I think at this point I understand all the transcendental arguments of CPR except this one - and probably this could considerably change my understanding of Kant as a whole.
Here is my confusion.
- Intuitions are a priori.
- Concepts are a priori.
- This might be a bit contestable but Kant in various locations does state pure appearances exist (intuitions without concepts). But they serve no value for us since it makes no judgement. There cannot be a true/false value attached to it, since it's not a judgement.
- However, we DO apply concepts without intuitions. This is the mistaken field of metaphysics as explicated in Transcendental Dialectic.
Now, if 1 and 2 are true, and Kant can grant 3 (however, this is not crucial to my argument), with what exact reasoning can he state that 4 is 'empty'. I can understand if he says they are 'useless'. But in the transcendental dialectic, his position is more than these metaphysical statements being useless - it is that they are incorrect/erroneous. His position is that it is a mistake to use concepts without intuitions.
For this, in my opinion, he will need to elucidate that concepts CANNOT exist without intuitions, i.e, their content is derived from intuitions. However, it cannot be 'derived' by definition, since they are a priori. On top of that we know they do exist independently in the field of metaphysics - Kant himself claims this.
So while, as a Nietzchean perhaps, I can grant Kant that concepts applied to things that cannot be present in intuition is probably useless (this is an ethical argument, not an epistemological one), I still cannot grant Kant an ontological reason for why I should not use concepts without intuitions in my discussions (which is what metaphysics did before him).
And yes, this argument is a part of Transcendental Deduction, but it's (a) quite unclear and (b) not satisfactory in my opinion because just because categories are a presupposition of experience which also involve intuition in no way mean applying them without intuitions is illegitimate.
I realized this question can be posed in a different way. (This is partly due to the fact the answer might just resort to the fact that synthetic a priori judgements require predicates that we can only get through intuitions).
- Are metaphysical judgements synthetic a priori? If not, how are they possible. They are not analytical, of course.
- If they are synthetic a priori, how can they be mistaken, given that they are a priori and the entire system of Kant depends universalizing that which is a priori.