Kant Again

 

 

Question:  What can we know by reason alone (i.e., what can we know a priori)?

 

          Logic:  (Whatever follows from the principle of non-contradiction)

                   Kant and Hume:  Yes.  (This is what Hume calls “Relations of Ideas”)

 

Metaphysics:     (For Hume, this would be knowledge of Matters of Fact known by reason alone.)

          Hume:  No.

                   Kant:  This is the fundamental question he raises in the Preface.

Answer:  Yes, but limited to what Reason (the mind) contributes to how things appear to us.

 

What does this mean?  How is this (metaphysical knowledge) possible?

 

 

Terminology

 

Hume's Division of 

All knowledge is of

Knowledge.

one or the other type.

 

(Kant)

Relations of Ideas 

Matters of Fact

(Kant)

analytic

negations are contradictory

 

negations not contradictory

 

synthetic

A priori 

knowable by reason alone,
without reference to sense
experience 

not knowable by reasons
alone, but instead requires
sense experience 

a posteriori

 

In his terminology, the question of how metaphysical knowledge is possible becomes:

How is synthetic a priori knowledge possible?  To understand Kant’s answer, recall:

 

“Reason has insight only into that which it produces after a plan of its own.”

 

That is, we can have knowledge (of matters of fact, i.e., metaphysical knowledge) only of those aspects of how things appear to us that are due to the nature of reason itself, i.e., that are due to the ways in which our minds work.

 

So, what does Reason (the mind) “contribute” to how things appear to us?  And why does it think it “contributes” anything?

 

Explaining this will take longer, but his answer to this question involves what he calls “Transcendental Idealism.”  There are interpretive differences as to what this really means, but as a first approximation, it is the view that (for reasons to be explained):

 

          We know only “appearances” and not “things in themselves.”

 

Another way that Kant puts this (in my view, a better way) is:

 

          We only know things as they appear to us, not as they are in themselves.

 

The interpretive question (in my view) rests on how we understand the word “as” in the two places it occurs here.  What is it to say that we do (or do not) know an object as ____?  On one view (which I think is mistaken), this means that the objects that we know are different than we previously thought: we used to think we knew things-as-they-are-in-themselves, but now we recognize (or Kant is telling us) that we only know things-as-they-appear.

          On another view (the one I believe is correct), Kant is not talking what kinds of objects we really know or experience, but is, instead, talking about how we experience objects.  It is not the objects that are different from what we previously believed, but rather that knowledge is different (is a different kind of relation between things) from what we thought before.  So, on this view, the clearest way to state Kant’s transcendental idealism is as the claim:

 

          We only know-as-they-appear things,

          We don’t know-as-they-are-in-themselves things.

 

On this interpretation, Kant is telling us that experience (consciousness or knowledge of objects) is always fundamentally interpretive.  If you like, you might say that consciousness is never “transparent” to its objects:  we never perceive an object as it is independently of our experience of it, but always only as it appears to us, given the necessary (and subjective) conditions of the possibility of conscious experience.

          Explaining why Kant believes this is a longer story, but it starts with his understanding of our knowledge of the truths of arithmetic and geometry.  Recall that Hume mentioned arithmetic and geometry as examples of knowledge of “Relations of Ideas.” That is, he felt that we knew these things by reason alone.  But, of course, for Hume, the only things we can know by reason alone are “relations of ideas,” i.e., claims whose negations are contradictory, claims where the predicate is contained in the subject.

          Kant agrees that we know arithmetic and geometry by reason alone.  In his words, our knowledge of them is a priori.  But, Kant claims, the truths of arithmetic and geometry are not such that their negations are contradictions—they are not such that their predicates are contained in their subjects.  In Kant’s terminology, these claims are not “analytic,” as Hume suggested, but are instead “synthetic.”  If this is true, then we have knowledge, by reason alone, of some statements that are true merely because their negations are contradictory, of statements where the predicate is not contained in the subject.  These are, in Hume’s words, statements of “matters of fact” that are known, contrary to Hume’s claims, by reason alone.  Hume’s rejection of metaphysics, recall, was based upon his claim that it was impossible to know by reason alone any statement whose negation was not a contradiction.  But Kant argues that the true statements of arithmetic and geometry are not such that their negations are contradictory, and yet, we do know them by reason alone.  So, Kant is saying, if we accept Hume’s reasoning for rejecting metaphysics, then we will have to reject arithmetic and geometry as well.  On the other hand, if we don’t want to reject arithmetic and geometry, then it must be possible for us to know on the basis of reason alone (at least) some claims whose negations are not contradictions. 

          In Kant’s terminology, to say that we know by reason alone some claim where the subject is not contained in the predicate is to say that we have “synthetic a priori knowledge.”  Using Kant’s language, Hume’s reason for rejecting metaphysics is that it is simply not possible to have a priori knowledge of synthetic statements.  But, if Kant is correct about arithmetic and geometry, then synthetic a priori knowledge must be possible because it’s actual—i.e., we actually have such knowledge.  But, and this is Kant’s task in the remainder of the Critique, if we can explain how—under what conditions—it is possible, then we may be able to explain how—and what kind of—metaphysics is possible.