70 THE SOCIAL THEORY OF GEORG SIMMEL
ality. In that case the demand for knowledge asks a type of
question which can be answered in terms of such abstract
and phenomenological categories. The answer does not have
to be more exact and does not have to go down to single
factors and causes. It may be changed if it is found to be
erroneous, but it will then be replaced by an answer which
consists merely of other contents of knowledge in a similar
form. The new answer will remain methodologically at
the same distance from causality as an ideal of exact science
 as the answer which it replaced.
History, whether in the form of a narrative or in the
form of historical laws, builds out of that which is given
of the historical actuality a theoretic structure on the basis
 of categories which have no place in an investigation
which aims at exact knowledge. But these categories
spring from fully autonomous needs of organization and
create a self-sufficient picture of the historical actuality.
That this picture is not a pure reproduction of the actuality
 is not a shortcoming, but its essence. It has dimensions
and a style of its own which bear no relation to the requirements
 for a scientific picture of the socio-historical
actuality.!
From the foregoing considerations it may be seen that
neither the philosophy of history nor the search for historical
 laws is identical and synonymous with sociology.
Whether one considers the historical law as a forerunner of
exact knowledge or as a self-sufficient interpretation of historical
 development does not influence the fact that it
does not reach objective validity.
Sociology is an exact science and has to apply the method
 of science. The philosophy of history and the search
for historical laws are forms of inquiry which result from
different and independent needs of cognition and produce
a type of knowledge that is speculative and not exact.
1 Probl. der Gesch., pp. 108-20.