6 
MUSEUM BULLETIN NO. 16. 
The field worker, is, as a rule, guided by very vague con- 
ceptions in predicating correctness for one version as against 
another. In the main he depends upon his informant. The 
informant has, it is true, very definite ideas on this subject. 
But what exactly is the value of these ideas ? After analysing 
the reasons advanced it will be found that in all cases the Indian 
bases his concept of a correct version on individual circumstances: 
the specific manner of narrating a certain myth, the individual 
who narrates it, certain characteristic expressions and intona- 
tions, etc. In the last analysis, then, the correct version of the 
Indian is not a dispassionately formed opinion but merely the 
expression of a purely local-emotional attitude. 
MYTH ANALYSIS. 
We will now turn our attention to the specific examination 
of Ehrenreich’s concept of a correct version. 
Let us assume for the sake of argument that only one correct 
version of a myth exists and that the variants represent devia- 
tions from this correct version. Are we to consider any constant 
element appearing in the “original” version and the variants as 
the essential nucleus of the myth, or are only certain common 
elements to be regarded as significant ? If, for instance, we were 
to find in a number of places a plot dealing with a warrior who 
goes on a warpath, is captured, and informs his captors that on 
a certain day, no matter how well he is guarded, he will escape— 
the probability of such a plot developing in a number of North 
American tribes quite independently is considerable. The 
plot, consequently, must be of a specific nature. If, however, 
we find in the same tribe two versions of a myth in which the 
plots are identical but the episodes quite distinct, which are we 
to regard as the original ? Or again, we may find two tales in 
which the episodes are identical but the plots different ; or, finally, 
two versions in which both plot and episodes are identical, but 
where the plot is developed in characteristically different ways. 
In other words, there are conceivably three kinds of identities: 
that of plot, that of episodes, and that of plot elaboration. 
What is the reason for this differentiation, and have we a right 
