14 THE SOCIAL THEORY OF GEORG SIMMEL

speculation on principle in certain cases, but one cannot
do this on the basis of standards which decide for empirical,
 scientific knowledge about validity and significance.
 These standards do not apply to metaphysics
because its problems and aims are different. It creates a
world-picture according to its own laws, with its own
methods, and by its own technique.
But apart from the value and significance which philosophic
 thought possesses in and for itself, as a special
function, as a special attitude toward the world, it has a
meaning and significance as a forerunner of exact science.
Philosophy was the mother of all sciences, and only
gradually have they obtained their independence. As the
forerunner of science, it takes in a tentative grasp what is
as yet unprovable and combines in inclusive concepts what
is as yet unobservable as single facts. With these it draws
pictures of the world which will be partially confirmed,
partially refuted by methodological empiricism, but which
are none the less the first approaches to knowledge.
Scientific thought begins with wide concepts and general
 reflections. It becomes more narrow in proportion as
it becomes more exact. With a few ideas it will try to
grasp the totality of existence, and only after innumerable
trials and errors in the highest abstractions does it begin
with analysis of complex concepts. It then follows the single
 threads of the weaving, which it formerly thought to
understand without knowledge of its structure.
The form of some occurrence which has often been
observed on the surface of appearances is postulated as a
general law until the incidental character of the coexistence
 of its composing factors is discovered. Then the
forms of the latter are postulated as the real general laws
until the process repeats itself. From a tracing of the
{ Probl. der Gesch., pp. 92-93.