The comparative method requires the definition of a type of story on the basis of a collection of folktales. The comparison is then between the type and the target myth. The comparison is not between a single folktale text and the target myth.
A type is defined in terms of motifs. It may consist of a single motif, but folktales usually have several motifs, i.e. several incidents. A folktale type is thus a constellation of motifs that constitute an independent story, i.e., a story that makes sense in itself and does not depend on its relation to some larger story. Different types may, however, have some of the same individual motifs. For example, in any number of folktales, the hero grows up and goes off into the world to seek his fortune. This motif by itself could hardly define a type. What defines a type is a recurring constellation of motifs.
In the preceding paragraph, "motif" referred to incident. The "term" motif can also be used for a stock character (e.g., wicked stepmother) or a thing (e.g., magic weapon).
Each example of a given type will be a variant thereof. As might be expected, the constellation of motifs or the story-pattern does not repeat itself exactly from one example of a type to another. The story-pattern varies either (1) in the pattern itself, through the choice of motifs to include or exclude, or (2) in the motifs, which may themselves present variation within sameness.
As for the first case, the question arises: by how many motifs can folktale A differ from folktale B and still be considered to have the same story-pattern as B? This question, which is mainly a theoretical one, usually does not cause problems; common sense is a sufficient guide. As for motifs, variation is usually easy to perceive. Someone is rescued. The motif is rescue. In one version of the folktale, the rescuer is human; in another, the rescuer is an animal. The motif remains basically the same (rescue); but it varies from one folktale to another (human or animal rescuer).
The application of the type to the target myth entails certain assumptions.
(1) None of the folktales constituting the type was derived from the target text. The hypothetically independent type must be constructed on the basis of folktales that are themselves independent of the target text. If the folktales are derived from the target text, they can hardly illustrate an independent type that can then be applied comparatively.
(2) The type was known to the teller of the myth and persisted in oral transmission down to modern times. This assumption has proved to be very uncongenial to some scholars of Greek myth. Even when they have recognized the relevance of folklore to Homer, for example, they have spoken as if Homer's "transformation" of a folktale into epic poetry somehow destroyed the original folktale, so that only the poetic form remained. There is a fundamental self-contradiction in this reasoning. The scholarly detection of the folktale origin of something in Homer depends upon modern folklore, because we have virtually no ancient folktales. (The ancients had no reason to make transcriptions of oral story-telling.) If the detection of the folktale origin of something in Homer has any validity, then the modern story-tellers who are the source of the folktale in question must belong to a tradition that goes back to antiquity. This situation is in itself distasteful to many classicists, because it means that peasants, tribesmen and -women, and other lowly people in the nineteenth and twentieth centuries, i.e., those from whom most folktales have been collected, were in possession of the same stories as the great geniuses of Greek literature.
(3) The existence of widely distributed folktales sharing the same story-pattern is due to geographical-historical diffusion and not to polygenesis (the spontaneous creation of the same story by different peoples in different places). Those who believe in polygenesis often use world mythology and folklore as a way of arguing for the fundamental sameness of the human soul everywhere: we are all one.
Finally, it is necessary to point out that the use
of a widely distributed folktale type does not entail the notion that that
type is found absolutely everywhere in the world. The belief in polygenesis
is often coupled with the belief (sometimes called "universalism") that
the same stories are found all over the world. The folklorist Alan
Dundes has written in Sacred Narrative (p. 257): "It is undoubtedly comforting
to think that all mankind shares common myths and metaphors, but the empirical
facts don't support such an illusion." Dundes' observations show
that universalism may really be a form of a ethnocentrism, because the
stories that the universalists claim are worldwide are often only the ones
that they have been able to find within the resources of Western culture.
A model of the application of a folktale type to an ancient story is William F. Hansen article in Lowell Edmunds, ed., Approaches to Greek Myth, The Johns Hopkins University Press, 1989. See also Edmunds' introduction to Hansen for reflection on problems of methodology.