摘要：Subjectivists, taking the “collapse” of the observation-interpretation contrast much too seriously, are led to imagine that even perceptual knowledge is active. And therefore subject dependent. Turning the tables on this popular trend, I argue that even conceptual knowledge is passive. Kant’s epistemology is conceptual. But if also active, then incoherent. If synthetic a priori truths are to follow upon our mental activity, they were neither true nor, far less, a priori before that activity. “A priori” and “active” are contradictory attributes of knowledge. As, indeed, are “a priori” and “subject-dependent” to begin with. Nothing a priori can be dependent on anything except itself, and least of all on the human subject. Kant does consider the active aspect of thought. The difference is that for him the more active it becomes, the less it is to be trusted. For we are no longer in the province of the Understanding, and its necessary truths, but in the realm of Pure Reason and its dialectical antinomies. Cognition activists who take a liking to Kant have simply mistaken Reason for the Understanding. And Reason is to Kant “the seat of all transcendental illusion”.
