490 THE AMERICAN JOURNAL OF SOCIOLOGY
son by the author, and especially the only one capable of escap- ing the double error into which he falls, in turn with many others the error, on the one hand, of making an arbitrary histori- cal selection, and consequently a bad concrete sociology, and, on the other hand, of at least partly disclaiming to found an abstract philosophy of societies, under the vain pretext that the concrete bases are insufficient. The true method, that of all the sciences, that also of their particular philosophies as well as their general philosophy, requires that these bases be sufficient. The illustrative schematic methods, the methods of historical sam- pling, are only the palliatives of a demi-science. How does one dare to recommend these methods for sociology when the scholar who would permit himself to recommend the same in the antecedent sciences would by that very^fact be disqualified? That the rigorous method is one of extreme difficulty no one denies, but this is always the case. Besides, the work to be undertaken is a work of co-operation, and the specialist in abstract sociology will have the privilege and the duty of borrowing the materials which will be furnished him by the specialists, properly so called. At the very least, it is necessary that he be acquainted with these materials, in order that the facts may not contra- dict the theory. Abstract sociology, even in its exemplative and schematic capacity, may be excused from recalling the facts to mind, but on condition of disowning any abstract law which is at variance, not merely with the concrete, but also with the ele- mentary facts of science. Nevertheless, from the point of view of clearness and of demonstration, abstract sociology will state with advantage the particular laws upon which it founds its gen- eral theory, and, as far as possible, the principal data underlying these particular laws.
Moreover, abstract social statics is connected with concrete sociology in this, that, although the latter studies societies in certain parts of space and time, yet it also seeks to disengage the social facts relative to particular civilizations from relatively general, constant, and necessary relations, neglecting the acces- sory variations. In brief, concrete sociology is an intermediate step between history, properly speaking, and abstract sociology.
