474 THE AMERICAN JOURNAL OF SOCIOLOGY
tionship of the " element " to the "whole " in all combinations. Social fact exhibits properties of its own, but what is its point of departure, if not the combination of individuals ? These latter undoubtedly are an essential factor of the social phenomenon, for the same reason that the elements of a chemical combination arc essential factors of the chemical compound. Durkheim com- pletely overlooks the fact that a compound is explained both by the character of its elements and by the law of their interaction. He tries to explain the " product " by the " product" itself, thus over- throwing the scientific conception of cause. It is a startling error of logic, all the more astonishing in a logician of Durk- heim's subtility.
Moving from such an erroneous conception of the social phenomenon, Durkheim is necessarily misled in his interpreta- tions of suicide.
True to his view of social fact as something extraneous to the individual and, therefore, independent of him, Durkheim tries above all to demonstrate the inefficiency of the individual factors in Suicide. But even if such demonstration were possible, the way in which he carries it out seems the least apt to lead to serious results. He examines successively the relationship of suicide to insanity, to alcoholism, and to other extra-social factors, taking each one in isolation in order to ascer- tain a constancy of interaction in which the true cause of suicide should be seen. He completely forgets that a social fact is never the result of a unique cause. Social fact is a complex prod- uct resulting from a combination of different elements, mutually modifying one another. It is impossible, under such circumstances, to find the " unique " or the " pure " cause in one individual factor to the exclusion of others. As Professor Bosco, the Italian statistician, justly remarks in an interesting review of Durkheim's book, 1 " social fact being always profoundly com- plex, it is difficult to ascertain in every instance the working of . a given cause. It suffices that its influence be ascertained in a certain number of cases, and that, above all, there be a possibil-
'See Rivista Italiana di Sociologia, November, 1887, pp. 378-9.
