In Kant's Power of Imagination, Rolf-Peter Horstmann defends the bold thesis that it is the imagination, not the understanding, that "carries the main burden in the laborious process" of cognition: it is the imagination that transforms "amorphous and unstructured physiological sense impressions" into intuitions (100); and it is the imagination-this time, working alongside and sharing "equal weight" with the understanding (78)-that generates "representations of full-blown cognitive objects," i.e., cognitions, from intuitions (100). 1 This interpretation casts not just Kant's theory of imagination but also the Critical account of cognition as a whole in a new light, and it usefully counterbalances the many readings of Kant that focus narrowly on the logical processes that underwrite cognition.There is a second way in which Horstmann's book is notably bold. For it demotes, if not eschews, some of the scholarly priorities that have become the norm in Kant studies. There is no exhaustive compiling and weighing of textual evidence and counterevidence, and only highly selective engagement with the secondary literature. 2 Instead, readers are given a comprehensive (if still schematic) analysis of the role of imagination in the cognitive process, as well as an account of the development of Kant's views on the subject from the 1781 Critique of Pure Reason through the 1790 Critique of the Power of Judgment-all in just 102 pages, set with a generously sized font and written from a perspective informed by current debates in the philosophy of mind. In Horstmann's words, the project is an "immanent and a reconstructive endeavor, relying solely on Kant's own resources" (3), 3 but it is perhaps more informative to note that he readily admits to the absence of unequivocal textual evidence for some of his claims, even allowing at one point that "one has to be quite imaginative in one's reconstruction in order to find support" for his view (45; also 32, 37, 46); and that he once introduces a topic by saying he will "elaborate how Kant might be thinking about this…, or, more precisely, how I would proceed, if I were Kant" (34n31). Horstmann thus clearly signals to the reader that his interpretation is at times liberally "reconstructive," and though Horstmann takes it to remain on solid ground textually, his priorities explicitly lie with "point[ing] out illuminating consequences" of the view "instead of defending[it] against potentially recalcitrant textual evidence" (46). This approach is an unconventional one, and it makes for a book that is thoroughly original and thought-provoking, but also hard to know how to engage with. 4 Traditional forms of textual bandying, for instance, seem ineffectual, or at least contrary to the spirit of Horstmann's undertaking. The difficulty of approaching this kind