Page:Folk-lore - A Quarterly Review. Volume 15, 1904.djvu/154

 136 From Spell to Prayer.

regards the inquiry we are now embarked on, we may say that, so far as he goes, Dr. Frazer is against the view that magic is capable of merging in religion so as to become part and parcel of it, but that he does not go very far into the question, and leaves it more or less open to further dis- cussion. Wherefore to its further discussion let us proceed.

Now in the first place it would clearly simplify our task if we could find sufficient reason for assuming that, what- ever it may afterwards have becom.e, magic was originally something wholly unrelated to religion, that, in short, it was originally siii generis. I may point out that this is by no means the same thing as to postulate, with Dr. Frazer, an " Age of Magic," when religion simply was not."* Our assumption would not exclude the possibility of some sort of religion having been coeval with magic. Which, let me add, might have been the case, even were it shown that magic can generate religion of a kind. For religion has all the appearance of being a highly complex and multifarious growth — a forest rather than a tree.

That magic was or igimWy sui generis might seem a doc- trine that hardly calls for establishment, so universally is it accepted by anthropologists. Its peculiar provenance is held to be completely known. Thus Dr. Frazer tells us that magic may be " deduced immediately from elementary processes of reasoning," meaning the laws of association, or, specifically, the laws of association by similarity and by contiguity in space or time. ''

Now it seems to me that, once more, these statements need to be construed liberally. The psychological purist might justly doubt whether Dr. Frazer is literally able to deduce magic immediately from the laws of association. He would, at any rate, deny Dr. Frazer's right to describe the laws of association as "processes of reasoning" or "laws of thought" in any strict sense of these terms.^ A

« See Cs. B.,^ i., 73. * G. B.,- i., 70. Cf. 62. " G. B.,- i., 70 and 62.