Major recent interpretations of Kant's first "critique" (wolff, Strawson, Bennett) have taken his transcendental deduction to be an argument from the fact of consciousness to the existence of an objective world. I argue that it is unclear such an argument can succeed and there are overwhelming reasons to believe kant understood his deduction as having a very different form, namely as moving from the premise that there is empirical knowledge to the conclusion that there are universally valid pure categories. Detailed (...) support for this contention is offered in an analysis of the second edition version of the deduction. (shrink)