500 THE AMERICAN JOURNAL OF SOCIOLOGY
interest categories necessary in working out those subdivisions of associations which will prove most important in the next stages of societary analysis.
We have thus sufficiently emphasized the difficulties of ana- lyzing and classifying associations. Progress in sociology depends, however, upon overcoming the difficulties.
The importance of classification as a stage in the progress of science has been common fame since Aristotle. The neces- sity of finding categories of classification adequate to the differ- entiation of social facts has been recognized, in the abstract, by all the important sociologists, and it has been insisted upon with great force by not a few. The demand for sociological classification has never been presented with greater clearness and energy than by Dr. Steinmetz in a recent monograph. 1 Cer- tain passages are so timely that we epitomize at some length :
The first fault that strikes every critical mind, even in the best works on sociology, with very rare exceptions, is the lack of universal and systemati- cally completed knowledge of their whole domain. We rarely get the impres- sion that the author is oriented on all sides or even upon many It
is exceptional for a sociologist, when trying to demonstrate the universal applicability of a law, to base his induction on more than a single category of
societies. He usually neglects with a liberality truly na'ive We may
sum up our charges in the following judgment : for all sociology which deals with humanity in general, or with both barbarous and civilized peoples, that is to say, which is not identical with ethnology proper, the comparative period is not yet begun. The comparative method is not yet an absolute essential in all the researches worthy of the name, which pretend to be more than rhetoric or chatter (causerie). Now, I believe the principal reason why this method is so little in favor is the absence of a classification and a catalogue of all peoples according to their social status and the degree of their civiliza- tion.
.... In point of fact, there are very few sociological propositions which are true of Germans and of Bushmen, of Athenians and Chinese and Esqui- maux. Yet sociologists do not confine their assertions to a special class .... nor do they trouble themselves to discover the domain within which a fact prevails. They seem indifferent to the capital question whether the phe- nomenon is only an exception, or a general fact common to numerous classes, if not to all. The reasonings of sociological works are almost always entirely vague and without sufficient foundation.
'"Classification des Types sociaux," in DURKHEIM'S L'Annte sociologique, III, 1900.
