Kant on the Sources of Metaphysics: The Dialectic of Pure Reason by Marcus Willaschek (review)

This book is about the Transcendental Dialectic in Kant's Critique of Pure Reason (CPR). Unlike most other treatments of this subject matter, it does not focus on Kant's criticisms of metaphysical arguments. Rather, it considers Kant's account of why "metaphysical speculation about the unconditioned"—for instance, about objects like God or the entire world—"arises naturally and inevitably out of the very structure of human reason" (8).

Willaschek posits "a three-part template underlying" Kant's account of how we are led to make unwarranted metaphysical judgments: "(1) a transition from the logical to the real or transcendental use of reason and its ideas and principles, (2) the tendency to misuse the latter by treating them as constitutive rather than regulative, and (3) the tacit assumption of transcendental realism as an explanation of that tendency" (264–65). Overall, Willaschek makes a plausible case for the explanatory potency of this template as a clue to reading the different parts of the Dialectic. I want to consider two features of his interpretation that seem to me especially interesting and controversial.

With regard to (1) and (2), Willaschek suggests (107–9, 120–24) that Kant's account of why various logical principles presuppose transcendental principles, in the Appendix to the Dialectic, allows us to understand his argument in the Introduction, when he makes his fundamental move from a merely logical principle to the Supreme Principle of speculative metaphysics: an unconditioned totality of conditions exists for every given conditioned object. However, it is not clear whether one can read the Appendix argument into the Introduction. In the Appendix, Kant accepts the transcendental assumption that nature contains (some indeterminate degree of) unity as legitimate. But he denies that nature could contain the unconditioned. Willaschek argues that the legitimate transcendental assumptions in each case are only regulative (160). For Willaschek, this means that we must not take these assumptions to be true: rather, we should assume the relevant features (the homogeneity of nature, or the unconditioned) only problematically, in the hypothetical use of reason (114–25, 131–34). However, for Kant, a merely hypothetical status is, arguably, not sufficient for the transcendental assumptions he considers in the Appendix (CPR A 649–54/B 677–81). Moreover, granting Willaschek's (controversial) view that the hypothetical use of reason does not involve (any degree of) assent, making a hypothetical assumption requires at least that it might be true. But, for Kant, the assumption that an unconditioned totality exists in nature is necessarily false. Thus, it is not clear whether Kant could assign the Supreme Principle a regulative-transcendental status that is modeled on the principles he discusses in the Appendix. This raises some doubts about Willaschek's interpretation of the status and origin of the Supreme Principle.

With regard to (2) and (3), Willaschek argues that human reason tacitly identifies empirical objects with things in themselves. For Willaschek, this amounts to the assumption that empirical objects are noumena in the positive sense qua objects of a divine intellect (141–45). We conflate a regulative with a constitutive use of rational principles because we assume that nature must exhibit a rational, divinely ordained order (161–62, 240, 245). This is a plausible take on how Kant conceives the road into rationalist metaphysics. But I am not sure how it could apply to Kant's conception of the empiricist (or, as we might say, naturalistic) metaphysics that underlies materialism, universal determinism, and atheism. Kant distinguishes between two basic kinds of realism (CPR A 781/B 809; cf. A 466/B 494): one posits special unconditioned objects outside nature, the other claims that there is nothing outside nature and that the principles of nature or experience govern all things in general. Willaschek tends to assimilate the latter to the former version of realism (148, 245–51).

This worry also affects Willaschek's suggestion that one can accept Kant's critique of metaphysics without accepting his...