HKB
shows himself indocile to certain lessons of experience, he does so because of other experiences which seem more demonstrative. The scholar does not do otherwise; only he introduces more method.
So magic is not, as Frazer has held,[1184] an original fact, of which religion is only a derived form. Quite on the contrary, it was under the influence of religious ideas that the precepts upon which the art of the magician is based were established, and it was only through a secondary extension that they were applied to purely lay relations. Since all the forces of the universe have been conceived on the model of the sacred forces, the contagiousness inherent in the second was extended to the first, and men have believed that all the properties of a body could be transmitted contagiously. Likewise, when the principle according to which like produces like had been established, in order to satisfy certain religious needs, it detached itself from its ritual origins to become, through a sort of spontaneous generalization, a law of nature.[1185] But in order to understand these fundamental axioms of magic, they must be replaced in the religious atmosphere in which they arose and which alone enables us to account for them. When we regard them as the work of isolated individuals or solitary magicians, we ask how they could ever have occurred to the mind of man, for nothing in experience could either suggest or verify them; and especially we do not explain how so deceiving an art has been able to impose itself for so long a time in the confidence of men. But this problem disappears when we realize that the faith inspired by magic is only a particular case of religious faith in general, and that it is itself the product, at least indirectly, of a collective effervescence. This is as much as to say that the use of the expression sympathetic magic to designate the system of rites which we have just been speaking is not very exact. There are sympathetic rites, but they are not peculiar to magic; not only are they to be found in religion, but it was from religion that magic received them. So we only risk confusion when, by the name we give them, we have the air of making them something which is specifically magic.
The results of our analysis thus attach themselves to and confirm those attained by MM. Hubert and Mauss when they studied magic directly.[1186] They have shown that this is nothing more nor less than crude industry based on incomplete science. Behind the mechanisms, purely laical in appearance, which are used by the magician, they point out a background of religious conceptions and a whole world of forces, the idea of which has been taken by magic from religion. We are now able to understand how it comes that magic is so full of religious elements: it is because it was born of religion.
III
But the principle which has just been set forth does not merely have a function in the ritual; it is of direct interest for the theory of knowledge. In fact, it is a concrete statement of the law of causality and, in all probability, one of the most primitive statements of it which has ever existed. A full conception of the causal relation is implied in the power thus attributed to the like to produce the like; and this conception dominates primitive thought, for it is the basis both of the practices of the cult and the technique of the magician. So the origins of the precept upon which the imitative rites depend are able to clarify those of the principle of causality. The genesis of one should aid us in understanding the genesis of the other. Now we have shown how the former is a product of social causes: it was elaborated by groups having collective ends in view, and it translates collective sentiments. So we may assume that the same is true for the second.