Kant's chapter on the ‘Transcendental Deduction of the Categories’ is a central piece in the puzzle of the Critique of Pure Reason. In it Kant attempts to answer, or at least start to answer, the question of with what right or justification we apply the categories, which he connects to the question of whether and to what extent the categories enjoy objective validity. His answer consists in arguing that the categories are objectively valid with respect to all and only objects of possible experience because they are conditions for the possibility of experience. In the process Kant discusses apperception, the ‘I think’, the nature of judgment, imagination, synthesis, inner sense, space and time, and more. Just what his claims and arguments amount to in this dense and winding chapter, and whether it is in any way successful, remain topics of high controversy, further complicated by the fact that Kant saw fit to completely rewrite it between the first, A-edition of the Critique of Pure Reason in 1781 and the second, B-edition of 1787. It is the subject of a truly enormous volume of commentary stretching back to Kant's own time and still going strong.

Kenneth R. Westphal's short book on the Deduction is quite unlike anything I have read. The bulk of it consists of a new English translation of most of the chapter's B-edition version, which appears columnized alongside the original text on the lefthand page, facing Westphal's elucidations on various aspects of the text on the righthand page (§3, 12–91). This core section is bookended by two short introductory sections (§§1–2, 1–11) and two short sections at the end, concerning space and time (§4, 92–98) and modality (§5, 99–104). There is also an analytic table of contents (§6, 105–6) and a list of references (§7, 107–11), which unfortunately is incomplete (most notably, the appropriate entries for Michael Wolff's cited works are missing). There is no index but happily a searchable electronic version of the book is freely available under the creative commons licence.

To the best of my knowledge, its format alone makes this book unique in Kant scholarship. Does it work? It has several laudable features. Bilingual German-English editions of Kant, even of parts of texts, are exceedingly rare, and any such is to be welcomed. Westphal's translation is genuinely new, and it is generally lucid and accurate. I might quibble over some of the details. ‘Erkenntniß’ (occasionally ‘Erkenntnis’), for instance, is here usually translated with ‘cognition’, but ‘knowledge’ is used at least once, whereas ‘erkennen’ usually takes ‘known’, but ‘cognise’ and ‘cognize’ are each used at least once. I do not mean to take a stand on which translation is preferable, or even on whether the difference matters, but consistency is a desideratum not lightly overcome. And there are other foibles. For instance, I'm not sure what happened to ‘unerkannt’ in the following: ‘Allein Erscheinungen sind nur Vorstellungen von Dingen, die nach dem, was sie an sich seyn mögen, unerkannt da sind’; ‘However appearances are only representations of things, which are there, altogether regardless of what they may be in themselves’ (B164, 82–83). Omitting ‘objective’ from the following, yet adding ‘with little difficulty’, is probably harmless but still noteworthy: ‘Wir haben oben die Begriffe des Raumes und der Zeit vermittelst einer transscendentalen Deduction zu ihren Quellen verfolgt und ihre objective Gültigkeit a priori erklärt und bestimmt’; ‘Above we have, with little difficulty, pursued the concepts of space and of time through a transcendental deduction to their sources and have explicated and specified a priori their validity’ (B119–20, 18). And where does ‘found’ come from in this: ‘We have already found two quite distinctive kinds of concepts…’; ‘Wir haben jetzt schon zweierley Begriffe von ganz verschiedener Art…’ (B118, 18)? Preserving word order is desirable, but at the cost of inventing a participle and turning what was originally the main verb into an auxiliary? Of course one of the great benefits of this kind of edition is that one can check the original easily, and overall I thought Westphal's translation admirable and illuminating.

There are also benefits to Westphal's very conscious eschewal of commentary on commentary on the Deduction. This allows him to focus on the text itself and to keep things relatively brief, avoiding almost altogether engagement with or even references to the vast secondary literature. Of course there are also pitfalls to such an approach. One thing that students will not get from this book is any sense of even the basic contours of debates around the Deduction or the place of Westphal's interpretation within them. Indeed there is little by way of argument for Westphal's interpretation here, either as against opposing interpretations or on its own terms. Westphal is upfront about this, albeit in a way that I found condescending and unduly negative:

Westphal then closes the introductory material to the volume by saying: ‘I respectfully submit that, even with its tactical omissions, the attention devoted here to Kant's B Deduction is more thorough and careful than what has become accepted in much contemporary scholarship’ (13). Yet on one of the few occasions when Westphal turns from generalizations to engage with a specific example of ‘commentary responding to commentary’, there is little that is respectful about his tone. Having accused his subject of ‘sheer negligence’, he concludes: ‘Regrettably, these blunders indicate what are now common failures of instruction, supervision and refereeing within the field’ (94n.). I respectfully disagree. But occasional slips in tone aside, I found the manner in which Westphal's interpretation is presented refreshing, and that interpretation is certainly interesting.

According to Westphal, Kant's aim in the Deduction ‘is to show that not even the minimal incomplete thought, “I think …”, is humanly possible without using (if implicitly) the categories in some actual context on some specific occasion to at least formulate a judgment about some (presumptive) particular(s)’ (2). The general themes of the interpretation are broadly familiar from his previous books on Kant. Westphal thinks Kant's project in the Critique of Pure Reason is one of cognitive semantics, in particular the semantics of singular cognitive reference. Kant's goal is to specify the sensible and intellectual, formal and material conditions of possible human experience and judgment, with a focus on the conditions under which it is possible for beings like us to cognitively refer to, identify, discriminate, and conceptually determine or specify particulars, in a way that is apt for knowledge and science. Kant examines issues of both cognitive process and cognitive validity, and although his account of the former is extremely sophisticated, it is strictly in service of his account of the latter, which concerns ‘truth, accuracy, justification or other such normative modalities’ (15). Moreover, for Westphal, Kant's treatment of these issues is in fact ‘neutral with regard to transcendental idealism’ (3). The arguments for transcendental idealism are no good but the enduring insights of Kant's cognitive semantics are independent of these arguments and the view itself.

Westphal's take on the Deduction will be of interest to admirers of his previous work as well as anyone working on the Critique of Pure Reason (I fall squarely into both camps). Exegetically, the single most striking claim is that §26 of the Deduction chapter alone constitutes the transcendental deduction of the categories proper—the preceding thirteen sections (§§13–25) merely set the stage, circumscribing and characterising what is to be proved in the deduction proper and why it is important (e.g. 4–5). Personally, I find this claim about as plausible as the claim that transcendental idealism plays no essential role in Kant's arguments, which is to say that I do not find it plausible at all. But it is a fascinating claim—one that was entirely new to me—and it deserves serious attention.

I am not sure whether the interpretation of the Deduction that is presented in this book will be fully intelligible to those unfamiliar with Westphal's previous work, which gives a much fuller account of many of its central issues and ideas. This holds of the introductory sections (§§1–2) and the comments on the text itself (in §3), which presuppose a lot of knowledge, for instance of what exactly Westphal means by ‘cognitive validity’ and a range of ‘-isms’; it holds also of the concluding sections (§§4–5), whose discussions of Isaac Newton, Thomas Reid and Johannes Nikolaus Tetens are erudite and intriguing but rather too condensed. Westphal seems to have intended otherwise. He says ‘By design, it [this book] is independent of my other two books [on Kant]’ (1). Yet he also makes much of the way in which those previous books provide extensive and essential ‘context’ and ‘stage setting’ for the ‘re-examination’ of the Deduction that he presents in this book (1–2); and he says that ‘Those seeking grounds for my remarks should please consult the two indicated books’ (3). It is not hard to see how such claims might come into tension. Not only what Westphal says but also what he chooses to focus on in such a short text will, I suspect, only become clear in light of his bigger picture. This is not really a criticism. Those interested in Kant's theoretical philosophy and its continuing relevance should read those other books and this one as well.

However, there is another, I think far more problematic sense in which this book is not self-standing. Westphal's edition omits substantial parts of the B-edition Deduction. In my view readers would be well-advised to have a full version of Kant's text with them when tackling this one. This goes for readers of all stripes but is especially important for those who would otherwise have little idea of what they are missing. The omissions occur throughout. Sometimes they cover part of a sentence or a couple of full sentences, sometimes whole paragraphs, and sometimes whole pages. Of fifteen sections (§§13–27), only four are presented in their entirety (§§15, 20, 22, 23). The subject matter of the omitted passages is partially but not exhaustively indicated : ‘Omitted are most passages overtly concerning Kant's transcendental idealism which do not concern issues of validity, and those concerning self-affection, i.e., how we can be sensibly aware of ourselves within inner sense’ (§3, 12; cf. §2.5, 6). Of course Westphal takes himself to have good reason for his omissions. In general: ‘The elided passages certainly are of interest, but as further comments by Kant on various topics, not as central components of, nor guides to, the Deduction itself’ (§3, 1 3). But we are to take this on trust. And we are given no proper guidance that omitted material includes, for example, any mention of Locke or Hume and Kant's discussions of the inadequacy of ‘physiological’ or ‘empirical derivation’ (B118–19, B127–28); several of Kant's discussions of contrasts and parallels with space and time as to their deduction and status as forms of sensibility (B120–21, B160); and any mention of Kant's contrast between intuitive and discursive intellect (B135–36, B138–39, B145). It is far from clear whether or why these fall under the topics Westphal is explicit about omitting. Sundry other matters fall by the wayside and the effect is sometimes quite awkward.

Contra Westphal's assertion that ‘All ellipses are marked’ (12), Kant's footnotes are often omitted without proper indication. For example, the footnote at B136 is omitted without any local marker whatsoever (42), as is the footnote at B133–34, except that this time Westphal introduces a paragraph break into the main text that is not in the original (38). When Westphal does include Kant's footnotes, he does not use footnotes—they are incorporated into the main text with a paragraph break and a star, ‘*’. This sometimes but not always involves Westphal introducing paragraph breaks into the text that are not in the original. Compare pp. 32–34 (B131) and p. 54 (B144), which introduce new paragraph breaks, with p. 80 (B162) and p. 86 (B166), which do not. There is also at least one occasion on which the ellipsis marking a missing portion of the main text is missing in the English column (80). I suspect it was misplaced on the previous page, where the translation has an ellipsis but no text is omitted (78).

Moreover, Westphal omits from the opening sentences of the Transition section just the first of Kant's ‘two possible cases in which synthetic representation and its object can come together, necessarily relate to each other’, to wit, ‘if the object alone makes the representation possible’ rather than vice versa (B124). Westphal denies that Kant's two cases exhaust the options and thinks neither is in fact relevant to the Deduction, a mistake related to transcendental idealism, for he thinks that Kant's notion of objects and representations making one another possible concerns production and process rather than validity (10–11, 25). This is intriguing but the result is an edition of Kant's text that begins this section like so: ‘Ist aber das zweite’; ‘In the second case’. No mention is made of the first case except in Westphal's commentary. Or again, when Kant returns to his dichotomy in the final numbered section of the Deduction (§27), raising and rejecting ‘a middle way’ that would be ‘a kind of preformation system’, Westphal sees fit to omit Kant's set-up of the issue altogether, both the restatement of the dichotomy and the introduction and characterization of the supposed third option. Westphal's commentary takes note but does not elaborate. And the result is an edition of Kant's text in which the second paragraph in the main text of §27 opens like this: ‘that in such a case [of a system of preformation] the categories would lack necessity’ (sic). No mention is made in this edition of the text of what such a system is. As an edition of the text, this is not only incomplete—it looks and feels incomplete.

This book has limitations but there are insights aplenty and it is always engaging. It is a very welcome addition to the bold and sophisticated interpretation of Kant's theoretical philosophy that Westphal has been developing in print for the best part of three decades.