he may know what to do under the circumstances, or he may take a
speculative interest in it and then he will devote himself to knowing
it simply for the sake of knowing. It is only, according to Bergson,
in the former case, when his interest is practical, that he will
attain his object by using the intellectual method of abstraction
which proceeds by analysis and classification. These intellectual
operations have such prestige, however, they ' have proved so
successful in discovering explanations, that we are apt to take it for
granted that they must be the best way to set, to work whatever sort
of knowledge we want: we might almost be tempted, off hand, to imagine
that they were our only way of knowing at all, but a moment's
reflection will show | that this, at any rate, would be going too far.

Before we can analyse and classify and explain we must have something
to analyse, some material to work upon: these operations, are based
upon something which we know directly, what we see, for instance, or
touch or feel. This something is the foundation of knowledge, the
intellectual operations of analysis classification and the framing of
general laws are simply an attempt to describe and explain it. It is
the business of science to explain and intellectual methods are the
appropriate ones for science to employ. But the business of
philosophy, according to Bergson, is not to explain reality but to
know it. For this a different kind of mental effort is required.
Analysis and classification, instead of increasing our direct
knowledge, tend rather to diminish it. They must always start from
some direct knowledge, but they proceed, not by widening the field of
this knowledge but by leaving out more and more of it. Moreover,
unless we are constantly on the alert, the intellectual habit of using
all our direct knowledge as material for analysis and classification
ends by completely misleading us as to what it is that we do actually