I want to address a vexed question in this essay: does Kant really need a transcendental deduction to justify his claim “man is evil by nature”? Transcendental deductions, Kant is the first to admit it, are notoriously difficult. In the case of the Religion within the Boundaries of Mere Reason, whose transcendental argument (if there is one) must be assembled through careful detective work, the difficulty is clearly compounded. I take up the gauntlet here because much of the current debate on this question is fueled, I suspect, by an insufficient grasp of the systematic character of Kant's doctrine of radical evil. Triggered by Kant's own lack of expository clarity at crucial passages, interpreters have tended to conflate the different notions of an “evil disposition” (böse Gesinnung) and a “propensity to evil” (Hang zum Bösen). A reader of the acuity of Henry E. Allison, for instance, says:

But surely this cannot be Kant's considered view. For he cannot possibly mean that the individual's choice of Gesinnung is equivalent to the species's choice of propensity. Otherwise, our personal wrongdoing would be explicated (and exculpated) by sheer membership in humanity.