nents
maintain. Where, then, on either side, does the mistaken narrowness begin,
and how far does the justification of each extend?

The conflict centers, first, about the question concerning the origin of
human knowledge and the sphere of its validity. Rationalism is justified
when it asserts that some ideas do not come from the senses. If knowledge
is to be possible, some concepts cannot originate in perception, those,
namely, by which knowledge is constituted, for if they should, it would
lack universality and necessity. The sole organ of universally valid
knowledge is reason. Empiricism, on the other hand, is justified when it
asserts that the experiential alone is knowable. Whatever is to be knowable
must be given as a real in sensuous intuition. The only organ of reality is
sensibility. Rationalism judges correctly concerning the origin of the
most important classes of ideas; empiricism concerning the sphere of their
validity. The two may be thus combined: some concepts (those which produce
knowledge) take their origin in reason or are _a priori_, but they are
valid for objects of experience alone. The conflict concerns, secondly, the
use of the deductive (syllogistic) or the inductive method. Empiricism,
through its founder Bacon, had recommended induction in place of the barren
syllogistic method, as the only method which would lead to new discoveries.
It demands, above all things, the extension of knowledge. Rationalism, on
the contrary, held fast to the deductive method, because the syllogism
alone, in its view, furnishes knowledge valid for all rational beings. It
demands, first of all, universality and necessity in knowledge. Induction
has the advantage of increasing knowledge, but it leads only to empirical
and comparative, not to strict universality. The syllogism has the
advantage of yielding universal and necessary truth, but it can only
explicate and establish knowledge, not increase it. May it not be possible
so to do justice to the demands of both that the advantages which they seek
shall be combined, and the disadvantages which have been feared, avoided?
Are there not cognitions which increase our knowledge (are _synthetic_)
without being empirical, which are universally and necessarily valid
(_a priori_) without being analytic? From these considerations arises the
main question of the _Critique of Pure Reason_: How are synthetic judgments
_a priori_ possible?

The philosophy of experience had overes