Thinking about Kant's reaction to Hume ... "awakened from my dogmatic slumber".
Kant realized, via Hume, the deep troubles involved in rationalism, all types. Kant's work in the Transcendental Deduction section of the "Critique of Pure Reason" is in some ways an extention of Humean skepticism ... a formalization there-of.
Kant's validation of reason is not like his predecessors ... there is no, like Descartes, looking for a ground for reason. Kant accepts that all such grounds will be problematic for the reasons Hume gave and for reasons refelcted still today in post-modernistic critique of rationalism.
"What can we know?" Kant really struggles with this question.
Kant does not begin the Critique right-off in validating reason. He starts talking about such things as "understanding". "Understanding" i take to be the deductions/inductions we make directly from our observations of nature ("objects of understanding" Kant calls the basic units). He takes what he takes to be the "best answer", and what most of us take today, the scientific world view. That is ... Kant sees in science, and its core matha nd logic, the most likely equivalent to what is actually going on in the real world.
AT THE SAME TIME, though, he is pointing out that "understanding" is a low level psychological event. Truth does not come from this route. He unifies the mesh of "objects of understanding" under a rational scientific unity, and then looks for the unifying elements in these and we have things like logic as the unifying principles.
But then Kant turns. He lets us know that logic and rationality are not tightly related to the "reason" he is going to be critiquing. This "reason" of his is not deduced or induced from nature. It is, at best, a unification of the unifying principles used in deducing nature.
That is ... reason is the ACTIVE human attempt to unify and the ability to come face to face with contradictions and unify beyond them ... where the contradiction are the result of going too far with the rationalistic approach.
I don't feel i am explaning this well, i think i need to re-read and try again.
But to get across my key thought here ... Kant does not think of "reason" as being a mirror of nature. Reason just is. Reason does what it does, it is active, but there is no justification in assuming that nature and reason are alike in any way. The project of reason may be doomed.
Rationalism, however, is from the basic understanding ... there is a tie between it and the real world and we expect them to mirror one another. But rationalism is not and cannot be a proper ground, it is limited by the limits of reason itself, whatever those may be. This is most obvious when contradictions and necessary contradictions (what is the term? ... antinomies).
So then, at heart, it is this thing, "reason" that Kant wants to critique as a "prelegomena" toward future work. His critique is a direct attack on the attempts at rationalistic grounding of science (or other modes of thought such as religion).
more later ...