This paper argues that there is an important respect in which Rae Langton's recent interpretation of Kant is correct: Kant's claim that we cannot know things in themselves should be understood as the claim that we cannot know the intrinsic nature of things. However, I dispute Langton's account of intrinsic properties, and therefore her version of what this claim amounts to. Langton's distinction between intrinsic, causally inert properties and causal powers is problematic, both as an interpretation of Kant, and as (...) an independent metaphysical position. I propose a different reading of the claim that we cannot know things intrinsically. I distinguish between two ways of knowing things: in terms of their effects on other things, and as they are apart from these. 1 argue that knowing things' powers is knowing things in terms of effects on other things, and therefore is not knowing them as they are in themselves, and that there are textual grounds for attributing this position to Kant. (shrink)

This is a defense of Kant against the allegedly neglected alternative in his formulation of transcendental idealism. What sets it apart from the contributions of others who have spoken for Kant in this regard is the construction of a general interpretive framework — a reconstruction of the one Kant provides for transcendental idealism — as opposed to the development of an ad hoc defensive strategy for refuting the charges. Hence, comprehensive clarification instead of pointed rebuttal. The difference is between focusing (...) on the text and focusing on the problem. No doubt, doing both is not only possible but also required, as the problem is supposed to be in the text, but the point is that it is not there, and further, that we need not go anywhere else to show that it is not there. Thus, the approach is constructive rather than defensive, or more accurately, constructive as well as defensive. And the construction rests on what Kant actually said rather than on what he might have meant or on what he should have said instead. (shrink)