Commentary on Kant's Critique of Pure Reason has focused intensively on the transcendental deduction of the categories – the pivotal chapter of the book that governs our understanding of much that precedes it and just about all that follows it. One simple way to understand the systematic function of the Transcendental Deduction is to appreciate that it provides an account of how the ‘two stems of human knowledge’ (A15/B29) – sensibility and understanding – must relate to one another in the production of knowledge. On Kant's view, these capacities are distinguished by their radically different modes of representation: intuition and concept. Although sensibility and understanding are fundamentally distinct – they ‘cannot exchange their functions’ – they must nevertheless cooperate in the production of knowledge: ‘Only through their unification can cognition arise’ (A51/B75–6). The task of the Deduction is to show how the categories – concepts that stem from the ‘nature of the understanding’ alone – apply necessarily to objects that can only be given in experience, and represented as given through sensible intuition.

Hans Vaihinger, in the late nineteenth century, posed a now famous trilemma for Immanuel Kant's theory of affection: (1) If things-in-themselves are the affecting objects, then one must apply the categories beyond the conditions of their application (space and time). (2) If one holds that appearances are the affecting objects, then one must hold that these appearances which are the effects of affection are themselves the causes of affection. (3) If one holds that things-in-themselves affect the noumenal self in parallel with appearances affecting the empirical self, then that which is a representation for the noumenal self must serve as a causally efficacious thing-in-itself for the empirical self's production of an empirical representation of the very same object (so-called ‘double affection’).

Like Kant's Critique of Pure Reason, we now have both an A- and a substantially revised B-edition of Henry Allison's commentary to and defence of that work: Kant's Transcendental Idealism: An Interpretation and Defense (Allison 1983, 2004). Unlike Kant's Critique, however, Allison's first edition is unlikely to persist as an occasional rival to the second, since reviewers rightly consider the second to have superseded the landmark scholarship of the first. Nevertheless, before Allison's first edition is entirely supplanted, I would like to save a neglected insight from its treatment of the B-deduction that is not retained in the second edition.

Upon arriving in Auschwitz Primo Levi discovered that rational discourse, in which actions are done for reasons, was left lying on the carriage floor together with his human dignity. By responding ‘Here one doesn't ask why’, the camp guard succinctly conveys the insight that evil defies reason. This paper examines two studies of evil that are predicated on that idea: Kant's and Augustine's. It argues that their theories share an underlying formation wherein evil remains incomprehensible, except in negative terms as an absence of the reasonable. This deep similarity in the structure of the concept of evil is exceptionally striking if we bear in mind the radical difference in the scope of its application: whereas for Augustine ‘evil’ is a general metaphysical concept that concerns everything that goes wrong in creation, for Kant it is limited to agents and what they do. My argument is that the privative structure which underlies their understanding of evil stems from a shared belief in the central role of rationality in ethical discourse, and in an absolute dependence of moral responsibility on free will.

This article analyses the teleological argument justifying historical progress in Kant's Guarantee of Perpetual Peace. It starts by examining the controversies produced by Kant's claim that the teleology of nature supports the idea of a providential development of humanity towards moral progress and the possibility of achieving a cosmopolitan political constitution. It further illustrates how Kant's teleological argument in Perpetual Peace needs to be assessed with reference to two systematically relevant issues: first, the problem of coordination linked to the necessity of realizing the ‘highest good’ as a historical end of practical reason, and secondly the problem of continuity posed by the temporal limitation of all individual efforts to cultivate moral dispositions. To illustrate the implications of both issues for the teleological argument in Perpetual Peace, the article draws attention to some important developments in Kant's analysis of teleology following the Critique of Judgment.