标题: How to approach 'thick corpus' in living oral traditions [打印本页] 作者: silver 时间: 2011-1-8 11:07 标题: How to approach 'thick corpus' in living oral traditions
Variation, textuality and thick corpus
(FFN 18, November 1999: 2-10)
The 5th Folklore Fellows' Summer School was held in Turku, Finland on August 8-23, 1999. The general topic of this international training course was "Variation and textuality in oral tradition". This theme was examined and discussed from many viewpoints in a number of plenums, keynotes and panels. The 1999 FF Summer School was the first at which both the participants and the teachers were given the chance to get absorbed in the questions of variation and textuality in advance as the articles on which the plenums and other sessions were based were published in the form of preprints. These preprints, numbering 26 in all, will be published as a volume called "Thick Corpus, Organic Variation and Textuality in Oral Tradition" probably by the Finnish Literature Society. In this review, I wish to offer you my interpretation of the preprints as well as three of the invited guest lectures presented during the two weeks at Turku. The review is organised according to three interrelated topics, the concepts of variation, textuality, and thick corpus.
Introduction
Variation is one of the key characteristics of folklore. As Lauri Honko (Preprint No. 1) notes, it is the life-blood of oral tradition. Studies of oral tradition always, either implicitly or explicitly, touch upon the problem of variation. Jyrki Pöysä (23) suggests that its seeming transparency and embeddedness in the scholarly concept of folklore have made variation all too easy to overlook. During the FFSS99 variation was looked into from several viewpoints, among them different types of variation, its manifestation in various folklore materials and rules of reproduction. The levels and categories of approach were many, but they all agreed that variation has much to do with meaning. I shall begin my interpretation with this, the concept of meaning.
In order to analyse the production and interpretation of meaning, it is, as Maria Vasenkari and Armi Pekkala (12) suggest, pragmatic to make a distinction between two types of modes of signification that are at play in all speech events. They concern the referential and indexical functions or dimensions of language. Referential meanings contain a correspondence between the content of the expression and a perceived state of affairs. They refer, for example, to persons, events or concrete objects, and are semantic by nature. Indexical meanings, on the other hand, are ultimately dependent on the context in which the utterance is expressed. This distinction between the referential (what is said or uttered) and the indexical (what the utterance tells about) is relevant in dealing with, and eventually understanding, the situated and contextual nature of oral tradition. Appreciation of the situated and contextual nature of oral tradition is the only a venue for getting to the indexical meanings inherent in it. The quest for indexicality makes certain requirements of the research data. The concept of thick corpus elaborated at the FFSS99 is about meeting these requirements.
Producing a thick corpus
The concept of thick corpus/data/material can be approached from various angles. When outlining the general characteristics of thick corpus Lauri Honko (1) states that the basis of thickness is intensive collecting. The focus in producing thick material is not on certain folklore items, such as a particular tale or genre, but rather on wider thematic fields. Such research most often concentrates on one or several informants in one particular community or region. The performance situation with its many dimensions is seen to constitute the contextual frame of the data and it is therefore an inseparable part of the material. In addition to these temporal and spatial characteristics of thick data, Honko points out that multiplicity, commensurability and comparability are important, too, in determining the general constituents of thickness.
Another angle on thick data is provided by looking at the way it is produced in field research. Tuija Hovi, Lena Marander-Eklund, Armi Pekkala and Maria Vasenkari have all conducted field research, specifically interview research, that has resulted in thick materials in the sense that the focus has been on a certain thematic field and the same informants of a certain community or region have been interviewed repeatedly over several years. Keeping the emphasis on the situated and contextual character of meaning has brought out the importance of the role of the researcher in the production of the data. Barbro Klein makes a challenging contribution to the discussion on producing thick data in that she has conducted research based on new kinds of materials.
Tuija Hovi examines (17) the way personal experience is built up into a narrative by means of which an individual becomes integrated into the community, and the way such narratives can serve as a source for research into the phenomenon of religious conversion. Her material consists of repeated thematic interviews among the members of a speech community, a certain Christian congregation. Hovi's analysis demonstrates how the conversion story is constructed as a situation-specific process that is framed by the narrator's current life situation, the narrative situation and the audience.
Lena Marander-Eklund (18) has interviewed a number of women on stories of childbirth. She interviewed the same women at certain intervals: one month before the birth, just after the birth and a year later. The time between the experience and the narration of it is the key to her examination of the variation in the women's accounts of childbirth. The analysis shows that variation takes place both on the level of events and on the levels of style and meaning. Marander-Eklund concludes that the stories of childbirth can be seen as a presentation of the self. Right after the birth the narrator tells what she went through while in the narrative told a year later it is a parent, a mother, who remembers her experience.
The paper by Barbro Klein (5) presented "The miracle in Södertälje, Sweden. Mass media, interethnic politics, and a profusion of texts and images", a case based on a type of data to which folklorists are rather unaccustomed. Klein's data consists of various types of materials: verbal and visual newspaper accounts, television newscasts, internet pages, interviews with journalists of different ethnic backgrounds and other interactions with various ethnic communities in Sweden. All are concerned with the miraculous events that took place in Södertälje, Sweden in 1992 when a saint appeared to a Syrian girl in a vision and blessed her with the ability to cure the sick. The events caused intense emotional involvement among many different kinds of people in Sweden: the number of testimonials, personal experience stories and anecdotes was great, and eventually the discussions came to concern larger issues, too, among them church and interethnic politics. With a thick data consisting of different types of materials, Klein presented an analysis of the events in Södertälje that clearly points out the extraordinary powers of the media in matters of belief and narration. It is something folklorists cannot, really, disregard.
Päivikki Suojanen (25) has carried out field research in three slums in Nairobi, Kenya. She has conducted numerous interviews that have concentrated on one particular theme, the cultural change and confrontation of different ethnic traditions. She discusses the questions of the self and the other, acculturation and adaptation in examining the survival strategies of the inhabitants of the slum.
Armi Pekkala and Maria Vasenkari (12 & 22) examine the production of data as based on their interviews with Northern-Ingrian women in Russia. The material consists of lifestories. From the viewpoint of dialogical methodology they see the data produced in an interview situation as characteristically intersubjective, time-, situation- and encounter-specific. They argue that the process of producing data should be comprehended as including the interpretation and representation in writing of the data in addition to the immediate act of acquiring it (e.g. in an interview). Notions of the researcher influencing the data are meaningless if that influence is not shown in the analysis, in the written research. Representation of the research process in writing is, according to Pekkala & Vasenkari (who on this point follow Clifford Geertz), the key to thick data. Only when the data is explicitly interpreted and contextualised, put into writing, does it become thick with meaning. Data is, thus, having never existed as such in the form of a pile of tapes, photographs or texts, actualised in relation to what is done with it, where, under what academic circumstances, from which theoretical viewpoints, for what purpose, in what manner, and by whom.
Thickness of archived materials
Thick corpuses as defined above consist of materials produced in the field; they are about "living" tradition systems. What about archived materials? As the analyses made by Patricia Nyberg, Marjut Huuskonen & Pasi Enges, Jyrki Pöysä, Senni Timonen and Ulrich Marzolph show, there are also archived material corpuses that can be considered to be thick. In determining the thickness of archived materials the criteria are, to a degree, different from those concerning field materials. The common denominator is the situated and contextual nature of the data.
Patricia Nyberg, Marjut Huuskonen and Pasi Enges (20) have all conducted research based on a vast collection of Saami folklore in the tradition archive of Turku University. The fieldwork for the Saami folklore project was carried out from the late 1960s up to the mid 1970s. During that time several folklore scholars repeatedly interviewed the inhabitants of one Saami village in Northern Finland. The aim of the project was to collect all the oral tradition known to the villagers and to examine its transmission in this particular community. The material accumulated in the village of Dalvadas is exceptionally abundant and well documented. Nyberg, Huuskonen and Enges present a project anatomy in that they examine the development of field methods during the project and analyse how the research strategies adopted actually functioned in concrete fieldwork situations. They look, in other words, at the process of producing data as it is manifested in the archived material of the project. Nyberg, Huuskonen and Enges debate the thickness of the data and end up by suggesting it is sought by examining the internal aims of the project: How well do the collected materials match the interests of the research? They answer the questions by concluding that the project material is, indeed, a thick corpus from the perspective of local conventions of narrating. The methods used during the project, such as having several different interviewers and repeating the same themes in repeat interviews, underline the variation by narrator and by situation.
In his analysis of variation in archived anecdotes (23) Jyrki Pöysä looks at the thickness of archive material and proposes that archive scholars should make use of the concept and distinguish between "thicker" and "sparser" corpuses. He defines a sparser corpus as a part of a research corpus formed on the basis of all the archive material, with the help of card indexes and the folklorist's own archive research. A thicker corpus, on the other hand, would be made up of a single folklore collection as it often ensures a more unified temporal and interpretive framework. Pöysä notes that archive data is usually temporally and spatially less dense than field data, but what it has to offer for the studies of variation is its emphasis on temporal, diachronic change. Pöysä points out, as do Nyberg, Huuskonen and Enges, that thickness ultimately depends on the correlation of the data with the research interests and aims.
The corpus analysed by Senni Timonen (26) is, as she points out, the thickest or densest body of poetic material in the Kalevala meter available in Finland. Larin Paraske, an Ingrian woman, was interviewed by Adolf Neovius in the latter half of the 19th century and the interviews, all of them well documented, resulted in 2,000 texts (songs), more than 30,000 lines. This material is exceptional in the sense that the collection of Kalevala-metric poetry in the 19th century did not, as a rule, focus upon a particular time, area, or individual. Timonen asks what scholarly approaches best serve the density of her material. She follows the Geertzian notion of thick description in that she states that the material is not, in itself, a description, but that a description must be created. Timonen sets herself the task of finding meaning "in the direction of exceedingly extended acquaintances with extremely small matters". Timonen asks Paraske how she comprehends poetry and reconstructs the concept of poetics from the singer's own perspective. Her analysis is a delicate example of how thick description is created, and of the range of interpretations permitted by archive material if the researcher has a thick corpus and the necessary interpretive skill.
Ulrich Marzolph, in his analysis of the constitution of meaning in the narratives of a Persian storyteller (19), demonstrates the significance of contextual information when working with archive material. He does so in a reversed manner. His material corpus consists of narratives told by a Persian storyteller to a British scholar in the mid 1940s. The material lacks almost all contextual information. This being the case, Marzolph asks where to look for meaning, "the specific meaning given narratives gain in specific situations for specific individuals". He sets out to do this by examining the language, the wording, and the use of imagery in the variants of certain narratives. Most interesting is his analysis of what the narratives tell about the woman's world in 1940s Iran, especially of power, its use and objects. The question of why the storyteller chose to narrate precisely those tales remains a mystery.
Variation in folklore research
Variation has been recognised as one of the most characteristic features of orally transmitted folklore for as long as folklore has been studied. The comprehension of variation (e.g. what it is that varies, what does not vary, what the variation indicates, how and why variation is significant) has been multifarious. The same can well be said of its study.
Ulrika Wolf-Knuts (13) examines the significance of variation in discussing the history of comparison in Nordic folklore studies. It is the historic-geographical method, constructed entirely around the idea of comparison, that she subjects to analysis. Beginning in the late 19th century and continuing up to the present day, she examines the development and application of the historic-geographical method in folklore research and asks in each case what is compared with what, how the comparison is conducted, what the purposes of comparing are, how the method has changed during the 150 years of its existence and what its future chances are. She examines in detail the main proponents and works of the "Finnish school", among them Julius and Kaarle Krohn, Walter Anderson, Anna Birgitta Rooth, Jan-Öjvind Swahn, Inger Lövkrona and Matti Kuusi. Wolf-Knuts concludes that the aim of comparison was to trace the origin and genesis of folklore phenomena. To find and reconstruct the basic form of the folklore item has from the very beginning been the goal of comparison. Variation has, accordingly, been the main methodological tool in reconstructing the history of the item in question. Recent applications of the historic-geographical method point in a new direction: the method has proved noteworthy in contributing to the study of meaning, in revealing the factors that have, for example, held a tale together in time.
As Lauri Honko notes (1), the true nature of real variation and the reasons for its existence were probably of no interest to the advocates of the historic-geographical school. Today the interest is in making the connection between variation and meaning. But how to go about it in a way that does not exhaust the concept of variation altogether? Honko proposes that the concept of variation, or the utilisation of form in narration, is combined with something else, such as cohesion and meaning in a narrative. The concept for this is, he suggests, textuality. Textuality calls for context, recognition of its intertextual, intratextual, social and cultural dimensions. As Honko states, without coherence there is no textuality. The question to be asked, then, is: From where does a performance, a narrative or a song, draw its coherence and meaning? From the performer, from the oral "text", from the context of performance, from the audience or from the combination of sources? Answers to these questions are sought, as demonstrated shortly, in two directions: by taking a look at the process of reproduction, the way oral tradition is produced and reproduced, and by examining the manifestation of variation in a particular text corpus. The interesting question of the relationship between the traditional, the collective and the individual is also raised in this connection.
Variation within a text corpus
Jyrki Pöysä (23) has studied variation in archived anecdotes. His data consists of written responses to the collecting competition on lumberjack tradition in Finland held in 1969. Pöysä distinguishes four sub-categories of variation. First, there is agent variation in which the characters may change but their significance and the significance of the narrative remain largely the same. Pöysä suggests that agent variation is one aspect of the ecotypification of tradition. Second there is variation on the level of abstraction. There the agent of a story can belong to an abstract social category or be a real, named individual. Pöysä links the changes in the level of abstraction with writing and collecting, i.e., lifting the anecdote from its original local context. This can be done by the archive respondent or by the folklorist. Both cases involve textualisation strategies and the solutions can be seen as situational variation specific to particular data. The third sub-category of variation is that of perspective, which depends on the narrator's attitude to the events in his or her story, and fourth is variation in the experiential distance from the narrated events. With a thick corpus at his disposal Pöysä is able to interpret and analyse the individual creativity inherent in the archived anecdotes.
Carola Ekrem and Ríonach uí Ógáin are both concerned with variation and the continuity of tradition. Carola Ekrem (14) has studied the counting-out rhymes of Swedish-speaking children in Finland, especially the continual transmission of the rhymes. Having examined the variation in the vast collection of rhymes, she notes that texts that admit variation on the level of structure in particular have the greatest chance of continual transmission.
Ríonach uí Ógáin examines the singing tradition in Irish (21) and her presentation is an example of the diachronic change of tradition that has taken place as a result of adaptation to the expectations and demands of the present day. She addresses changes in the song tradition and examines how the changes are due to historical, social, economic and traditional influences.
Dell Hymes' paper on "Variation and narrative competence" examined variation in narrative texts as it appears in the organisation of lines and groups of lines. The narratives analysed belong to the Native American traditions. Hymes' focus was on how narrators make use of relations of two and four, or of three and five, e.g. four scenes in an act, four stanzas in a scene, four verses in a stanza, or those of three or five. He took up cases where the usual pattern of, say, three and five is replaced by its alternative, i.e., two and four, and as the analysis demonstrated, such alternation is associated with gender. Hymes' presentation revealed that the gendering of texts can sometimes be ascribed to the gender of the speaker, sometimes to the gender and social status of the audience as perceived by the speaker, and sometimes to the gender of the actor or speaker within the story, whose presence commands a scene or a whole episode or tale. He suggested that in order to understand oral narratives, we need to analyse all the known repertoire of individual narrators, so as to discover their styles, the range and occasions of formal variation in them, and meanings they may individually have.
This is what Lauri Honko and Anneli Honko (16) have done in Southern India, where they have concentrated on the epic repertoire and idiolect of one singer. In their article Lauri & Anneli Honko report on the empirical findings on certain forms of variation and their impact on the textuality, meaning and cohesion of oral narrative. In analysing the different types of variation manifest in the Siri-epic, they argue that variation is not an end in itself. It is neither striven for nor thought of by the performer, who goes not by form but by meaning. The "correct" meaning is, however, inextricably linked to the poetic form. This being the case, attention is turned away from the various forms of variation toward the work of the singer in performance, his construction of the path of composition. The composition, in this case producing the epic text, requires, so the argument goes, the existence of a prenarrative, a pre-textual frame. Lauri and Anneli Honko call it a mental text and it is seen to consist of storylines, textual elements (e.g. episodic patterns, images of epic situations), their generic rules of reproduction and contextual frames (e.g. remembrances of earlier performances). Mental texts are individual and go on changing during the career of the singer or storyteller. As the Honkos suggest, the construction of the path of composition represents the main work of the singer in performance. The analysis confirms the need to look at the repertoire and performances of one individual performer and the variation within it in order to understand textuality in oral tradition.
Rules of reproduction and domains of variation
Annikki Kaivola-Bregenhøj (4) takes as her starting point in studying variation the definition of folklore as a performance in the here and now, moulded by factors conducive to both invariability and to variation according to the specific performing context. These factors are sometimes called the rules of reproduction, and Kaivola-Bregenhøj examines them as they evolve from her research on the repertoire of one narrator. There are four such general rules. First, there is the process of assimilating, memorising and reproducing folklore. Second, there are the genre-specific norms, and third, the personal attributes of the narrator. Fourth, there is the performing situation, the context of the reproduction. These rules cover the various dimensions inherent in reproduction and well serve as a general outline to the topic.
Anna-Leena Siikala (11) emphasises the significance of genre in understanding variation. She examines the genre of korero, the "Truth and Tradition", basic knowledge of the organisation of society, of the Southern Cook Islands. Her analysis of korero, which does not fit into any of the established Western definitions of an oral genre, proves the supposition that views based on oral narration in European communities are insufficient for dealing with all the aspects of variation. Siikala's starting point for understanding variation is the argument that oral narratives should be seen to exist in a process of social intercourse. It is research on the different genre practices, their social ties and contexts, in other words, situated and contextual oral tradition, that opens up the possibility of understanding the nature of variation. Siikala lists crucial domains of variation: the habitus of the narrator, memorising and narrating, the cultural significance of the discourse field, the political, moral, religious, etc., values of narratives, the degree of institutionalisation of the performance practices, the performance settings, and the shared and individual ways of giving meaning to narratives. Examination of these domains of variation as they are manifest in the narration of korero shows that the variation in narration is caused by the internal logic of the genre.
The contribution of theories on the psychology of memory has been remarkable in understanding the process of assimilating, memorising and reproducing folklore. Lauri Harvilahti and Ilkka Pyysiäinen look at the problem of variation from a cognitive perspective. Pyysiäinen (9) elaborates the concept of mental text from the viewpoint of cognitivist theories, whereas Harvilahti (3) sets out to ponder on the application of theories of the psychology of memory to folklore materials in general. He focuses on the schema theory in particular. Its detailed examination suggests that from the viewpoint of schema theory, folklore appears to form schematic structures which are used both to evoke a general picture of the overall structure, and as an aid in constructing specific details. The production process of oral poetry utilises a system of linguistic and poetic strategies, among them the "pre-existing" overall structures and themes, core subject matter, episodes and substitution, groups of words, meter, alliteration, and formulae. Harvilahti suggests that each folklore performance is an instance of reproduction, not of composing or creating. According to this view, the singers interpret a song which they reproduce as they experience it. This process of interpretation is influenced by the singer's competence, interests and aims, the demands of the collective tradition and performance situation. The singers use ethnopoetic strategies in giving clues and conceptual information (e.g. by using a specific register) through which others (e.g. the listeners) guided by their horizons of expectation, can interpret the discourse. Harvilahti proposes, along the same lines as Siikala, that the traditional definition of genre be revised and elaborated to underline the significance of the specific context of expression. An orally performed epic poem, for example, is emphatically a culturally established form of discourse having its own conventions, rules, norms and history, which are actualised in social intercourse. Harvilahti makes a strong and well-justified case for the culture- and genre-specificity of comprehension, assimilation, interpretation and reproduction of oral tradition.
Variation within limits
Perhaps the most difficult aspect of textuality is establishing the relationship between the individual and tradition, and elaborating on their interplay. This is what John Miles Foley (2) does when he examines the significance of the legendary singer figure in South Slavic oral epic tradition. Foley shows how the proverbial figure of the Guslar embodies the tension between the individual and tradition. He argues that the Guslar (as well as Homer to whom he makes the comparison), is an anthropomorphisation of tradition, and a way of designating the poetic tradition. Each individual singer legitimises himself by claiming professional descent from the Guslar, the great bard. When a real-life guslar performs the songs, he appropriates the Guslar as a direct ancestor and makes himself the present-day embodiment of the greatest of all bards. As Foley's analysis so aptly demonstrates, in that event the individual and tradition combine, the tangible instance standing pars pro toto for the larger, intangible whole of the poetic tradition. The Guslar must, in order to achieve this, assume a double identity, in that he must have a credible biography (although consisting of tell-tale characteristics) and yet his deeds must outstrip anything that might be expected of a real-life singer. He represents, eventually, the finite instance and the ongoing tradition. As Foley states, neither the legendary singer figure nor the poem can ever have a single fixed form until he or it is instanced, whether in a certain singer's individualised account of his famous forebear or in a single given performance of a song. Each account and performance is in itself an instance informed by tradition and collectively they illustrate the fundamental compositional feature of South Slavic (and Homeric) epic poetry: variation within limits.
Seppo Knuuttila (6) approaches the interplay of individual and tradition from the viewpoint of mentality studies. Mentality may be understood here, as Knuuttila suggests, as constituting the "mental equipment" which enables the formation and expression of collective opinions. The study of mentalities enables us to make sense of the past as well as of the present and of the two-way relationship between them. Knuuttila emphasises the importance of observing the variation in the structuring of the past caused by individual and collective factors. In folklore studies attention should be paid, specifically, to the intra- and internarrative variation of different forms of mentalities. As the study of mentalities does not have any special methodological bonds, Knuuttila suggests some approaches to guide the research from the indication of meanings to their interpretation. One of them is the analysis of meaning based on symbol anthropology developed by Clifford Geertz, and the other is the clue-method developed by Carlo Ginzburg. In both the emphasis is on the public and visible dimensions of culture. Another approach to the study of mentalities can be found in looking at the likeness of the concepts of identity and mentality. Knuuttila calls for the interest of contemporary research in identity to focus on such questions as the constitution of the different dimensions of sameness and remaining the same; how are the criteria for sameness and difference selected; how are expressions of identity localised in social and cultural interactions; how is the relationship between I and self mediated in time and space; and how are self and other distinguished as subjects? Studies on the construction and deconstruction of identity provide a firmer grasp and understanding of mentalities.
Textualisation and contextuality
One matter on which most folklorists are agreed is the complementary relationship between textualisation and contextuality. During the FFSS99 this relationship was examined primarily as it is manifest in the process of textualisation, which is seen to consist, first, of the situated production of the "texts" and second, of expressing them in writing.
Margaret Mills (7) read a paper on "Women"s tricks: Subordination and subversion in Afghan folktales" in which she dealt with the various methods of textualisation and their implications for the interpretation of the texts created in that very process. Her data consists of Afghan folktales on women's tricks, which as a cultural theme is a widely discussed topic in Arabic, Persian and Turkish languages, cutting across both oral and literary genres. Mills addressed the problem of textualisation in asking how many different aspects of text and textualisation one needs to look at in order to explore the realisations (and possible receptions) of the theme. She illustrated the problem by presenting three different styles of textualisation two of which were verbatim translations of recorded oral Persian texts and one presented the narrative in a summary form. Mills pointed out the true problem of textualisation that all folklorists must consider over and over again: very few approaches to writing down oral text allow us to track the oral effects, and no written form fully renders all the techniques of dramatic presentation used by oral performers, eye contact being one of them. When looked at from this point of view, verbatim translations supplemented by notes on the paralinguistic features best serve the purpose of understanding the interpretations made by the narrator. Mills addressed some central questions of the contextuality of data when she stated that the noncommensurability of texts or procedures of textualisation is more likely to be a problem if one is trying to compare texts created at different times by different authors, under different conditions and for different purposes.
Ríonach uí Ógáin (8) analyses the way folklore collectors in Ireland in general, and one of them in particular, have interpreted the concept of context. Full-time folklore collectors in Ireland were, when the post was founded in the early days of the Irish Folklore Commission in the 1930s, instructed to keep diaries in which they documented their day-to-day lives in the field. uí Ógáin examines the diaries of one collector. Her argument is that both the context and the text are heavily influenced by two factors: the collector and the informant's attitude to the collector. uí Ógáin introduces the concept of subtext in order to deal with the silent, undocumented effects which the cultural background bears on the collection. She skilfully illustrates how the interaction of the collector with the informants, the collector's innate understanding, background and dedication, all of which contributed greatly to the creation of the collecting context and to the textualisation process itself.
Textualisation of archived materials
Tradition archives in the Nordic countries, especially Sweden, Norway and Finland, have to a large degree acquired their materials through questionnaires. Anders Salomonsson, Ann Helene Bolstad Skjelbred and Barbro Klein all concern themselves with the textualisation of archived questionnaire materials and the problems, challenges and questions it raises.
Ann Helene Bolstad Skjelbred (24) examines the questionnaire as a method based on her research on the collections in the Norwegian Ethnological Archives. She analyses the change in the method that has taken place during the decades of this century. She illustrates how the focus of questionnaires, the type of information sought and the dominance of the male gender and certain social groups have changed with time, as has the model of questions and answers. In the earlier decades the aim was to cover wide contentual and geographical areas in a neutral and objective manner. Now the emphasis is on self-oriented personal experiences. Skjelbred debates the generalisability of questionnaire materials and suggests that in most cases the generalisations that can be drawn are not valid outside the particular group of respondents because of all the unavoidable reservations for time, space and social context. Skjelbred suggests that archived questionnaires be treated not as oral but as written material, analogous to letters, not interviews.
Anders Salomonsson (10) is concerned with connecting documentation and research to textualisation. He asks what has happened to the basic material when it is presented as the scholar's results. His perspective on this question is research historical in that he examines the collection and archiving methods of Swedish ethnologists to see how they have changed with time. He takes up the political and ideologial dimension inherent in all research in stating that the production of data is always influenced by certain values and ideologies that determine what is considered important, credible and true. It is the researcher's responsibility to examine and explicate this aspect of the situatedness in the process of textualisation. As Salomonsson points out, questions of politics are always questions of ethics, too. He concludes by saying that we documentalists, archivists, antiquarians and researchers are producers of history, conveying a picture or a version of the past that is based on the fragile thread consisting of interviews, questionnaire responses, or artefacts. The version produced by the researcher or exhibition builder, for example, is hardly questioned by its consumers but most often accepted, and as such that version easily attains the status of truth. This is something researchers seldom think of. There are many ethical dimensions to be considered.
Barbro Klein presents an example analysis of the textualisation process. Her starting point is a view of the centrality of folklore archives in the construction of symbolically powerful national heritages. She analyses the archiving practice by looking at one text, a legend about a Swedish king, examining the ways in which the author of the questionnaire, the respondent, the archivist, and the editor of the published text use the written medium. As her analysis clearly points out, the questions that arise are ethical and ideological. They are ideological in the sense that in this case the questionnaire, the excerpt cards made by the archivist and the printed text edited by the researcher all contribute, ultimately, to the representation of the national heritage. They are ethical in that the result of the textualisation process is a sanitised folklore heritage in which people are deprived of agency and power. The question that must be asked, not only in this case, but in every instance of textualisation, is: How can we find responsible ways to conduct our research and at the same time represent the views and the lives of the people we study with some kind of honesty? Klein suggests that we carve out ethical spaces and look for solutions that differ from the conventional ones and thereby contribute to the shaping of critical folkloristics.
Margaret Mills dwells on the problem of truth which she takes to be the link between the concepts of textuality and variation and that of ethics. She presents five meditations on the topic, which examine the problem of truth in different contexts. Among these are the regimes and registers of "truth" in belief studies. Mills raises discussion on the meanings evoked by the concepts of "truth" and "true" as they have occurred in the study of folklore texts and suggests that such keywords as fact/factual, real/reality, authentic/authenticity, objective/objectivity and belief/knowledge deserve special attention. Another issue is the truth in folktales. As Mills points out, storytellers and their audience, the researcher included, take the stories on various levels of understanding, which involves various notions of the truth possible in the narratives. It is important, as she says, to do research on the various kinds of truth that the tellers and the audience discern in different stories in and out of a performance context. Mills also examines the genre of testimonio and the vexations of truth, and truth as embodiment. In conclusion she the remarks that texts always have more potential meanings than any one receiver can perceive or codify. She suggests that instead of asking "Is this text true?" we should ask "To what is this text true?" or "How is this text true? How is it not true".
Conclusion
In conclusion I would like to take up some consistent lines of thought that kept coming up as I listened and read through the papers delivered during the FFSS99.
It became clear that the concepts of variation and textuality go together. In a sense and to a degree they are interdeterminate. Variation can be thought of as a key to textuality, and vice versa. The question that needs to be answered is: Where does the coherence and meaning come from in this particular case? Several scholars suggested that the study of variation should focus on organic variation, to use Lauri Honko's term, in living tradition systems. This would mean putting more emphasis on individuals, on individual creativity. It would turn the focus on the quality of the research material, because the study of living tradition systems and individual creativity requires a unified temporal and interpretive framework.
To my mind the issue touched on in all the papers, either implicitly or explicitly, is the situated and contextual nature of oral tradition. I think it is justified to say that the concept of thick data is worth noting in approaching and appreciating the situatedness and contextuality of oral tradition. Thick corpuses, by definition, make organic variation visible and open up arenas to the study of textuality.
A thick corpus can consist of either field or archive materials. Thick corpuses often focus on a particular thematic field and as such concentrate on one or several informants in one community or region. They are, in other words, temporally and spatially situated. Additional characteristics are multiplicity, commensurability and comparability. Thickness, by definition, also imposes other criteria on the data. It trains the focus not only on the material itself, but on the whole research process: the production of the data, its interpretation and analysis as well as writing up the research. Appreciation of the active involvement of the researcher in all phases of the research process is an epistemological requisite of thickness. If looked at from a Geertzian point of view, as I would like to do, the ultimate criterion of thickness has to do with the textualisation of the data. As the argument goes, data becomes thick with meaning only when it is interpreted and put in writing.
FFSS99 Preprints
1. Thick Corpus and Organic Variationby Lauri Honko
2. Individual Poet and Epic Tradition: Homer as Legendary Singer by John Miles Foley
3. Variation and Memory by Lauri Harvilahti
4. Varying Folklore by Annikki Kaivola-Bregenhøj
[5. The Miracle in Södertälje, Sweden. Mass Media, Interethnic Politics, and a Profusion of Texts and Images by Barbro Klein]
6. How to Seize Mentalities by Seppo Knuuttila
[7. Womenís Tricks: Subordination and Subversion in Afghan Folktales by Margaret Mills]
8. Some Comments on Context, Text and Subtext in Irish Folklore by Ríonach uí Ógáin
9. Variation from a Cognitive Perspective by Ilkka Pyysiäinen
10. Documentation and Research by Anders Salomonsson
11. Variation and Genre as Practice by Anna-Leena Siikala
12. Dialogic Methodology by Maria Vasenkari & Armi Pekkala
13. On the History of Comparison in Folklore Studies by Ulrika Wolf-Knuts
14. Variation and Continuity in Children's Counting-out Rhymes by Carola Ekrem
15. Story-Pattern as Sêma: The Odyssey as a Return Song by John Miles Foley
16. Variation and Textuality in Oral Epics: a South Indian Case by Lauri Honko & Anneli Honko
17. Textualising Religious Experience by Tuija Hovi
18. Variation in Repeated Interviews: Stories of Childbirth by Lena Marander-Eklund
19. Variation, Stability and the Constitution of Meaning in the Narratives of a Persian Storyteller by Ulrich Marzolph
20. Observations on Interview in a Depth Study on Saami Folklore by Patricia Nyberg, Marjut Huuskonen & Pasi Enges
21. Aspects of Change in the Irish Language Singing Tradition by Ríonach uí Ógáin
22. Producing Thick Data: an Ingrian Finnish Case by Armi Pekkala & Maria Vasenkari
23. Variation in Archived Anecdotes by Jyrki Pöysä
24. Possibilities and Limitations: a Critical Look at a Norwegian Tradition Archive by Ann Helene Bolstad Skjelbred
25. The Encounter of Rural and Urban Traditions in Nairobi Slums by Päivikki Suojanen
26. Thick Corpus and the Poetics of a Singer by Senni Timonen.
Maria Vasenkari
Department of Cultural Studies/Folkloristcs, University of Turku 作者: silver 时间: 2011-1-8 11:08
Thick Corpus and Organic Variation
(FFN 15, April 1998: 11-12)
The 5th Folklore Fellows' Summer School will study variation and textuality in folklore. The word "variation" brings to mind the fact that variation is the life-blood of oral tradition. It is the main characteristic which differentiates oral culture from literary culture. Oral "works" are fluid, whereas literary works show a permanent form. In oral tradition, a single realisation of a theme can never claim the position of a master form which dominates over all other realisations of a theme or motif. During his/her performance career, a ballad singer or storyteller may be convinced that he/she has changed nothing in a song or story, yet variation creeps in all the time as the chain of performances continues. The performances are not identical, variation is constantly present and part and parcel of the oral composition itself, but the "thing" that varies remains a problem for the observing outside analyst. The performer may regard his last performance as the very best realisation and the "whole truth" of a theme, yet he will change many things, large and small, in his next performance. Flexibility rather than stability seems to be the key to the continuity of tradition, to the kind of "invariability" typical of oral tradition which prevents variation from going astray and distorting the song or story completely.
Variation is not an end in itself, although occasionally the analyst may feel that it is just that, when he observes the verbal fluency and prosodic joy of a master of words in the retellings of a theme: the performer does not seem to repeat himself, nor does he copy others. Instead, he seems to be more than happy to perplex the text-oriented collector through the diversity of linguistic form which still remains capable of conveying a theme. It is not that the performer could not repeat or copy, if he wanted, it is just that he is not primarily occupied by the strict maintenance of texture. There are other concerns which put him to think and work.
Variation is thus neither striven for nor thought of by the performer. It is a by-product of something else. Tentatively one may say that the performer, despite his mastery of poetic form, or perhaps just because of it, goes not by form but by meaning. Yet the "correct" meaning is inextricably linked to the poetic form. It is not merely a matter of what is said, but of how it is said, and what is left unsaid. In a sense, the poetic form may be ready and available, whereas the meaning is not. It is something that must be produced over and over again, never quite the same and always within a sequential web of expressions governed by several performative styles and a particular situation.
There are at least two things that the performer must continuously observe: the storyline and the context of performance. The construction of the actual meaning takes place somewhere between a pool of pre-extant plot structures and multiforms, pre-worked phrasal and formulaic expressions and the often unique preconditions of a particular performance which determine what can, must or may be expressed, in what performative style, for how long and directed at whom. Adaptation of the story into a unique sequence of events preceding the performance to an audience of a particular kind sets the frame for "the art of the possible" in the construction of the actual meaning to be derived from a story. Nothing is self-evident at the outset: the old story is treading a new land.
Stories, again, tend to create worlds of their own and impose themselves on human beings. They are not stable, either. The most consistent picture of a story may be gained through a postulated mental text in the mind of an individual performer. It becomes visible through different realisations in performance, never quite the same and indicative of mental editing which may have taken place between consecutive performances or in the actual situation of performance. Mental texts are not shared but individual. Singers and storytellers may seem to have a similar story to tell but the outside observer soon comes across salient differences between performers. In the learning phase, there may even be a close similarity between the teacher and the pupil, but this proves to be a temporary phenomenon, because the teacher will continue to develop and edit his story and the pupil will create his own predilections and cultivate his own system of expressions. In the case of two competent performers, their mental texts of the "same" story may show affinity and may even affect each other, but they will never become identical.
Saying this is easy, but proving it in detail is not. The problem lies with customary folkloristic materials, which, despite all the proud claims by the founding theoreticians of basing their conclusions on a comprehensive study of all the available variants of a song or story, do not reflect the variation discussed above. The keyword may be "available". The historic-geographic school aimed at collecting, if not all, then at least a representative number of variants from all the cultures which possessed a particular folktale, ballad or proverb. Representative of what? Certainly not the "organic" variation of an oral product within a competent narrator's performances during a relatively short span of his career. Certainly not the organic variation apparent in a tradition community where performers and audiences know each other, are in contact or at least by hearsay share certain traditions, take and give each other impulses. And certainly not, despite the "geographic" focus of the method, the organic variation to be found not only in cohesive communities and social groups but also in cohesive regions where people sustain channels of communication and share traditions.
What the earlier comparative method was able to create was a bird's eye view on the "variation" of individual folktales, ballads, proverbs, etc., in time and space. The variation observed was used as a tool in designing the history of origin and dissemination of an oral product. Much criticism has been levelled at such ambitious goals as the reconstruction of the "earliest form" or the "routes of diffusion". Less attention has been paid to the fact that the picture created of variation in folklore is in many ways misleading. The materials used were a haphazard collection of often incommensurable documents scattered widely in time and space. A few hundred variants, at best, purported to delineate a picture of the extant forms of a tale, song or proverb through continents, scores of countries and hundreds of years. The network of variants remained thin and full of holes, and there was no real "dialogue" to be constructed between the variants, which practically never belonged to the same cohesive community. In short, the variation was artificial and not representative of organic variation. I once suggested that such variation be called "tradition-phenomenological" (Journal of Folklore Research 23/1986: 113-15) and kept separate from other forms of variation.
Behind each item included in the list of variants of a historic-geographic monograph there was a world of organic variation which remained invisible because it was represented by only one item. The key to that world would have been fieldwork. In the classical method, however, fieldwork was rarely resorted to. It was enough to cull the libraries and archives. Fortunately, during recent decades many folklorists have become involved in creating thick material corpuses through intensive collecting of the repertoires of one or several performers/ informants in one community or region. "Repertoire" need not refer to a particular tale or even genre; it can be a thematic field and, most importantly, it implies the documentation of performance situations, relevant audiences and the social background of expressions of folklore.
Recent fieldwork projects have been producing and analysing material corpuses that are quite different from the classical historic-geographic ones. Their time span may be short and their geographic scope limited, but the frequency of the folklore expressions studied may be high, i.e. the same story will be documented several times from one or more performers. Such thick corpuses make organic variation visible. Studies concentrating on one performer, one village or one thematic field in a community or region are the most common. The distribution of tradition in a cohesive community, individual repertoires and collectively shared knowledge, variation of folklore according to social role, performance context, audience and the socio-psychological ambience are topical elements of many field projects.
The multiplicity of data is not the only criterion of a "thick" corpus. The commensurability and comparability of the data are just as important. They are part of living tradition systems maintained by individuals and groups having the possibility of social exchange. Even as archived specimens, unable to change and develop, they may be said to sketch the scope and limits of real variation in a textual universe which existed at the time of their documentation.
There is a plan to produce a book for the FF Summer School 1999 on the topic of "Thick Corpus, Organic Variation and Textuality". Most of the teachers on the course will contribute articles which may not be formally read but will be intensively discussed by a panel, with questions from the participants. These articles will be of two kinds: 1) examples of analyses based on thick materials and 2) more theoretical treatises on the key concepts. The book will be sent to the participants well before the course.
