The practice of comparing substantial numbers of cognate items of folklore to establish their similarities, differences, and probable course of development. Comparison has always been characteristic of folklore studies. Whether we consider the interest in foreign cultures that flourished during the Renaissance, the citation of parallel texts beginning with the Grimm brothers, the interest in cultural evolution championed by late-19thcentury anthropologists, or the stemmatic model of the historic-geographic method to be the beginning of “modern” folklore scholarship, we find that more or less rigorous comparison has always been essential. Only through careful comparisons of superficially similar traditional forms can the relative contributions of replication and variation, or the likelihood of independent origins (polygenesis), be assessed. In addition, comparison of traditions in adjacent cultures can identify particular developments that have occurred in each culture or subculture. Anthropologists have argued that, because comparison is essential to human thought, there is no single comparative approach (Lewis 1970; cf. Boas 1940:270–280). Folklore does not suffer from this ambiguity. Here, “comparative” is a technical term, borrowed from philology, as evidenced in such phrases as Antti Aarne’s title Leitfaden der vergleichenden Märchenforschung (Guide to Comparative Folktale Study, [1913]). Insofar as they are practicing the comparative method, folklorists restrict their comparisons to variants of a single tradition: Their purpose is to define that tradition. The Historic-Geographic Method. The study of Indo-European linguistics began in 1786 when Sir William Jones established the relationship of Sanskrit to Greek and Latin. On the basis of the similarities among these languages, a parent language was inferred. The simplicity and objectivity of this approach made it seem both suitable and desirable for the reconstruction of early forms of oral traditions. Intrigued by work of Elias Lönnrot, who “reconstructed” the Finnish epic The Kalevala from ballads and songs, Julius Krohn began, in the latter part of the 19th century, to codify procedures for reconstructing oral traditions based on philological principles: The variant texts should be analyzed to see how they led back to a single original text (in German, the Urform; in English, the “archetype”). At this time, scholars were arguing about the diffusion of traditional literary forms, including folktales: Some believed that it was primarily prehistorical, while others thought that most of it occurred during the Middle Ages. Julius’ son Kaarle Krohn set out to test the idea that European folktales came from India in historical times by reconstructing the original forms of individual tales and, at the same time, determining their original homes and their paths of migration. His method was called historic-geographic because early (literary) variants are arranged historically and later (oral) variants are grouped geographically, in order to achieve the reconstruction as objectively as possible. Ideally, the versions of the tale will fall into a stemmatic pattern—all of the later forms of the tale can be seen to have developed from the hypothetical original form. This stemmatic pattern is analogous to that used by philologists to reconstruct the Indo-European language.
This method works very well for traditions, like language, that change slowly or in regular patterns. It is most useful for folktales that are relatively stable, like “The Clever Peasant Girl” (AT 875), “The King and the Abbot” (AT 922), and “The Kind and Unkind Girls” (AT 480). For tales that are subject to more variation, the method is more difficult to use. The original form of the tale is less obvious, and may even be the subject of a debate. Kaarle Krohn’s Übersicht über einige Resultate der Märchenforschung (Survey of Some Results of Folktale Study) (1931) offers an alternative historical development for several tales. Nevertheless, the fact that scholars using the method assembled large numbers of variants and paid careful attention to their contents advanced the respectability of folklore studies as an academic discipline. The American folklorist Stith Thompson contributed greatly to the success of the historic-geographic method. His Motif Index (1995) and his revisions of Aarne’s The Types of the Folktale (Aarne and Thompson 1964) helped considerably to make the large numbers of variants necessary for historicgeographic studies easily available. Since this classification began with European tales, and since the Americas are peripheral regions of European culture, folklorists studying American folktales have not taken full advantage of the possibilities the method offers. Yet, regardless of whether it is able to establish an archetype for a particular tale, the method often discovers features that are unique to particular regions. Thus, the tale “The Taming of the Shrew” (AT 901) has developed an American joke form (Brunvand 1991:213–228). Two other genres to which the historic-geographic method has been applied successfully are the ballad and the riddle. Following Svend Grundtvig’s work on Danish ballads (Danmarks gamle Folkeviser, beginning in 1853), Francis James Child classified English ballads, adding notes that are useful even for tale studies. Neither of these scholars attempted the rigorous laying out of variants and careful reconstruction of the original text that Krohn advocated, but they did comment on the relationships among their texts, noting which were derived from which. Thus, these notes were the foundation of 20th-century historic-geographic ballad studies (Taylor 1927–1928). In addition to his many historic-geographic studies of folktales, Aarne wrote several historic-geographic studies of riddles, which are interesting because they identify regional characteristics. The method depends on the scholar’s ability to find variants. Genres that have defied indexing, such as legend, where the texts are too variable, are difficult if not impossible to study in this way. In the case of simple material, such as folk beliefs and single narrative motifs, each item is too brief to permit conclusions about its history. Based as it is on the idea of texts being diffused from a single place of origin, generally by land routes through territory inhabited by settied populations, the method is not suitable for Old World traditions brought to America by immigrants from different countries. In this case, Old World ethnic origins, not the location in America, may be the appropriate criteria for the arrangement of variants. Nor is the method suitable for modern traditions that are carried by national or international media, or even in letters or telephone conversations; in such cases, geography is irrelevant to the development of the item. The Comparative Approach in America. “While the American folklorists Thompson and Archer Taylor were prominent in the development of the indexes that are essential to the historic-geographic method, less-stringent comparative approaches have characterized the study of folklore in America. For example, annotators of folktales, such as Herbert Halpert and Leonard Roberts, in comparing American variants to their European counterparts, have occasionally remarked on peculiarly American alterations to certain tales. American ballads and proverbs have received more attention. For her comparative study of “The Maid Freed from the Gallows,” Eleanor Long (1971) found that grouping the texts according to their verbal correspondences was more helpful than using their places of origin. Tristram Potter Coffin (1977) noticed that American ballads tend to be shorter and more lyrical than their British antecedents. They concentrate on the climax of the story and replace some of the supernatural motifs with rationalistic ones. Ballads composed in America have also been cataloged according to story types (Laws 1964), which has aroused some opposition from those who believe that indexing ballad themes would be more effective (Wilgus 1970). Since, for practical reasons, folklore indexes must be made before each entry can have been thoroughly studied, they often separate items that are, in fact, interconnected. While such lapses do not necessarily prevent the indexes from leading to the relevant material, they make them cumbersome and they are often considered defects. Taylor was instrumental in advancing proverb scholarship to the point where comparative studies (both American and international) could be accomplished. American proverbs have been collected more thoroughly from literary sources than from oral ones. Comparative studies show that some become more stylized with time, evolving into forms with more pronounced poetic features such as rhythm, rhyme, and well-defined structure (Taylor 1975; Mieder 1987:157–228). Taylor, in his English Riddles from Oral Tradition (1951), annotated English (including Anglo American) riddles with references not only from Europe but also from other parts of the world; however, subsequent studies of riddles have been more interested in their structures and functions than in their types and variations. Finding-lists of children’s rhymes (Abrahams 1969; Abrahams and Rankin 1980) provide access to that material. In the case of superstitions, however, indexing has been frustrated by the variety of combinations of motifs. Many superstitions are spread rather evenly throughout the country rather than confined to particular regions. In the realm of material culture, comparative research has managed to do well even in the absence of indexes. Comparisons of artifacts reveal cultural regions within the United States, sometimes showing how European forms were altered as they were adapted to the requirements and fashions of America (Glassie 1968). For studies of oral tradition of preliterate cultures, a geographic arrangement of the material is often feasible. Franz Boas took such an approach to the study of Native American folklore. He was interested in the regions that such studies reveal and in the borrowing, lending, and adaptation of culture elements. Thompson (1965) followed up his early interest in Native American folktales by using the geographic pattern of one tale to determine its history, as a test case to prove that the historic-geographic method would indeed work even without old variants. The comparative approach has been important in the study of the origins of African American folklore. Melville Jean Herskovits, following Boas’ example, pursued what he called “the comparative study of cultures within a given historic stream” (Herskovits 1956:141), as opposed to the global comparative approach of earlier anthropologists such as Tylor and Frazer. Richard M.Dorson (1967) was able to cite European antecedents for many folktales told by African Americans. As more African folktales, and indexes to folktales, have been published, many of those same tales can be seen to have African antecedents as well. If this situation makes it seem as though origins are wherever you look for them, that is not the fault of the comparative method, but rather of the imperfect state of folklore collection and classification. In the 1920s and 1930s, the historic-geographic method was presented as the method of folklore scholarship (Krohn 1971). Its present advocates have come to believe that other approaches can supplement it to good effect. Beginning in the 1960s, it became unfashionable and was quickly overtaken by other methods. Several of these, however, including structuralism (in which a single pattern is found in several different items) and the oral-formulaic approach (in which type-scenes and story types replace motifs and tale types), are derived at least in part from folklore’s comparative approach or from its antecedents or cognates in other disciplines such as linguistics and classics. Also at about the same time, many American folklorists became interested in genres of folklore that had not been indexed: They were too new (such as jokes and toasts) or too variable (such as legends). Since variants could not be located, comparative studies were impossible. Even so, as long as folklore studies continue, the comparative approach will not become obsolete because tradition can not be defined without careful comparison. While the particular details of traditional items such as folktales and riddles are usually culture specific, more general features are often found in diverse cultures. Thus, the comparative approach is well suited to showing basic, common beliefs and traditions held jointly by people of many different cultures. With it, folklorists can explore both the unity and the diversity of mankind. Christine Goldberg References