While many of Kant’s positive doctrines in the Critique of Pure Reason are contained in the Transcendental Aesthetic and Transcendental Analytic, Kant scholars increasingly recognize that a full understanding of Kant’s critique of (theoretical) reason requires a close reading of the Transcendental Dialectic. A ‘critique’ of reason is an investigation of the limits of its cognition, and the Dialectic is the part of the Critique that reveals pure reason’s susceptibility to ‘a natural and unavoidable’ illusion; this illusion, according to Kant, will ‘not cease to lead our reason on with false hopes, continually propelling it into momentary aberrations that always need to be removed’ (A298/B354–5). The crucial task of the Dialectic is to show how we can keep pure reason in check and prevent ourselves from succumbing to the errors of dogmatic metaphysics.

Proops’ new book is an extremely welcome contribution to our growing understanding of the Dialectic and offers an exceptionally comprehensive discussion of the Dialectic’s arguments. Proops defends the following core claim about the role of the Dialectic in Kant’s ‘critique’ of reason: rather than aiming simply to debunk and discard the arguments of rational metaphysics, in the Dialectic Kant carries out a winnowing exercise aimed at uncovering the valuable core of the dogmatic metaphysician’s arguments. This winnowing exercise reveals two ways in which rational metaphysics is a partially legitimate enterprise. First, Kant discovers that theoretically grounded arguments in rational metaphysics fail to support knowledge (Wissen) of supersensible objects, but (suitably modified) they succeed in supporting certain doctrinal beliefs (doktrinale Glauben), especially concerning the soul and a wise Author of Nature. Second, they also provide arguments for Transcendental Idealism that supply Kant’s signature doctrine with even greater evidence (Evidenz) than is contained in the Transcendental Aesthetic (p. 13).

Proops’ strategy in the book is to examine each of the Dialectic’s arguments in the order in which Kant presents them in the Critique. Accordingly, the first main part of the book is devoted to Kant’s assessment of the arguments of rational psychology (in the paralogisms), the second to his assessment of the arguments of rational cosmology (in the antinomies) and the third to his assessment of the arguments of rational theology (in the ideal of pure reason). Prior to these three main sections (which span chapters 3–16 and contain impressively careful commentary on virtually all of the Dialectic’s main arguments), Proops provides three important preliminary chapters. The Introduction offers an initial defence of two core claims: first, that we should read the Dialectic as a test of rational metaphysics analogous to a metallurgical assay (Feuerprobe) and, second, that Kant advocates a kind of sceptical method according to which we should lengthen inquiry via consideration of opposing arguments (as in the antinomies), not to avoid judgement but rather to achieve certainty. In chapter 1, Proops offers an account of transcendental illusion. In chapter 2, he summarizes the development of empirical and rational psychology from Wolff to Kant. Finally, two concluding chapters provide an account of the regulative use of reason’s ideas (chapter 17) and further reflections on Kant’s conception of transcendental realism, his philosophical method and the overall success of his efforts in the Dialectic (chapter 18).

The breadth of Proops’ project makes it impossible to comment on everything of interest in his book, but scholars working on almost any aspect of the Dialectic will find in it an invaluable resource: the book is exceptionally sensitive to the historical and philosophical context in which Kant was working and includes substantive and interesting interpretative claims on a truly impressive range of issues (it is difficult to find any argument of dogmatic metaphysics discussed in the Dialectic where Proops does not offer some explanation of its failure in Kant’s eyes). In the remainder of this review, I will focus on just two parts of Proops’ analysis that I think deserve further discussion. The first concerns his account of doctrinal belief. The second concerns his discussion of how the Dialectic – and especially Kant’s account of transcendental illusion – can lead to a better understanding of the core commitments of transcendental realism.

Beginning with the first issue, recall that one of the book’s key aims is to show that the ‘fiery test’ of critique is not a wholly negative endeavour, since ‘Kant sees two branches of traditional speculative metaphysics in particular, namely rational theology and rational psychology … as in part legitimate enterprises’ (p. 5). They are partially legitimate, according to Proops, because pared-down, non-dogmatic versions of some of the arguments of rational psychology and rational theology support doctrinal belief in an afterlife and a wise Author of Nature (according to Proops, no such arguments survive in rational cosmology). How do these pared-down arguments function, and how should we understand the notion of doctrinal belief? As Proops has it, both pared-down arguments start from an assumption that is itself a case of doctrinal belief. Namely, they begin by assuming that ‘nature is suffused with purposes’, a claim justified by its indispensability to inquiry into nature (p. 419; see also pp. 185–7). As Proops argues, Kant thinks that we could decide not to inquire into nature, but if we decide to inquire into nature, then we must believe nature is infused with purposes (p. 185). Once this assumption is granted, analogical reasoning supports the conclusions that we survive our mundane deaths and that a wise Author of Nature exists. More specifically, the argument distilled from rational psychology concludes that we must survive our mundane deaths because otherwise we would be unlike other living things in expiring before we realize our potential; the argument distilled from rational theology says that nature must have a wise and powerful author because otherwise natural purposes would be unlike human artefacts in lacking a designer (pp. 184–5 and 418–19). According to Proops, these arguments do not provide demonstrative proofs of their conclusions, but they support doctrinal belief understood as ‘somewhat akin to high-credence partial belief states’ (p. 4); for the analogical argumentative form can support conclusions with ‘empirical certainty’ (p. 418).

Proops himself recognizes that doctrinal beliefs have some unusual characteristics. One such characteristic is that they are analogous to pragmatic beliefs (insofar as they are justified by a means-end relation) but nonetheless count as theoretical rather than practical. As Proops explains, doctrinal belief in the purposiveness of nature can be ascribed to each of us because ‘we would have to assent to the judgment that nature is suffused with purposes if we were to be pursuing natural scientific inquiries’ (p. 187). And this belief is theoretical rather than pragmatic because ‘it is a belief that we would have to hold if we were to pursue what is for most of us in fact a merely imagined or conceivable project’ (pp. 187–8). From these facts, Proops concludes that our doctrinal beliefs in an afterlife and wise Author of Nature are also theoretical rather than pragmatic or practical. For both beliefs ‘inherit’ their status as doctrinal from the belief in nature’s purposiveness on which they depend (p. 185), and their further supporting premises likewise do not undermine their status as theoretical. Doctrinal belief in an afterlife ‘rests on the empirically-justified theoretical belief that human beings die before they realize their potential’ (p. 188). Doctrinal belief in a wise Author of Nature depends on the empirically grounded theoretical belief that there are designed things (artefacts) which things in nature resemble insofar as they too seem to be designed (p. 419).

Here are three concerns about this picture. The first is that Proops’ argument seems open to the objection that belief in the purposiveness of nature would be pragmatic rather than doctrinal for those actually engaged in scientific inquiry. As Proops argues, the reason belief in the purposiveness of nature is theoretical rather than pragmatic for most of us is that it serves a ‘merely imagined or conceivable’ project (p. 188). But if this is correct, then it seems that for the natural scientist, belief in nature’s purposiveness is pragmatic. After all, for the natural scientist, it is justified by its indispensable role in her actual project, not her merely imagined one. From this, however, it would seem to follow that belief in afterlife and a wise Author of Nature also is not doctrinal for the natural scientist (rather, it is pragmatic). At least on the face of it, this presents a challenge to Proops’ core claim that doctrinal belief in an Author of Nature and afterlife survives as unqualified achievement of theoretical argumentation in the Dialectic.

Second, related to this, given that our two doctrinal beliefs are partly empirically grounded on Proops’ account, there is a case to be made that they are not pared-down versions of arguments found in rational psychology and rational theology at all; rather, they are different, empirically grounded alternatives to the arguments of rational psychology and rational theology. Admittedly, Proops himself explicitly takes aim at this traditional conception of rational psychology and rational theology, arguing that ‘the arguments Kant endorses for the doctrinal beliefs in a god and an afterlife … are non-a priori arguments that nonetheless belong (on Kant’s conception of these disciplines) to rational theology and rational psychology respectively’ (p. 453). But it seems to me that if this is correct (and rational metaphysics includes empirical arguments), then pressing questions remain about Kant’s understanding of pure reason and the scope of legitimate criticisms of its proper use. Is analogical reasoning (such as Proops says supports Kant’s doctrinal beliefs) a mode of reasoning belonging to pure reason? Does a critique of pure reason require scrutiny of reason’s employment in other empirical argument forms? Questions such as these could have received more attention in Proops’ discussion, given the importance of doctrinal belief to his overall account.

Third, and finally, it would have been nice to hear more about why belief in an afterlife and a wise Author of Nature are the only two doctrinal beliefs that survive the Dialectic’s ‘fiery test’. Proops acknowledges that doctrinal belief in the purposiveness of nature ‘is not formed on the basis of observation or other evidence: its justification consists in the fact that it is an indispensable means to a contingent end’, namely, scientific inquiry (p. 419). But Kant seems to believe that other regulative principles and ideas play this same role in inquiry, and Proops himself allows that the Ideas of reason in general ‘serve as indispensable tools in natural-scientific inquiry’, a fact that ‘constitutes their justification’ (p. 424). Given this, why are there not doctrinal beliefs corresponding to each idea or regulative maxim? Though the breadth of his project imposes unavoidable restrictions on how much it is feasible to say on any topic, this question seems especially deserving of further attention (cf. Chignell Reference Chignell2007, for the suggestion that doctrinal beliefs are more numerous than Kant explicitly acknowledged).

The second main comment I would like to make concerns Proops’ account of transcendental illusion and its relationship to our understanding of transcendental realism. Proops recognizes that Kant gives different glosses on transcendental illusion throughout the Dialectic, and he points out (correctly) that Kant does not always explain how these different glosses relate to one another. In particular, Proops argues that Kant’s core account of transcendental illusion cashes out illusion in terms of mistaking something subjective (like a principle governing theory construction) for something objective, i.e. a principle describing the world (pp. 46–9). But as Proops argues, Kant also suggests that transcendental illusion is identical to the appearance that transcendental realism is true, and this conception of illusion is the dominant one in the antinomies (pp. 212, 259 and 333). As Proops writes, ‘[t]he phenomenon of antinomy is the appearance that reason is inherently contradictory’, and this leads to the discovery of a transcendental illusion, ‘which, in the context of the antinomies at least, [Kant] identifies with the illusion that Transcendental Realism is true’ (p. 212). While this might lead one to think that Kant’s account of illusion is too disjointed to form a coherent picture, Proops argues that the picture is brought into harmony by the fact that transcendental realism can be understood as equivalent to the view that the sensible world is ‘given as an absolute whole’ (p. 248, n. 7; see also p. 253). According to Proops, this conception of the sensible world generates the antinomies (or at least the mathematical ones) and it explains why transcendental realists are especially susceptible to regarding subjective principles as objective ones. So Kant’s two characterizations of illusion can be unified.

Rather than delving into the details of Proops’ argument about the coherence of Kant’s account of transcendental illusion (but see pp. 257–61 and 270–1), I want to raise some questions about his conception of transcendental realism (and, correspondingly, transcendental idealism). In particular, although Proops stresses early on that the antinomies can teach us ‘a great deal about what Kant takes the commitments of Transcendental Realism to be’ (p. 33), it is sometimes difficult to tell exactly what kind of interpretation of transcendental realism and transcendental idealism he takes his discussion to require. For instance, should we take the transcendental realist’s claim that the sensible world is ‘given as an absolute whole’ as a metaphysical claim or an epistemic claim? Likewise, is the transcendental idealist’s assertion that the sensible world is given as a merely ‘relative’ whole epistemic or metaphysical (p. 271)? Proops also devotes relatively little attention to how this characterization of idealism relates to other ones. For instance, he says elsewhere that transcendental idealism is the doctrine according to which appearances are ‘partly constituted by one or more of the two forms of human sensible intuition’ (pp. 261–2), but he does not explain whether or how it follows from being partly constituted by the forms of intuition that appearances are a merely relative whole. Similar questions are left open about how different characterizations of transcendental realism elucidate the epistemic overreach characteristic of dogmatic metaphysics. To point to just one question of this sort, Proops says both that transcendental realism ‘entails’ the possibility of knowledge of things in themselves (p. 458) and that transcendental realism asserts that ‘the world exists as an absolute totality’ (p. 56). Does the possibility of knowledge of things in themselves follow from the absolute totality claim, or is it a consequence of a more fundamental characterization of realism? Once again, although it is perhaps unreasonable to ask Proops to do even more in a book that is already so ambitious, it would have been a helpful addition had he explained more directly exactly which account of transcendental realism and which account of transcendental idealism emerge from the Dialectic (especially given that the Dialectic provides such fruitful territory for understanding how the epistemic and metaphysical components of idealism and realism relate to one another).

Let me close by again emphasizing that Proops’ book covers an extraordinary amount of ground, and as such, I have been able to comment only on a small fraction of it. Among numerous other topics, I have not been able to comment on such interesting (and occasionally controversial) theses as the following: (i) that Kant embraces a methodological thesis he (mistakenly) takes to be a form of Pyrrhonism (pp. 15–29); (ii) that transcendental illusion helps empirical inquiry (pp. 450–2); and (iii) that thinkability and logical possibility must come apart for Kant, since transcendental realism is (on Proops’ reading of Kant) logically impossible but nonetheless thinkable (pp. 459–60). I hope readers will explore these and other themes in Proops’ extremely valuable contribution to the scholarship, which is, as far as I am aware, the most comprehensive exploration of the Dialectic currently available.