It is hard to read Kant’s Critical theory of freedom as a stable theory with major claims consistent across the board, or even as one displaying a linear development. For Kant’s theory appears to have taken at least two significant turns. In the first phase, constituted by the Critique of Pure Reason (1781/1787), Groundwork for the Metaphysics of Morals (1785), and the Critique of Practical Reason (1788), Kant puts forward a ‘source’ conception of freedom, according to which freedom is a capacity for ‘absolute causal spontaneity’ (A446/B474) and consists in the agent’s being the ultimate causal source of an action. In this phase, Kant firmly states that this causality must be governed by a law, though not a natural but a normative and intelligible sort of causal law, and that the moral law is this very causal law of freedom. This leads him to both offering a positive conception of freedom as autonomy or capacity to self-legislate, as ‘a will’s property of being a law to itself’ (G 4: 447), and espousing a striking identification: ‘a free will and a will under moral laws are one and the same’ (G 4: 447). The second phase seems to display a sharp turn from source to ‘leeway’ freedom or liberty of indifference, when Kant formulates good and evil in the Religion within the Bounds of Mere Reason (1793) as both imputable (thus, free) radical deeds and exercises of power of choice, differing only in the order of priority between two opposing fundamental incentives, the moral law and self-love (Rel 6: 36). Finally, the third phase implies a return back from leeway to source, through Kant’s explicit denial in the Metaphysics of Morals (1797) that freedom can be defined in terms of a two-way ability to choose for or against the moral law (MM 6: 627).

A tempting and historically informed explanation of Kant’s apparently wavering views on freedom involves two objections that his early critics levelled against him. The ‘objection from necessity’, first voiced by Ulrich in 1788, states that understanding the moral law as the causal law of freedom would undermine freedom itself and reintroduce determinism at the noumenal level, while removing it at the phenomenal level. This would be a damning consequence for an incompatibilist like Kant. The remedy, as Reinhold argued in 1789, would have to be a redefinition of freedom as a leeway or two-way capacity to choose for or against the moral law. Kant’s baffling shift towards leeway in the Religion is an effort to accommodate this objection. Yet, such a shift has its own unwelcome consequences. The ‘objection from chance’, as raised by Creuzer in 1793, is that conceiving freedom as a two-way capacity to choose between good and evil actions would remove the explanatory ground of why the will chooses one way rather than another. The threat of noumenal determinism would just be replaced by that of noumenal chance. Kant’s reverse shift from leeway to source freedom or from moral libertarianism to moral necessitarianism in the Metaphysics of Morals is motivated by this objection.

Owen Ware’s contribution to the Cambridge Elements series, Kant on Freedom, takes issue with this narrative. While Ware admits that Kant’s texts display awareness of the ‘freedom controversy’ in the 1780s and 1790s, he defends the rather bold position that Kant remained consistently committed to moral necessitarianism and a source conception of freedom, never shifting towards a leeway conception of freedom. The Element is composed of three main chapters. The first chapter lays out the freedom controversy. The second offers a rigorous defence of Kant’s moral necessitarianism against the objection from necessity. The third and final chapter considers the early reception of Kant’s theory of freedom, immediately after 1793, including Fichte’s contribution to the freedom controversy, as well as contemporary accounts by Paul Guyer, Iain Morrison, and Marcus Kohl, attributing a leeway conception of freedom to Kant.

Ware argues for a number of related claims to support his view that Kant’s theory of freedom does not espouse leeway freedom and moral libertarianism in any of its developmental stages: (i) Kant’s theory is not susceptible to the objection from necessity; (ii) Kant’s doctrine of radical evil in the Religion does not commit him to a leeway conception; (iii) Kant’s moral necessitarianism (or his rejection of leeway) is compatible with the imputation of evil or immoral acts to free agents. I agree with the essence of the programme Ware pursues. So, my reflections below are coming from a rather sympathetic place.

Ware’s defence of Kant’s moral necessitarianism against the charge that it reduces free volitional activity to a noumenal sort of determinism is aptly based on the distinction between natural and practical notions of necessitation (pp. 16-18). Ware, again, rightly notes that while natural necessitation is a feature of heteronomous causality and thus applies to temporal relations between phenomena, practical necessitation is a feature of autonomous causality and applies to intelligible relations in the noumenal domain (p. 19). He observes that the objection from necessity, charging Kant with noumenal determinism, mistakes autonomous causality for heteronomous causality. Only the latter kind of causality, he argues, leads to determinism and thus is incompatible with freedom. Ware holds that autonomy or self-determination is not the kind of determination that would undermine freedom. A will autonomously restricting itself to the moral law is not less, but more free (pp. 19-20).

It is not clear to me this decisively blocks the objection from necessity. For Ware also insists that practical necessitation, like natural necessitation, consists in the exclusion of alternate possibilities (p. 18). In fact, however, the modal difference between practical and natural necessitation is that the former does not exclude alternate possibilities but the latter does. While a practical-normative law obligates without determining the consequence, a natural-empirical causal law determines the consequence to the exclusion of alternate possibilities.Footnote ¹ As Ware himself also acknowledges later on, it is indeed really possible for a free will to deviate from the moral law, or more precisely, for Willkür to disobey Wille’s legislation, whereas it is not really possible for an empirical event not to follow the natural law of causality (p. 40).

This modal difference between practical and natural necessitation is what effectively blocks the charge of noumenal determinism against Kant’s moral necessitarianism. Ware’s ambivalence about recognising this might be due to the fact that this difference gives the appearance that Kant endorses leeway freedom. Ware takes on the challenge of incorporating the possibility of a free will’s deviation from the moral law into his account, though somehow without abandoning his talk of exclusion of alternate possibilities and accepting leeway as part of Kant’s theory of freedom. Ware’s strategy is to locate Kant’s conception of an idealised form of moral agent, a perfect or ‘holy’ will at the centre of his account of Kantian freedom. The idea is that such a perfect will would never deviate from the moral law because it lacks any opposing influence that would be able to distract or tempt an imperfect will like ours (p. 11). The moral law is both objectively and subjectively necessary for such a being (G 4: 412). Alternate possibilities are then genuinely excluded for a perfect will. Kant’s famous footnote in the Religion does indeed emphasise this necessity or lack of contingency for the kind of free moral agent God is (6: 50n). Yet one worry here is that for such a holy will, the difference between natural and practical necessitation, which is supposed to be the remedy for the determinism charge on Ware’s account, essentially disappears. In fact, Kant’s footnote empathetically rejects indeterminism and predeterminism, but leaves it open whether a holy will is not deterministic.

Now, Ware is certainly right in suggesting that the ‘can’ in ‘ought implies can’ does not, as some think, point to a two-way possibility or capacity to do otherwise. For Kant refers to the principle solely to express the positive possibility of complying with the moral law, when the agent fails to do so.Footnote ² Nevertheless, the principle still entails the abundant actuality and thereby the possibility of deviating from the moral law as pertaining to imperfect wills like ours, for which the moral law is subjectively contingent. So, what to do with this undeniable possibility? Ware’s argument is that this possibility cannot be conceived as a positive power or ability so as to warrant a two-way freedom of indifference (pp. 42-43). Here he follows Allen Wood’s clever suggestion that ‘Not every possibility is a power. Some possibilities are in fact due to a lack of power’ (Wood Reference Wood and Wood1984: 81). Accordingly, deviating from the moral law is not a proper exercise of freedom but a misuse or abuse of this power. There is clear textual evidence that Kant rejects defining freedom in terms of a two-way choice: ‘Freedom of choice cannot be defined […] as the ability to make a choice for or against the law (libertas indifferentiae) […]. Only freedom in relation to the internal lawgiving of reason is really an ability; the possibility of deviating from it is an inability’ (MM 6: 226-7). Yet, Kant does not appear very consistent on this point, since he also defines ‘choice’ (Willkür) as a two-way faculty (Vermögen), ‘a faculty to do or refrain from doing as one pleases’ (MM 6: 213), and states that only Willkür, and not the will (Wille), can be called free (MM 6: 226). While Ware’s position is still a plausible one, the distinction between a possibility or ‘option’ (as Ware puts it) to act in one way and an ability to do so, which he heavily relies on, seems to beg the question and require additional argument, a metaphysic of powers or abilities perhaps, at least, to show that it is more than a semantic subtlety. For from a purely metaphysical and normatively neutral point of view, the distinction does not hold up.

Let me conclude with a few notes on Ware’s claim that the problem of the imputation of immoral acts does not force Kant’s theory to allow leeway freedom. Ware admits that the ground of imputation is nothing but freedom itself such that an act is imputable to an agent if and only if it is free. Now, provided that Ware argues that any deviation from the moral law is in fact not a power but a misuse of freedom, the question that he has to answer is how and on what grounds this misuse of freedom is imputable. Ware asserts that a culpable or imputable misuse of freedom (in the form of frailty, impurity, self-conceit, etc.) is possible in a moral necessitarian model (p. 45) but seems to evade answering how this is really possible. Especially crucial is that if freedom is the condition of the possibility of imputation, the imputable misuse of freedom must itself be free, which is Kant’s position in the Religion too: ‘Every evil action […] whatever the natural causes influencing him […] can and must always be judged as an original exercise of his power of choice’ (Rel 6: 41). First, this means that a misuse of freedom is, after all, a use or exercise of it too, which seems to cut against Ware’s claim that evil, construed as a misuse of freedom, is an inability and would diminish or corrupt freedom. Second, Ware rightly recognises, as an inevitable consequence of his strategy of taking a perfect will as the paradigm of a free will, that freedom comes in degrees: the more perfect the will is in living up to the demands of the moral law, the freer it is, and the more morally deviant it is, the less free it is (pp. 45-46). This seems to suggest that imperfect freedom is a consequence of an imperfect will, whereas Kant’s core argument in the Religion is that the imperfection of will or character should be the consequence of the exercise of freedom. But more importantly yet, this reasoning about degrees of freedom opens the backdoor for the problem of imputation: the most evil or imperfect will would be the least free, and thus, the least imputable.

As should be evident from my discussion above, this concise Element is quite a thought-provoking contribution to the ever-intensifying literature on Kant’s theory of freedom. The challenge it takes on is serious, and it helps our grasp of the philosophical nuances of Kant’s theory of freedom and why it excited so much interest and controversy both at his own time and today.