Dieter Henrich ‘s “Notion of a Deduction” (1989), opened up approaches to both Deductions in terms of legal as opposed to syllogistic reasoning. Since the CpR is shot through with juridical metaphors and analogies, many points of connection suggest themselves. In this paper, I extend and modify Henrich’s approach, in order to extract a particular logic of evidence. I argue that the three syntheses of the A-Deduction correspond to parts of a deductive procedure, and that their names have been chosen (...) to indicate this connection to the reader. Nonetheless, the principal aim of the paper is not to develop and defend these historiographical claims, but to explicate the structure of the logic of evidence in question and link it to Kant’s intended refutation of Hume. Since the procedures Kant describes are part of the law of evidence of many nations and are equally well at work in contemporary information-theory, a precise reconstruction can map directly onto contemporary problems in philosophy, physics, and informatics, without any loss of historical accuracy. (shrink)