to inquire whether it is possible to lay down any general principles which may enable us to assign its proper value to such evidence. It is the purpose of this paper to formulate one such principle and inquire to what conclusions it may lead us.
I must first say something about the sense in which I propose to use the term 'myth.' At the present time this word is used with many different meanings. By some it is limited to narratives which give an account of the doings of gods or of those who possess in some measure divine characters; to others a narrative is only a myth if it stands in a definite relation to ritual and serves to explain and justify this ritual; by others the term is used more widely, but is limited to narratives which give an account of or explain natural as opposed to social phenomena.
In the present state of our knowledge the chief justification of classifications and definitions lies in their usefulness, and a classification which may be useful from one point of view may not necessarily be so from another. I approach the subject of myth in this paper in its relation to the history of culture, and from this point of view, if from no other, the limitations implied by the usages I have mentioned are neither helpful nor necessary.
The general class of which myths form one group are narratives which give a concrete account of events. By thus laying stress on the concrete nature of "narratives" I intend to exclude the abstract accounts of events given by science.
From the point of view of the history of culture the first grouping of narratives depends on whether they are or are not historical, whether they are records of events which have actually happened or the work of the human imagination. We can be confident that even those narratives of peoples of rude culture which come nearest to history contain imaginative elements, but this only makes it the task of the student to distinguish the two elements from one another.
