[22] The rationalism which is imminent in the sociological theory of knowledge is thus midway between the classical empiricism and apriorism. For the first, the categories are purely artificial constructions; for the second, on the contrary, they are given by nature; for us, they are in a sense a work of art, but of an art which imitates nature with a perfection capable of increasing unlimitedly.
[23] For example, that which is at the foundation of the category of time is the rhythm of social life; but if there is a rhythm in collective life, one may rest assured that there is another in the life of the individual, and more generally, in that of the universe. The first is merely more marked and apparent than the others. In the same way, we shall see that the notion of class is founded on that of the human group. But if men form natural groups, it can be assumed that among things there exists groups which are at once analogous and different. Classes and species are natural groups of things.
If it seems to many minds that a social origin cannot be attributed to the categories without depriving them of all speculative value, it is because society is still too frequently regarded as something that is not natural; hence it is concluded that the representations which express it express nothing in nature. But the conclusion is not worth more than the premise.
[24] This is how it is legitimate to compare the categories to tools; for on its side, a tool is material accumulated capital. There is a close relationship between the three ideas of tool, category and institution.
[25] We have already attempted to define religious phenomena in a paper which was published in the Année Sociologique (Vol. II, pp. 1 ff.). The definition then given differs, as will be seen, from the one we give to-day. At the end of this chapter (p. 47, n. 1), we shall explain the reasons which have led us to these modifications, but which imply no essential change in the conception of the facts.
[26] See above, p. 3. We shall say nothing more upon the necessity of these preliminary definitions nor upon the method to be followed to attain them. That is exposed in our Règles de la Méthode sociologique, pp. 43 ff. Cf. Le Suicide, pp. 1 ff. (Paris, F. Alcan).
[27] First Principles, p. 37.
[28] Introduction to the Science of Religions, p. 18. Cf. Origin and Development of Religion, p. 23.
[29] This same frame of mind is also found in the scholastic period, as is witnessed by the formula with which philosophy was defined at this time: Fides quærens intellectum.
[30] Introduction to the History of Religions, pp. 15 ff.