Page:American Journal of Sociology Volume 5.djvu/638

 622 THE AMERICAN JOURNAL OF SOCIOLOGY

Among the followers of Comte there has not been due observance of the limitation just suggested. Descriptive analy- sis is logically presupposed as a condition of validity in genetic classification, or in causal analysis, which is another aspect of the same thing. Social facts and forces have been arranged in classes by sociologists whose haste to reach genetic classification has made them neglect necessary descriptive analysis. This criticism may be applied at once to De Greef. His famous schedule of social phenomena involves a thesis about the order in which those phenomena emerge." That hypothesis turns the schedule, to a certain extent at least, into a genetic classifica- tion. In that character De Greef's proposition is more than questionable. As a descriptive analysis for certain purposes it has not been excelled. We may then at once set down to the credit of the sociologists of this group a commendable beginning of the process of grouping like social facts. This is a necessary preliminary in all science. The "classifying sociologists" have been criticised not so much because they did not do their part well as because the critics did not see that this part was worth doing at all. Such judgments condemn the critics rather than the criticised. Classification is not the whole of science, but it is an essential stage in the scientific process. The men who belittle it tend to disregard the authority of facts, and to claim scientific authority for their lucubrations independent of facts.

The processes that have given the group-name to the "clas- sifying sociologists" have sometimes been called collectively "descriptive sociology." This term stands for all that is involved in arranging the material facts in classified order, without attempt to enter upon the next step, namely, interpreta- tion. Whether this designation is to be permanent experience alone can decide.

A passage from Barth is pertinent at this point :

According to Comte, sciences must be parallel with things. When we arrange the latter according to their decreasing generality, and their increas- ing complexity, we have at the same time their actual correlation. Just so, when we arrange the sciences according to the same principle, we have the

' Introduction a la sociologie. Vol. I, p. 217.