Despite its brevity, Stephen Howard’s Kant’s Late Philosophy represents a very significant contribution to advanced scholarship on Kant’s metaphysics of nature in general and the Opus postumum in particular. At the same time, non-specialised readers with no previous knowledge of these dimensions of Kant’s thought will likely endorse the following assessment of the book’s use value: it is easily the most serviceable introduction to the Opus postumum available in English (or, arguably, in any other language) to date. As such, it is an eminently worthy addition to the Cambridge Elements series.

The book’s introduction briefly surveys the editorial history of the bundle of thematically related manuscripts, composed by Kant between 1786 and 1803, that we conventionally call the Opus postumum. Howard also provides a compressed summary of the two main styles of interpretation that have informed scholarship on the Opus postumum over the last 125 years: the systematising approaches characteristic of commentaries from the 1890s to the 1970s and the ‘historical’ approaches taken in the wake of Burkhard Tuschling’s groundbreaking study of the Opus postumum manuscripts written before the middle of 1799 (Tuschling Reference Tuschling1971).

Chapter 2 is primarily concerned to establish the proper relationship between Kant’s self-confessed problem of a ‘gap (Lücke)’ in his system of natural philosophy and the notion of a mediating science designated by the term ‘transition (Übergang)’, i.e. the scientific undertaking whose basic stated task in the Opus postumum is to mediate between the a priori principles of a special metaphysical science (namely, principles of the type presented in Kant’s Metaphysische Anfangsgründe der Naturwissenschaft [1786]) and the empirical part of physics. Traditionally, the notions of ‘gap’ and ‘transition’ in Kant’s late philosophy had been regarded as two sides of one and the same theoretical coin until Eckart Förster challenged this kind of account of this relationship as an unjustifiable dogma of Kant scholarship (Förster Reference Förster1989, Reference Förster2000), thus opening the way for the plethora of often conflicting interpretations of the gap/transition problematic that we encounter in the more recent secondary literature. After carefully considering the uses of ‘transition’ in Kant’s major published works and correspondence between 1781 and 1796, however, Howard argues that the interpretive controversies can be resolved if we take care not to conflate two distinct meanings of ‘gap’ when this term is used with reference to the Opus postumum’s varying characterisations of the transition project, which also puts us in a position to understand exactly what Kant wanted to achieve when pursuing this project to its destination.

Chapter 3 traces this pursuit across all phases of Kant’s concern with the problem of transition in the Opus postumum manuscripts. Focusing on ‘the question of the unity or disunity of Kant’s project’ (p. 24), Howard maintains that the crucial unifying thread is furnished by the ‘stable form’ (ibid.) of the problem itself, i.e. the sheer persistence with which Kant ceaselessly attempted to make the transition from the metaphysical foundations of natural science to physics. We must indeed recognise that Kant’s use of the phrase ‘metaphysical foundations of natural science’ in the Opus postumum’s assorted drafts no longer straightforwardly aligns with the systematic account of foundational principles provided by the Metaphysische Anfangsgründe of 1786. Yet the core meaning of ‘transition’ remains stable despite the multiple variations in thematic emphasis that we encounter when following the development of Kant’s late philosophy of nature into the final fascicles of the Opus postumum. Having clarified this key feature of the work’s underlying systematic stability, Howard comes to grips with the ‘major shift’ that ‘takes place with regard to Kant’s conception of physics in fascicles 10/11’ (p. 28), i.e. the decisive shift which represents the ‘arrival point’ (ibid.) of Kant’s entire transition project.

In Chapter 4, taking issue with influential views in the secondary literature, Howard argues that Kant works with a new conception of physics in Fascicle 10/11, a conception whose scope is expanded in a way that emphasises the ‘subjective side’ of physics by incorporating ‘psychology in a broader notion of “physiology”’ (p. 41). This means that ‘[b]y the time he is writing fascicles 10/11, Kant’s conception of physics has broadened so that it is indistinguishable from physiology’ (p. 42), which in turn leads to the view of physics as ‘a science that treats the sum total (or complex) of not only moving forces but also perceptions, that is, representations accompanied by consciousness’ (ibid.). Moreover, this view of physics also involves the transition project’s portrayal in cosmological terms since the whole of such perceptions is, for Kant (see OP, 22: 308.19–20), ‘combined and connected under a principle into a world-whole (Weltganze)’ (p. 53). Howard therefore holds that the new conception of physics at work in Fascicle 10/11 points towards the role played by the concept of ‘world’ in the ‘system of ideas’ that Kant treats in the last-written manuscript of the Opus postumum (i.e. Fascicle 1), thus making his final reflections on ideas of reason ‘a comprehensible development in the transition project’ (p. 55).

Howard’s concluding chapter retrospectively frames the overall argument presented in Chapters 1–4 and then considers a number of factors motivating Kant’s transition project during his final years. The Appendix helpfully provides a description of Kant’s writing process when composing the Opus postumum manuscripts, followed by an annotated table that clarifies the chronological order of pages in the two volumes (21 and 22) of the Prussian Academy edition text and, finally, a note on the currently available editions and translations of this opus. Apart from providing the relevant bibliographical details on these sources, the extended (though not exhaustive) list of references at the end of the book covers secondary sources in English, German, and (in one instance) French that will be of primary interest to those approaching the Opus postumum with serious scholarly intent.

The primary topic of the following comments will be the conception of physics that, as Howard argues, represents the arrival point of Kant’s transition project in Fascicle 10/11.Footnote ¹ Howard’s account of this new conception is indeed persuasive, especially his position that Kant’s systematic turn towards the subjective side of physics indicates a fundamental parting of ways with respect to the various classificatory concerns that underlie his attempts to determine the content of his transitional science in earlier phases of the Opus postumum. That said, however, we still need to assess the ground-level philosophical plausibility of the conception itself before we can work out its full significance and definitively establish its broader systematic import for Kant’s late philosophy of nature.

While this task goes far beyond what can be accomplished in a short book review, we can at least get a clear sense of what a detailed assessment will require by focusing on a key passage in Fascicle 10/11 that Howard quotes (p. 46) in support of his interpretation of Kant’s conception of physics:

Consider, then, ‘the moving forces directed toward the subject, which are the cause of perception’ (lines 4–5) and thus the a priori determinable forces which affect perception ‘as anticipations of sensible representation in empirical intuition’ (lines 6–7). Since sensible representation in empirical intuition is perception, i.e. ‘empirical representation with consciousness’ (lines 1–2), the a priori determinable moving forces by which sensible representation in empirical intuition is anticipated must be forces that can bring about consciousness as ‘one in which there is at the same time sensation’ (cf. A165/B207). (I say ‘must be’ because consciousness in which there is at the same time sensation is the definition of perception specifically relevant to the first Critique’s Anticipations of Perception, the transcendental principle of which is quite transparently in play in the Opus postumum passage here at issue.) Now if this is the case, then we are presented with a number of interesting issues concerning the architectonic coherence, not to mention the very nature, of Kant’s theory of knowledge when we take account of the causal efficacy of the subject with respect to the object of outer sense and its a priori determinable moving forces. Two fundamental questions emerge in view of what Kant says about the forces just mentioned.

According to the line of argument pursued in the quoted passage, the moving forces of the object of outer sense furnish conditions of outer perception because they are causal conditions of empirical representation with consciousness (i.e. consciousness in which there is also sensation). As such, they are clearly a priori cognisable conditions just because they are determinable as conditions without which perception of the outer-sense object would not be possible. Moreover, it is in relation to this epistemic function of these perception-causing forces, i.e. the outer-sense object’s forces directed towards the subject, that the perceiving subject specifies a priori forms of action of its moving forces as those forces whose representation is identical to that of perception (lines 9–10), i.e. empirical representation with consciousness (lines 1–2). In the passage under consideration, then, this can only mean (or at least ought to mean) that the representation of these a priori specifiable forces of the subject is the same as the empirical representation which comes about when the subject acts upon the object of outer sense, thus representing this object in appearance with the subject-directed, a priori determinable moving forces that are causally responsible for its perception as an object of empirical intuition. Hence, the two basic, and closely related, questions raised by Kant’s late conception of physics: (1) How would it be possible to distinguish between a priori and empirical cognition – and, by implication, between the a priori and empirical parts of physics – if perception is understood in the way that Kant in fact understands it in the context of Fascicle 10/11 (namely, as consciousness involving sensation caused by the outer-sense object’s a priori determinable moving forces the representation of which is identical to that of empirical representation with consciousness)? (2) How can we make sense of the notion that (sensation-causing) moving forces of matter are a priori determinable conditions of perception, if (i) the representation of such material conditions is the same as the sensible representation caused by the subject acting on the object of outer sense and (ii) this subject-caused sensible representation is the same as empirical representation with consciousness caused by moving forces of the object of empirical intuition? (Needless to say, there is a lot going on here between the inside and the outside of the subject’s inside-out and outside-in sides, both inside outwards and outside inwards. The underlying problem, of course, is to sort out just what is going on with Kant’s argument without spinning the subject’s outside-inside side(s) off and away into the far reaches of space causally determined by the moving forces of matter (i.e. cosmic space regarded as a universal continuum of matter-constituting moving forces) or else collapsing all sides together into the subject’s self-caused representational activity, thus ending up with a form of either transcendental realism or empirical idealism. Kant was well aware of the problem (see, e.g., OP, 21: 439.2–443.6, 467.10–469.13, 492.27–493.16, 504.18–505.1).

The first of the two questions posed above raises the spectre of an ‘assimilative transition’ between the a priori and the empirical levels of our scientific knowledge of corporeal nature. If the true destination of the Opus postumum’s transition project is, as Howard persuasively argues, Kant’s conception of physics in Fascicle 10/11, then it seems to be one that requires the absorption of a theory of the a priori necessary conditions of empirical cognition into an account of empirical physics itself.

The second question ultimately leads to an even more basic set of issues. How could Kant’s Fascicle 10/11 portrayal of the conditions of perception furnished by moving forces (i.e. the sensation-causing moving forces whose representation is identical to sensible representation caused by the subject’s effect on the object of outer sense) be made consistent with a theory of a priori cognition involving the programmatically pivotal assertion put forward in the very first sentence of the first Critique’s Transcendental Doctrine of Elements: ‘In whatever way and through whatever means a cognition may be referred to objects, that way through which it is immediately referred to objects, and at which all thought as a means aims, is still intuition’ (A19/B33)? If intuition related to an object through sensation (i.e. empirical intuition; see A20/B34) is taken to depend on the existence of a priori determinable moving forces, then how can this sort of representation plausibly be said to be immediately referable to its object, even if the representation of those sensation-causing forces is identical to sensible representation caused by that very same object when the subject self-affectingly acts upon it? And if the relation between empirical intuition and ‘object’ is necessarily a mediated relation by virtue of that causal dependency on a priori determinable moving forces, then what exactly would be the point of the distinction between ‘intuition’(as representatio singularis) and ‘concept’ (as a discursive representation) that Kant had employed when accounting for the role played by the understanding’s categorial functions in our cognition of objects of empirical intuition? Put differently, what could be the point of Kant’s classic distinction between the discursive character of human understanding and the nature of an intuitive intellect if there can be no non-mediated relation between intuition and object in the case of intuition through sensation caused by a priori determinable moving forces?

This last line of questioning, I take it, is especially germane to one’s choice of interpretive approach to the Opus postumum’s later manuscripts for two main reasons. First, Kant’s theoretical focus shifts well away from his new conception of physics in the Opus postumum manuscripts composed after Fascicle 10/11. Second, while there is good reason to think that Kant, during the 1790s, would have been willing to refine his treatment of the intuition/concept distinction’s metaphysical ramifications – see, e.g., his reply to J. S. Beck of 3 July 1792 (Br 11: 347) – there is no indication in the Opus postumum that he was prepared to rethink the way in which he had drawn and employed the distinction itself from the 1780s onwards.Footnote ²