Abraham Anderson’s 2025 The Skeptical Roots of Critique advances the bold and intriguing thesis that Hume’s scepticism convinced Kant that metaphysics is impossible and led him to the ‘quasi-sceptical’ tenets of transcendental idealism (xii). It does this by advancing the even bolder thesis that Hume inspired Kant’s discovery of the antinomies. Notoriously, Kant wrote both that ‘the Antinomy of pure reason … first woke me from dogmatic slumber and drove me to the criticism of reason itself’, and that ‘it was the objection [Erinnerung] of David Hume that first … interrupted my dogmatic slumber, and gave my investigations in the field of speculative philosophy an altogether new direction’.Footnote ¹ Rather than see these as conflicting pronouncements, Anderson intriguingly seeks to integrate them.

Anderson admits, Hume ‘is not mentioned in the Antinomy, nor does the objection …, as reported in [P], correspond to any of the arguments put forward there’ (xi). But he plausibly and cleverly proposes that Kant was not inspired by Hume’s attack on ‘the principle that every event has a cause’, but by his attack on the principle of sufficient reason (xii, xiii), specified as the principle that where the conditioned is given the sum of all its conditions must be given (1). The latter, of course, is the foundation of the Antinomies (5). For Anderson’s Kant, Hume’s attack on this principle did not present an epistemological challenge that needed to be met. It offered a refutation of rationalist metaphysics that needed to be accommodated (xii). This is an original, noteworthy, and compelling line of interpretation. It is backed up by a series of chapters divided into short, punchy sections, making this book not only an important but an engaging and enjoyable read.

In an Introduction devoted to a survey of secondary literature on Kant’s ‘awakening’, Anderson argues that Kant came to conceive the Antinomy as arising from the conflict of the principle of sufficient reason with itself (5). He further argues that since P earlier describes Hume as having awoken him from his ‘dogmatic slumber’ (4: 260) and somewhat later wrote that the cosmological ideas awaken philosophy from its dogmatic slumber (4: 338), ‘[P] thereby challenges us to consider how the Antinomy is connected with the objection of David Hume’ (6). He also adds that since Kant described the Antinomy as avoiding ‘the slumber of fictitious conviction’ (A407/B434), he ‘thereby poses, in the Critique itself, the question of how Hume’s effect on dogmatic slumber is related to that of the Antinomy’ (7). However, that question cannot have been posed in A. To my mind, the implications of what Kant would have found in the Sulzer translation of U4 and U7 make a sufficiently strong case.

Chapter 1 engages Michael Forster’s proposal that Kant was initially wakened to apply an antithetic method in DSS. Anderson maintains that the antithetic found in DSS was not the expression of a sceptical crisis that Kant eventually overcame but ‘the source of the whole Critical perspective’ (20). It contains a true antinomy concerned with the possibility of human freedom (32). Kant’s approach to this problem was guided by an engagement with Rousseau’s ‘Profession of Faith of the Savoyard Vicar’. This text opposes reasons for thinking that we are inert material beings to reasons for thinking that we have spontaneously acting immaterial souls. The opposition is resolved with a fideistic declaration that, while we cannot understand how freedom is possible, it is not a contradictory concept. We are therefore entitled to faith that it exists (22–3). Anderson shows that DSS follows a similar trajectory (24). It also goes on to initiate the critical project of drawing of the boundaries between what we can know and what lies beyond our capacities. That suffices to make it a proto-critical text.

Not satisfied with this substantive result, Anderson claims that ‘The need to restrain speculation through a philosophy which judges its own proceeding recalls Hume’s remark’ (26) at U1.12 about the need to discern the limits of the understanding. But did Kant recall that remark? Might he have thought it on his own or gotten it from some other source? ‘[I]t came into my Thoughts, that … before we set our selves upon Enquiries of that Nature, it was necessary to examine our own Abilities, and see, what Objects our Understandings were, or were not fitted to deal with’ (Locke Essay, Epistle:7).

Anderson further appeals to U7.25 to argue that Hume’s objection underwrites what Kant found in Rousseau (35). As Anderson glosses it, U7.25 asserts that since we cannot understand agency in minds any more than in bodies, ‘we must reject the action of God on bodies or we must cease rejecting any cause simply because we cannot understand it’ (33–4). But the second disjunct is not stated at U7.25. Kant cannot have been inspired by U7.25 unless he was reading more into it than it contains. Perhaps he needed to look no further than Rousseau or within.

Chapter 2 brings up a passage from Blomberg’s notes on Kant’s 1771 logic lectures (24: 217) that suggests that Kant did not have a sophisticated understanding of Hume’s philosophy as of that date. Anderson shows that ‘there is at least one section of [U] in which Hume argues in the way Kant says’. More ambitiously, he argues that Kant was so ‘fascinated’ with this one passage that he generalized its method ‘to Hume’s way of arguing in … the [VS] as a whole’ (39).Footnote ²

Anderson’s passage is U8.36, where Hume addressed the objection that determinism makes God the author of sin. Anderson maintains that this passage ‘suggests that there is an insoluble conflict between human freedom and necessity and that Hume’s [compatibilist] “reconciling project” has failed’ (41).

This time, Anderson omits rather than embellishes. His citation of U8.36 (41) contains ellipses where Hume wrote:

As Hume made clear, and as is underscored when the subsequent sentence (which Anderson does cite),

is read in this context, there is an irresolvable problem ‘whatever System’ we embrace. It is not that freedom is inconsistent with necessity. It is that freedom is as inconsistent with Christianity as necessity. If we opt for ‘freedom’ we cannot ‘reconcile the Indifference and Contingency of human Actions with Prescience’. If we opt for ‘necessity’ (absolute decrees), we make God the author of sin. Either way there are not just ‘Mysteries’ but (contrary to the Savoyard Vicar) ‘Contradictions’ that can lead us to only one conclusion. Hume was not ‘throwing up his hands’ (Anderson, p. 41). He was telling his libertarian critics that they are just as compelled to accept the consequence. Kant would have found Hume laying out that consequence at VS Vol.3’s translation of M1751 App.1.21: Morality, being a matter of taste and sentiment, can only exist for the creatures of an impassive and imperturbable creator, not for that creator itself.

It could be that Kant misread U8.32 along the lines that Anderson suggests. But that presumes that he paid any attention to U8.32. Anderson considers Blomberg’s logic notes to be ‘good reason’ for this (42). But that requires that we accept Anderson’s reading of U8.32, without which there is no case for taking the former to represent good knowledge of even that part of U.

There remains good reason to accept Anderson’s claim that the Kant of this period could have thought of U ‘as a Pyrrhonian endeavour to suspend the authority of rationalist theodicy’ (48). Contemporaries like the author of the ‘Specimen’ (L 4–19) and James Beattie (Anderson, pp. 54–7) thought so, so why not Kant? I am just not convinced that U8.32 had any special role to play in giving Kant this impression.

Chapter 3 is devoted to drawing on what Kant said at A473/B501 to make a case that the Antinomies were inspired by Beattie’s attack on Hume (translated into German in 1772). Anderson reads the Beattie/Hume conflict as focused on whether the concept of cause is robust enough to allow us to infer that a singular event (the creation) could have had a cause. This traces ‘Kant’s awakening by the Antinomy to a source in Hume’s critique of theology’ (64) and ‘[indicates] that the transcendental idealism which Kant initially adopted to respond to Hume’s problem was confirmed by the Antinomy’ (67).

Chapter 4 brings up what Anderson considers to be an inconsistent or at least ambiguous passage, Hume’s note to U12.29. There, Hume claimed that his philosophy rejects ‘That impious Maxim of the antient Philosophy, Ex nihilo, nihil fit’ (U1750 12.29n) because his philosophy allows that, a priori, anything may cause anything else. As Anderson reads it, this note countenances the possibility of creation of the universe by a ‘free cause’ (75), that is, what Hume would have described as an agent exercising ‘liberty of indifference’. Anderson finds tension with both Hume’s rejection of the intelligibility of free causality (U8.23–5) and the argument of U11.30. But U12.29n only says that ‘the Will of the supreme Being may create Matter’ and ‘the Will of any other Being might create it’. There is no suggestion that those wills may be indifferent to ‘Motives, Inclinations, and Circumstances’ (U1750 8.23). And if they are all wills, they are of a species. There is no suggestion that the cause may be ‘of so singular and particular a Nature as to have no Parallel … with any other Cause’ (U1750 11.30). The author of U12.29n could still insist on the unintelligibility of both liberty of indifference and unique causes.

Building on its reading of U12.29n, Chapter 4 goes on to argue that Hume anticipated and influenced Kant’s solution to the dynamical antinomies (78–9). Chapter 5 observes that the ex nihilo principle was both used by Clarke to prove the existence of God and taken by Bayle to have caused Spinoza to deny it. Kant is supposed to have studied Clarke and U12.29n is supposed to be an echo of Bayle, all of which underwrites Kant’s thought in the cosmological antinomies. Chapter 6 is devoted to further specifying the roles of Clarke, Spinoza, and U12.29n in the formulation of the thesis and antithesis arguments of the antinomies. Chapter 7 takes on the question of why Kant did not present the antinomies any earlier. Anderson intriguingly argues that this final impetus had to come from D.

Anderson’s book offers profound insights into Kant’s aims and meaning in the Antinomies and into the nature of his critical project as a whole. It also offers conjectures about influence based on multiplied possibilities, extended associations, assumption of ‘echoes’ and interpretations Kant may not have shared. Readers will find a compelling case for a bold thesis on the development and core meaning of Kant’s critical philosophy even without these embellishments. If I had to offer any serious concern with Anderson’s book, it would only be that there is too much Kant in his Hume.