Spread of folklore motifs as a proxy for information exchange: contact zones and borderlines in Eurasia.

Berezkin, Yuri

1. The database and its analytical unit

Hundreds of thousands of folklore texts have been published since mid-19th century. To systematize this huge volume of data, two international systems of classification were created, using either tale-types (Aarne 1910, Aarne and Thompson 1961, Uther 2004) or elementary motifs (Thompson 1955-1958) as basic units.

The tale-type was originally understood as narrative plot with a more or less precise origin in space and time. This idea was severely criticized (Jason 1970), so now the ATU (Aarne-Thompson-Uther) tale-types mostly play a role of reference points in search of parallels for particular texts. There are several reasons why ATU index is impossible to use for historical studies, i.e. for assessing a degree of similarity/dissimilarity between folklore traditions. This index is Eurocentric so its use for sub-Saharan Africa, Siberia, Southeast Asia and Oceania is restricted while Australia and America are completely beyond its scope. Ethnic attribution of texts is systematically provided only for Europe. For other areas it is absent or practically absent not only in the reference index itself (Uther 2004) but even in some regional indexes that use ATU system (e.g. El-Shami 2004, Thompson and Roberts 1960, Ting 1978). In many cases sets of episodes found in particular variants of the same tale-type are so different that it makes impossible to assess the degree of similarity between particular texts without consulting the original publications. There are relatively many mistakes, and to correct them we need a database that would contain detailed abstracts of texts and not only a list of publications cited for every tale-type.

Concerning the index of elementary motifs, it was created with a declared aim to hold aloof from any historical problematic (Thompson 1932:2). The aim was to reduce any text to a kind of standard combination of 'characters', i.e. elementary motifs. Descriptions of motifs were intentionally deprived of details, wordings like 'origin of frog' (A2162), 'dwarfs in other world' (F167.2), 'self-mutilation' (S160.1) being typical. It is symptomatic that an expert can easily extract a set of registered motifs from a given text but it is usually impossible to restore a content of any real text on the basis of the set of motifs extracted.

Because my purpose was not to suggest another universal typology but to apply the mass folklore material for study of past migrations and interregional contacts, a classification unit adequate for such a research had to be found. Not to coin a completely new term, I named such units 'motifs' which are defined as any features or combinations of features in folklore texts (images, episodes, sequences of episodes) which are subject to replication and found in different traditions. Those motifs that are known universally or widespread chaotically across the world have no interest for our research and have not been included into the catalogue (http://www.ruthenia.ru/folklore/berezkin). On 15.01.2015 the database contained ca. 50,000 abstracts of texts providing information on the spread of 1963 motifs according to 914 traditions from all over the world, each tradition usually corresponding to particular language or dialect. For Melanesia and partly for the Northeast India, Indonesia, Tropical Africa and Amazonia several ethnic traditions are sometimes merged together either because of the shortage of data or because groups in question are small and share similar culture.

2. Categories of motifs and representativeness of the data

Among units of folklore texts subject to replication are motifs-images and motifs-episodes. Motifs-images can correlate with some elementary motifs of S. Thompson and are mostly related to cosmology and etiology. Motifs-episodes which sometimes find parallels in ATU tale-types are mostly related to narratives about adventures and tricks. All correspondences with existing folklore indexes are, however, approximate. My wordings of motifs are composed in such a way that all the texts cited under a corresponding item should contain all the details noticed in the definition. Otherwise statistical processing of the material would give distorted results. Motifs related to cosmology and etiology are ascribed to category A and related to adventures and tricks to category B. There is no precise borderline between both groups of motifs but the processing of the data demonstrated that this two-fold division makes sense. These two sets of motifs often show different tendencies of areal distribution and were probably disseminated with different speed and under the influence of different factors (Berezkin 2005).

One of the most important conditions for the creation of the global database suitable for the study of prehistoric migrations and cultural interactions is a uniform degree of intensity of research. It is practically impossible to consider all texts that had been ever published but the completeness of data must be more or less uniform. This uniformity should be on a regional scale if we compare sets of ethnic traditions selected for these regions. And if we compare just separate traditions and not their regional agglomerations, every tradition should be equally well represented in the database. The number of registered motifs for particular tradition depends on its objective richness, on intensity of the field research and on the availability of published or archive materials. For example, at the moment there are 234 registered motifs for the French, 279 for the Estonians (without the Setu), 234 for the Mari, 267 for the Armenians, 192 for the Tajik, 252 for the Buryats but only 49 for the Maltese, 50 for the Tabasarans (with the Aguls, Dagestan), 50 for the Yazghulami (Pamir) and 72 for the Shors (Southern Siberia). In such a situation it would be incorrect to compare the Tabasarans with the Armenians or the French with the Maltese but possible to compare the Western Mediterranean Europe with Eastern Europe, the Caucasus, Central Asia and Southern Siberia. Because the purpose of this paper is the search of the major tendencies of the spread of the motifs across the entire Old World, the uneven amount of available data on particular traditions cannot significantly influence major conclusions. Our catalogue based on data extracted from more than 6000 publications in a dozen European languages contains comparable information on folklore traditions of the whole world. Two major areas that remain underrepresented (big Dravidian-speaking nations of South India and traditions of Africa along the borderline between Sahel and Sahara in Niger, Chad and Sudan) are not crucial for topics discussed in this paper.

3. Results of factor analysis: 1st and 2nd principal component: 'Europe' against 'Asia'

For the present study the areal distribution of 615 motifs was statistically processed. These are adventure and trickster motifs typical for Nuclear Eurasian and North African tales. Motifs widespread in sub-Saharan Africa, Siberia, Southeast Asia and Oceania but absent in Nuclear Eurasia were ignored. The New World materials are also beyond the scope of the present research. Only traditions with 20 or more adventure and trickster motifs were considered, so the number of traditions processed (339) is less than the total number of the Old World traditions in the database (503).

As mentioned above, motifs that correspond to episodes of adventure and tricks form the category B while motifs related to cosmology and etiology (the origin of the universe and its different parts and elements) form the category A. Because most tales that contain the B-motifs (animal tales, fairy tales and realistic tales of Western Eurasia as well as different sorts of adventure and trickster tales in other regions) are not strictly related to particular ethnic worldviews, they are relatively easily borrowed moving across all kinds of social and natural borders. Patterns of their areal spread can be taken as proxies for the intensity of communication and information exchange between groups of people. Such an exchange always took place but in the Old World it probably greatly intensified during the last two thousands years or so because of demographic growth, the rise of ever bigger states and empires, intensification of transcontinental trade, etc.

Factor analysis is a statistical method that allows evaluating the variability among the observed, correlated variables, in our case the intensity of mutual combination of motifs within many traditions. From all the totality of the motifs every principal component (PC) selects two groups of motifs that are the most different from each other, one with positive and another with negative mathematical values. The higher the absolute value, the better the tradition in question represents the corresponding tendency of the areal spread of motifs. The overall number of the PC can be as big as the number of traditions processed, but with large and heterogenic material practically only the first two or three PC reveal important tendencies. In our case these first three PC are responsible for less than 17 percent of total information. However, namely this information reflects major transcontinental tendencies and not relations between traditions on a local level.

The 1st PC reveals the most dominant tendency and contrasts Europe, North Africa, Western and Central Asia with the rest of the Old World. This result was predictable because materials not represented in Nuclear Eurasia simply have not been computed. Attention should be drawn to some details, however. The hypothesis about India as a 'homeland of fairytales' (Benfey 1859), popular during decades, is not supported by the present research. Even such rich and well studied South Asian folklore traditions as the Santali, the Sinhalese, the Marathi and some others including the Hindi- and Chhattisgarhi-speaking groups of Northern and Central India, contain fewer motifs known in Europe than the traditions of the Steppe Belt of Eurasia. Were South Asia an important center of the formation of the narrative plots which later spread across the world, we should expect a greater diversity of the regional folklore and more correspondences with materials from Western and Central Eurasia. China shares even fewer motifs with Europe and Western Asia than India and much fewer than Korea and Japan. Japanese folklore contains clear European borrowings absent in China but the folklore exchange between the Chinese and the Altaic traditions was also rather weak.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The 2nd PC selects two major groups of traditions inside Nuclear Eurasia. Oral stories of these groups contain sets of motifs most different from each other.

Traditions located between the Caucasus and Mongolia with adjacent Siberia form one group while traditions of Western Europe and the Mediterranean with adjacent Africa form another group. In sub-Saharan Africa corresponding stories are almost certainly borrowed from the North (Berezkin 2012, 2014:349-350), mostly from North Africa--Near East. In coastal areas of Guinea and Congo some European borrowings dated to early Colonial epoch are also possible. At the Near East the Iranian traditions are divided from the Arabic ones.

In Eastern Europe and the Baltic region the traditions of the Finns, Estonians, Livonians, Latvians, Lithuanians, Byelorussians and Ukrainians are strongly 'European', the folklore of the Crimea Tatars and especially of the Bashkir is strongly 'Asiatic', the folklore of the Gagauz, Volga Tatars, Mari, Udmurts and Komi moderately 'Asiatic'. The Russians and the Mordvinians are slightly on the 'European' side while the Chuvash are slightly on the 'Asiatic' side. This tendency is especially interesting because in other respects Russian, Ukrainian and Byelorussian folklore has much in common. The situation with Baltic Finnish folklore is similar. Though Finnish, Livonian and Estonian groups are typically 'European', for the Setu who live on Estonian-Russian border and belong to the Russian Orthodox church as well as for the Karelians the 2nd PC registers only slight predominance of 'western' motifs over the 'eastern' ones.

Sets of motifs in tales recorded among western and eastern Sami demonstrate dichotomy of the same kind. In particular, the Sami of Kola Peninsula have Western Siberian parallels absent among groups living in Finland, Sweden and Norway (Berezkin 2008).

Absolute mathematic indexes for the most richly represented 'Asiatic' traditions (Kazakh, Buryat, Altai, Tuvinians, Mongols, Georgians) are higher than indexes for the best represented 'European' traditions (German, Spanish, French, Estonian, Greek. etc.). The 'Asiatic' complex looks like being innovative and expanding and the 'European' one as more neutral. It should be noticed that the traditions of Oceania which could not be influenced either by 'Europe' or by 'Asia' stand slightly closer to European and Mediterranean set and farther from the Caucasian--Southern Siberian one. All Medieval and Ancient traditions that have enough B-motifs of our list to make their processing meaningful (Ancient Egypt, Ancient Greece, Old Testament, Edda. "One Thousand and One Nights") also stand nearer to the 'European' complex. Concerning Southern Siberia, Central Asia and the Caucasus, some traditions of these regions are only moderately 'Asiatic' but all such cases are conditioned by the insufficient amount of published data. Unlike such groups as Kara Kalpak, Tofa or Yazgulyami, all Eastern European traditions are among the best represented in our database, so the transition from the 'Asiatic' set of motifs to the 'European' one in the territory between the Ural Mountains and the Baltic is well confirmed.

4. The 3rd principal component: the 'North' against the 'South'

The opposition between East and West is the most important one inside the Nuclear Eurasian folklore zone. Another important tendency revealed by the 3rd PC is the difference between northern (forest and tundra belts of Eurasia) and southern (Mediterranean--South Asian) traditions. Here all Baltic peoples, Byelorussians and Ukrainians stand near to the Russians with Swedes, Poles, Germans and Hungarians far behind and Italians or Scotsmen even farther. This Northern Eurasian folklore province includes almost all Siberia with highest indexes for the Samoyed, Tungus and Chukchi groups. As for the Southern province, its core area lies between the Caucasus and South Asia but the Mediterranean and African traditions are also part of it.

The Mediterranean--South Asian complex correlates pretty well with the spread of Islam. At the same time the spread of Islam could not be the only responsible factor and the corresponding information network should have older roots. Ancient Egyptian and Ancient Greek traditions stand slightly nearer to it than to Northern Eurasian complex while not only "One Thousand and One Nights" but also the Old Testament clearly belong to this southern complex. As for the Northern Eurasian complex, processes that took place in different time periods (up to the formation of the Russian Empire) are likely to overlap producing similar patterns of areal distribution of motifs. However, the earliest of these processes could be many millennia deep in time.

First of all, the Northern Eurasian complex of the B-motifs to a large degree overlaps with the spread of some A-motifs, in particular with the interpretation of some celestial objects seen in the night sky (Berezkin 2010) and with motifs typical for the cosmogonic earth-diver myth (Berezkin 2007). Considering the North American parallels, Northern Eurasian motifs in question had to exist already in Terminal Pleistocene--Early Holocene though their spread to the Baltic could have taken place somewhat later.

Secondly, this complex has parallels in Indonesia, New Guinea and Oceania that initially looked as a result of an 'incorrect' statistics explained by insufficient data. To check this suggestion, the trickster and the adventure motifs were computed not together but separately (Figs. 3 and 4). The results proved to be similar in both cases and the Indonesian-Melanesian-Polynesian parallels certain. Such parallels can be understood if we consider Northern Eurasian set of motifs as an archaic one while the Mediterranean--South Asian complex as an innovative development in the core zone of early civilizations. Both trickster and the adventure motifs registered in Siberia and Eastern Europe quite often have parallels in the New World while for the Anatolian, Iranian or Indian traditions such parallels are very rare.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

The Northern Eurasian trickster stories are mostly animal tales with zoomorphic protagonists while the Mediterranean--Central Asian--South Asian trickster stories usually have anthropomorphic protagonists. The episodes themselves are also different, the ones of the northern complex having some parallels in America and the ones in the southern complex never.

5. Conclusions

A short period of historically oriented folklore research in the late 19th--early 20th centuries in the U.S., Germany and Northern Europe ended with general disappointment in corresponding methods and theory and with a shift to functional, structural or psychological interpretations of the material. Besides many specific faults and mistakes, there were some capital problems. Unlike archaeological and to a lesser extent linguistic and genetic data, the folklore material does not have its own internal chronology and uses the data of other historical disciplines for checking areas of the spread of particular motifs against areas of prehistoric archaeological traditions, language families, etc. A hundred years ago our information on human past was absolutely insufficient. Besides, the processing of thousands of texts without computers was technically too complicated and troublesome. Now when the situation has changed dramatically the mass material on mythology and folklore can become an important source of data on the human past.

This paper is a kind of a by-product of my long-term research on a distant past like peopling of America, out-of-Africa movement of modern humans, Austronesian dispersal and the like. Mass materials on the adventure and trickster motifs in Eurasia have been statistically processed for the first time and some results proved to be unexpected. First of all it is the position of Eastern Europe as a contact zone and border belt between Western Europe and Asia. I would stress that the border in question is not a counterpart of S. Huntington's borders between civilizations. My 'European' complex of 2nd PC includes not only the European Christian traditions but also Muslim traditions of North Africa and the Near East, while my 'Asian' complex includes Christian Georgian tradition as one of its most typical cases. Areal patterns of particular folklore complexes and the borderlines between them seem to correspond to the intensity of communication between people. This intensity sometimes correlates with patterns of the spread of world religions and sometimes is stipulated by other factors.

With a rapid increase of the amount of data included into the database more detailed and multifaceted analysis of the cultural interaction in the Eurasian past becomes possible. In particular separate analysis of the spread of different categories of motifs beyond their basic division into two groups looks promising.

DOI: 10.3176/tr.2015.1.01

Acknowledgments

This article was supported by the Russian Scientific Fund, project 14-18-03384

References

Aarne, Antti (1910) Verzeichnis der Marchentypen. (Folklore Fellows Communications, 3.) Helsinki: Suomalainen Tiedeakatemia.

Aarne, Antti and Stith Thompson (1961) The types of the folklore: a classification and bibliography. (Folklore Fellows Communications, 184.) Helsinki: Suomalainen Tiedeakatemia.

Thompson, Stith (1955-1958) Motif-index of folk literature. Vol. I-VI. Copenhagen and Bloomington: Indiana University Press.

Benfey, Theodor (1859) Pantschatantra. Funf Bucher indischer Fabeln, Marchen und Erzahlungen. Leipzig: Brockhaus.

Berezkin, Yuri (2005) "Continental Eurasian and Pacific links in American mythologies and their possible time-depth". Latin American Indian Literatures Journal 21, 2, 99-115.

Berezkin, Yuri (2007) "'Earth-diver' and 'emergence from under the earth': cosmological tales as an evidence in favor of the heterogenic origins of American Indians". Archaeology, Ethnology & Anthropology of Eurasia 4, 32, 110-123.

Berezkin, Yuri (2008) "Sibirsko-saamskie paralleli v oblasti mifologii na fone siuzheta ATU 480". [Siberia--Saami links in mythology against the ATU tale-type 480.] Natales Grate numeras? Sbornik statei k 60-letiiu Georgiia Akhillovicha Levintona, 119-143. [Natales Grate Numeras? Collection in honor of the 60th birthday of Georgii Akhillovich Levinton.] Saint Petersburg: European University at Saint Petersburg.

Berezkin, Yuri (2010) "The Pleiades as openings, the Milky Way as the path of birds and the girl on the Moon: cultural links across Northern Eurasia". Folklore (Tartu) 44, 7-34.

Berezkin, Yuri (2012) Afrika, migratsii, mifologia: arealy rasprostranenia fol'klornyh motiviv v istoricheskoi perspective. [Africa, migrations, mythology. Areas of distribution of folklore motifs in historical perspective.] Saint Petersburg: Nauka.

Berezkin, Yuri (2014) "Three tricksters: world distribution of zoomorphic protagonists in folklore tales". In Scala Naturae. Festschrioft in Honour of Arvo Krikmann for his 75th birthday, 347-356. Anneli Baran, Liisi Laineste, and Piret Voolaid, eds. Tartu: ELM Scholarly Press.

El-Shamy, Hasan M. (2004) Types of the folktale in the Arab world: a demographically oriented tale-type index. Bloomington: Indiana University Press.

Jason, Heda (1970) "The Russian criticism of the 'Finnish school' in folktale scholarship". Norweg 14, 285-294.

Thompson, Stith (1932) Motif-index of folk-literature. Vol. 1. (Folklore Fellows Communications, 106.) Helsinki: Suomalainen Tiedeakatemia.

Thompson, Stith and Warren E. Roberts (1960) Types of Indic oral tales. (Folklore Fellows Communications, 180.) Helsinki: Suomalainen Tiedeakatemia.

Ting, Nai-Tung (1978) A type index of Chinese folktales in the oral traditions and major works of non-religious classical literature. (Folklore Fellows Communications, 223.) Helsinki: Suomalainen Tiedeakatemia.

Uther, Hans-Jorg (2004) The types of international folktales. Vol. 1-3. (Folklore Fellows Communications, 284-286.) Helsinki: Suomalainen Tiedeakatemia.

Yuri Berezkin

Museum of Anthropology and Ethnography (Kunstkamera), Russian Academy of Sciences and European University at Saint Petersburg

Address:

Museum of Anthropology and Ethnography (Kunstkamera)

Universitetskaya emb., 3

Saint Petersburg 199034, Russia

E-mail: berezkin1@gmail.com

Tel.: +79218743569