2Ö2 Kant: A Biography
questions that can — indeed must - be asked, but that cannot be answered
in metaphysics itself. Hume's "empiricism" may be a good strategy for
science (at least up to a point), but it is a bad one for metaphysics. In any
case, not all metaphysical questions can be answered. The boundary prin¬
ciple is meant to restrict Hume's principle so as to prevent us from going
too far along the road Hume wanted to travel.
Put differently, Hume's principle simply tells us to refrain from doing
something, while the boundary principle also tells us to do something. In¬
deed, the boundary principle seems to tell us to do something that Hume's
principle, taken by itself, might prohibit. It suggests that we should look
beyond possible experience, so as to modify or restrict Hume's principle
in particular cases, namely in those cases that have to do with the bound¬
aries of experience, whatever they may be. We must admit that appearances
do not exhaust all of reality. Appearances presuppose something that ap¬
pears, which is "distinct from them (and totally heterogeneous)," namely,
a thing in itself. While we cannot know what is beyond experience, we can
still think it. In fact, Kant claims that we must think about such things, and
that reason itself forces us to do so. We must, Kant argues, at least assume
things external to reason. One of these is the existence of God as a de¬
signer, for "without assuming an intelligent author, no comprehensible
ground for design and order can be stated without falling into patent ab¬
surdities. Although we cannot prove the impossibility of such design
without an original intelligent author ... there yet remains ... a sufficient
subjective ground for assuming such an author."S0This "subjective ground"
is the "need of reason" that we encountered earlier. It is the reason why Kant
had to "deny knowledge to make room for faith."
The answer to the fourth question — "How Is Metaphysics Possible as
a Science" - should now also be clear. Metaphysics, which exists as a human
need and a "natural disposition of reason," is possible as a science of our
necessary conceptual framework.
In order that a science of metaphysics may be entitled to claim... insight and convic¬
tion, a critique of pure reason must exhibit the whole stock of a priori concepts, their
divisions - according to the various sources (sensibility, understanding, reason), to¬
gether with a clear table of them, the analysis of these concepts, with all their conse¬
quences, and especially the possibility of synthetical knowledge a priori by means of a
deduction of these concepts, the principles and the bounds of their application, all in
a complete system.^51
Kant's Critique and Prolegomena are today usually read as works by an
antiskeptical philosopher. Accordingly, the question of whether or not Kant