While it is not obvious from the main title, the focus of this book is squarely on Kant’s transcendental deductions (hereafter simply Deduction) of the categories in both the 1781 and 1787 editions of the Critique of Pure Reason. The reference to a circle in the title follows from Séguy-Duclot’s interpretation of the function of the Deduction as the foundation of knowledge in the human subject. He thereby understands Kant as pursuing a similar task to Descartes with the cogito. Insofar as Descartes regarded the latter principle as insufficient to ground any claims about being, the subject must resort to a second principle, that of a non-misleading God who ensures that the epistemological evidence reveals truths about being. But the reliance upon a second principle would appear to amount to a circularity making the Cartesian project vulnerable to the sceptic.

Kant’s project, insofar as it abandons the ‘ontological project of traditional metaphysics’ (p. 9), is set up to avoid such circularity by grounding knowledge exclusively upon the subject. At its heart lies the Deduction, which must ‘prove the legitimacy’ (p. 21) of the categories as conditions of the possibility of objective experience. Central to Séguy-Duclot’s interpretation is a focus upon the role of the imagination. Absent from the Duisburg Nachlass (dated 1775), the imagination forms a crucial part of the A-Deduction but is again rather ‘occulted’ in the B-Deduction (p. 23). Neo-Kantian interpreters such as Hermann Cohen view this faculty as addressing psychological questions irrelevant to the main thrust of the Deduction, but Heidegger on the contrary understands the imagination, together with temporality, as key to transcendental subjectivity (p. 23). This interpretative bifurcation is reflected in divergent attitudes among scholars to the dichotomy of the so-called objective and subjective deductions. In the preface to the first edition of the Critique, Kant writes that while the main task of the Deduction is showing the objective validity of the categories in that objects cannot be conceived other than by presupposing the categories, the Deduction will also shed light upon the very possibility of the understanding (Axvi–xvii). But the operative question here is: is this second task, that is, the subjective deduction, a mere speculative add-on (Cohen) or rather essential (Heidegger) to the Deduction, despite what Kant seems to be saying? All depends upon whether it is understood as merely psychological or of transcendental import. Séguy-Duclot argues for the latter interpretation insofar as it is required for the objective reality of the categories (pp. 28–30) and indeed for the chapter on schematism (p. 32), while not siding with Heidegger who thinks that the objective deduction is thereby made superfluous – that is, both deductions have their essential function within the whole (p. 33). In fact, Séguy-Duclot argues that the subjective deduction is foundational to the objective one (p. 164).

In analysing the three syntheses which form the core of the subjective deduction in the A-Deduction – the synthesis of apprehension, of reproduction, and of recognition – Séguy-Duclot sheds important light upon the synthesis of reproduction (pp. 45ff) whose structure is misleading insofar as Kant begins by describing the empirical synthesis, whereas all he needs is the pure one which is to ground the pure synthesis of apprehension. To explain how it grounds it, Séguy-Duclot draws upon the Axioms of Intuition and the description of the process of generating a line. A further problem with Kant’s inclusion of the empirical synthesis of reproduction is that while the example he takes pertains to the connection of a heterogeneous manifold (e.g. events in time), what is at stake here can only be the composition of a homogeneous manifold (p. 50). However, in concluding upon the pure synthesis of reproduction, Kant deliberately focuses upon the latter (mathematical) case, insofar as the concept of objectivity has not yet been introduced (p. 54). The latter concept is however essential to the dynamical case which requires the introduction of objective time. This leaves the dynamical synthesis of reproduction unexamined at this stage: Séguy-Duclot thus claims a distinction between the treatment of the mathematical and dynamical syntheses, although this is not spelled out by Kant. The synthesis of recognition leads to the introduction of the unity of consciousness as necessary for the unity of the synthesis (A106–7). Séguy-Duclot further argues that the unity of consciousness is so far merely subjective: to get to the transcendental unity of apperception, the notion of object must first be introduced at A105 as a condition of rule-governedness (pp. 77ff.). And with this, Kant can now consider dynamical connections and thereby provide the missing ground for the dynamical synthesis of reproduction and the law-governedness of phenomena (pp. 87ff).

Séguy-Duclot’s reading of the structure of the A-Deduction is one of the intriguing aspects of the book, of which scholars should take note. There is hardly any scholarly consensus about the proof structure of the A-Deduction, but it is noteworthy that unlike most readings, for Séguy-Duclot the subjective deduction is not just the second section of the Deduction, running from A98 to A111 or as some argue until A114, but crosses over into the third section until A119 (pp. 67–71), while the actual objective deduction of the categories only takes effect well into the third section (towards the end of A119) (p. 113). It is at A119 that the faculty of understanding is first introduced, where it is defined as the relation of the unity of apperception to the imagination, and since it is thereby related to sensibility, this is where the subjective deduction ends. Séguy-Duclot’s reconstitution of the structure of the A-Deduction clearly shows the intertwining of the subjective and objective deductions which is to be expected insofar as the ground of objectivity lies in the unity of apperception (p. 114). Kant speaks of two deductions merely provisionally for the deduction will have been completed only at the end.

One of Séguy-Duclot’s principal claims is that the principle of objective deduction is to be found in the subjective one. The subjective one ‘becomes’ objective (p. 114). To illustrate the importance of acknowledging the exact place of the juncture between the subjective and objective deductions in the A-Deduction, Séguy-Duclot addresses two historical interpretations that reflect the diverging readings of the relation between the subjective and objective deductions. On the one hand, Cohen struggles to ensure that the imagination’s productivity belongs to the understanding and turns to idealism in rethinking the role of the given in Kant’s theory: givenness is an activity of sensibility upon the matter of phenomena (pp. 129ff, 135). By contrast, Heidegger introduces a notion of pre-giving of this matter in space and time, again an activity of sensibility (pp. 144–5). As an alternative, Séguy-Duclot proposes that the subjective deduction leads to the ‘principle of the necessary unity of synthesis’ which expresses the ‘necessary reciprocal presupposition of the unity [of apperception] and synthesis’ (p. 156). The unity of apperception and the synthesis of the imagination thus condition each other ‘reciprocally’. From Séguy-Duclot’s perspective, Cohen’s error is to take the unity itself as ground, leaving no room for the distinction between unity and synthesis since the latter is resorbed under the understanding, and Heidegger privileges synthesis, thereby making it impossible for him to appreciate the independence of the unity of apperception with respect to temporality (p. 157).

However, Séguy-Duclot’s verdict is that the relation between unity and synthesis is problematic insofar as the very units which are brought together by the imagination’s synthesis of apprehension have a unity which could not, on pain of regress, arise from a synthesis under the unity of apperception; this problem arises ultimately because of the unresolved status of the imagination between the understanding and sensibility flagged at the outset (pp. 168–73). The B-Deduction must therefore address these issues if the Deduction is not to flounder in a vicious circle between unity and synthesis.

Although the threefold synthesis no longer features in the B-Deduction, intriguingly, Séguy-Duclot submits that Sections 15–17 in the B-Deduction effectively amount to a subjective deduction, with the ‘possibility of the understanding’ established at B137, where similarly to the A-Deduction the subjective deduction ends. But, as Séguy-Duclot observes (pp. 180ff), the subjective deduction seems much simpler here than in the A-Deduction, and this is because Kant abstracts from the way in which the manifold is given. What is given in sensibility is thus considered from a purely logical point of view, whereby compared to the A-Deduction account of the subjective deduction temporality no longer plays a role and the understanding is solely responsible for the synthesis. It follows that the reciprocity between unity and synthesis is now no longer between transcendental apperception and the imagination as in the A-Deduction but between apperception and the understanding, two faculties that are homogeneous and even identifiable (pp. 187–9). Whereas in the A-Deduction, the understanding still served as ‘third intermediary term’ between unity and synthesis, as a ‘principle of necessary unity of the synthesis’, in the B-Deduction, this latter principle is identified as the synthesis itself: ‘the understanding becomes identified with the synthetic unity of apperception’ (p. 188).

The equivalent of the objective deduction in the B-Deduction extends from the third paragraph of Section 17 to the end of Section 21. Séguy-Duclot argues that the B-Deduction is incomplete insofar as human knowledge requires taking into account how the manifold is given to our particular type of sensibility, that is, a spatiotemporal one (pp. 197–8). Here, it is somewhat disappointing to see that the author does not engage with the multifarious interpretations (e.g. Allison Reference Allison2004; Brouillet Reference Brouillet1975; Evans Reference Evans1990; Henrich Reference Henrich1969) that address the way in which the so-called ‘two steps’ of the B-Deduction are related, if the second step is not to be seen as dispensable insofar as it can just be inferred from the argument of the first step. Nevertheless, Séguy-Duclot converges with Allison (Reference Allison2004) in understanding the objective deduction as establishing the objective validity rather than the objective reality of the categories, that is, their validity for objects of thought rather than the knowledge of real objects (pp. 202ff).

The issue of not specifying in what way the second step is not merely based on an analytical inference from the argument in the first step appears to resurface when Séguy-Duclot describes the key type of synthesis of the second step, namely, the figurative synthesis, which is responsible for the ‘knowledge’ (Erkenntnis) of an object, as a ‘specification’ of the intellectual synthesis, which is concerned with the ‘thought’ of an object only (pp. 207–8). If the knowledge of an object is just a specification of the thought of an object, why could it not be inferred from the first step once the type of sensible intuition, that is, a spatiotemporal one, was specified? But it is important to see, as Séguy-Duclot indeed argues (p. 212), that the intellectual synthesis of the first step is not just any synthesis but is already a synthesis that requires a sensible intuition; it is merely the case that in the first step the argument abstracts from the specific human intuition, as Kant himself indicates at B145, in the transitional Section 21. This abstraction is subsequently lifted in the second step in that it is now shown how the synthetic unity of apperception which defines the concept of an object is applied in human sensibility itself, namely, to the human-specific forms of space and time. It is thus, Séguy-Duclot argues, that the synthesis of the imagination is a ‘specification’ of the ‘non-sensible’ (p. 208) intellectual synthesis in sensibility.

Séguy-Duclot’s focus is however rather upon the change in Kant’s theory of imagination, with its synthesis now serving the understanding, while in the A-Deduction it was independent of it (p. 209). The conclusion of the objective deduction follows insofar as the a priori synthesis of apprehension is a figurative synthesis, thereby guaranteeing that all synthesis of apprehension conforms to the categories (pp. 215f). This involves an account of the unities of space and time as presupposing synthesis as Kant claims in the footnote to B160–1, which has generated much debate. Séguy-Duclot examines this by stressing the Cohen/Heidegger interpretative bifurcation (pp. 217–20) which he sees as reflected in the ‘Anglo-Saxon literature’. It also involves the two examples Kant uses at B162–3 of which Séguy-Duclot shows the key role in relation to the unity of time (pp. 224–7).

For Séguy-Duclot, the B-Deduction as a whole is incomplete however as it only shows the possibility of the understanding taken in a narrow sense as a faculty of concepts or rules which can be identified with the unity of apperception, whereas in the objective deduction Kant draws upon a notion of the understanding defined as the ‘power of cognition’, which is not identifiable with the synthetic unity of apperception but rather dependent on it (pp. 230–1). It would appear that his argument for this difference is rather flimsily based on a passage at B137, which Séguy-Duclot reads as indicating a difference between the unity of consciousness and the understanding. He takes Kant here to be arguing for a different view of the understanding that is not identifiable with the synthetic unity of apperception, in contrast to what Kant said in the preceding section of the B-Deduction (which argues for the identity of the understanding and the unity of apperception). The ‘enlarged definition’ of the understanding (p. 231) is said to avoid any dependence of the unity of apperception upon that of the imagination, and thus the problem of circularity can be addressed (pp. 231–4). Séguy-Duclot seems to believe that the act of the understanding, in its narrow sense, presupposes as ‘its own condition of employment’ ‘some empirical representation’, and thus the B-Deduction, despite the imagination being relegated to an act of the understanding, ‘does not succeed in eliminating the presupposition of the operationality of the transcendental imagination’ altogether. The imagination continues to be present but in an ‘occult’ manner (p. 234).

Now Séguy-Duclot argues that, while if a concept is assumed given judgement is determinative and the role of the imagination is thereby subordinated to the understanding, for the sake of completeness reflective judgement – when no concept is given – will also have to be considered, thus requiring a role for the imagination independently of the understanding. The problem with the relation between the understanding and the imagination is therefore not only that of an apparent circularity that was flagged earlier but also of the need to account for the pre-conceptual unities that will be synthesised under the categories (pp. 237–40). To address these lacunae, Séguy-Duclot’s interpretation takes the reader into an unexpected examination of aesthetic judgement in the Critique of Judgement (pp. 243ff).

But is Séguy-Duclot right to demand that the deduction of the categories address the question of the applicability of the categories when no concept is presupposed? To show the objective reality of the categories is to show that they apply to real objects, and there is only an object insofar as there is a concept of it, so is addressing the question of the possibility of forming such a concept a task that belongs to the Deduction? It is arguably not, in which case the B-Deduction has addressed the issue of a circle between the unity of apperception and the imagination’s synthesis from the A-Deduction by showing how the imagination is subordinated to the understanding.

Séguy-Duclot thinks not, and his excursus into the third Critique creates an opportunity for interesting novel interpretations of key sections of the Analytic of the Judgement of Taste (pp. 243–61). His overall verdict about the Deduction is that, in systematic terms, it avoids circularity but at the cost of a restricted ambition, which does not explore the sources of knowledge lying outside consciousness (pp. 272–3).

Whether one agrees with this overall verdict, there is no doubt that this book presents Kantian scholars with a novel interpretative approach which is solidly grounded in a careful analysis of especially the A-Deduction. Though on a first reading the structure of the book itself appears somewhat fragmented with lots of reprises, the overall systematicity of Séguy-Duclot’s interpretation is impressive, as several problems concerning the details of the Deduction are related to the issues he identifies in the overall structure of the Deduction.