This new Cambridge Companion is formally restricted to comments on the Critique of Pure Reason although earlier and later philosophical references are also given, especially on the historical background to (Part I), and later influence of (Part III), the Critique. The book provides full coverage of almost all the major sections of the Critique and is for that reason described as ‘the first collective commentary’ (cover) on Kant's book. Major sections which do not have a chapter to themselves include the two Prefaces, Schematism, Phenomena and Noumena, and the Amphiboly of Concepts of Reflexion.

Almost every contributor marks issues in the discussion which need further development and the same must hold for any limited review. I, therefore, outline the central, distinctive, character of each contribution, noting sometimes issues that remain open. With so much material it is impossible to offer an evaluation of the contributions, individually or collectively, but at the end I make two comments on the general character of the discussion.

Paul Guyer's editorial introduction sets the scene for the subsequent contributions with accounts of the historical background to the first Critique, of its central aims, and of its relation to the later Critiques. It contains also a summary of the way each contribution fits into the structure of Kant's work.

Two chapters on the immediate historical, rationalist (Desmond Hogan) and empiricist (Kenneth Winkler) antecedents to the Critique provide a summary of the central doctrines against which Kant eventually rebelled. Desmond Hogan examines relations between Kant's and rationalists’ accounts of the a priori. He identifies a ‘doctrine of absolute contingency’ as an item in some rationalists’ views of freedom and in Kant's thought from the 1760s through to the first Critique. He links this doctrine with Kant's account of ‘noumenal ignorance’, but notes that this needs further elaboration (38). The main point of the discussion, raising issues that recur throughout the contributions, is to demonstrate the ‘persistently rationalist orientation of Kant's mature thought’ (40), and to place a moral issue about freedom at the heart of Kant's metaphysics in the first Critique. Kenneth Winkler gives an extensive survey of earlier empiricism in Locke and Hume, rather than Berkeley. He shows how much Kant owed to those two empiricists and what his main disagreement with them amounted to. There is an account of Kant's novel, more complex vocabulary with which he enlarged the empiricists’ account of knowledge acquisition, and a classification of different kinds of empiricism against which Kant reacted. At the end echoing Desmond Hogan's contribution there is reference to a relevant ethical dimension in Kant's response to empiricist epistemology (72).

In Part II, the section which considers the sequence of Kant's arguments throughout the Critique, Lanier Anderson begins with a survey of Kant's analytic apparatus from the Introduction. There are familiar problems with Kant's attempt to separate, and then bring together in the ‘synthetic a priori’ classification, the ‘a priori/a posteriori’ and ‘analytic/synthetic’ distinctions, and these are noted along with the difficulties in Kant's argument (or non-argument, as it is represented (89)) for the synthetic a priori character of ‘7 + 5 = 12’. What is distinctive in the discussion is the claim that of the various criteria for separating analytic from synthetic judgement (containment, contradiction, information-bearing) it is ‘containment’ which has priority. That criterion is understood in terms of Kant's conception of logical concept-relations in taxonomies, and its claimed priority contrasts with a common view that the contradiction criterion is more fundamental. It allows an explanation for Kant's classification of ‘7 + 5 = 12’ as synthetic, however, on the ground that the containment criterion is too weak to license such arithmetical judgements as analytic.

Lisa Shabel provides an extensive account of the text of the Transcendental Aesthetic from the metaphysical and transcendental expositions, and especially of the role of the transcendental exposition and its relation to Kant's view of mathematics (geometry). In the concluding sections there is a reference to types of ‘mind-(in)dependence’ in relation to Kant's classification of empirical/transcendental realism/idealism, a discussion of the ‘neglected alternative’ debate, and a brief account of Kant's sizeable final section of ‘General Observations’. All of those issues are recognised as open for further discussion (114).

Paul Guyer tackles the core of the Analytic of Concepts, the Metaphysical and Transcendental Deductions. As in other contributions a summary is offered of the texts together with discussion of ambiguities in Kant's central aim. At almost every step in the given Kantian argument there is claimed to be an extensive series of inadequacies. What is distinctive in the contribution is a survey of three stages in Kant's grasp of the central notion of ‘apperception’ and its relation to judgement. Paul Guyer distinguishes the (A) version of the arguments, a transitional version on the way to (B), and the latter revised (B) version itself. In addition there is a strong emphasis on the need for additional support from the Principles section and especially the Refutation of Idealism, both considered elsewhere in the volume. There is a robust criticism of Henrich's ‘two-stage’ interpretation of the Deduction, but in the end many of the noted philosophical issues are admitted to be left unresolved (148).

Eric Watkins provides a full survey of the central part of the Analytic of Principles, that is, the texts offering proofs of the mathematical, dynamical and modal principles. There is little on the Schematism, or on the Refutation of Idealism, or on the two concluding sections, Phenomena and Noumena and the Amphiboly of Concepts of Reflexion. These are sometimes referred to elsewhere, but only the Refutation of Idealism is given a later chapter to itself. The discussion of the proofs of the principles emphasizes the scientific, Newtonian, background and involves reference to the later Metaphysical Foundations of the Natural Sciences.

Dina Emundts's chapter on the Refutation of Idealism offers a detailed formal reconstruction of Kant's argument together with a short, 3-page, concluding section on Phenomena and Noumena. What is distinctive in the account is its aim of providing a defence of Kant's argument once various ambiguities, especially in its concepts of ‘persistence’ and ‘outer’, are resolved. The defence rests on the, sometimes questioned, claim that the argument requires an appeal to the Analogies, particularly the first Analogy. It turns crucially on the question whether Kant's requirement for a persistent object in outer experience properly excludes all other options. There is discussion, and rejection, of the idea, said to be the most commonly held, that the proof is designed to establish the reality of things in themselves.

Of the five chapters on the Dialectic the first by Michael Rohlf offers an overview of the background to Kant's critical examination of particular traditional metaphysical doctrines. The discussion is sympathetic to Kant's classification of Ideas of pure reason and attempts to answer objections to the intended separation of reason and understanding and its underlying architectonic. The constitutive/regulative distinction is given prominence and the emphasis on both negative and positive aspects of Kant's discussion can be usefully read in conjunction with Frederick Rauscher's account of the conclusion to the Dialectic in its Appendix. Three chapters deal with the Paralogisms (Julian Wuerth), the Ideal (Michelle Grier) and the Antinomies (Allen Wood).

Julian Wuerth's account of the Paralogisms is distinctive in spending considerably more time on the pre-Critical background to the issues than on the two (A) and (B) versions from the Critique. The motive for this lies in the claim that Kant's position in the Critique was already extensively anticipated in the earlier works, and so marks a consistency in Kant's thought potentially at odds with his radical change of direction between 1770 and 1781. The conclusion is that Kant maintains an ‘ontologically significant’ belief in the soul as a thing in itself and so a ‘deeply engrained empathy with rationalism’ (243–4). That conclusion raises, but does not resolve, the issue of what such significance amounts to, and so of the extent of Kant's sympathy with rationalism in the Critique.

Allen Wood, in discussing the Antinomies, provides a summary of the arguments in the theses and antitheses, and of Kant's strategy in attempting to resolve the disputes, as well as a historical overview of the traditional problems themselves. At virtually every stage in the arguments for thesis and antithesis, and of Kant's own response to them, Allen Wood indicates faults and inadequacies. Some emphasis in the discussion, echoing Allen Wood's earlier work, is placed on the arguments in the third Antinomy over freedom and on Kant's resolution of its conflict. In this case Kant's central error is claimed to rest on the ambiguity of Kant's strong (committed) or weak (non-committal) account of noumenal (transcendental) freedom. The claim is that such an ambiguity, but more particularly the strongly committed part of it, contains a ‘shameful violation of (Kant's) own critical principles’ (204–5).

Michelle Grier offers a full account of the Ideal, and its discussion of traditional proofs for the existence of God, in which the central issue is represented as the connection (or lack of connection) between the ideas of a ‘most real being’ and of a ‘necessary’ being. There is a close attention to the detail of the three related arguments from the ontological, cosmological, and physico-theological proofs of the existence of God, as well as a subordinate reference to the historical background. In the conclusion attention is drawn to Kant's distinction between on one side a natural inclination (which Kant thinks understandably powerful) to believe in God via our responses to the sublime in nature, and on the other side the metaphysical attempts to provide a formal proof for that belief (which Kant regards as failures).

Frederick Rauscher's contribution sums up the conclusion to Kant's treatments of the disputes in the Dialectic and also provides a discussion of Kant's sometimes neglected account of the Canon of Pure Reason from the Doctrine of Method. He emphasizes the extent to which in the Appendix to the Dialectic Kant makes essential room for positive as well as negative responses to the debates and the central concepts in them. The contribution disentangles, on Kant's behalf, the various (scientific, religious, metaphysical) contexts in which the contentious concepts operate, and underlines Kant's commitment to their role, not as designating ‘extra-mundane’ objects, but only as rules for unifying or systematizing our grasp of human life in immanent nature (309).

Adrian Moore's account of the official second part of the Critique, the Transcendental Doctrine of Method, complains of its relative neglect but provides only a limited survey of its argument. Attention is focused on the structure of transcendental arguments, here identified through Kant's reference to ‘transcendental proof’, and on debate following Strawson's appeal to such arguments as an anti-sceptical strategy. At the centre of the discussion is the idea, associated with both Wittgenstein and Strawson, that there is something self-defeating or incoherent in Kant's attempt to ‘think both sides of the boundary’ between what we can know (appearance) and what we cannot know in principle (things-in-themselves). To resolve the issues raised by such an account is, however, admitted to be beyond the scope of the discussion (325–6).

In Part III four contributors examine Kant's subsequent influence on philosophical development in accounts of early German idealism (Rolf-Peter Horstmann), later neo-Kantians (Konstantin Pollok), Continental philosophy, especially Heidegger (Daniel Dahlstrom), and analytic philosophy (Kenneth Westphal). Rolf-Peter Horstmann's account is divided into three parts; first are some of the immediate reactions to Kant's work in Reinhold, Jacobi and Maimon, second the responses of Schelling and Fichte, and finally Hegel. It is made clear that all these responses contained, often unclearly, some aspects of Kant which the commentators thought essential and important, others which they rejected, together with the various aims of publicizing Kant's views, or rewriting them, or using them to develop their independent work. Two of the central issues involve an attempt to locate a fundamental ground, or premise, from which Kant's conclusions might be validly derived, and a related strategy which Kant either followed, or might better have followed, in refuting traditional scepticism. At each stage in the interaction between Kant and his commentators, or interpreters, there are succinct comments on the shortcomings both of the interpretations offered of Kant and of the preferred alternatives to his views.

Konstantin Pollok offers a detailed overview of neo-Kantian responses which focuses on Cohen, Natorp and Cassirer associated with the Marburg school, and Rickert, Windelband, Lask and Bauch from the South-west school. Many other philosophers associated with these groups are fitted into the picture but not extensively discussed. Two particular issues stand out throughout these responses. First is a preoccupation with the development of science and the status of scientific theories, even, among later commentators, with post-Kantian developments in mathematics (e.g. non-Euclidean geometries) and physics (e.g. relativity theory). Second is an anxiety about the relationship between metaphysics or epistemology on one side and psychology on the other. Many of the neo-Kantians are represented as no less hostile to psychologism in philosophy than Frege.

Daniel Dahlstrom considers the response of Continental philosophers to Kant's work but focuses primarily on Heidegger. Other Continental philosophers, such as Husserl, Jaspers, Merleau-Ponty and Derrida, are noted but not extensively discussed. The claim is that although these, and other, Continental philosophers were strongly engaged with Kant only Heidegger offered, like some of the neo-Kantians, an interpretation of the whole Critique of Pure Reason. Even Heidegger's response is focused specifically on his Kant and the Problem of Metaphysics though it is recognized that Heidegger himself criticized that early work, regarding it as an ‘over-interpretation’ of Kant. Heidegger's interpretation is represented as resting essentially on accounts of the imagination, specifically ‘transcendental imagination’, in Kant and particularly of the related role of schematism and its connections with time determination. Within Heidegger's phenomenological realism the central thesis is that transcendental imagination has to be the basis for all the faculties, especially sensibility and understanding, which Kant regards as fundamental.

Kenneth Westphal's account of the reaction to Kant among analytic philosophers discusses in detail three examples from three different periods of the development of analytic philosophy, namely C. I. Lewis, P. F. Strawson, and Wilfred Sellars. All three shared some views about Kant but disagreed about others; it seems clear that they do not represent one single identifiable ‘analytic’ response. All three rejected Kant's account of space and time as a priori forms of intuition. Lewis and Sellars, but not Strawson, acknowledged weaknesses in empiricism associated with Kant's criticisms. As a consequence Sellars, but not Lewis, was more sympathetic than Strawson to Kant's ‘synthetic a priori’ classification of judgements. Lewis and Sellars had a particular interest in the implications of Kant's approach for the sciences, and in Sellars's case for his contrast between the ‘scientific’ and the ‘manifest’ image, but Strawson's account focused on the more general conditions for possible experience exemplified in ordinary life. Sellars's response owes much to Carnap's influence, to a conception of ‘intentionality’, and to a more sophisticated account of mind and language than Kant offers. Kenneth Westphal's own discussion places an emphasis on the Amphiboly of Concepts of Reflexion, which does not feature elsewhere in the volume, and on Kant's conceptions of a ‘transcendental logic’ and a ‘transcendental topic’.

It will be evident, even from these brief summaries, that the Companion contains a huge amount of information and of material for further discussion. Some of the latter concerns philosophical issues, raised but not resolved, in the various contributions, but even the historical and exegetic material leaves unresolved tensions and apparent disagreements which are left for the reader to deal with. Perhaps the most striking example is that of Kant's proximity in the Critique to empiricism or rationalism. Some contributors emphasize Kant's allegiance to rationalist principles even in his mature philosophy (e.g. Hogan, Wuerth) while others stress his hostility to such principles deriving in part from his limited acceptance of aspects of empiricism (e.g. Winkler, Rauscher). Others (e.g. Allen Wood) regard Kant as simply inconsistent in clinging to both sides in the relevant disputes. Such conflicts cast doubt on the appropriateness of calling the volume a ‘collective commentary’, for it is questionable whether a ‘commentary’ can properly contain such conflicts. Conventional commentaries properly note divergences of interpretation but it is unusual for them actually to advocate them.

Most of the issues left for further discussion have a philosophical rather than historical aspect, largely because the dominant orientation of the volume is towards a historical survey of Kant's text and its background. Characteristically the contributions carefully outline Kant's text in some designated section, comment on the historical background and then note philosophical issues arising in the discussion with, however, no time for their resolution. It may be that current work on Kant emphasizes historical rather than philosophical discussion and there may be many understandable motives for this. One may reflect a continuing interest in newly published Kant texts which undoubtedly raise questions about the historical development of his own thought. Another may result from recognition that earlier commentators interested primarily in philosophical rather than historical questions sometimes exhibited a cavalier attitude towards both the text and its historical background, which needed to be corrected. That historical emphasis is, however, a limitation in the volume for those who would prefer to read more philosophical discussion, though it cannot count against the richness and interest of the material that is presented.