SIX

Approaches to Folklore

The Literary and
the Anthropological

In Philosophy in a New Key, Susanne Langer stresses the importance of “the nature of our questions.” These questions taken together comprise our basic concepts, our intellectual framework. As she says,

A philosophy is characterized more by the formulation of its problems than by its solution of them. Its answers establish an edifice of facts; but its questions make the frame in which its picture of facts is plotted. They make more than the frame; they give the angle of perspective, the palette, the style in which the picture is drawn—everything except the subject. In our questions lie our principles of analysis, and our answers may express whatever those principles are able to yield.

(Langer 1970, 4)

Using Langer’s metaphor, it is possible to examine the angles of perspective, the frames of reference used by the literary folklorists and by the anthropological folklorists in viewing folklore material. What questions did they bring to the material? And how did the nature of the questions asked shape the material studied?

The literary folklorists classified folklore into genres which were further divided into major and minor genres. And certain literary folklorists specialized in one area: Child and Kittredge in the ballad, Thompson in the folktale, Taylor in the proverb and riddle. The anthropological folklorists studied folklore as part of culture—a way of learning more about the culture history, as with Boas, or a way of learning more about cultural patterns, as with Benedict. For the literary folklorists, the frame for the study of folklore was the written tradition. Folklore was studied as it existed within the literate civilizations, mainly those of the Indo-European world. For the anthropological folklorists, the questions were not directed to folklore per se, but rather to culture: what was the nature of culture? and how was this reflected in folklore?

It is a matter of import to single out culture as the basic concept for the anthropological folklorists. If this is not established as the point of origin for the questions about folklore, the boundaries between the literary folklorists and the anthropological folklorists might be blurred. Both spoke of the oral literary aspect of folklore. But this oral literary aspect had a different point of reference for the anthropological folklorist than it did for the literary folklorists. The anthropological folklorist chose as their frame cultures without writing. They could not, then, delineate the body of material by saying it was oral, since all that was studied in a nonliterate culture was oral. But they could and they did stress the oral literary aspect of folklore. Erminie Voegelin called it “orally transmitted prose”; Herskovits, “unwritten literature”; and Bascom, “verbal art” (Leach 1949, 400, 403, 398).

The material of the anthropological folklorist and the literary folklorist was, indeed, common ground. But it was plowed and plotted in different ways, to different purposes. The cycle of inquiry for the literary folklorist was ultimately for the elucidation of written literature; and the cycle of inquiry for the anthropological folklorist was to augment the understanding of a particular culture, and/or culture in the universal sense.

In addition to their varying approach to the lore, the literary folklorists and the anthropological folklorists also had a different frame of consideration for the folk. For the literary folklorist, the folk were, for the most part, members of the peasant community. Thompson was the exception to this, since he, although a literary folklorist, studied native American material. For the anthropological folklorist, the folk were members of non-Western, tribal cultures. Yet, the literary folklorist and the anthropological folklorists did agree on one folk group, the black American. As Dundes notes in “The American Concept of Folklore,” black folklore was a combination of European and African elements (Dundes 1966a, 230). Still, the literary folklorist emphasized the European elements and the anthropological folklorist emphasized the African. So in essence, they maintained their divided orientation, even though working with the same folk group.

In both cases, for the literary and for the anthropological folklorist, the folk were set apart from the scholars themselves. The folk were to be found in bucolic enclaves for Child, untouched by time and education—in life, on the Shetland Islands; in print, in unpublished ballad manuscripts. And for the anthropological folklorists, the folk (and they were likely to be referred to as the primitive) were to be found on the Northwest Coast with the remains of a once lavish culture, or in the Southwest in the pueblos or in the hogans. The folk were for both the literary folklorists and the anthropological folklorists the other.

In their scholarly work, the literary folklorists adopted a European orientation to their subject matter, theory, and form of presentation. Frequently this link between the American scholar and the European approach was clearly stated. In his preface to The English and Scottish Popular Ballads, Francis James Child said, “In the editing of these ballads I have closely followed the plan of Grundtvig’s Old Popular Ballads of Denmark” (Child 1962, ix; see also Dundes 1966a, 240). Child corresponded with Grundtvig and sought his counsel for the five volumes of his ballad work. He also maintained extensive correspondence with Andrew Lang in Britain, Rein-hold Kohler in Germany, Kaarle Krohn in Finland, Giuseppe Pitré in Italy (McNeil 1980, 541), and was influenced by Cecil Sharp of England (Dundes 1966a, 240). Kittredge, too, identified with European standards of scholarship and maintained a correspondence with Europe’s leading folklorists. For Thompson, European folklore scholarship was crucial. As he said, “It is only by an emulation of these European colleagues that our folklore scholars can proceed with certainty and skill” (Thompson 1949, 244). As noted, both Stith Thompson and Archer Taylor traveled to Finland to meet with Kaarle Krohn. In fact, as Thompson emphasized, Finland was regarded as the center of folklore studies (Upadhyaya 1968, 112). In his memoirs Thompson noted that the trip to Finland was like a scholarly pilgrimage for American folklorists to visit Krohn, the leader of folktale studies.

The European orientation to the study of folklore was clearly evidenced at the Midcentury International Folklore Conference held at Indiana University from July 21 to August 4, 1950 (Thompson 1953). Of the four symposia, three began with addresses by European folklorists. In the first symposium, Stith Thompson, the organizer of the conference, introduced Sean O’Sullivan who spoke on the work of the Irish Folklore Commission in collecting folklore. In the second, Maud Karpeles from the International Folk Music Council, London, England, introduced Ake Campbell of the Dialect and Folklore Archive in Uppsala, Sweden, who spoke on the European systems of archiving folklore. In another, George Herzog of Indiana University introduced Sigurd Erixon of the University of Stockholm who spoke on the European study of folklore. In only one symposium, entitled “Making Folklore Available,” was the lead speaker an American. Alan Lomax, listed with the affiliation of New York City, spoke about his work with radio programs and the use of folklore.

In “The Future of Folklore Research in the United States,” Stith Thompson laid out the sequence for the study of folklore: “1. it must have its field defined; 2. the materials for study must be assembled; 3. the materials must be arranged and preserved; and 4. the materials must be interpreted” (Thompson 1949, 244). Thompson devoted a good deal of attention to the second and third category, touched on the first, and ignored the last. And in truth, this was generally true of the literary approach to folklore. The stress was on collection, classification, and preservation of text. The collection was viewed as crucial—to save the material from oblivion, to pass it on to future scholars. The classification provided a means of organizing a growing body of texts. This was done with genre classification, tale-type identification, and motif analysis. These three categories of the literary approach—collection, classification, and preservation—ultimately contributed to the establishment of folklore archives.

The Midcentury International Folklore Conference provided just such a forum for the study of folklore. The symposia encompassed the following: the collecting of folklore, archiving folklore, studying folklore (definition and theory), and making folklore accessible to the public. Referring to Thompson’s four categories listed above, one notes that three of the four were included at the Midcentury International Folklore Conference: the definition of field, the assembling of the material, and the preservation of the material. The fourth category, the interpretation of the material, was not included. In its place was another category, “making folklore available.”

Child’s ballad scholarship was predicated on a complete collection of all versions. His emphasis was always on gathering together manuscripts of ballads—collecting, annotating, and editing them. Thompson was explicit about the importance of collecting.

For folklore as for other subjects this is a continuing process and will never end. Collectors have now been busy for a century and, especially in ballads and folksongs, have published a great mass of material. They have filled many volumes of folklore journals and many cabinets with phonograph records.

(Thompson 1949, 245)

Archer Taylor noted, “The collecting of tales is, obviously enough, the prerequisite to their study, and in this direction scholars have been very active indeed” (Taylor 1940, 1).

Dan Ben-Amos, in the introduction to Folklore Genres, remarks on the link between collection and classification. “The genre orientation in folklore research is a direct continuation of the first stage of research—the very act of collecting” (Ben-Amos 1976, xi). And Dundes, in “Folk Ideas as Units of Worldview,” notes, “Once any corpus of folklore has been collected, it is to matters of genre classification that folklorists invariably turn. Obviously the exigencies of archiving have forced the folklorists to think in terms of classification and genres” (Dundes 1978a, 106). Further, the genre orientation carries over from collecting to publication, and it provides the framework for most collections, anthologies, and indexes (Ben-Amos 1976, xi). Since the classification by genre framed the work of the literary folklorists, let us examine Child’s concept of ballad, Taylor’s concept of proverb, and Thompson’s concept of folktale.

Walter Morris Hart, in an essay on “Professor Child and the Ballad,” observed, “it is obvious that Professor Child already had in mind the conception of ‘a real traditional ballad,’ a ‘specimen of authentic minstrelsy’” in his early work (Hart 1906, 799-800). Hart contrasted Child’s work on the ballad over the forty year period from his first publication in English and Scottish Ballads (1857-1859) to the final volumes of The English and Scottish Popular Ballads (1882-1898). Though the earlier conception of the ballad was not so complete and rigorous as the latter, “the significant fact is that for at least forty years Professor Child retained without essential change his conception of the traditional ballad as a distinct literary type.” For Child, the ballad was a “very important species of poetry” which preceded “the poetry of art,” a term Child used to refer to written poetry The ballad told a story in lyric verse; it was a narrative song (Hart 1906, 800, 756, 781).¹

For the literary folklorists, the genres of folklore designated certain literary types. Though these types were at times hard to define, the scholar developed a “feel” so that one genre could be distinguished from another. Archer Taylor, whose book, The Proverb (1931), is a classic, was recognized as the authority on proverb scholarship. Yet he never defined the genre: “The definition of a proverb is too difficult to repay the undertaking.” He felt that no definition would “enable us to identify positively a sentence as proverbial” (Taylor 1931, 3). Stith Thompson, the international expert on the folktale, remarked that “no attempt has ever been made to define [the folktale] exactly.” He saw this lack of definition as a “great convenience . . . since it avoids the necessity of making decisions and often of entering into long debates as to the exact narrative genre to which a particular story may belong” (Dundes 1978a, 24-25).

The literary folklorists adopted both an evolutionary and a devolutionary explanation for the origin of folklore. As Dundes explains in “The Devolutionary Premise in Folklore Theory,” while people were said to evolve, folklore was said to devolve or to degenerate: it passed from the higher to the lower classes (Dundes 1969a). And this is precisely Francis James Child’s view of the evolution of society and the devolution of the ballad: “Whenever a people in the course of its development reaches a certain intellectual and moral stage, it will feel an impulse to express itself, and the form of expression to which it is first impelled is, as is well known, not prose, but verse, and in fact narrative verse” (Hart 1906, 756).² At this stage, according to Child, the people are not divided into classes by political organization or book culture: “There is such community of ideas and feelings that the whole people form an individual.” The popular ballad, Child maintained, was “an expression of the mind and heart of the people as an individual.”

Child did not concur with Wilhelm Grimm who said that the ballads “write themselves.” An individual, after all, composed them, “still the author counts for nothing” and the ballad comes to us as an anonymous creation. This position he stated succinctly, “The fundamental characteristic of popular ballads is therefore the absence of subjectivity and of self-consciousness.” Further, the popular ballad had its origin “in that class whose acts and fortunes they depict—the upper class.” It “is not originally the product or the property of the lower orders of the people.” Society, then, reached the right stage of development, and the ballad was created—by the upper class. At this stage, there was “no sharp distinction of high and low ... in respect to knowledge, desires, and tastes.” However, with “an increased civilization” and the beginning of “book-culture,” there arose a division between the high and low classes. The high class turned away from the popular ballad to written poetry (what Child called “the poetry of art”). “The popular poetry is no longer relished by a portion of the people, and is abandoned to an uncultivated or not over-cultivated class—a constantly diminishing number” (Hart 1906, 756-57).

Taylor, while less explicit than Child, also adopted an evolutionary frame for certain genres. He identified jests as part of the “primitive narrative tradition.” And in reference to Arthur Beatty’s efforts to establish the marchen as prior to the ballad, Archer Taylor remarked, “I hasten to say that I, too, believe the ballad to be much younger than the marchen as a genre” (Taylor 1940, 20-21). In his caution, he remarked, “The question of ways and means to prove or disprove the correctness of such a belief needs serious thought.” Ralph Steele Boggs’s explanation of the development of folklore was succinct and evolutionary: “As folklore comes down through history it is modified by the cultural level in which it is found” (Boggs 1929, 9).

The European orientation which was adopted by the literary folklorists focused on the story of the text (Dundes 1966a, 243). The text was catalogued by type and broken down into components or motifs. Dan Ben-Amos discusses this early orientation to the study of folklore: “The formative years of folklore have been devoted to the construction of research tools such as classification systems, indexes, bibliographies, and annotated collections” (Ben-Amos 1973, 115). Certainly Thompson was active in these endeavors. He worked on The Types of the Folktale and the Motif Index of Folk-Literature. His work was directed, he said, toward facilitating archival work, toward building classificatory systems. Child’s work on the ballad came to be associated with a type classification—the Child Ballad Number.

Ben-Amos notes that these organizational systems were originally developed to enable the folklorist to carry out research in a professional manner. Thompson in his early work on American Indian narratives recognized the need for a general reference work on motifs. As he said in “Narrative Motif-Analysis as a Folklore Method,” there had been a lack of “adequate tools” for the study of traditional narrative. “Exact terms of reference have been needed, and a great deal of labor has been wasted because people were talking about different things under the same name” (Thompson 1955, 4). Thompson’s works on folktale classification and motif elements have become standard tools of reference for folklorists. The Types of the Folktale and the Motif-Index of Folk-Literature have enabled scholars to communicate about narrative folklore with greater facility.

The Motif-Index of Folk-Literature catalogues the basic elements of narratives from all over the world. For example, from volume 1, part A, “Mythological Motifs” comes the following:

A1.4. Brahma as creator. (Thompson 1955, 66)

Motifs are cross-referenced. Thus “A220. Sun-god” is listed in small print with “A1.1. Sun-god as creator.” Volume 1 covers “Mythological Motifs,” “Animals,” and “Tabu.” It also includes a “General Synopsis of the Index,” a bibliography and abbreviations. Volume 6 provides an index, which facilitates the search for specific motifs.

Given this elaborate six volume reference work, still the question remains, what are scholars referring to when they talk of motifs? In his article “Narrative Motif Analysis as a Folklore Method,” Thompson asked this same question. And he answered, “To this there is no short and easy answer” (Thompson 1955, 7). In order to select the material to be included in the Motif-Index of Folk-Literature, he used the following guidelines: “Certain items . . . are the stuff out of which tales are made. It makes no difference exactly what they are like; if they are actually useful in the construction of tales, they are considered to be motifs” (Thompson 1955, 7). The motif, then, cannot be defined, but is the basic element out of which a narrative is composed.

Thompson’s intent for the Motif-Index of Folk-Literature was to provide a sensible arrangement of motifs. “The Index is merely an application of well-known principles of division and subdivision and is made so that it can be enlarged at any point without limit. . . . The principal end of the Index is to find a place for everything” (Thompson 1955, 7). Even Thompson’s method of work illustrated his desire to find a place for every motif. As Dorson recalled:

[Thompson] would sit at his work table in Room 40 of the old Indiana University library, piles of motif slips laid out on the table before him, and spin a poker chip container around in lazy Susan fashion until he found the right slot for the slip under consideration. But if it seemed an isolated and questionable motif he discarded it.

(Dorson 1977c, 5)

The Motif-Index of Folk-Literature, a reference work that can expand without limit, provides the ultimate in the literary orientation to collecting and preserving folklore. Every element of every narrative from anywhere in the world can find a place in this reference work.

The reviewers of the Motif-Index of Folk-Literature recognized its strengths and its limitations. Gerould, in his review of the first edition of the Motif-Index of Folk-Literature, praised Thompson for undertaking a task that required not only fortitude but “scholarly intelligence of a high order” (Gerould 1936, 275). Kurt Ranke gave the second edition of the Motif-Index a mixed review. From his European perspective, it was “the typical product of an American mind as well as of American possibilities.” The work had been done “in grand manner in a piece of humanistic research.” However, a larger number of type indexes and handbooks had not been consulted and “no new motifs of great importance have been found” (Ranke 1958, 81, 83). Ranke remarked on the potential and the problem, “Thompson’s great achievement is the planning, organization, and execution of his opus magnum, without question an enormously respectable work which places him among the greats of our discipline. But posterity must solve the problems which he has raised thereby” (Ranke 1958, 83). And to do this, the regional indexes would have to be completely incorporated to produce “a great universal index.”

Another major contribution of the literary folklorists was the establishment of the tale-type indexes. Stith Thompson’s translation and revision of Antti Aarne’s Verzeichnis der Märchentypen (1910) is the foremost type index. As Thompson said in the preface to the second revision of The Types of the Folktale, Aarne’s original classification was directed toward “arranging the great Finnish collections of tales.” He further states that it served well for the tales of northern Europe. The countries of “southern and southeast Europe and of Asia over to India” were not included. Though attempts were made to extend the coverage in the 1928 revision, still, as Thompson himself said, the index was not inclusive. The appearance of “regional and national indexes of folktales from Russia, Spain, Iceland, and Lithuania brought suggestions of many new types” (Thompson 1961, 5). Thus the research for the 1961 revision started almost as soon as the 1928 work was published.³

The basic form of The Types of the Folktale was established by Aarne’s work in 1910. Thompson’s contributon was the translation and expansion of the index. The index categories—or as Thompson referred to them, the index numbers—establish the framework for the classification. In the first revision of The Types of the Folk-Tale, Thompson retained Aarne’s original “Outline of the Classification of Tales”: I Animal Tales—Nos. 1-299; II Ordinary Folk-tales—Nos. 300-1199; III Jokes and Anecdotes—Nos. 1200-2499 (Thompson 1928, 20-21). In the second revision of The Types of the Folktale, the outline follows the same numbering and categories, except that the third is further divided: from 2000 to 2399 fall the formula tales of part IV; and from 2400 to 2499 is part V, the unclassified tales (Thompson 1961, 20).

Each tale-type was assigned an Aarne-Thompson or AT number. On occasion, there was a further division by letter. Thus AT 314A is the tale of “The Shepherd and the Three Giants,” which is summarized as follows: “He overcomes three giants, gets three horses at their castles and with these wins a tournament three times, defeats three ogres or helps the king thrice in battle” (Thompson 1961, 110). The tale summary is followed by a listing of the motifs which are found in the tale-type. For AT 314A the following entries appear:

And finally, there is a listing of the major bibliographic sources for the tale, the countries in which the tale was located, and the number of versions found.

Thompson’s revision of Aarne’s 1910 Verzeichnis der Märchentypen was received as a welcome expansion. Spargo said of the 1928 work, “In spite of whatever faults can be found, The Types of the Folk-Tale is the most useful single work for the student of the popular tale” (Spargo 1930, 444). And Warren Roberts called the 1961 revision a “primary research tool,” a model for future indexes (Roberts 1977, 6).

Literary folklorists adopted a method for the study of folklore that combined their emphasis on the collection, annotation, and dissemination of the text. The name of this approach also indicated their European orientation. It was called the Finnish historic-geographic method. In this study of narrative folklore, the attempt was made to arrive at the original, or Ur form, and to trace the route of diffusion. The method involved the formidable task of collecting all available versions of a tale in print and ordering them chronologically, thus encompassing the historical aspect. The oral versions were then plotted according to geographic order. The scholar examined the corpus of texts and broke them down into principal traits, which were often represented by code to facilitate comparison of numerous versions. When the archetype for each trait had been determined, they were put together to comprise the hypothetical Ur form of the narrative. Basic to the historic-geographic method was the assumption that an item diffuses from the center of creation in a wave-like movement. Also implicit is the assumption that age can be correlated with distribution. A tale on the edge or periphery of the geographic mapping would be derived from and younger than the tale at the center. Thus, a trait that is more widely distributed is also assumed to be older.

Archer Taylor’s The Black Ox: A Study in the History of a Folk-Tale illustrates that approach. At the time of writing in 1927, Taylor’s intent was to introduce this “systematic method” from Finland to the English readers (Taylor 1927, 3). The principles of the historic-geographic method had been presented in Kaarle Krohn’s Die Folkloristische Arbeitsmethode (1926). Taylor wrote an article on the historical background of this approach in 1928, “Precursors of the Finnish Method of Folklore Study.” As his corpus, Taylor used more than a hundred recorded Finnish versions of the tale. “This number,” Taylor said, “is sufficiently large to enable us to determine with some degree of accuracy both the details of the normal form and the various centers of distribution.” Since there was not a great deal of variation in the form of the recorded tale, “it is possible to construct any prototypic form with more than the usual confidence.” In order to reconstruct the original form of the narrative and to arrive at “its primitive form,” each incident or trait was to be examined. The next step, then, was to put together the original incidents: “After establishing each element in this manner, we may by simple addition of the elements arrive at the outlines of the primitive form of the whole.” Taylor warned the scholar to choose versions from one country and only then gradually to extend the geographical consideration. Such a tale from one country would yield “a local ‘normal form’” and would eliminate “certain recalcitrant and apparently anomalous variations” (Taylor 1927, 1, 4, 6, 5).

Taylor emphasized that there were certain criteria that established a trait “as belonging to the normal or the primitive form”: “A trait which appears in many variants is likely to be old and to be intimately associated with the tale. A trait’s wide distribution weighs heavily in its favor” (Taylor 1927, 6). The nature of the text also gave credibility to the determination of the age of a trait. If it occurred in a “well preserved” narrative, one that was not “fragmentary or otherwise known to be corrupt or contaminated” then there was added assurance of its antiquity. Taylor noted that a trait could be considered as demonstrably old if it occurred in early versions or “implied ancient custom or superstition.” And if a trait were useful in the management of the story it was likely to be older. Another indication of age had to do with the cumulative effect of traits, those established earlier influenced those later added to the narrative. As Taylor said,

The readiness with which the varying traits may be derived from the one selected as primitive or normal may give a further indication of an earlier state of affairs, particularly when such derivation points to development in one direction and cannot be readily construed in a reverse sense.

(Taylor 1927, 6-7)

Taylor addressed the hostile critics of this method and noted that they were mainly opposed to the criteria used for determining the age of the traits. He admitted that no single factor alone was sufficient to determine age, and that “a numerical preponderance” would not establish age. But, he said, even the most hostile critic would have to concede

that a trait which is widely distributed, which appears frequently in the variants, particularly in the fuller and better ones, which is attested at an early period in the tale’s recorded history, which is per se old (involving some ancient religious or superstitious idea), which is useful in the tale’s economy, which permits competing forms to be derived from it by some natural and readily explicable alteration or substitution, which yields evidence in agreement with that deducible from other traits, and which shows a development in accord with a known or a probable cultural trend, must belong to the earliest ascertainable form of the tale. I insist upon the word must. Granted that the trait meets these requirements, one is compelled to accept it as original; no option is conceivable.

(Taylor 1927, 7)

After providing an explanation and a defense of the method, Archer Taylor presented the most common version of the tale of the Black Ox.

When he is plowing, a farmer throws a knife into a whirlwind, for he had heard that this causes the whirlwind to disappear. A Lapp orders him to come after his knife or all will not be well. The farmer arrives Christmas eve. While eating, he sees the knife and recognizes it. He wants to return for the holiday dinner. The Lapp promises to take him for the ox that stands at the door-post. The Lapp says, “Will you go like thought, like a bullet, or like a black cock?” The farmer answers, “Like a black cock.” He is put in a trough drawn by reindeer. The trough bumps into a steeple. The farmer’s cap flies off, but it is not recovered. ... At home the farmer orders the maid to substitute a smaller black ox. But the Lapp and the ox are in the trough and the bellowing of the ox is audible as they disappear.

(Taylor 1927, 16)

Abstracts of variants were given in the appendix. Taylor broke the narrative into traits, analyzed them, and determined the point of origin. “Ultimately the story is of Scandinavian origin; a closer localization is impossible with such scanty evidence, although indications favor Norway. From Sweden it was borne to Finland, where it attained a characteristic, highly elaborated form, and in Finland it passed from west to east” (Taylor 1927, 63-64).

Stith Thompson was also enthusiastic about the use of the Finnish historic-geographic method for the study of folk narrative. In his article, “The Star Husband Tale,” Thompson attempted to still the objections of two leading European folklorists. Albert Wesselski argued that literary versions had such a powerful influence on oral versions as to make the reconstruction of the original oral version impossible. And C. W. von Sydow stressed the importance of linguistic and national boundaries and opposed the notion of wave-like diffusion of the tale (Thompson 1965b, 418). Thompson examined the eighty-six reported versions of the Star Husband Tale, and in his conclusions postulated the place of origin, path of diffusion, and age of the tale. He said, “The Central Plains would seem the most reasonable place of origin for the simple tale or basic type.” As to the age, Thompson remarked, “It is, of course, quite impossible to tell just when this tale began to be told.” Then after reviewing the types of the tale and the accompanying dates of the recorded versions, he added, “It would seem from these facts that this tale in its basic form must go back at least to the eighteenth century. But that is as close as we can come to an estimate of its age” (Thompson 1965b, 455-57).⁴

The Finnish historic-geographic method was not accepted unanimously and without reservations. Both Archer Taylor and Stith Thompson in their articles on the application of the method were attempting to counter criticism and to gain approval of a wider audience. Certainly there was opposition. And at the Midcentury International Folklore Conference, some of this opposition was articulated. Thompson lead the session devoted to a discussion and a debate about the Finnish historic-geographic method (Thompson 1953, 267-86). Many of the participants expressed their reluctance to accept the validity of the reconstructed archetype, to dismiss the complicating factors of differing linguistic groups and the effect of the printed versions on oral renditions. Alan Lomax was perhaps the most candid:

I should like to ask a question which I am sure many people here would like some information on. What so far . . . are the achievements of the so-called historic-geographic method? From what has gone so far I can’t make out what the method has brought forth in terms of scientific conclusions.

(Lomax in Thompson 1953, 276)

What, indeed, were the achievements of the Finnish historic-geographic method? From the review of Taylor and Thompson’s work, it is apparent that rigor is required to collect and study all the available versions of the item under scrutiny. But after this was done, what were the results? Taylor was able to suggest that the story of the Black Ox was of Scandinavian origin, probably from Norway. And Thompson was able to tell us that the Star Husband tale probably originated in the Central Plains sometime around the eighteenth century. These results were in keeping with the stated goals of this form of study. As Thompson said,

The student using this method is attempting to find what the tale probably looked like at the beginning, before it started on its long wanderings. He is trying to find out vaguely in what part of the world it started, and also to make some guesses, not too wild guesses it is hoped, as to about what period of the world’s history it may have originated.

(Thompson 1953, 269-70)

Perhaps it is not fair to fault Taylor and Thompson and other advocates of the Finnish historic-geographic method for failing to consider that which was not part of their stated concern—the meaning of the material. However, a determination of origin and of age was among the goals of this method, and this was not accomplished with any degree of certainty. The method, though rigorous, was sterile and did not accomplish its purpose.

The final, and perhaps more serious, problem with this approach was the separation of the folk from the lore. Scholars talked of tales that migrated from country to country, but they excluded from the discussion the tellers of the tales and the effect they had on the creation and dissemination of folktales. This was part of the orientation of the literary folklorist, toward a study of the text and an elimination of the folk. Thompson said,

One assumption I am making is that the student of folklore is resolved to study the history of the particular item that he is interested in. Now this does not preclude an interest in the people who own the item or the people who tell it. It means merely that for the particular study the scholar is making, his interest is focused at least for the moment on the particular item he is working on.

(Thompson 1953, 268)

Thompson viewed this separation of the folk from the lore as a definite advantage. He referred to it as a dispassionate reading of the material; the scholar did not have to be distracted by people.

There were those among the literary folklorists, however, who did not let this exclusion of the folk occur without objection. During the symposium devoted to a discussion of the Finnish historic-geographic method at the Midcentury International Folklore Conferènce, Albert Lord, a member of the Slavic Department at Harvard University, raised an objection. “I wonder whether it is possible to arrive at any archetype of a tale or a song or an epic, if we consider that in every performance of an art form in oral tradition, whether it be a tale or an epic, the individual singer introduces variations” (Lord in Thompson 1953, 275).

For the anthropological folklorists, the concept of culture provided the crucial framing for their enterprise. As Kroeber said, “If there is a subject matter specific to our science, it is culture” (Kroeber 1940, 4). And Wirth, in his review of social science contributions to American scholarship, remarked, “Perhaps the most significant contribution which anthropology has made to social science and to popular intelligence centers around the concept of culture and the independence of culture from biology” (Wirth 1953, 64-65).

In 1871, Edward Burnett Tylor wrote what is now regarded as the classic definition of culture: “Culture or Civilization, taken in its wide ethnographic sense, is that complex whole which includes knowledge, belief, art, morals, laws, custom, and any other capabilities and habits acquired by man as a member of society” (Tylor 1871, vol. 1:1). George Stocking, in his detailed and penetrating examination of the culture concept, points to the evolutionary framework of Tylor’s definition. Culture existed in the singular and was marked off in stages. Stocking directs us to Boas for the development of the culture concept, the idea of “cultures” in the plural, representing a way of life and not a stage of development. (See Stocking 1968, 69-90, 195-233.) Boas arrived at this view of the plurality of cultures through his early work on Northwest Coast Indian folklore. Here, as he noted in 1887, he observed tribes that were distinct linguistically, but shared “so great a similarity in myths and beliefs”; and tribes with the same language that had a different body of folklore (Stocking 1968, 206). This was the beginning of Boas’s concern with the necessity for separating race, language, and culture—a concern that would become a central motivation in his work and in his teaching. As Boas explained it, his method was “to inquire into the peculiarities of the single tribes, which are obtained by a thorough comparison of language, customs, and folklore” (Stocking 1968, 206). In his introduction to the Publications of the Jesup North Pacific Expedition, Boas said, “before we seek for what is common in all culture, we must analyze each culture by careful and exact methods” (Lowie 1911, 604).

In a letter to Hodge, Boas stressed the holistic nature of culture: “the culture of each person forms a unit, all features of which are interdependent” (Darnell 1969, 179).⁵ Thus, an aspect of culture could not be meaningfully studied outside the cultural context. Boas strongly opposed Otis T. Mason’s arrangement of museum exhibits according to item, without regard for cultural integrity. As Boas said in 1887, “By regarding a single implement outside of its surroundings, outside of other phenomena affecting that people and its productions, we cannot understand its meaning” (Lowie 1948, 69; see also Jacknis 1985).

Though there were different domains for the literary folklorists and the anthropological folklorists—one deriving from literature, the other from culture—still there were shared concerns and similar approaches. For the anthropological folklorists, as for the literary folklorists, the collection of material was of primary importance. From the multitude of texts, the literary folklorists attempted to reconstruct the original text. And from the detailed linguistic texts, the anthropological folklorists tried to reconstruct the traditional culture—or as Lowie called it “reproduce [the] ancient culture” (Lowie Papers, Lowie to Risa Lowie, 6/23/1914). For both, the study of the material was largely devoted to a discussion of diffusion and an analysis of traits or motifs.

Gladys Reichard noted the importance of texts for Boas: “The strongest rocks in Boas’ self-built monument are his texts, his belief that what people record of themselves in their own words will in the last analysis reveal their motivations and ideas most accurately” (Reichard 1943c, 55). Boas recorded the tales phonetically in the native language, and translated them interlineally. He was emphatic about the importance not only of collecting the texts in the native language, but also of publishing them. As he remarked, he came to his first disagreement with Newell over just this issue:

I consider the publication of the Indian texts essential. Just as little as you would be satisfied with having only translations of European material, can we be satisfied with having Indian material in translation only. Even if the present use of Indian texts is limited to a few people, they are indispensable.

(Boas Papers, Boas to Newell, 3/28/1906)

Boas’s primary emphasis was on collection, or what he would term the complete description of a culture, which had to precede the formulation of general theories. In “The Growth of Indian Mythologies” (1896), Boas said, “If we want to make progress on the desired line, we must insist upon critical methods, based not on generalities but on each individual case” (Boas [1896] 1948, 435). The study of Indian mythology, and indeed by extension, the study of anthropology was based on the accumulation of individual cases, or information on single cultures.

However, Boas’s stress on the particulars did not mean that he was inordinately opposed to theory. As he wrote to Kroeber, “all through my life I have discussed theoretical questions in connection with definite problems” (Kroeber Papers, Boas to Kroeber, 7/24/1917). Boas advised Kroeber to present the theory separate from the “detailed investigation,” because “evidently no one reads this kind of material, but expects to get his theoretical point of view from general theoretical works.” On this theme, Walens remarked, “it is a perennial mistake of anthropologists to believe that Boas’s eclectic idea of culture prevented him from making any summary statements about Kwakiutl culture” (Walens 1981, 18).

Boas’s early work in folklore shaped his later work in anthropology, especially in linguistics. (See Hymes 1963; Rowe 1962; Stocking 1974c; Darnell 1969, 338.) This is apparent in two articles in which Boas sets forth his method for the study of distribution: his 1891 article, “Dissemination of Tales among the Natives of North America”; and his 1896 article, “The Growth of Indian Mythologies.” Ruth Benedict in “Franz Boas as an Ethnologist,” said that his 1896 writing “was basic to all his work in this field. Dissemination of traits was one of the processes that always had to be taken into account before it was possible to understand the working of the human mind in its cultural creations.” As Benedict said, the study of diffusion “was not an end in itself nor did it by itself furnish the key to the understanding of culture.” But for Boas, it was a necessary precedent in order to understand how “each human group built up its own version of life and code of behavior” (Benedict 1943, 29).

In “Dissemination of Tales among the Natives of North America,” Boas discussed narratives that were similar but widely dispersed among the native peoples of North America and that were shared among the Eskimos and Indians. After the presentation of the material, he remarked, “From these facts we conclude that diffusion of tales between the Eskimo and Indian tribes of the western half of our continent has been quite extensive.” And if tales have spread, then, Boas suggested, other cultural elements as well “have spread from one center over the Arctic and the North Pacific coasts” (Boas [1891] 1948, 441, 443).

“The Growth of Indian Mythologies, a Study Based upon the Growth of the Mythologies of the North Pacific Coast,” was delivered first as an address to the seventh annual meeting of the American Folklore Society, in Philadelphia, on December 27, 1895. In this, Boas had two concerns. First, he was arguing against the theories that explained similar elements in widely dispersed folklore, that is, solar mythology, radical diffusion, parallel development, and psychic unity. And second, Boas was presenting his own position, that “similarity of traditions in a continuous area is always due to dissemination.” While willing to concede to the solar mythologists that “the grandeur of nature upon the mind of primitive man” might influence thematic elements of myths, he emphasized that it could not explain the form that myths would take (Boas [1896] 1948, 432, 433). Boas was also willing to admit the possibility of independent origin or parallel development in certain circumstances. For instance, when a tale, composed of a single incident, was found in numerous cultures from around the world, then this tale might have been invented independently in each culture. As he noted in his earlier article on the dissemination of tales, “The tale of the man swallowed by the fish, or by some other animal, which has been treated by Dr. E. B. Tylor is so simple that we may doubt whether it is due to dissemination” (Boas [1891] 1948:437, 438). However, when a tale was composed of several elements that were combined in a similar manner, then the explanation could not be parallel development or independent invention, but must be diffusion of the tale from one culture to another. Boas’s emphasis was on diffusion—or dissemination as he called it—within a contiguous or limited area. He was not interested in discussing diffusion on a global scale, such as Schmidt and Rivers attempted with the Kulturkreislehre.

The diffusion of tales and other aspects of culture would, according to Boas, ultimately illustrate the connection between tribes, including their migration and their place of origin. In his later work in folklore, Boas developed the idea of the culture reflector method, that folk narratives mirrored aspects of the culture. As he said in the section on “Mythology and Folklore” in his textbook General Anthropology, “If it is true that myths are built on the experiences of everyday life, we may expect that the dominant cultural interests are reflected in them. The incidents mirror the life of the people and their occupations, and social life may in part be reconstructed from these tales” (Boas 1938, 622). Thus, through studying folklore, one could see a portion of the culture. This was especially valuable for cultures that were so changed by extensive white contact as no longer to be traditional. In his Tsimshian Mythology and Kwakiutl Culture as Reflected in Mythology, Boas used narrative folklore to reconstruct the past.

Unlike the scholars who codified the Finnish historic-geographic method, Boas did not formulate an explicit method for the study of diffusion in narratives. He did, on occasion, provide an indication of his assumptions about diffusion, as is evident in the following statement: “Whenever we find a tale spread over a continuous area, we must assume that it spread over this territory from a single center” (Boas [1891] 1948, 439). This assumption was basic to the age-area hypothesis in anthropology. In this approach, as in the Finnish historic-geographic method, the movement or diffusion of traits was conceptualized as concentric circles or waves, like ripples on water made when a rock is thrown. And just as with ripples, a trait diffused from a central point, unless an obstacle lay in the path.⁶

In order to study the diffusion of narratives, the anthropological folklorists used another approach similar to that used by the literary folklorists. It was necessary for Boas to break a narrative down into elements. As he explained, “A single element may consist of a number of incidents which are very closely connected and still form one idea” (Boas [1891] 1948, 438). Boas listed the elements of the dog-husband story of the Dog-Rib Indians from the Great Slave Lake as follows: “1. A woman mated with a dog. 2. Bears pups. 3. Deserted by her tribe. 4. Sees tracks of children. 5. Surprises them. 6. Takes their skins.⁷. They become a number of boys and one girl.⁸. They become the ancestors of a tribe of Indians” (Boas [1891] 1948, 438). These eight elements from the narrative of the Dog-Rib Indians have been combined in a similar way by the Indians on Vancouver Island. Boas remarked that “The elements may have arisen independently in various places, but the sameness of their combination proves most conclusively that the whole combination, that is, the story, has been carried from Arctic America to Vancouver Island, or vice versa” (Boas [1891] 1948, 438).

Boas was determined that others would work on this project. On August 31, 1907, he wrote to Kroeber, “Have you been able to do anything about the catch-words concerning which I wrote last spring? I am very anxious that our committee should present a report at the Christmas meeting and should be very much indebted for contributions” (Kroeber Papers, Boas to Kroeber, 8/31/1907). Boas mentioned Ehrenreich’s Mythology of South America, Stucken’s Astral Myths, and Frobenius’s Das zeitaltendes Sonnengottes as works that Kroeber might consult. Boas again corresponded with Kroeber on the subject, “It is my idea to try to devise a number of catch-words for discussion by the Committee when we meet this winter, or, probably better, later on by correspondence. ... I will send you a short list which will indicate perhaps more clearly what I mean” (Kroeber Papers, Boas to Kroeber, 9/17/1907).

The concept of narrative elements was developed further by Robert Lowie in an article entitled “Catch-Words for Mythological Motives”: “The advantages of uniform terminology—of brief, unequivocal designations for wide-spread elements which are constantly referred to in mythological discussions—are obvious” (Lowie 1908a, 24). Lowie had compiled the list “at the suggestion of Professor Boas.” His intention was to use catch-words which were already sanctioned by usage. In one category, “Catch-Words in General Use, or already Suggested,” Lowie gave twenty-nine categories. Included among these were the following: Orpheus, visit (journey) to the sky, bear and deer, theft of fire, rolling skull (head), rolling rock, evil father-in-law, tar-baby, world-fire, turtle’s war party. He also gave thirty-two “proposed catch-words” (Lowie 1908a, 24-27).

Lowie had hoped to stimulate “revision and collaboration” (Lowie 1908a, 24). Kroeber responded in “Catch-Words in American Mythology,” a paper delivered to the California Branch of the American Folklore Society. He referred to Lowie’s article and then credited Boas with the stimulus for pursuing this attempt at classification: “The idea has for some years been agitated by Dr. Boas, to whom credit is due for the realization of the value of this method of approaching the problem of handling so large a body of material” (Kroeber 1908a, 222). Kroeber offered a few modifications to Lowie’s catch-words, and then provided a “designation of concepts occurring in myths and tales of the California Indians.” Kroeber’s catch-words include a description:

1. Theft of sun (luminaries, light).
Analogous to theft of fire. The Yurok tell also of the theft of night, of water, of food.

2. Creation in a vessel (under cover).
Creation of men, or animals, in a basket, from a bundle, under a blanket.

3. Skunk-shaman.
The skunk, pretending to be a shaman, kills his patient by shooting.

4. Abandonment on tree.
Son or younger brother abandoned on tree which grows upward or the branches of which are blown away.

(Kroeber 1908a, 223)

His list continued with a total of fifty-three concepts that he had found useful in his study of California mythology.

The work of Kroeber, Lowie, and Swanton on catch-words of American Indian narratives was part of a larger plan, that of compiling a myth concordance. Lowie referred to “Professor Boas’ ” suggestions as spurring him on in his work. Kroeber noted Boas’s “agitation” on this subject. Actually the plans for a reference work on American Indian mythology had been present from the early days of the Bureau of American Ethnology. As Darnell notes, “One of the Bureau’s first projects was an abortive attempt to classify the mythology of North America. The First Annual Report included a summary of American mythology written by Powell. . . . The concordance was pursued by Jeremiah Curtin between 1883 and 1892” (Darnell 1969, 59). Curtin’s plan for his work on mythology was to “reveal the mental systems of various groups of American Indians” (Darnell 1969, 59). He would use his work in mythology to aid in his classification of “these nations at successive periods.”

Frank Boas suggested a collaboration between the American Folklore Society and the Bureau of American Ethnology on a general work in American mythology (BAE, Boas to Henshaw, 12/6/1890; Darnell 1969, 290). Boas envisioned this on a grand scale with the following people responsible for specific groups: John Murdoch (Eskimo), Garrick Mallery (Atlantic Algonkian), Daniel Brinton (southern and central Algonkian), Reverend McLean (Blackfeet), Horatio Hale (Iroquois), James Mooney (Cherokee), A. S. Gatschet (southeastern tribes), J. O. Dorsey (Sioux), John W. Powell (Shoshone), Jesse Walter Fewkes (Zuni), Washington Matthews (Navajo), John Bourke (Apache), Jeremiah Curtin (California), Alice Fletcher (Sahaptin), Boas (Northwest Coast). As Darnell noted, “The plan was abortive for a combination of reasons, including lack of publication funds and difficulty of eliciting appropriate material from the would-be contributors” (Darnell 1969, 290).

Though this work did not materialize, still Boas did not give up the idea. He headed a committee of the American Anthropological Association which included Dorsey, Kroeber, and Swanton. Working also through the American Folklore Society, Boas appointed Swanton as head of the committee composed of Boas, Swanton, and Dixon. Though serving on both committees, Swanton did not know how to proceed. He wrote Boas, “As yet I haven’t a clear idea of the method of identifying myths by catch words, and should like an example” (Boas Papers, Swanton to Boas 8/3/1907). Boas expanded on the concept in a letter to Swanton:

The principal thing is to find a very simple catch-word which represents either an incident in a myth or a whole type of myths; for instance, the numerous tests of son-in-law or nephew or suitor might be called “test myths,” or the tales of the ascent to the sky by a chain of arrows might be called “arrow-chain.”

(Boas Papers, Boas to Swanton, 8/10/1907)

Dixon caught on and compiled “a practically complete list” with ease: “Cannibal-head; Canoes of enemy bored with holes; Time miraculously shortened; House miraculously enlarged; Pitch on belly to imitate fat. ...” However, there was still the question of judgment: “how minute do you wish the analysis to be?” (Boas Papers, Dixon to Boas, 9/30/1907). Dixon asked whether the catch-words should be large elements or small.

The project was overwhelming. The quantity of material made the organization difficult. And, even more central to the problem, there was disagreement as to the method of classification. Swanton despaired of finding a “connection between more than a thousandth part of our myths”; but even such minimal resemblance would be “sufficient for some sort of classification that is vastly better than our present chaos” (Kroeber Papers, Swanton to Kroeber, 9/23/1910).

In his 1907 article, “A Concordance of American Myths,” Swanton addressed both the benefits and the difficulties of such an undertaking. He foresaw “great additions ... to the fund of human knowledge” should a suitable method be devised for organizing all North American myths into appropriate categories. “The accomplishment of such a work would involve the compilation of a concordance in which all the genuine unaffected American myths should find a place.” He anticipated three obstacles: “(1) the difficulty of classifying accurately and uniformly throughout, (2) the difficulty of knowing where to draw the line in admitting myths and mythic elements, and (3) the danger of making the work so cumbersome as to be practically useless” (Swanton 1907, 220).

In 1909, Swanton delivered his presidential address to the American Folklore Society on “Some Practical Aspects of the Study of Myths.” He referred again to the importance of compiling a concordance of American myths: “Here it is proposed to classify all myths under types, each type to have some suitable catch-word; i.e., its technical term in the science, such as ‘magic flight,’ ‘Potiphar,’ ‘rolling-stone,’ ” (Swanton 1910, 5).

Along with the classification of “myths and larger elements,” Swanton stressed the need to study “mythic formulae” (Swanton 1907, 221). These were “conventional modes of expression and conventional mythic ideas” that were unique to a tribe.

Thus, among the Haida a mythic town is described by saying that it was “a five-row town,” while the neighboring Tlingit call it “a long town.” Among the Haida, again, an incarnate deity indicates his supernatural origin by clamoring for a copper bow and arrows, but his Tlingit counterpart is content to “hunt all the time.”

(Swanton 1907, 222)

Swanton called the compilation of a myth concordance “a crying need in the further study of folk-lore and anthropology.” He was enthusiastic about the prospects, but recognized the difficulties in bringing together such a work. He concluded his article, “A Concordance of American Myths,” with the following questions: “(1) is such an undertaking practicable; (2) how comprehensive can a concordance be made without destroying its usefulness, or in other words how much shall be included in it; and (3) granted its desirability, what steps can be taken by anthropologists and students of folk-lore to make it an actuality” (Swanton 1907, 222).

Waterman, who had written his dissertation at Columbia on the explanatory element in American mythology, hoped to pursue a myth concordance. Waterman was under the impression that the Bureau of American Ethnology had been working on a classification of myths, and he wrote to Hodge to inquire. He suggested “a concordance of occurrence and diffusion of every tale” (Darnell 1969, 293).⁷ Kroeber learned of Waterman’s interest and wrote to him: “Your plans for a handbook on mythology interest me greatly but I fear you will find it a bigger undertaking than you imagine. I have played around the fringe of the problem enough to be impressed by it” (Darnell 1969, 293).⁸

In 1913, Stith Thompson wrote to Boas to inquire about the myth concordance. He told Boas of his dissertation topic on “European Borrowings in American Indian Tales,” and asked if there was “any reason why the investigation of this topic would not be advisable” (Boas Papers, Thompson to Boas, 5/29/1913). Boas responded to Thompson, “I am interested to hear that you are taking up the question of borrowing of European folklore in America” (Boas Papers, Boas to Thompson, 5/29/1913) and advised him concerning current research.

The anthropological folklorists continued their interest in a catch-word concordance. On November 5, 1926, Boas wrote to Parsons, “Since you left we have been considering the question of the Concordances” (Boas Papers, Boas to Parsons, 11/5/1926). Parsons took this up as a project and directed it specifically toward a concordance of Southwest mythology (Boas Papers, Boas to Reichard, 11/4/1926). However, in spite of the enduring enthusiasm, the plans for a concordance of Indian mythology were never realized, perhaps because of the nearly insurmountable difficulties (Kroeber Papers, Kroeber to Swanton, 9/3/1910).

Thompson carried on in a tenacious fashion. And in a sense, Thompson’s work was a continuation of the anthropological work on catchwords. Thompson on several occasions acknowledged his debt to anthropologists. In his letter to Kroeber, he had characterized his dissertation as “about equally in the fields of comparative literature and anthropology” (Kroeber Papers, Thompson to Kroeber, 7/5/1914). He also noted that in his attempt to make a distribution map for North American Indian tales he was aided by Clark Wissler’s map of culture areas which arranged “tribes by culture rather than linguistic families” (Thompson 1953, 271).

While the literary folklorists compiled collections and anthologies according to genres, the anthropological folklorists were likely to compile their material either by focusing on a single culture or on several cultures. Boas’s Kwakuitl Culture is an example of the encyclopedic approach to the study of one culture, with folklore as an integral part of the study. Gladys Reichard in writing on “Franz Boas and Folklore” stressed this nature of his work:

The title Franz Boas as Folklorist might seem proper but would actually be inappropriate. Professor Boas did not study folklore because he was a “folklorist” nor by studying it did he become one. He used it as an important part of the whole which he envisioned; a description of the tribes in which he was interested and an interpretation of their culture.

(Reichard 1943c, 52)

The anthropological folklorist might also study a single aspect of folklore as it was manifested in several cultures. Elsie Clews Parsons’s Pueblo Religion and Ruth Benedict’s “The Concept of the Guardian Spirit in North America” exemplified this comparative approach.

On occasion, the anthropological folklorist focused on a genre of folklore. For example, Paul Radin wrote The Trickster, a Study in American Indian Mythology (1956) and Ruth Benedict produced two volumes on Zuni Mythology. However Radin’s and Benedict’s work were not intended as investigations of folklore genres—a concept that was not used by the anthropological folklorists—but as explorations in cultural patterning. In the prefatory note to The Trickster, Radin said, “Manifestly we are here in the presence of a figure and a theme or themes which have had a special and permanent appeal and an unusual attraction for mankind from the very beginning of civilization” (Radin 1972, xxiii). Benedict’s intent in Zuni Mythology was to explore the pattern or theme in Zuni culture through a study of their mythology. And mythology was conceptualized more as a system of belief, that is, customs and practices and sacred narratives, than as a collection of narratives.

The emphasis on cultural patterns was critical for the shift that took place in American anthropology, a shift from an earlier focus on detailed description to a focus on theme and meaning. Thus folklore manifested certain themes that resonated in other areas of culture as well. In Benedict’s terms, it was part of the personality of the culture.

Ruth Benedict, Paul Radin, and Melville Jacobs were the innovators in this new approach to folklore. And their contributions had an impact that is still reverberating in folklore theory. The three were characterized by Lowie—rather coolly, it should be noted—as scholars of folklore who were “not averse to psychoanalytic interpretation” (Lowie 1960, 467). Together they sparked the psychological interpretation of folklore.

This innovation in folklore studies combined three approaches: the psychological, the cultural, and the folkloristic. Radin’s remarks in the opening to The Trickster clearly show this:

Our problem is thus basically a psychological one. In fact, only if we view it as primarily such, as an attempt by man to solve his problems inward and outward, does the figure of Trickster become intelligible and meaningful. But we cannot properly and fully understand the nature of these problems or the manner in which they have been formulated in the various Trickster myths unless we study these myths in their specific cultural environments and in their historical settings.

(Radin 1972, xxiv)

Fieldwork has long been regarded as a distinctive part of the anthropological endeavor. And the stress on the importance of fieldwork has been present in American anthropology from the early years: “The Bureau took the fieldwork of its staff extremely seriously. When Frank Cushing needed a scalp for his initiation at Zuni, it was provided from the museum collection” (Darnell 1969, 68).

Franz Boas, in his own research and in the training of his students, also encouraged fieldwork. He worked among the Eskimo of Baffinland from 1883 to 1884, and among the Indians of the Northwest Coast for nearly two and a half years (Rohner 1966, 159). Alexander Lesser notes another of Boas’s contributions to fieldwork. Referring to the Torres Straits Expedition of the British under W. H. R. Rivers and A. C. Haddon, Lesser says, “somewhat overlooked in this historical view are the monumental field researches of the Jesup North Pacific Expedition, which was mounted under Boas’ inspiration, direction, and editorship while he was at the American Museum of Natural History and at Columbia” (Lesser 1981, 11). The Jesup Expedition, which was contemporary with the Torres Straits Expedition, resulted in seventeen volumes on Siberia and the Northwest Coast. Triloki Pandey noted the importance of the American anthropologists to the development of fieldwork. “One should not belittle the importance of earlier field workers such as Boas and Cushing among others, in making anthropology an observational science” (Pandey 1972, 322, n8).

If culture was, as Kluckhohn said, like water for fish, then fieldwork was cultural immersion. The assumption was made that through the process of immersion, the anthropologist would learn the culture through the voices, eyes, and lives of the people. For the anthropological folklorists, this meant studying the folklore in context. As Voegelin said, the anthropological folklorist was “interested in the function of folk tales, in the various other interpretations of folk tales in their setting, and in their meaning to the teller” (Voegelin in Thompson 1953, 283).

While fieldwork has been long associated with the anthropological approach to the study of folklore, still it would be a skewing of intellectual history to use this as the major point of division between the literary and the anthropological approaches. There has been just such a tendency to equate the literary approach with library research and the anthropological approach with field research. While this might frequently be the case, still it does not take into account those literary folklorists who recognized the importance of fieldwork for capturing the life of the material. Phillip Barry was such a person. He emphasized the need to go to the folksingers themselves. And because Barry did this, he knew that, contrary to what the academicians were saying, ballads were still present as songs, and were not just recited as poems. He was also cognizant of more pressing problems for ballad study than the tired consideration of the communal theory (Alvey 1973, 71).

Another scholar who belied the equation between the literary approach and library research was Milman Parry. As Bynum noted in his work on oral literary studies at Harvard, “Though a Classicist by profession, he preferred to think of himself as a professional hybrid—a ‘literary anthropologist.’ It was an apt expression” (Bynum 1974, 28).

From the appraisal of the literary and the anthropological approach, one can see the shared ground. The anthropological folklorists were early concerned with the dissemination of narratives. In order to study this dissemination, they attempted to compile an index of catch-words and a concordance of myths. Thompson, along with other literary folklorists, began a study of the diffusion of folktales and an analysis of motif elements. In his work, Thompson was influenced both by the work of the American anthropologists and the European folklorists. His work on American Indian narratives was a continuation of the work done by Boas and his students. Contemporary with the work done by the anthropological folklorists in the United States, was the work in Finland on tale type. While Boas and his students attempted to compile a myth concordance, Antti Aarne was compiling his tale-type index. In 1913, Thompson wrote to Boas about the possibility of pursuing work on a myth concordance. And Thompson did just that. In 1924, he was invited by Kaarle Krohn to continue the work of Antti Aarne. In his work on the Motif-Index of Folk-Literature, he pursued in expansive detail the beginning attempts by Boas, Lowie, Swanton, Waterman, Dixon, and Kroeber to index elements or catch-words for American Indian narratives. Thompson chose the folk literature of the entire world for his corpus. Thus, in a real sense, he carried on work in the early tradition of the American anthropologists and the Finnish folklorists.

Both the literary folklorists and the anthropological folklorists were concerned with the collection of a large quantity of material. For the literary folklorist, the emphasis had been on the collection of folklore and the classification according to genre or tale type or motif. For the anthropological folklorists, the emphasis was on the collection of the material—in the earlier years, on the reconstruction of culture history; and in the later years, on pattern and meaning. The starting point for the two was different. For the literary folklorists, it was literature; for the anthropological folklorists, it was culture. But the cognitive approach to the material was strikingly similar: it was the collection and classification of folklore.

The literary folklorists in their study of the diffusion of a narrative focused on the narrative itself. It was the text that interested them. And it was the origin of the text that concerned them. They asked: What is the Ur form and from whence did it come? For the anthropological folklorists, the text was a conduit. And their questions were phrased: What does the narrative tell about the people and from whence did they come? The focus was different: for the literary folklorists, the text; for the anthropological folklorists, the people. But the underlying framework sustained the same assumptions. Through the study of diffusion, age and origin could be ascertained. The original narrative could be reconstructed; and the traditional culture could be described.

The anthropological folklorist and the literary folklorist have shared more common ground in their study of folklore than has been recognized, either by intellectual historians or by the folklorists of the two persuasions themselves. It cannot be denied that there are very significant differences between the two. However, under close scrutiny, one realizes that there are also shared concerns and shared orientations. The two roots of American folkloristics feed from a common source, the material of folklore. And both the literary and the anthropological approaches shared in the intellectual orientation of the time, which were the concerns with the point of origin, with the past, and with the original.

To close the consideration of the orientations to folklore, Archer Taylor’s critique of the work in folktales and Melville Jacobs’s suggestion for a combined scientific and literary approach should be examined. Archer Taylor in “Some Trends and Problems in Studies of the Folk-Tales,” concluded, “The quality of the essays devoted to describing the collecting of tales and to interpreting the significance of tales in folk-life is depressingly low and rarely rises above the level of conversation at an afternoon tea” (Taylor 1940, 23). He suggested that a freshness might be gained if “the study of the forms of tales is allied to the classification of tales.” And he added, “enlargement of our knowledge of the forms of tales would improve our systems of classification” (Taylor 1940, 12-13). Taylor wanted to take the work in narrative beyond Thompson’s Gerüst or framework (Upadhyaya 1968, 113), beyond a mere classification of motifs and types. He was suggesting a combination of type and form, a combination that Vladimir Propp had accomplished in his work Morphology of the Folktale. Though written in 1928, this was unknown to the English-reading public until the 1958 translation. Taylor, attuned to structure through his work on formula tales, was also anticipating the approach taken by Lee Haring in Malagasy Tale Index, a classification of tales according to form or structure. Using von Sydow’s work in oikotype as an inspiration, Taylor suggested that comparisons could be made between countries for differences in meaning (Taylor 1940, 14). Taylor’s critique indicated his dissatisfaction with the work of the literary folklorists and his desire to move beyond mere classification to form and meaning.

Melville Jacobs posed another challenge for the anthropological folklorists. He asked, “Can both scientific and literary function be combined in publication of folktales and still adhere to rigorous standards?” And he answered, “I believe that they can, if interpretive comments supplement the stark translations and if the additions are written so as to include, readably and pleasurably, associations and sentiments which the native audience experienced” (Jacobs 1959b, 124). Jacobs in this statement was proposing a blending of humanities and science, a combined approach suggested by Ruth Benedict a decade earlier. He was also anticipating the development of performance theory in folklore. The text could be combined with the context, and both could add meaning to the other. His query and response provide a tantalizing suggestion of what is to be gained through a combination of the literary and the anthropological approach to the study of folklore.