49 2 THE AMERICAN JOURNAL OF SOCIOLOGY
abstract than the qualitative sociological laws. Any science may be considered as formed when it is in possession of its spe- cial methods, when the boundaries of its domain are well marked, and when the formulae of its abstract laws can be quantitatively expressed. Thus the law of gravitation has a completely positive formula : all bodies attract each other in direct proportion to their mass and in inverse proportion to the square of the dis- tance. In sociology, an example borrowed from circulatory phenomena may be given ; the useful effect is in inverse propor- tion to the dead weight and in direct proportion to effort.
The most exact abstract sociological laws are, then, both qualitative, or descriptive, and quantitative ; hence the neces- sity of statics that is to say, the analytic study and the classi- fication of the social elements as a basis. The method of observation with its numerous inductive processes, variable in different sciences, but identical at bottom, is the only possible method. Deduction can be applied only in sciences already formed, and from one science to another; in the latter case, great caution must be exercised in order to avoid error, espe- cially in passing from biology, with its psychic dependence, to sociology.
Sociology, even abstract sociology, was naturally descriptive in the beginning. At least this statement seems to us to be true in the case of Auguste Comte and Herbert Spencer, to cite only the most eminent founders of social science. But positive soci- ology will be established in reality only when it shall rest upon the statistical and concrete facts of all the special social sciences, particularly economics. Auguste Comte was in error when, inverting the order of the positive method, he contended that in sociology and biology the whole is known before the parts. This pretended knowledge of the whole before that of the constitu- ent elements was uniquely empirical, superficial, and plausible. It is to this deplorable error of method that we must ascribe his tendency to attribute essential and permanent characteristics to transitory historical forms. His entire hierarchic and absolute conception of societies has for its point of departure this error in method, the consequences of which have been further amplified
