120 THE AMERICAN JOURNAL OF SOCIOLOGY
of the component elements of the substance examined, without reference to the more precise conditions of their combination. In these complicated cases, however, elementary analysis tends irresistibly to pass over into the more advanced stages of the analytical method.
The second stage of analysis is causal analysis. It consists in the sepa- ration of a fact into its component parts with reference to the causal relations of the same. Thus tree, swaying motion, sound, may he analyzed as above in a merely descriptive or elementary fashion ; but when there is advance to a different logical plane the process changes. Or I may say, when the process changes, the analysis is evidently proceeding on a different logical plane. Thus, " the god in the tree and the god in the air are angry with each other, and, wrestling together, they make what we see," viz., the details analyzed into the other terms above. Here the components of the distin- guished parts are gods, passions, contest — in turn causes and effects. This analysis for purposes of explanation presupposes elementary descriptive analysis. Yet the descriptive analysis may be very summarily performed or combined with explanation, so that investigation really begins at once with the latter. Instances are physics, psychology, and history; while, on the other hand, chemistry, physiology, political and social science are parts of knowl- edge in which descriptive analysis is a stadium of intelligence with a distinct importance of its own.
The ground of this difference lies in the different tasks of these sciences. Physics and psychology concern themselves with the explanation of general phenomena, the one of objective, the other of subjective, occurrences. To this end, each begins analysis with the simplest facts, in which, without any descriptive preparation, a consideration of causes becomes at once necessary. The investigation of more complicated phenomena may then follow close upon that of these simpler causal analyses, and there is at once connected with these simple analyses an attempt to test the causal principles, analytically reached, by seeing whether they are applicable as explanations in a synthesis. Such simple points of departure are wanting in history. As a compensation, and logically a similar affair, history employs a comprehensive process of abstraction, which permits seizure upon certain chief factors of historical events, which may be easily referred back to certain psychical motives.
Quite in contrast with this is the situation of the other group of sciences. In their case the simple facts are for a long time beyond reach — thus the facts which qualitative and quantitative analysis must discover in chem- istry, the morphological and chemical qualities of an organ in physiology, the facts about occupations and customs in sociology. Hence the initial necessity of independent descriptive analysis.
I speak of this at length, because in the social sciences, with the single exception of political economy, men have hardly begun the necessary process of descriptive analysis, at the basis of all credible science or philosophy or theory of society. That is, they have hardly begun to propose the necessary
