(4) Rules for explaining social facts, (a) It is common to find the reason of a fact in its utility. This is to confound final with efficient causation and is no more admissible here than in natural science. Function and cause must be examined separately. (But as M. Bernès insists in his article, a distinguishing mark of society is intelligent action. While then it by no means follows that the actual use was foreseen and so consciously aimed at by the social group, it is nevertheless quite possible that this may have been the case. To exclude all teleological explanation from history is to assume that social facts are in no wise different from physical facts; it begs the question in advance.) (b) the cause of a social fact must always be sought in preceding social facts, not in states of individual consciousness, (c) The function of a social fact should be sought in the relation it sustains to some social end. (d) The first origin of every social process should be sought in the constitution of the internal social medium. This will depend upon two factors: (α) the number of social units, the "volume" of society; (β) the degree of concentration, the "dynamic density," which is in turn a function of the number of individuals who are in commercial and social relations. If we do not adopt this plan we are reduced to explain social progress by "tendencies" instead of by real causes; and further we are forced to treat all as one species in greater or less stages of advancement.
(5) Methods of induction, (a) The doctrine of plurality of causes, bound up with Mill's philosophic presuppositions is to be rejected. The same effect is not produced by different causes. (b) The most valuable of the inductive methods for our purpose is that of concomitant variations, for which we may draw our facts either from a single unique society at different times, or from several societies of the same kind, or from several of different kinds.
The three articles of M. Bernès are devoted in part to a criticism of Durkheim but develop also the author's own views of the province and method of the science. The two dominant characteristics of contemporary sociology are declared to be, (a)
