aesthetic is brought up again in Groundwork III as the standpoint from which we regard ourselves as part of the spatiotemporal world of (human) sense; and the possibility of a more than human way of considering things in themselves that is suggested in the Aesthetic is, at least, a necessary condition for the intelligible-normative standpoint delineated in Groundwork III. Moreover, proponents of both readings can agree that the contrast between a normative-intelligible and a non-normative, sensible standpoint plays an especially significant role in Kant’s philosophy. I now examine in greater detail what this contrast amounts to.

2. The Normative and the Empirical

On the most common interpretation, the standpoint that invokes empirical laws of nature is a theoretical standpoint that Kant contrasts with a practical standpoint from which we determine our free will under moral laws: Kant seeks to distinguish between ‘the point of view of empirical causal explanation; and the point of view of freedom and morality’ (Allais Reference Allais2015: 305);Footnote ⁷ between the ‘explanatory standpoint of theoretical reason’ and the ‘normative standpoint of practical reason’ (Korsgaard Reference Korsgaard1996: 173); between ‘the theoretical point of view, where the concern is with explanation rather than action’ and the ‘practical point of view’ that concerns ‘ourselves as accountable moral agents’ (Allison Reference Allison2004: 48); between the ‘naturalistic and practical standpoints’, which yield ‘two accounts of action’, where ‘the first, theoretical account … consider[s] acts as natural events and … aim[s] to explain their occurrence … the second, practical account … consider[s] acts as expressing certain determinations of the will, and moral action as expressing certain sorts of determination of the will’ (O’Neill Reference O’Neill1989: 61, 67; cf. Frierson 2011: 84, Hill Reference Hill1992: 135–40).

This interpretation assumes that the normative laws that we cognize from the intelligible standpoint coincide with the practical laws of morality that govern our faculty of choice (the will). There are passages that support this assumption (see GMS, 4: 387–8). However, when Kant says that a rational being can cognize non-empirical laws ‘for the exercise of its faculties’, he refers to ‘faculties’ in the plural (4: 452); this suggests that he is concerned with more than just practical reason or the will here. This suggestion can be confirmed as follows. Kant states that, from the non-empirical standpoint, a rational being regards itself ‘as an intelligence, hence not with respect to its lower faculties’ (4: 452). Our ‘lower faculties’ characterize us as receptive, passive beings. No normative prescriptions can be given for the use of those faculties: ‘The lower faculties cannot be instructed, because they are blind’ (V-MO/Collins, 27: 244). The active, spontaneous ‘higher’ faculties that we possess ‘as intelligences’ include more than just practical reason or the will: ‘The higher faculties of cognition are: understanding, power of judgment, and reason’ (A130–1/B169–70). Accordingly, Kant claims that a rational being ‘must regard itself as belonging to the intelligible world … in respect of whatever there may be of pure activity in it’, and he adds that the faculties that are capable of purely spontaneous activity are the understanding and reason (4: 452), including theoretical reason that directs (‘determines’) the understanding (A546–7/B574–5).Footnote ⁸ Since the spontaneous ‘higher faculties’ that rational beings ascribe to themselves ‘as intelligences’ include capacities for theoretical judgement, the normative laws that intelligences can cognize for their actions include not only practical laws for the exercise of the will but also theoretical laws for the exercise of the capacity to judge, i.e. for cognitive acts of judgement. Judgements result from ‘actions’ of the understanding (A294–5/B350–1).

To be sure, Kant pronounces the contrast between the empirical and the normative, between what is and what ought to be, most emphatically in his practical philosophy: practical reason ‘declares actions to be necessary, even although they have never taken place, and perhaps never will take place’ in the empirical world of appearances (A548/B576). His conviction that moral norms have the highest normative priority for us perhaps explains why he tends to equate the contrast between the empirical and the normative with the contrast between empirical and moral laws. However, sometimes Kant makes explicit that the empirical-normative contrast can also be drawn with regard to acts of judgement. This is a constant theme in his introductions to the third Critique where he discusses the status of principles of reflection:

Kant characterizes the laws of logic in the same terms. These laws cannot be derived from psychological observations of how the understanding actually works (Log, 9: 13–14) because: ‘In logic, we do not want to know: how the understanding is and thinks and how it has thus far proceeded in thinking, but how it ought to proceed in thinking’ (Log, 9: 14).Footnote ⁹

Thus, against the standard interpretation, I contend that the non-empirical perspective from which we ‘find ourselves’ under normative laws is not just a standpoint of practical reason but a broader normative standpoint that also involves theoretical reason.

However, here one might object that the issue of normativity is so pervasive in Kant’s conception of human reason that we cannot single out just one rational perspective as the normative standpoint. When we engage in empirical thought, we exercise our higher, norm-governed rational faculties. Accordingly, the conclusion of a piece of reasoning about the empirical world can be put as: ‘I ought to believe that p.’ Thus it may seem that there is no informative contrast between empirical and normative thought. If so, the standard practical-theoretical contrast remains as the only viable alternative to conceptualize Kant’s distinction.

In response to this worry, I want to take a clue from Kant’s suggestion that the difference between the two standpoints rests on two different ways in which we can ‘regard’ a rational being (GMS, 4: 452). From the empirical standpoint, one regards a rational being ‘as nothing more than a part of the world of sense’: one views its behaviour ‘like all the other appearances’, as ‘the inevitable outcome of nature’ (A540/B568). Kant stresses that this standpoint precludes normative judgements: ‘ought has no meaning when one has only the course of nature in view’ (A547/B575).Footnote ¹⁰ This claim should be taken literally, as relating to what one has in view (vor Augen hat) from the empirical standpoint: a causally determined natural being and a purely descriptive subject matter.

Thus my interpretation can happily grant the important insight that normativity, for Kant, infuses all our conscious thinking including our judgements about the empirical world. If I am engaged in empirical thought, I seek to figure out what I ought to believe, but this is only my self-conception as the norm-governed subject of empirical thought. I do not represent the intentional objects of my empirical thought (what I ‘have in view’) as standing under normative principles. For Kant, empirical thought is non-normative in the striking sense that one cannot intelligibly raise normative questions about its purely natural object: as long as we focus only on empirical phenomena and natural causes, ‘we cannot ask what ought to happen in nature’ (A547/B575).Footnote ¹¹ The empirical perspective on rational beings seeks to answer descriptive, i.e. observational, explanatory or predictive questions without considering how these beings ought to act. By contrast, from the non-empirical standpoint one reasons about norm-governed agents and considers how they, rationally speaking, must act irrespectively of how they do, did or will act in the natural order of things. ‘The ought expresses a kind of necessity … which is found nowhere else in the whole of nature. The understanding can cognize in nature only what is, what has been, or what will be’ (A547/B575).

However, here another difficulty arises. If the distinctive feature of reasoning from the normative standpoint is that it cannot settle its questions by appeal to what is, was or will be the case in the natural order of things, then, one might argue, only practical reasoning satisfies this condition: theoretical reasoning always aims at judgements about the natural order, and thus its questions must be settled by deferring to descriptive empirical facts. If that is correct, then the only genuinely, irreducibly normative standpoint is that of practical reason. This may be the root of the traditional view that Kant’s contrast is between the ‘explanatory standpoint of theoretical reason’ and the ‘normative standpoint of practical reason’ (Korsgaard Reference Korsgaard1996: 173).

Here my response is that Kant’s system includes judgements of theoretical reason that do not defer to representations of descriptive empirical facts. Consider the regulative principles of reflective judgement or theoretical reason, such as the principle that ‘prescribes that we ought to study nature as if systematic and purposive unity, combined with the greatest possible manifoldness, were everywhere to be met with, in infinitum’ (A701/B729). Kant stresses that this principle is not based on the cognition of any descriptive fact; in particular, it is not obtained from the empirical standpoint ‘through observation of the accidental constitution of nature’ (A651/B679). Theoretical reason prescribes that (e.g.) biologists ought to presuppose a systematic unity of natural phenomena in their attempts to cognize the properties of a newly discovered animal species, even though they cannot know that such a unity obtains in nature as a matter of fact. More generally, via its normative ‘principles of judgement’ theoretical reason directs empirical cognizers to ask appropriate, fruitful questions about the natural world: reason here takes the role ‘of an appointed judge, who compels the witness to answer questions which he himself has formulated’; the ‘witness’ is ‘nature’, i.e. our empirical perceptions of the natural world (Bxiii). Hence, normative principles of theoretical reason have priority over, and are thus irreducible to, our observation of empirical facts.

Another case in point here is reason’s cosmological idea of a complete totality, which is, properly understood, a rule that prescribes ‘what we ought to do’ as cognizers of nature:

When we cognize some empirical condition that has explanatory power, we should always take it to have further conditions that yield an occasion for further enquiries. This principle derives from pure reason’s commitment to explanatory completeness: theoretical reason propels us to strive for a normative ideal of consummate empirical understanding that always exceeds our finite experience of nature.

Hence, judgements that prescribe how we ought to exercise our rational capacities, which are irreducible to descriptive judgements about empirical facts, are not the sole privilege of practical reason. The popular view that for Kant theoretical reason is concerned only with what is rather than with what ought to be is thus incorrect (or at least requires qualification).Footnote ¹² This vindicates my suggestion that Kant’s standpoint distinction is between a normative and an empirical perspective rather than between a practical and a theoretical perspective.

3. More than Human Normativity versus Empirical Anthropocentricity

I have argued that Kant’s standpoint distinction revolves around two central contrasts. The first contrast is between (1a) representing rational beings in accordance with the conditions of human sensibility, through empirical judgements, and (1b) representing rational beings through purely intellectual representations. The second contrast is between (2a) empirical laws of nature and (2b) normative (but not merely practical) laws for the exercise of our higher faculties. To fully explicate Kant’s conception of the two standpoints, we must understand the connection between the corresponding pairs: (1a) and (2a); (1b) and (2b).

Consider first the connection between (1a) and (2a). For Kant, empirical judgements that locate rational beings in space and time always, however implicitly, involve the thought that such beings are subject to laws of nature. The domain of nature is identical to the domain of objects of experience that we can intuit through our spatiotemporal sensibility (MAN, 4: 467); it is a fundamental principle of Kant’s philosophy ‘that all events in the sensible world stand in thoroughgoing connection in accordance with unchangeable laws of nature’ (A536/B564). Now our human sensibility is not sufficient to explain the law-governedness that is essential to the empirical perspective: this law-governedness also has a purely intellectual source in the categories of the understanding. But the character of human sensibility is indispensably necessary for, and partly constitutive of, the law-governedness of the empirical world. This is clear in the case of time: the principle that appearances are subject to lawful generalities and causal necessities is essentially a rule of time-determination, i.e. a rule for ‘determining something with respect to its temporal sequence’ (A200/B246) which is only ‘valid … under conditions of temporal succession’ (A202/B248). The intellectual concept of causality requires a temporal schema, succession, for its empirical application to objects; without that schema the empirical standpoint cannot identify any causal relata (A144–7/B183–7, A243/B301). This does not explain why space is a necessary condition for the lawfulness of the empirical world. Here I can only indicate that spatial intuitions are required to determine the temporal properties that give objective empirical reality to the categories.Footnote ¹³ Since mere inner sense presents us with no abiding substratum of change, we can only represent an empirical world of permanent causal agents and patients by representing them as parts of one space.

Thus, combining (1a) and (2a) yields the notion of a stance that is both anthropocentric and empirical. This standpoint aims at empirical cognition, i.e. it seeks to answer observational, predictive and explanatory questions by appeal to causal laws of nature. In order to achieve such cognition, we must intuit things through our human sensibility, as beings that exist in one law-governed spatiotemporal framework. Or, conversely: in order to cognize the determinate spatiotemporal properties that things have insofar as they conform to our human forms of sensibility, we must represent things as part of one law-governed causal order of nature.

But why is Kant’s conception of a distinctively normative standpoint (2b) tied to the idea of a non-anthropocentric point of view (1b)? Let me first confirm that Kant indeed posits this link. He is explicit about the link in the case of moral normativity. A moral norm, such as

My argument that practical and theoretical reason share one normative perspective requires that there is a similar connection between normativity and non-anthropocentricity in the case of theoretical norms. Now there is an important disanalogy here: for Kant the moral law applies to all rational beings, including a divine holy will (GMS, 4: 414; KpV, 5: 32). By contrast, theoretical norms such as the prescription to trace every condition to some further condition have no bearing on a divine intellect because such an intellect has an immediate grasp of the whole of reality (B71–2; KU, 5: 401–10). Nonetheless, the epistemic norms that derive from the a priori ideas of pure theoretical reason apply to all finite cognizers, i.e. to human thinkers as well as ‘other thinking beings’. A priori ideas, being ‘imposed by the very nature of reason itself’ (A327/B384), must be independent of sensibility: since these ideas are ‘simply categories extended to the unconditioned’ (A409/B436), and since neither the pure categories nor the representation of the unconditioned involve a reference to space and time, these ideas stake their claim on any finite understanding. For instance, the above-mentioned epistemic norm that prescribes a regress ‘in the series of the conditions of given appearances’ (A508–9/B536–7) does not mention space and time. The bare notion of a given appearance – just like the bare notion of the unconditioned (or of complete composition, division, etc.; see A415/B443) – is not confined to the human standpoint. Every finite understanding ought to aim for ever greater completeness in the synthesis of what is given to the senses, regardless of the specific character of the forms that structure the reception of the sensible manifold. Even more obviously, every finite theoretical understanding is governed by the norms of logic.

Having established that for Kant the link between normativity and non-anthropocentricity is characteristic of both practical and theoretical normativity, I now want to consider why these two properties are linked together in a distinctive standpoint on rational beings. Here I want to suggest that the property of non-anthropocentricity is a necessary condition for two features that Kant deems essential to normative principles: their a priori necessity and their role as autonomous laws of freedom.

Concerning the first point, consider how Kant argues in the passage cited above (GMS, 4: 389): moral norms must be founded in a priori concepts of pure reason that are shared among all rational beings because otherwise these norms would depend upon empirical conditions. This would be fatal to the claim to necessity that, for Kant, is integral to normative laws: a rule ‘which rests on principles of mere experience’ and thus ‘on empirical grounds’ cannot be necessary law (4: 389). Kant uses essentially the same line of reasoning when he defends the necessity of theoretical norms. He argues that ‘if a judgment … asserts a claim to necessity, then … it would be absurd … to justify it by explaining the … judgment psychologically’ i.e. by appeal to an ‘empirical origin’ (EEKU, 20: 238). He then points out that the principles of reflective judgement do make a claim to normative necessity and hence must rest on a priori principles: they ‘do not state that everyone judges in this manner – if they did, they would be a task for the explanations of empirical psychology – but rather that one ought to judge in this manner, which is to say as much as: they have an a priori principle for themselves’ (EEKU, 20: 239). Thus, whether he is concerned with moral or epistemic norms, Kant presents the a priori necessity of the relevant rules as a non-negotiable desideratum, which he secures via the assumption that these rules are grounded in a priori concepts of rational faculties that we would share with other rational beings rather than in empirical (psychological, anthropological) conditions that would limit the validity of these rules to human beings.

This explanation of the link between (1b) and (2b) raises a complication. From the above passages, it seems that Kant argues directly from the a priori necessity of normative laws to their non-anthropocentricity. However, for Kant a priori necessity is not per se incompatible with anthropocentricity, as is clear from mathematical laws that expound the a priori conditions of human intuition. I believe that we can explain Kant’s direct inference from a priori normative necessity to non-anthropocentricity as follows. While the non-anthropocentric, purely intellectual character of a principle is not required for a priori necessity as such, it is a strictly necessary condition for a priori normative necessity. That is, whenever a principle purports to prescribe how one ought to act (or what actions would be good and right), then the principle makes a claim to a kind of a priori necessity which can only be grounded in purely intellectual (non-sensible) sources. This qualification may seem ad hoc: if a priori necessity is not as such incompatible with sensibility, why should there be an incompatibility when and only when the a priori necessity has a normative character? Because normative laws concern only the use of our higher faculties: such laws apply to a rational being ‘as an intelligence, hence not with respect to its lower faculties’ of sensibility (GMS, 4: 452). Normative rules are essentially rules for instructing the intentional activities of rational beings, whereas ‘the lower faculties cannot be instructed, because they are blind’ (V-MO/Collins, 27: 244). Because the a priori conditions of our human sensibility that are examined in mathematical laws constrain the exercise of lower, passive or receptive faculties, Kant does not consider these a priori laws of human sensibility when he infers directly from the a priori necessity of normative rules to their non-anthropocentricity.Footnote ¹⁴

The fact that normative laws are rules for the use of higher, active intellectual faculties brings us to the second central idea that sustains the link between (1a) and (2a). For Kant, normativity is inseparably tied to the idea of freedom as autonomy: normative laws are not imposed on us by ‘foreign causes’ but spring from the spontaneous activity of our higher faculties. Kant famously holds that moral normativity rests upon the autonomy of the rational will (GMS, 4: 448; KpV, 5: 132), but the idea of autonomy has a wider scope: we can view ourselves ‘in respect of the faculties of the soul generally, insofar as they are regarded as higher faculties, i.e. as faculties containing an autonomy’ (KU, 5: 196). Theoretical reason or the power of judgement, as ‘higher faculties’, are autonomous (or ‘heautonomous’) sources of norms that prescribe how we ought to reflect upon nature. Because practical and theoretical norms derive from the autonomy of higher intellectual faculties, they do not depend on species-relative forms of sense and (thus) possess a more than human validity. Accordingly, the normative standpoint represents norm-governed beings via the a priori idea of freedom or absolute spontaneity qua capacity to conform to self-legislated standards of right and good (though not necessarily practical) agency. This idea has a more than human validity because it designates a super-sensible character that is shared among all beings that possess practical and theoretical reason.Footnote ¹⁵

4. The Coherence of Kant’s Standpoint Distinction

Kant’s standpoint distinction is intriguing because it brings together different commitments that are central to his overall system. Unfortunately, it thereby also invites some of the central problems that arise for that system. In this section, I discuss two striking worries about the coherence of Kant’s standpoint distinction.

The first worry is that the two standpoints seem incompatible: from the empirical standpoint one judges that human beings are spatiotemporal and causally determined, whereas from the normative standpoint one judges that human beings are non-spatiotemporal and free from causal determination. A discussion of this issue inevitably leads into debates about Kant’s idealism. I must confine myself to a few relevant remarks.

Some suggest that the above judgements are not in conflict because they relate to different entities, non-spatiotemporal noumena and spatiotemporal phenomena. However, this two-world reading undermines the crucial point of Kant’s standpoint distinction: a rational being has two standpoints from which it can regard ‘itself’ (GMS, 4: 452).Footnote ¹⁶ Kant frequently claims that the phenomenal and the noumenal human being are numerically identical, one and the same (see e.g. Bxxvii; GMS, 4: 456; KpV, 5: 95). This does not commit him to the view that, for every given object that we individuate as an appearance, there is exactly one corresponding thing in itself. This qualification aligns with one of my earlier suggestions (see section 1): Kant’s claim that we can regard the very same being from two different standpoints is restricted to a specific class of entities, namely, rational beings to which we attribute higher intellectual faculties.Footnote ¹⁷

But how can one and the same rational human being coherently be judged as spatiotemporal, causally determined and as non-spatiotemporal, free from causal determination? For some, the two types of judgements do not reflect ontological classifications but only two different ways in which human beings can consider themselves, namely either in accordance with the conditions of human cognition or apart from those conditions. Since the qualifications ‘as considered in accordance with conditions of human cognition’ or ‘as considered apart from the conditions of human cognition’ are (implicitly) built into predicates such as ‘causally determined’ or ‘free’, there is no contradiction between the judgement that the same human being can be regarded both as causally determined and as free.Footnote ¹⁸

There is another type of one-world reading that construes Kant’s idealism as a metaphysical doctrine that rests on a contrast between two types of properties.Footnote ¹⁹ One type of properties derives from our mental faculties, namely, from the cognitive conditions of human experience that, in Kant’s idealism, are also ontological conditions of the objects of such experience (A158/B197). Representing things in spatiotemporal terms and as subject to deterministic laws of nature are necessary conditions of human experience. Since we are objects of human experience, it follows that we have spatiotemporal properties and are subject to deterministic laws of nature. But since these features depend on our finite representational faculties, they have no absolute, mind-independent metaphysical status: they belong to the phenomenal part of our constitution (Beschaffenheit). By contrast, freedom is a property that we possess independently of any finite forms of cognition: it belongs to the intelligible part of our constitution. Thus being causally determined is a property that pertains to us not absolutely, but only in a certain relation, namely, the relation in which we, as objects of human experience, stand to the formal conditions of human experience (B67–9), whereas freedom is a property we possess apart from that relation. The claim that a being is P (only) in relation to X is compatible with the claim that this same being is not-P apart from this relation (cf. Allais Reference Allais2015: 73).

I cannot discuss here which of these interpretations is preferable.Footnote ²⁰ The important issue is that there are two ways of construing Kant’s insistence that ‘there is not the least contradiction’ between the two standpoints because one and the same (eben dasselbe) human being ‘must … represent and think itself’ in a ‘twofold manner’ (GMS, 4: 457). Although a deflationary, anti-metaphysical reading is sometimes simply equated with a two-standpoint interpretation, such a reading is just one way of interpreting Kant’s standpoint distinction.

I now want to discuss a second worry about the coherence of Kant’s standpoint distinction, which arises regardless of whether one adopts a metaphysical or an anti-metaphysical two-standpoint reading. I have argued that Kant’s notion of a normative standpoint designates a perspective that considers human beings via purely intellectual representations (1b) and that considers how human beings ought to act (2b). One might worry that these two components do not work together: the normative standpoint cannot realize its function, namely, to determine how human beings ought to act, unless it defers to facts of human experience and thus uses sensible representations. This problem manifests itself in (at least) three different ways. (a) How individual human agents ought to act depends on their spatiotemporal circumstances: if I have promised to meet a student at 3pm, then I ought to leave the house at 2pm to arrive in time on campus. (b) What one ought to do depends on causal laws of nature: in addressing the question of when I should leave the house, I must think of myself as an embodied agent who is subject to the laws of physics, which prevent me from flying to campus in just a few seconds.Footnote ²¹ (c) The specific obligations that we incur collectively, as a species, depend upon empirical conditions, such as anthropological facts regarding human nature.

There is no analogous problem for the coherence of Kant’s empirical, anthropocentric standpoint. The empirical standpoint altogether renounces the concepts that are central to the normative standpoint: empirical thought does not represent human beings as free, spontaneous, autonomous, or as standing under normative laws. Moreover, though empirical thought does apply the intellectual categories, it does not apply them in their purely intellectual form that transcends the merely human perspective; rather, it schematizes these concepts in accordance with our human sensibility. Hence the ‘intelligible ground does not in the least impugn the empirical questions’ (A545/B574); ‘in [our] experience no question is ever asked in regard to’ (A30/B45) purely intelligible features such as our free noumenal character.

There are two different ways to defend Kant’s normative standpoint against (a)–(c). One proposal, call it the austere interpretation, is to purge the intelligible, normative perspective from all empirical or spatiotemporal representations. This renders the normative standpoint extremely abstract, but not completely empty: this standpoint can still give us a representation of norm-governed agency by invoking only the most general forms of prescription (such as the moral law, or the epistemic directive to continue in the regress of given conditions) and by invoking only the idea of a free, spontaneous subject endowed with ‘higher’ faculties such as will or understanding. The benefit of this proposal is that it provides the most straightforward way of preserving the integrity of a purely intelligible, non-empirical perspective on rational human agents. It also accords with some of Kant’s formulations in Groundwork III (especially at 4: 452), which may encourage the idea that considering ourselves as parts of the intelligible and sensible-empirical world is a mutually exclusive, either-or disjunction. However, the cost of the austere reading is that Kant’s notion of a normative standpoint can no longer be understood as a perspective from which individual persons can reach concrete judgements about how they ought to act in the empirical world that they inhabit. Austerely defined, the normative standpoint becomes so abstract that it seems far removed from our ordinary, engaged self-conception as free norm-governed agents. I regard this as a non-trivial cost because I see no evidence that Kant considers the normative standpoint an artifice of abstract theorizing. For instance, he says: ‘All human beings think of themselves as free according to their will. Hence arise all judgements about actions such as ought to have been done, although they were not done’ (4: 455). Here Kant describes how every ordinary human being (‘even the commonest understanding’: 4: 452) adopts a normative-intelligible standpoint on concrete spatiotemporal actions such as thefts or lies (cf. KpV, 5: 98–100).

An alternative, non-austere reading is suggested by Kant’s remarks that ‘insofar as we consider ourselves obligated, we regard ourselves as belonging both to the sensible world and yet at the same time to the intelligible world’ (GMS, 4: 453). Perhaps Kant here articulates what adopting the normative standpoint really amounts to for beings like us: sensible creatures endowed with higher rational faculties. To consider oneself obligated is to consider oneself subject to norms that one apprehends in the imperatival formula ought, which is suitable for finitely rational, that is, embodied and sensibly affected beings. On this reading, the normative standpoint for us is nothing over and beyond a perspective from which we regard ourselves as members of both worlds, that is, as free, norm-governed beings that are also spatiotemporally situated and subject to natural laws. This reading has the benefit that Kant’s idea of a normative standpoint captures our ordinary sense of a perspective from which we consider how we ought to act. But if occupying the normative standpoint involves considering ourselves in accordance with the conditions of human sensibility, and as subject to empirical laws of nature, does this not contradict the conception of that standpoint which I have worked out in sections 1–3? There it seemed that from the intelligible-normative standpoint one uses only purely intelligible representations and views rational beings as entirely free from empirical laws of nature. However, I want to suggest that there is a way of preserving the coherence of Kant’s notion of a normative-intelligible standpoint even on the non-austere reading.

Consider first the above-mentioned problem (a): I must consider spatiotemporal, empirical conditions to reach a concrete normative judgement. Now, even though the judgment that I ought to leave the house at 2pm so that I can keep my promise is infused with empirical content, Kant can insist that my reaching this judgement requires that I retain an (at least implicit) awareness of purely intellectual representations that have strict priority over what I can only represent in sensible terms as being in space in time. This point has a twofold meaning.

First, I must suppose that how I act in space and time ultimately depends on some aspect of my agency that is altogether insensible: I must regard my act as originating in, or as being grounded upon, a non-sensible, atemporal exercise of my higher intellectual capacities (here, of my free will). This point raises difficult issues, which I cannot pursue here. But it should be uncontroversial that Kant makes this point, i.e. that on his view a non-sensible ground of our empirically observable actions is indispensable to our self-conception as free agents (A539–41/B567–9; A550–9/B578–87; KpV, 5: 94–8; Rel, 6: 31). Thus, even though from the non-empirical standpoint we still represent our free, norm-governed agency as unfolding within a spatiotemporal framework (e.g. I determine myself to leave the house at 2 pm), we must also represent norm-governed agents as having features that are not themselves part of that framework. Moreover, we must regard these purely intelligible features as prior to, or more fundamental than, anything that we can represent empirically as unfolding in space and time.Footnote ²² We subordinate the sensible to the intelligible (GMS, 4: 461).

Second, even though I represent my concrete obligation (that I ought to be on campus to meet the student at 3pm) in spatiotemporal terms, I must take this obligation to be based upon a normatively prior principle that makes no reference to empirical conditions at all: namely, the principle that it is unconditionally good to keep one’s promises, which in turn refers back to the purely intellectual moral law that is the basis of all practical normativity, and that has objective reality or validity quite apart from its application to sensible conditions. This is also the key to mitigating the third of the above-mentioned problems (c): how can a standpoint that allegedly transcends our anthropocentric perspective establish a range of obligations that refer to empirical facts of human nature? Kant’s answer, I think, is that any attempt to establish concrete norms that apply to human beings only given certain facts about their empirical nature must be grounded in more fundamental, absolutely necessary intellectual principles that completely transcend our anthropocentric perspective and thus apply to all rational agents (MS, 6: 215–17). Without such a purely intellectual foundation, specific norms would be arbitrary and contingent.

This leaves the worry (b): Kant’s normative standpoint is supposed to view agents as free from empirical laws, and yet in figuring out (e.g.) when I ought to leave my house I must regard myself as subject to natural laws that prevent me from flying to campus. Here, I think, Kant should say that the normative perspective can concede that the physical acts of free agents are constrained by laws of nature: this perspective does not regard our body as free from causal determination because the body is not one of the higher faculties that we attribute to ourselves as an intelligence. From the normative standpoint, we deny that we are subject to empirical laws of nature only with regard to rational faculties such as the understanding or the will. We regard one aspect of ourselves (our higher mental capacities) as free and governed by rational norms, and another aspect of ourselves (our body) as unfree and governed by deterministic natural laws.

Of course, we think that typically our physical actions depend on our free will, i.e. on our freely chosen maxims or intentions. But our normative perspective on ourselves as morally obligated agents must acknowledge that the free control we exercise over our physical capacities is limited, precisely because these capacities are also constrained by natural causes. This does not mean that our will always lacks causal efficacy with regard to what happens in nature. But it does mean that the extent to which our will has such efficacy is not entirely up to our own, free volitional efforts. Accordingly, for Kant the freedom of will that we possess as moral agents is compatible with our inability to realize our freely chosen maxims (KpV, 5: 36–7, 45). A morally good will can ‘sparkle like a jewel’ despite its incapacity to realize its intentions (GMS, 4: 394). For example, I regard myself as having the freedom to choose to be on campus at the promised time, regardless of how strongly I desire not to go. But I am also aware that my morally good choice might not, despite my best efforts, have the intended physical effects: when I seek to execute my intention, my body might be causally determined to suffer a stroke. More generally, what I can achieve through my free agency is limited by determining empirical conditions that are outside of my free control.

Here one might wonder what accounts for the difference between cases in which our free choices do and cases where our free choices do not initiate the intended bodily movement. I think that Kant would respond that this question is both inscrutable and irrelevant to us when we adopt the normative standpoint on free rational beings: ‘Whether the causality of the will suffices for the reality of its objects or not is left for the judgments of … theoretical reason … In the practical task of reason … what matters is only the determination of the will and the determining ground of its maxim as a free will, not its success’ (KpV, 5: 45). Part of the allure of Kant’s standpoint distinction is that it systematically defines the contexts in which we can legitimately pursue explanatory questions: it is futile to raise such questions unless we can answer them via reference to laws of nature (GMS, 4: 459). Since the impact of our free volition on our bodily activity cannot be understood via such laws, we cannot explain or comprehend why our free volition sometimes results and sometimes fails to result in intended bodily movement.Footnote ²³ But for Kant this incomprehensibility should not matter to us when we adopt the normative standpoint, since our aim in adopting this perspective is to figure out how we ought to act as free agents rather than to understand how free agency works. Given this aim, we can ‘happily admit being unable to comprehend how’ the noumenal causality of reason influences the sensible world (KpV, 5: 49).

The non-austere strategy for defusing the three worries concerning Kant’s normative standpoint involves a concession, and an insistence that the concession does not touch upon Kant’s main point. The concession is that, unless we adopt the austere reading (which renders the idea of a normative standpoint of free agency artificially abstract), we cannot sustain the claim that a normative standpoint on our agency completely abstracts from the representations of space, time and empirical laws of nature. But perhaps Kant’s notion of a normative standpoint does not require such a complete bracketing of sensible and empirical representations: it might require only the weaker, but still highly significant claim that insofar as we regard ourselves as free, norm-governed agents, we must retain an awareness of certain features of ourselves which are irreducible to spatiotemporal, empirical features; and we must retain an awareness that these purely intelligible features are more fundamental than anything that can be sensibly represented.