Frederick Rauscher’s Naturalism and Realism in Kant’s Ethics (Rauscher Reference Rauscher2015) is a wonderful book. It does not give in to the temptation of focusing on the exotic. Instead, Rauscher draws his reader back to familiar, perennial questions on the ontological and epistemological status of morality in general, and God, freedom and immortality in particular. This is a daunting task, and Rauscher knows it. Every reader of Kant has views on these issues; we are not likely to break new ground, and consensus is not to be expected. Rauscher’s book can be read as an attempt to revive and improve upon Hans Vahinger’s fictionalism (p. 150). The ideas of God, freedom and immortality are fictions that are ‘not meant to have ontological implications’ (pp. 207–8). In contrast to Vahinger, however, these fictions are ‘not merely part of the empirical psyche’: they have an ‘a priori’ ground (p. 151). Like these ideas, moral values do not exist mind-independently and are instead dependent ‘upon the transcendental moral agent’. Rauscher calls this position ‘moral idealism’ (p. 245). Moral idealism shares with constructivism the view that the practical standpoint has a certain priority over the theoretical. Rauscher resists the ‘label’ constructivism, however, because Kant uses the term ‘construction’ rather differently than constructivists do (p. 24). Since the practical perspective ‘demands no ontology’, there is ‘no possible clash [with] the ontology provided by theoretical reason’ (p. 95). Hence Rauscher believes that his moral idealism is compatible with a ‘metaphysical naturalism’: a type of naturalism ‘which accepts as real only entities studied by the sciences – for Kant, nature in space and time’ (p. 241).
I would wager that many readers already have a number of passages in mind that call Rauscher’s idealism and naturalism into question. Rauscher is aware of this. Turning this discussion into a battle about passages from Kant would be unfair, however. His book is not an attempt to capture the ‘historical Kant’ (pp. 5, 150–1). Instead, he offers an interpretation that ‘encompasses the main claims in Kant’s ethics and shows how Kant either did or could have, using resources available within his practical philosophy, presented an entirely metaphysically naturalistic ethics’ (p. 242). Hence, if we want to engage with Rauscher’s Kant productively, we should not single out passages that do not support his interpretation. Instead, what must be shown is how Rauscher’s metaphysical naturalism and moral idealism are incompatible with ‘the main claims in Kant’s ethics’. This is what I shall attempt to do here.
This paper is divided into five sections. Rauscher and I both agree that Kant is a moral idealist, yet we disagree about the nature of his idealism. In the first section, I show that Rauscher neglects the common ground of Kant’s theoretical and practical philosophy, i.e. reason as a capacity for knowledge. As a consequence, Rauscher does not do justice to Kant’s practical cognitivism, and it remains an open question how his moral idealism can account for strict objectivity. In section 2, I show why Rauscher’s naturalism turns categorical imperatives into mere fictions that are either not really categorical or in fact overly demanding. In section 3, I turn to the notion of transcendental freedom, which seems to pose a problem for Rauscher’s naturalist interpretation. Rauscher makes room for his naturalism by restricting Kant’s notion of transcendental freedom to the act of self-legislation and operates with a compatibilist notion of freedom at the level of choice. I argue that this view is not only incompatible with Kant’s account of acting from duty but also cannot do justice to Kant’s definition of freedom of choice as the ‘capacity of pure reason to be by itself practical’. In section 4, I turn to what Rauscher calls the priority of theoretical reason over practical reason, or, as he also puts it, the ‘priority of belief’ over ‘determination of action’. I argue that the distinction between ‘belief’ and ‘action’, as Rauscher draws it, cuts right through Kant’s notion of a practical belief and a practical postulate. If Rauscher is willing to let these notions go, he must also give up on Kant’s doctrine of the highest good. Finally, I return to Rauscher’s moral ontology. In particular, I discuss two of his claims: (i) moral realists run into Kant’s heteronomy objection, and (ii) moral idealists should give up on the notion of intrinsic moral value properties. I argue against both claims and offer an alternative account.
1. Why Kant is a Moral Idealist
To make room for naturalism in Kant, Rauscher draws on Kant’s distinction between theoretical and practical philosophy. According to Rauscher, these are distinguished by their subject matter: ‘Practical philosophy is practical not by form, but by the object. … The theoretical is knowing, and the practical is behaving’ (p. 74).Footnote 1 ‘The main meaning of that divide is the difference between knowledge and action’ (p. 77). If the claims of practical reason are not about the world (‘what exists’) but are instead only determinations of our behaviour, there can be no conflict between the claims of practical and theoretical reason. Hence, Rauscher concludes, ‘Kantian morality then, must be prima facie compatible with a naturalistic metaphysics’.
Rauscher’s distinction is deeply rooted in Kant scholarship. Christine Korsgaard, for example, accuses the moral realist of confusing epistemology with ethics (Korsgaard Reference Korsgaard1996: 44). This distinction strikes me as problematic, however. The faculty of reason is a capacity for knowledge. Knowledge pertains not only to what is the case but also to what ought to be the case. Kant’s Critique of Practical Reason(!) is Kant’s moral epistemology. It is his attempt to establish the fundamental principle of practical knowledge, i.e. the categorical imperative. The object of practical knowledge is not simply behaviour but, as the table of the categories of freedom tells us, knowledge of principles and actions as good or evil. These objects have a formal aspect (practical universalizability) and a material aspect (our desires, which ought to conform to the form of practical cognition).
This conception of Kant’s ethics as a practical epistemology is relevant to Rauscher’s naturalism and idealism. Rauscher defines idealism as the view that ‘moral principles, properties, or objects of the world are dependent upon the transcendental … moral agent’ (p. 245). I agree with Rauscher that Kant is a moral idealist (Bojanowski Reference Bojanowski2012, Reference Bojanowski2016), but I think we should avoid the term ‘constructivism’ because it does not do justice to Kant’s cognitivism. In other words, Kant is not an idealist merely because moral principles are dependent on the moral agent. He is an idealist because these principles are dependent on practical reason. I agree with the realist that it is not some particular cognition or desire of mine that makes a principle good. This does not mean, however, as the realist believes, that moral values are out there, mind-independently. Without the form of practical knowledge, there would not be moral values. This form is independent of our particular minds, but it is not mind-independent in general; it is the form of reason as such. According to Rauscher, ‘[p]ractical reasoning is simply the a priori formal structure of reasoning in deliberation when one attempts to choose freely’. But it is not clear why it is this ‘a priori structure’, and not some other structure, that guides us in our moral thinking. What is it about this so-called ‘a priori structure’ that makes it a capacity for morality? That makes our judgements objective and not merely intersubjective? It is not clear why Rauscher’s idealism does not turn the moral law into a mere ‘figment of the imagination’. In resisting the common distinction between theoretical and practical philosophy, we might be in a better position to appreciate practical reason as the true origin of moral values.
2. Why Kant is Not a Naturalist
I turn to Rauscher’s naturalism. According to Rauscher, Kant’s priority claim amounts to the idea that practical reason has priority because ‘speculative metaphysics is irrelevant to practically oriented action, so that even a speculative denial of the postulates would not affect practical reason’ (p. 82). What Rauscher calls the priority claim is perhaps more appropriately called an ‘independence claim’. He believes that ‘even if there were theoretical disconfirmation’, everything would remain the same from the practical standpoint (p. 106). This strikes me as an overstatement. Kant holds that practical reason has priority because it gives cognitive content to the ideas of God, freedom and immortality, which were shown to be merely problematic concepts for theoretical reason (KpV, 5: 121).Footnote 2 Again, I do not want to quarrel about passages here; Rauscher’s independence claim strikes me as too strong for reasons that go beyond textual considerations. Even if we knew that we were not transcendentally free, we would still need to make choices. If we knew theoretically that pure reason could not be causally efficacious, however, we would no longer think of ourselves as categorically obliged. Categorical obligation presupposes transcendental and not merely comparative freedom. If a moral sceptic (or a compatibilist) could deliver a conclusive argument against transcendental freedom, the categorical imperative would become void. This is why Kant holds that transcendental freedom ‘constitutes the real moment of the difficulties in [practical freedom]’ (A533/B561). In other words, practical freedom is a kind of transcendental freedom, a first causality. Rauscher wants to read Kant as a compatibilist (p. 181), but the compatibilist notion of freedom is a merely relative or comparative notion of freedom. Kant rejects it because it cannot properly account for categorical obligation. And even though the ‘difficulty’ is speculative, our speculative knowledge that we are unable to act as first causes would also mean that we know we are unable to act from the representation of laws. This would lead us to believe that categorical obligation does not apply to us. The inference from the categorical ‘ought’ to ‘can’ is merely conceptual. There is still room for scepticism, and this scepticism has consequences for our practical point of view. This is why Kant always needs to refer the sceptic to his solution to the Third Antinomy (GMS, 4: 455–6; KpV, 5: 30, 48 ff., 94–106; KU, 5: 175; RGV, 6: 39–40; A462/B490ff.). Rauscher underestimates the enlightening effect that Kant’s moral philosophy is supposed to have on its reader. Is not the contemporary situationist literature precisely a sceptical attempt to fundamentally question categorical obligation? Empirical research is supposed to put our ability to act from principles into doubt, and by doing so it wants to disqualify the categorical imperative as overly demanding. This is the kind of claim the Third Antinomy is supposed to fend off.
Perhaps Rauscher underestimates the significance of the Third Antinomy because he believes that the first Critique revealed that the ideas of reason do not ‘have any reality of their own’ and are instead ‘illusory’ (p. 231). It is true that speculative reason is not entitled to ascribe reality to the idea of freedom. Yet Kant does not move from this claim to the claim that this idea is ‘illusory’. The result of the Third Antinomy is that the concept of first causality is an ens rationis: it can be thought without contradiction but does not have any positive content. Again, if we knew that transcendental freedom were an illusion, we would not consider ourselves to be categorically obliged. We would act not under the ‘idea’ of freedom in Kant’s technical sense, as Rauscher has it, but merely under the representation of comparative freedom. In short, if transcendental freedom were a mere fiction, we would not be categorically obliged. Kant would be an error theorist of some sort: categorical obligation is an invention, but it is not really justified. Indeed, there would be no way to justify it.
The solution to the Third Antinomy shows that the idea of a transcendentally free cause can be thought without contradiction. Our cognition of moral obligation gives us a reason to assert this idea by ascribing it to ourselves. If this assertion were merely a fiction, it would coincide with Rauscher’s naturalism. However, turning transcendental freedom into a mere fiction leaves too much room for the moral sceptic. Rauscher wants to restrict Kant’s ontology to ‘entities studied by the sciences’. I believe that we should ascribe a more nuanced view to Kant. We can only have positive knowledge of phenomena, but practical cognition gives us reason to believe that there is more to this world than ‘entities studied by the sciences’. Kant’s critical philosophy can be understood as a critique of this kind of scientism; the bounds of theoretical knowledge are not also the bounds of the world. If transcendental freedom (a spontaneous cause) were not real, it would be wrong to hold us accountable to categorical imperatives. I take Kant’s claim to be that we do not have theoretical access to its reality, but we have to assert its reality on practical grounds. In short, my main worry is that Rauscher’s naturalism turns categorical imperatives into mere fictions that are either not really categorical or in fact overly demanding.
3. Why our Faculty of Choice Must Be Transcendentally Free
Rauscher wants to defend a compatibilist notion of freedom, but he is aware that there is a stronger notion of freedom in Kant. Rauscher wants to ascribe this stronger, transcendental notion not to our faculty of choice (at least not for the most part) but, as he puts it, to reason: ‘[Kant’s] argument about freedom concerns not a transcendentally free power of choice or decision but only a transcendentally free power of reason to produce the moral law’ (p. 134).
Rauscher’s idea is that transcendental freedom applies to the act of self-legislation, while the particular judgement or choice is ‘not free but is determined’ (p. 134). ‘Instead of the judgment being free, “reason must regard itself as the author of its principles independently of alien causes … it must be regarded of itself as free”’ (p. 134). I will only mention here that I do not believe that this passage in Kant (GMS, 4: 448) proves Rauscher’s point. For one, the term ‘principle’ is ambiguous. It can mean the fundamental law or the particular maxim we act on. Maxims are practical judgements. Hence if ‘principle’ is to be understood as ‘maxim’, our power of choice is indeed involved. But again, let me try to explain why Rauscher’s account is incompatible with ‘the main claims in Kant’s ethics’ (p. 242).
My argument is similar to that sketched above: if transcendental freedom is only about the ‘production’ of the moral law, then acting from duty is impossible. I take it that Rauscher does not want to give up on the idea of acting from duty. Hence, transcendental freedom must be about the production of the moral law. Acting from duty requires that we perform an action for the sake of its universalizability. If our actions were fully determined by given desires, the representation of the law-likeness of our actions would still be possible, but it would be a mere epiphenomenon that played no causal role in their genesis. Acting from duty would be impossible. Hence categorical obligation requires that we believe that our actions are naturally undetermined. Rauscher’s attempt to restrict transcendental freedom to the act of self-legislation and to operate with a compatibilist notion of freedom at the level of choice does not do justice to Kant’s definition of freedom of choice as the ‘capacity of pure reason to be by itself practical’ (MS, 6: 214). Rauscher wants to ascribe transcendental freedom to the legislative function and compatibilist freedom (at least for the most part) to our faculty of choice. However, applying the notion of transcendental freedom (an uncaused cause) to legislation is misleading. If this were correct, then even the generation of the categories or transcendental laws of nature would have to be understood as an act of transcendental freedom. Rauscher does indeed say, in the passage quoted above, that reason ‘produce[s] the moral law’ (p. 134). Yet we should not apply the relation of cause and effect to reason and its fundamental laws. Reason does not ‘produce’ the categories and the moral law; they are its constitutive features. Transcendental freedom as a first causality only comes into play when we consider the possibility of an act of reason that has an effect in the empirical world. This explains why the problem of transcendental freedom only arises within the context of the Third Antinomy (a first event in nature) and not within the context of the Analytic. When Kant defines freedom as the ‘capacity of pure reason to be by itself practical’, this practicality should be understood in a causal sense. Freedom is the capacity not simply to give a law to oneself but to act from that law. Again, if we were unable to act from the law, we would always need some extra incentive; acting from duty, and hence categorical obligation, would not apply to us. Rauscher is right to say that we can consider our ‘judgments the causal product of chemical and electrical events in the brain’, but the question is whether the chemical and electrical events in the brain are the only causal factors, or whether our practical cognition influences them causally. In Kant’s words, the question is whether ‘the cause in appearance … was not so determining that there is not a causality in our power of choice such that, independently of those natural causes and even opposed to their power and influence, it might produce something determined in the temporal order in accord with empirical laws, and hence begin a series of occurrences entirely from itself’ (A534/B562). I take this to be freedom of choice in an incompatibilist sense. We would not be entitled to ascribe to ourselves freedom of choice in a transcendental sense if we could not cognize the categorical imperative. Only the cognition of unconditional obligation makes us aware that we can act not only from the desire for happiness but also from the representation of the universalizability of our maxims. Thus we should think of our capacity of volition as a unified capacity that is free in an incompatibilist sense because it can act from the representation of laws. This does not mean, as Rauscher reads Kant, ‘that only actions caused by reason are free’. Rauscher moves too far in the Reinholdian direction he attempts to bypass. Kant does not hold that our actions are free if and only if we act from (and not against) the moral law. His view implies only that a rational capacity of practical volition can be understood as transcendentally free if and only if reason can be practical by itself. If we did not have this capacity, we would have no reason to ascribe transcendental freedom to our will (as a capacity of volition, not an act).
4. Why Speculative Reason does Not Have Priority over Practical Reason
I return to the relationship between theoretical and practical reason. Rauscher claims that there is also an implicit claim in Kant about the priority of the speculative over the practical:
Kant’s explicit doctrine must presuppose a higher priority of speculative reason as reason’s interest in knowing about given objects. Kant is operating with an unstated assumption that reason must have an interest in certain objects, namely, the objects of the postulates, not for determining right acts but for understanding the relation between right acts and the rest of our experience. Hence the postulates appear not as determinants of right acts in the process of deliberation but only when reflecting on the nature of morality, the fact that right acts do occur, and the fact that agents believe those acts ought to occur. And for this reflection, Kant must be assuming that speculative not practical reason is at work, and hence that there is a higher priority of speculative reason. (p. 85)
This account does not put Kant’s practical postulates in their proper place. Consider immortality. Kant’s main point is that we have to believe in our immortality because this belief makes possible a certain action that we cognize as practically necessary, namely striving for moral perfection, i.e. making ourselves ‘holy’. This is implied in the obligation to promote the highest good in the supremum sense of the term, which is complete agreement between one’s moral dispositions and the moral law. Rauscher holds that ‘the necessity of belief [in the postulates] is a product of theoretical reflection not a prerequisite of action’ (p. 88). Yet we do not believe in our immortality because of some theoretical interest. Instead, our belief in our immortality is absolutely practically necessary if we want to do what we ought to do: to perfect ourselves morally. Only under the assumption of the immortality of the soul is doing what we are morally obliged to do possible. The reason for believing in our immortality is practical, not speculative or theoretical.
Rauscher puts his argument in terms of agency and action. At one point, he suggests that ‘the structure of reason does not need to exist in order for agents to act at all because they can act on the basis of sensible impulses alone. … Kant does not claim to know that reason must exist in order for human beings to act’ (p. 127). It is important to stress that the belief in our immortality is not necessary for all kinds of agency. Only finite moral agents, agents who have inclinations and the capacity of pure practical reason, need to practically postulate their immortality in order to be able to do what they morally ought to do, i.e. morally perfect themselves. In fact, if practical philosophy is understood independently of practical reason, simply in terms of mere behaviour, neither the postulates nor the moral law come into the picture.
To be sure, ‘the soul is immortal’ is, as Kant says, a ‘theoretical proposition’ (cf. KpV, 5: 122). It is a proposition about what is the case rather than what ought to be the case. But this does not turn the postulate into a theoretical postulate. Rauscher holds that there is a ‘priority of belief’ over ‘determination of action’ in Kant’s Groundwork and his second Critique. Thus we are ‘justified in labeling both of them works of speculative philosophy’ (pp. 86–7, 93). However, ‘the soul is immortal’ is a proposition to which we are entitled, even though there is no theoretical reason to assert it. ‘The soul is immortal’ has its justificatory ground in pure practical reason. This is the difference between it and the theoretical proposition ‘ghosts are pink’, which we do not have reason to assert even on practical grounds. Hence the belief ‘the soul is immortal’ should be understood as a ‘practical belief’ (A823/B851). Rauscher’s distinction between ‘belief’ and ‘action’ cuts right through the notion of a practical belief and a practical postulate. We must hold on to these notions, however, if we are to hold on to Kant’s doctrine of the highest good. If Kant is right, giving up on the highest good would also make the categorical imperative ‘fantastic and directed to empty imaginary ends and … therefore in itself … false’ (KpV, 5: 114). There is no indication that Rauscher would be willing to pay this price.
5. Why Kant is Not a Value Realist
Kantian moral realists believe that our humanity is a mind-independent property that we perceive as valuable. Rauscher criticizes this idea because it undermines our autonomy: ‘An independent value of humanity as the basis of the moral law would violate autonomy if that value were to be seen as shaping the actual legislation of the categorical imperative by reason’ (p. 216). I agree with the general direction of Rauscher’s criticism of Kantian moral realism, but I am not confident that this particular argument succeeds. Kant attempted to show that all traditional moral theories are heteronomous. They all presuppose some end as given – an end that can ultimately be reduced to our desire for happiness. This is true not only of empiricist theories but also of rational perfectionism. According to Kant, only his own moral philosophy is based on the principle of autonomy. However, Kantian moral realists could claim that our will is determined by our cognition of a metaphysical moral fact, the value of humanity, of which we have intuitive knowledge. Hence moral realism does not presuppose some end as given. Our volition is guided and caused by cognition of what is objectively good or bad. Thus, moral realism is not subject to Kant’s heteronomy objection (Stern 2012: 25–6).
This is not to say that Kantian moral realists are correct. I believe that a response to the moral realist can be found in Kant. Briefly, the concepts of autonomy and heteronomy relate to our capacity of volition. They must relate to this capacity if cognition is to be practical. Moral realism does not qualify as an account of practical reason because it decouples moral value from volition. The terms autonomy and heteronomy do not even apply to this view. Hence it is no surprise that moral realism is not part of the table of ‘possible principles of morals from the fundamental concept of heteronomy’ (KpV, 5: 40). The concept of heteronomy only applies to theories that are fundamentally practical. A theory that is not practical is not even a moral theory. I cannot go into this in further detail here (see Bojanowski Reference Bojanowski2016); let me simply emphasize that I fundamentally agree with Rauscher that the idea that we cognize some mind-independent value property through ‘inner [or] outer intuition’ is incompatible with Kant’s moral epistemology and ontology (p. 218). I do not think, however, that the heteronomy charge against contemporary Kantian realists hits its target.
There is a second point at which Rauscher’s criticism of moral realism differs from mine. Rauscher believes that, in rejecting moral realism, we must also reject intrinsic value properties. He holds that ‘there is no place in Kant’s philosophy for any intrinsic value properties. … Moral value cannot be an intrinsic property of objects but instead is an order imposed by reason’ (pp. 209–10). I agree with Rauscher’s view that value is in some sense ‘imposed by reason’, but I disagree that this forces us to abandon the idea of intrinsic value properties. Or, to put it slightly differently, I believe that we can be moral idealists and still hold on to the idea of intrinsic value.
Rauscher rightly wants to resist the view that humanity is a mind-independent value property perceived through sensible or rational intuition. Yet I do not find his alternative position entirely convincing either. Here is what he says:
Pure practical reason as the faculty of transcendental moral agents holds humanity to be an end in itself; this constitutes the value of humanity. Reason does not insist that humanity is an end in itself because it has some other value. The value of humanity is just being held by reason as an end. Value as an end in itself is not an intrinsic property of beings with humanity but demands a relation to practical reason. (p. 221)
Reason does not recognize any pre-existing value of humanity but reason assigns that value to humanity through its ordering of ends. All value, even absolute value, is thus for Kant ideal rather than real. In a transcendental sense all value depends upon the transcendental moral subject as a rational being subject to the categorical imperative imposed by reason itself. (p. 224)
I find this mysterious. Why does reason ‘assign’ this value to humanity? What makes humanity valuable? Rauscher’s answer seems to be that reason wants it so. But this means accepting the voluntarist horn of the Euthyphro dilemma. Embracing this horn clearly contradicts the spirit and the letter of Kant’s non-voluntarist ethics. Again, I agree with Rauscher that it is practical reason that makes humanity valuable. But how? Rauscher’s answer is that ‘[t]he value of humanity is just being held by reason as an end’. But this does not explain why humanity has value. I think we can agree with Rauscher that humanity is not a mind-independent value property without accepting the view that reason simply wants humanity to be valuable. When Kant speaks of humanity, he does not mean the sum of all human beings. ‘Humanity’ is supposed to pick out our essential feature qua human. The essential feature of human beings, from a moral point of view, is their capacity to act from the representation of laws – or, what amounts to the same thing, the capacity to have a good will.
If human beings were not able to act from the representation of laws and only had the capacity to set ends, they would be guided by the concepts of the agreeable and the disagreeable alone. Yet agents who can only act from the concepts of the agreeable and the disagreeable can only pursue actions because they are conducive to their happiness. They cannot pursue actions because they are good in themselves. Humans, by contrast, are essentially moral agents. They do not only act because their actions contribute to their personal happiness; they can make the realization of their personal happiness dependent upon its universalizability. An action done for the sake of its universality is good not merely for me but in itself. I take it that Kant’s main claim is that only a being who can pursue actions because they are good in themselves, who can act from the representation of laws, has inner (‘intrinsic’) value, i.e. dignity. I take this to be the inner or intrinsic value of our humanity.
In short, I agree with Rauscher that humanity is not a value property that exists mind-independently and that we cognize as valuable in intuition. This does not mean, however, that there is no room for intrinsic value. Rauscher seems to presuppose that we must be value realists if we are to accommodate the notion of ‘intrinsic value’. His alternative, idealist view is that reason simply ‘assigns … value to humanity through its ordering of ends’. Yet if we understand ‘humanity’ as the essence of our moral being, and if this essence is our capacity to act from the representation of laws, then our humanity is practical reason. Reason does not confer or ‘assign’ value to humanity; humanity, as practical reason, is itself the source of moral value.
