Peter J. Claus

California State University, East Bay

In the annual Siri rituals of Dakshina Kannada District ⁽²⁾ in Karnataka, southern India, hundreds of women sing the story of Siri and become possessed by the siri   spirits. ⁽³⁾ As the story describes them, these spirits are three generations of women of a matrilineal family: Siri; her daughter, Sonne; and her twin grand daughters, Abbaga and Daraga. The story recounts the tragedies which befall them.

In its various contexts, singing the Siri PaaaDdana  is a performance in which the singing is also a form of discourse, a specialized kind of speech act, which is capable of transforming women who utter it. In one context, that of singing it in the rice fields as women transplant the paddy seedlings, the spirits themselves are attracted by the singing -- an audience for the performance -- and can possess the women, turning the performance into an other-worldly experience. ⁽⁴⁾ Singing in this context is usually avoided since the work of transplanting is disrupted until the presence of the spirit is controlled and the spirits are politely asked to leave.

In the context of the annual Siri rituals, though, the song is expected   to elicit the spirit's presence, and there are those present who can control them. As women sing the story they are transformed into spirits and their interaction is as the characters of the story. Their entire speech -- not just the recitation of the story line - is sung and all speech acts are to be understood in the context of the story-ritual which inextricably links mythical and present-day worlds..

The ritual performance is regarded as curative, although I do not wish to reduce its function to this narrow definition. Before I can describe in what sense it is regarded as curative, I must briefly describe how it is the women come to the rituals. The typical life history episode begins with a woman, usually a young woman, at home and among her relatives, "acting out" - speaking inappropriately, or not speaking at all - in such a manner that her behavior is identified as being "troubled by" a siri   spirit. ⁽⁵⁾ Her relatives are advised to take her to the next Siri rituals. There, she will stand among women who are adepts, who once had a similar condition and have been coming to the Siri rituals for years to help other women learn to control the siri   spirits. As they begin to sing the story of Siri they become possessed by the spirits, individually transformed into the characters of the story they sing, and the context itself is transported to that of the story.

The new member begins to exhibit the behavior she had been exhibiting at home. The adepts call upon the novice to identify which spirit (which character of the story) she is. They call to her in kinship terms: "Sister is that you?" "Mother why don't you speak to us?" "Daughter what is troubling you? Speak to us." In the process of the impromptu   sung discourse which follows, she identifies her spirit-self and her relatives are taught to accept her in this identity and learn how to control the transformation so that it is no longer so disruptive of normal life, and, indeed, can be beneficial to it. Among the requirements of this new identity is that she is to return to the Siri rituals annually and join the ranks of the adepts.

In Tuluva culture ritual control is a male role, and in this story there is a male character whom I have not yet mentioned. He is Kumar, Siri's son, whom (as the story relates it) she had transformed into a spirit as an infant. In the ritual Kumar is the term used for the central role of the male priest who initiates the singing and controls the spirits of the story once the context is transformed/transmuted into that of the story.

Some of the locations where the annual Siri rituals are held have instituted a remarkable drama to initiate the possessions. The head Kumar leads two girls - in reality adult women, but in the myth, girls - to a platform (kaTTe  ) on which the performance takes place. The girls (the spirits of Abbaga and Daraga) sit at opposite ends of a cenne   game-board, with Kumar (the priest-medium) between them. All three are possessed, their personae   that of the spirits; the priest alternates between the character of Kumar and the character of another spirit, Bermer. Prompted by Kumar, the girls commence singing (in the first person and using the present tense) the final portion of the Siri Epic in which the three generations of Siri's matrilineage, begun as a boon from the god Bermer, comes to an end when Abbaga hits and kills Daraga with the game board, throws her into a well and commits suicide by jumping in after her. The drama reenacts that fateful event, but Kumar, in the person/role of the priest, stops the action before it comes to a violent conclusion.

I have already explored this ritual drama using the analytic tools of performance studies in a previous paper, "Ritual Transforms a Myth" (Claus 1991) ⁽⁶⁾. There I focus on the details of the discourse in which this male role, in the multiplicity of the mythic identities in the ritual character of Kumar, comes to acquire its place in the structure of the performance, a place which enables Kumar to control the text of the discourse. I summarized my argument (analysis) thus:

"The Siri PaaaDdana , a story-song in a woman's genre, forms the reference text for the drama. Its story content and even a great deal of its actual text is borrowed into the performance. But it is modified in important ways: it becomes a personal history narrative [of the spirits], recounted in the first person. Later, even this is replaced by actual speech [present tense] and we are witnesses to the mythological event. We are there. The woman's fieldsong recedes to a mere historical account, a legend, superseded by the drama.

"In this dramatized form, the priest-medium, Kumar, a man, controls the performance of what was once the song in a woman's genre. The drama is part of a public ritual performance. In fact, it is not so much a development within the myth   tradition as an exploitation of that tradition in the on-going and highly dynamic development of the ritual   tradition. The particular drama is predicated, historically and operatively, on preceding ritual events.

"Although exploiting the women's tradition and superseding it, claiming (when asked directly) superiority over it, the men-centered ritual serves for the benefit of women. The major function of the cult's rituals is to save defenseless women from quarrels and jealousies which arise within kin groups. While the women's song tradition revolves around these sorts of problems and presents them as fatefully tragic truth, the ritual tradition, through the intercession of fictive male kinsmen, attempts to alter and [re]solve them.

"Women participants feel the rituals are good and sacred. Men participants feel the same way. Non-believers, mostly modern men, however, see the cult as gross gender exploitation. Indeed, the ritual activities could (and at places, probably do) easily devolve into this if the drama-dimension and entertainment function overwhelm the ritual-dimension and social functions." (Claus 199152-3; italics and bracketed inserts added for clarity.)

Possessing and Controlling Texts

In my present paper I want to extend this analysis to our own collection, analysis and interpretation of this oral text. I do so, in part, in an attempt to reduce the sense of "Us" and "Them" in the performances in which we, too, when we collect, analyze and interpret it, take control of the text and its capacity to transform reality and construct meaning. We must realize that the "text" - as discourse - is already used by others, and our discourse about it is part of an historic as well as, now, a transcultural sequence of uses. It is precisely the moment when we say "Yes, but that (ours) is different," that we create an unbridgeable divide between Us and Them, and that we create the "Other" that has been said by some to plague the discipline of anthropology. What I want to do in this paper is focus on the use of the text which is common to all of us by retaining the concept of performance, and ultimately to explore the path toward a new relationship between anthropology and the communities, texts, and performances we study.

At the same time, I also want to contrast two academic approaches to the study of kinds of oral textual traditions that exist in ritual contexts. Among the several scholars who have introduced the Siri story and Siri ritual to a larger, international audience, are myself, an American anthropologist, and Lauri Honko, a Finnish folklorist. ⁽⁷⁾ We have each, through a number of conference presentations and publications -- both of which, along with the field research stage of our work - I will consider as "performance." Some of the parallels between these two Western scholarly traditions and the two Tuluva performance contexts I mentioned above (singing in the fields and the "curative" ritual context) are striking. But in drawing these parallels, I won't call ours "ritual performances"; instead, I would like to retract, temporarily at least, my labeling the Tuluva traditions as "ritual." By doing that, I had already created a divide between our own "serious" discourse and theirs on a basis for which I had no clear justification, no clear definition in mind. Of course I might have conscientiously done so, but that would have brought me back along an historical journey through the stages by which the social sciences objectified human interactions (specifically through textual discourse) and established the preeminent role for the scientific observer. Instead, I merely presumed that background to establish my authority and legitimacy. In short, I wanted only to assert my control over the text in a context of my own making: academic science. Defining it and recontextualizing the discourse, I took possession of the text.

One of the admirable features of Laurie Honko's use of the text is the remarkably detailed way that he has described not only the process by which he, too, took possession of it, but contextualizes this within the long history of oral text collection and representation in the European intellectual context, and more specifically that of his own Finnish folklore scholarship. Essentially the entire 633 pages of Textualizing the Siri Epic (itself one of three volumes in a set) is devoted to the intellectual history and immediate collection background for the publication of The Siri Epic (transcribed and translated in the other two volumes). ⁽⁸⁾

I realize the awkwardness of a reflexive exercise of this nature using my own work in contrast to that of another Western scholar. I shall try to curb my natural tendency toward professional rivalry and jealousy by keeping in mind that in this new context, the intention of the performance, like that of the Siri ritual, need not emphasize conflicts of interest, but rather a common interest in utilizing the discourse of the text toward the beneficial result ("curative" as in the case of the ritual) of generally improving the epistemological assumptions of Western scholarship which both anthropology and folklore share. I will thus not use our individual names, but rather identify us as "the Anthropologist" and "the Folklorist." ⁽⁹⁾ Doing so is meant to force our individual scholarship into intellectual roles associated with the collectivities we represent, and from which, on one hand, I might be criticized as fulfilling or not the collective goals of anthropologists, and, on the other hand, anthropology is represented in the work I have done.

The Anthropologist and the Folklorist

When we control (use) the text, there is a vast difference between the sense in which the Folklorist and the Anthropologist do so. ⁽¹⁰⁾ The Folklorist strives to use the complete and accurately recorded oral recitation. The Anthropologist is satisfied with using a synopsis of the story. In this case, the Siri Epic presented by the Folklorist is an astonishing 15,683 lines, recited over six days by a man, Gopala Naika, who is a Kumar at one of the locations of the Siri rituals. ⁽¹¹⁾ The Folklorist takes pains to acknowledge the "authorship" of the singer and establish (grant him) the "ownership" of his text (the recitation), in part, perhaps, because it is so rarely (if ever) sung as a continuous narrative in this fashion, and only in the context of folklore collection. The Folklorist has documented with even greater detail his own activities, in diary-like fashion, replete with photographs, of the recording sessions.

The Anthropologist had encountered the sung narrative on a number of occasions, mostly in naturally occurring ritual contexts, and had also recorded artificially contrived recitations provided by both men and women who were identified as people who knew the epic (paaaDdana  , sandi  ). ⁽¹²⁾ In publications the Anthropologist presented only brief synopses of the narrative in story form. In comparison to the text collected by the Folklorist, all but one of the texts he collected would have to be regarded as fragments. Nevertheless, in some of these cases the singers themselves regarded what they recited as the "complete text," a fact which suggests that variation in the tradition exists, making it difficult to treat one person's recitation as more authentic or more authoritative than another. ⁽¹³⁾ While the Anthropologist notes the name, gender and caste, place of residence, and other identifying social characteristics of the singer, there is no implication of individual authorship. The text is presented as tradition and as community property.

In a number of ways both the Folklorist and the Anthropologist recognize that knowledge of the Siri story is more widespread than its recitation as a text qua   text. Even amongst those who do recite it in performance contacts, it exists within their knowledge, their minds, in such a way that they can relate it in several narrative forms. When Gopala Naika recited the "complete" epic in sung verse for the Folklorist he did so for the first time in history, drawing upon what the Folklorist called a "mental text," a concept "forced upon" the analyst by his very attempt to collect the "complete text" (p. 94-5). How the analyst coped with what was brought about was to focus on the elements and the process of his and Gopala Naika's text production (p. 94) following a variant of Albert Lord and Milman Perry's theory of "oral composition." That, and from the Anthropologist's point of view, an exceptionally expanded role given to Gopala Naika's own individuality in the Siri text tradition (passim  ).

The Anthropologist also recognizes a "mental text," but is far less concerned with the literary process by which the recitation is constructed out of it. Instead, the mental text is seen as a sort of story summary, associated with a kind of moral import, a paradigm for living, the essence of a community's moral order. ⁽¹⁴⁾ Anthropologists are interested in any and all of the utterances which emanate from this. But the Anthropologist's focus is not so much on the content of utterance as it is on the context, particularly the social context, and how the utterance is received. Whenever the text is recited we want to know the relationship between the text and the individual (or perhaps the individual's status): what legitimizes their use of the text in this context and by what right might they transform it in the act of its utterance? And, in the Tuluva traditions, since singing it acquires a special status as a speech act, one which asserts itself as "truth"(see Claus 1991b: 168), who gives them the right to sing   the story? ⁽¹⁵⁾ Since in the ritual context, the sung "text" and the spirit become one and the same, and both identified as truth (satya  ), ultimately, it can only be the spirits themselves, by their willingness to enter the woman. The singer is merely the vehicle for the spirit's voice. It is they who initially question the individual by "troubling" them, but once they are satisfied with the purity of the woman's intentions, their presence in her is regarded as "protection."

The Folklorist, drawing upon the Anthropologist's previous work on contextual variation of the Siri text, also sought out ritual contexts in which the text is uttered by Siris and Kumars. As it happens, one of these was a ritual dramatization similar to that described in the introduction to this paper, above (see Honko 1998: 409-418). But rather than seek to understand the performative dimensions of the recited text as speech acts, the Folklorist's attention is on compositional questions: the degree to which Naika's "complete text" is utilized, what motifs are kept and in what sequence, and whether the lines of text are altered. His concern is with the short time in which the "main content" of "the epic" could be dramatized in ritual. Because the dramatization and recitation did not complete the text to the point expected by the Folklorist it is regarded as something of a failure (Honko 1998: 417). When the ritual as a whole, with dozens of possessed men and women producing a cacophony of recitation, did not produce even a single coherent portion of "the Epic" the Folklorist was "not impressed with the performance." The Folklorist also wonders at the "... habit of 'beginning at the end' of the epic," and speculates why this might be so in other ritual contexts as well, but only to conclude that "... the epic story serves as a spring board for further narratives." In any case, what attention to performance the Folklorist does provide is in the nature of description and the identification of units of literary composition (themes, motifs, etc.) written in the style of extended fieldnotes, a backdrop for text collection and comparison, not the analysis of the performance per se   in a phenomenological sense.

If the Folklorist was not impressed with the performance, the Anthropologist certainly has been. Indeed, like many temple festivals in India, the Siri festivals are spectacles, with hundreds of possessed women, several public possession events and a variety of other religious performances occurring concurrently during the same night and over successive days; these festivals are designed to draw a crowd. But even spectacles must be justified.

The inclusion of every new event in any temple festival is based not only in mythico-religious history, but also complex social, political and economic considerations and negotiations. Even the inclusion of a particularly popular musician in a deity's procession (a common element in Tulunad's temple festivals) entails a potential elimination of some other event, and then a selection from amongst competing performers, each with their promoters and patrons. In the same way, the inclusion of a Siri "event" is one amongst many seen as potentially desirable by the temple management. It's inclusion in the multi-day celebrations has undoubtedly been instrumental in the enormous growth in popularity of some festivals, and at some locations it has become that for which the temple is known in many segments of the population. However, the manner in which the Siri cult is given representation varies considerably from place to place. At one temple Siri cult representatives are merely a part of the temple procession; at another, the Siri cult activities take place at a separate shrine several miles from the main temple and the ritual which takes place is reminiscent of a bhuta   ceremony in form. Where the full cult group activities are included as a prominent part at the temple location, the temple managers must negotiate with the Kumars just what format the ritual will take and how long it will last. In negotiating with the Kumars, the temple managers potentially relinquish their own control over a significant portion of the festival activities. The Kumars, under the local leadership of a chief Kumar, who himself has connections with Kumars at other festival locations, stand to gain a role of prominence.

To situate the Siri cult performance within this larger set of negotiations, the text of the Siri Epic itself plays a central role. There has to be some myth-based justification for the cult's inclusion within the larger festival activities, a "charter" as Malinowski called it. Such justifications are often based in local interpretations of the paaaDdana   text and by associating elements of the myth with elements of the local cultural and physical landscape, which, given the relative uniformity of the Tuluva countryside, are not overly difficult to come up with. What Malinowski sums up with the idea of charter is itself actually a matter of public discourse with a history and a host of competing local interest groups. Honko describes his collection of the local chartering traditions at Nidegal, but is apparently unaware that each of the other major locations where the Siri festival occurs also have similar traditions, and they are somewhat at odds with those at Nidegal. The Siri tradition is such that local Siri cult groups also participate in the activities at particularly famous central locations, notably Kabitar (primarily for the southern areas of the district) and Nandolige (primarily for the northern area). There, it seems likely, the Siri participants, and notably the Kumars, negotiate their versions of the text through combinations of a variety of discourse media, genre and styles. The search amongst these participants is for truth (satya  ), because each component of the story is a testimonial, of sorts, of the existence of the siri   spirits and their powers. One of the ways conflicting testimonial is resolved, it appears, is by compounding several different, independent story traditions into a larger set. Other, less contradictory elements can be added as additional episodes, or even scenes and events, by utilizing the distinctive phraseology and singing style of the Siri PaaaDdana  . Altogether, this might account for the growth in the length of the paaaDdana  , as well as its continued existence in shorter "complete" versions. Gopal Naika's remarkable text - the longest so far recorded - appears to have been the product of his long and active participation in this public discourse arena (See Naika's personal account of how he acquired the text in Honko 1998: 519-47).

Thus, in contexts ranging from the sung speech exchanged between cult participants in the personae   of the siri   spirits, to the dramatic enactment of the Siri PaaDdana   which initiates the cult activities, to the women who sing the paaDdana   in the fields, to the village level negotiations through which the performance is included in the larger temple festival, the "text," in its various forms and genres are intertextually linked together, and each performance which utilizes the text expropriates its meaning and significance from a previous source and is dependent upon that source for its own legitimate use. In performance analysis we identify intentions, contexts and rules governing use of speech acts, and break the performance down into stages as control passes from one individual to another, usually accompanied by relatively standardized markers. But I don't want to lose sight of the what we are doing when we do this. Unlike the Folklorist, it is not the text itself we are interested in, but the community and how its modes of communication establishes its structure through discourse. The text, we recognize, is a dynamic and fluid one which acquires new meanings and significance in its various contexts, old and new. Our interest shifts from the text to the control of the text as a discourse in a community's arenas (social, political, economic, etc.) of activity.

Performing in the Western Context

Over the years since the time Max Mueller and E. B. Tyler debated the course of human history, when the Folklorist (textualists) and the Anthropologist (culturalists) began their quest for the "Other" together, they drifted apart precisely over issues regarding texts (see Dorson 1971). Now, to their surprise, they encounter one another in the same villages, examining the same phenomena: orally performed texts  . I have tried to show that what motivates their encounter is different, that the text they collect, often from the same individuals, is different, and that what they see as significant in the texts is different. I have only instantiated one dimension of the differences, Lauri Honko's and my own work with the Siri texts, while the range is much broader, but I think the motivations are similar across that range, which includes, besides folklorists and thinking of my own regional specialty of India, Indologists, who are text specialists par excellence  .

I want to end by questioning why the Folklorist and the Anthropologist are there in the first place. What, exactly, are we doing there, what is the motivation behind our efforts? What significance do we see in the texts we collect and bring home to use in our performances? What is that     performance all about? At this point I have to speak only for myself and let the Folklorist do the same. For me, aside from the personal gain (a job, promotion, some small royalties) which I share in common, and in competition, with my anthropological peers, there is another quest which I feel has to do with meaning in the broadest sense, my existence in the cosmological order that my intellectual traditions have bequeathed to me. While it may seem naive and presumptuous to try to recall this, ultimately I think it is necessary to do so. Otherwise, as I pore over these long texts, my interpretations keep shifting, a basic set of contradictions arising out of my "etic" perspective of the texts and the "emic" perspective of those people who produce them: at one time I work on them from the perspective of an objective Western scientist, at another from the cultural perspective of a person who believes in spirits. For purely practical reasons, it seems reasonable to try to remove the contradiction and rise above the duality to a common position in which both I and the singers wish to possess the song.

In any case, I search these texts of possession performances, specialized speech acts, for a meaning I suspect I shall never understand through my own intellectual traditions. Why? For two reasons. Firstly, because the anthropological tradition is, after all, a societal one, each performative event in my research is contextualized to my society's larger intention to control the world (indeed, the universe) through the performance of a "science" which denies it even is   a performance, and instead insists it is "reality;" and secondly, because within this "scientific" performance tradition I am not given a means to identify my individual quests for meaning. All of my efforts to witness, collect, analyze, explain and then report to my fellow social scientists are contained within a tradition which acknowledges truth only in externalized objects. This is why we collect texts in the hopes that within the text lies the truths about the tradition. But then we (the Folklorist in particular) are forced to limit ourselves to what we can understand about the text. While we have done a remarkable job of collecting texts, using ever-increasing advancements in recording and storing technology and a constantly evolving comparative methodology, what it is we are hoping to find may not lie solely within the text at all. If this were the case, perhaps we would be satisfied once we collected the complete text of a tradition or discovered the means by which a singer produced the text through a variety of literary devises we could then label. Even if we could, the source of tradition would still remain hidden in the singer's mind (the "mental text") since only resultant traces of a literary process become objectified in textual components of the "completed text." In any case, what fascinates me is not the "complete text," but the text performed during the Siri festival and other contexts in which its recitation actualizes the siri   spirits - brings the spirits to possess the singer. But I am not satisfied when I learn that the ultimate legitimacy behind the text are the spirits themselves. A performance analysis falls short of comprehending what this actually means. Essentially, the performance analyst is an outsider and a non-believer, and can never understand the desire of the singer to sing, nor the source of her gratification. To know that which grants legitimacy to the singer's song has to be to know the spirits.

Yet their traditions continue to attract and fascinate me, to the point of obsessiveness. ⁽¹⁶⁾ When, during the 19th C. and in anthropological (and folklore) evolutionary discourse in particular, religion was superceded by science, the Primitive Other replaced Western religious texts as the well-spring of our meaningful existence. As we placed ourselves in relation to the Other in this evolutionary scheme we (particularly those who gave rise to the textualist traditions of folklore and philology) also placed our scientific (textual) traditions in relation to religious texts as something pertaining to the ineffible past. Subsequent eras of anthropological discourse revolved around classification of societies, putting each in their proper place from our perspective, and assigning common sense functions to their institutions. Performance analysis shifts attention to ongoing processes of construction and reconstruction, Theirs and Our own. But I am afraid it doesn't get us any closer to what we seek. We are still only translating and labeling, providing our own common-sense explanations of what we witness.

One direction we might look to find a resolution to this duality is Marcel Mauss' identification of "total prestations" which rest on the obligations to give and to receive as well as the obligation to repay. In a sense, the Siri PaaDdana   can be looked at in this way. When a male Kumar appropriates the Siri text from a woman's fieldsong tradition and in the ritual context gives it back to an individual woman to acknowledge her possession by a siri   spirit he does not, in fact, "appropriate" the PaaDdana  , but was given it during a previous Siri ritual in much the same way he and the other adepts give it to the novices who are brought to them. By receiving the spirit (of Kumar) he is receiving the PaaDdana   in any and all of its variations. The same is true of other (female) novices and this is why they must return, to repay (give) the spirit/song to others.

Perhaps I didn't need to emphasize objectivity and objectification as I had done; this created a distance, a barrier which very quickly became unbreachable. During my first stint of fieldwork specifically focused on understanding this remarkable ritual, after months of recording, translating and interviewing women who had an intimate experience with the siri   spirits, I felt I had gained for myself (albeit somewhat vicariously) a certain experiential knowledge of the spirits. As it happened, just at the moment the hundreds of others became possessed, I momentarily lost consciousness during a Siri festival. I cannot say that I was possessed, because I did not pursue the performative ritual definition of this experience as such, but rather, in my name, my assistant, himself by then an experienced intermediary between his foreign employer and the tradition he too was learning, brought me to the priest and begged to be excused from joining the ranks of the possessed.

I think, now, after 25 years of pursuing the course I have -- the objective study of the Other's tradition -- I made a fundamental mistake. I should have allowed myself to become possessed and to have been given   the text in the form of the words of the spirit -- the really few words necessary to perform the speech act of possession -- and the authority (indeed, the obligation) to use it. As it is, I think the reason I have not gained a satisfactory understanding of the performance and its text is that my ownership of it lacks legitimacy (of course I conscientiously followed the etiquette of good field practice and acquired the usual permissions to gather and use the data). What I came away with was a text and a ritual performance, not a text which was a spirit.

1. This paper rests on a long series of others I have written about the Siri cult and its associated texts, and upon fieldwork which was carried out in the Tulu-speaking region of Karnataka over many years on a wide variety of topics. The Siri songs (PaaDdana  ) and their associated rituals were among the central topics of two of these fieldtrips: as one of three possession cult rituals associated with PaaDdana   textual traditions in a year-long project investigating the textual sources of the speech of spirit-possessed mediums carried out in 1975, and in 1996 as one of five PaaDdanas studied in a systematic collection project carried out over a six-month period which spanned the entire Tulu-speaking region. In addition to these two projects, I have had the occasion to witness a number of Siri Cult performances and interview a number of participants during several shorter visits to this region from the period 1987 to 1998. My first collection of the Siri PaaDdana   was done in 1968 during my dissertation research on Bant-Nadava kinship organization.

2. Dakshina Kannada which had been previously known as South Kanara (or South Canara) District has now been separated into two districts, Udupi District and Mangalore District, with Udupi and Mangalore, respectively, as their district seats.

3. In the spiritual cosmology of Tuluva culture, the term siri   can refer to either a class of spirits or a particular spirit (or deity) named Siri. In this paper I will capitalize the term when it is used for the song tradition (the Siri Pana  , the Siri Epic), the annual ritual (the Siri ritual) and in other instances where the term explicitly refers to the proper name of this spirit or when the reference is ambiguous. I use the lower case and italics when the term explicitly refers to the class of spirits. Throughout most of the Tulu-speaking region the siri   spirits are frequently said to be seven in number (veru sirikulu  ) and include the five members of Siri's matrilineal family mentioned in this paragraph. The identity of some of these spirits in some of the song traditions varies, however. In the northern portion of the Tulu-speaking region, and further north, into the Kannada-speaking portions of Udupi District, some song traditions (called hogalikke  ) and rituals devoted to cikku   spirits include the names and stories of spirits which to the south and in Tulu are called siri   spirits, suggesting that the class of spirits is cosmologically cognate. Also to be considered is the reference to siri   spirits in other Tuluva oral traditions and those of neighboring regions (both Tulu- and Malayalam-speaking) to the south (for further discussion see Claus 1993b). Until a more complete picture of folk cosmology is researched, it is probably best to consider the Siri Epic as one still being developed out of this larger song and ritual tradition and its development a discourse with a variety of religious, political and economic overtones.

4. While considered bothersome in this context by some women, possession by the siri   spirits is not usually considered dangerous. There is, however, the possibility that other, more dangerous spirits might also be attracted to the singing and could desire to possess the singer.

5. I must stress, though, that this is only one "typical" route by which a woman becomes a siri   medium and that there are probably other significant regional and situational patterns. One of the flaws of a methodology which draws its case study samples from the Siri ritual is that it ignores regional and family traditions where the individual siri s are drawn from other means of selection, such as hereditary succession (mother to daughter), aptitude for possession, and female leadership qualities. In my interviews with siri   mediums in the villages in Udupi District I encountered women who claimed to be siri   mediums who had never attended the Siri festival and said only that their mothers were siris. It may be that the context of their practice is primary that of their family or neighborhood (kare  ) and that the ritual in which they become possessed is similar to that of an agel  . It is still unclear as to what connection there is between the Siri rituals described in Claus 1975 and 1991 and many other ritual traditions in which possession mediums play a central role in Tuluva culture. There is evidence, though, that the Siri rituals are a very dynamic set and that the large, popular annual festivals may be relatively recent developments - a response to some of the many changes in Tuluva society over the past 50 years. Its emphasis on "curative" functions, in particular, may be related to these changes.

6. This paper was first presented at the South Asia Conference, Madison, WI, 1991 and published in the Journal of South Asian Folklore in 1997. A version of the paper, along with color photographic illustrations, is also currently available on the Internet at http:\\class.csueastbay.edu/anthropology/claus/index.htm.

7. Other scholars include our Indian academic colleagues: B. A. Viveka Rai (1996), Amritha Someshwara (1962), and S. Upaddhyaya (1986), Haridas Bhat, S. A. Krishnaiah and N. V. Rao (1984) and others. Taking into account their participation in the textual and ritual traditions entails several additional dimensions of translation and interpersonal relations which I exclude from the present consideration.

8. Few anthropologists can match this record of published documentation of research assumptions and procedures. In addition to the three-volume publication mentioned above were several conference presentations, workshop sessions, and throughout the several years of fieldwork, a number of orientation meetings during which the research team discussed research and interpretive methodologies and strategies.

9. I realize that there are numerous schools of thought in both disciplines, both historically and at the present time, and that these often see themselves opposed to one another on many issues. I should note in particular that there are also schools of folkloristics which stress both context and performance, and thus approximate that which is represented by the Anthropologist in this paper. In fact, as I discuss at a later point in this paper, both folkloristics and anthropology ultimately derive from the same early 19^th Century philological traditions. What I want to contrast here is the century-long emphasis on the text (written and oral) by folklorists and the equally long emphasis by anthropologists on culture and society.

10. I mean here, of course, our presentation (and archiving), classification, analysis, and interpretation of this material in our Western knowledge base which, since at least the days of Colonialism, reaches out to, and deeply affects, those who maintain the indigenous traditions. As has been frequently pointed out in the post-colonial era, our control over the text in this sense is an exercise of our power over not only Others, but the very stuff (discourse) of their culture identity.

11. Gopala Naika is a Kumar at Niega, a village in Belthangadi Taluk, in the recently formed Mangalore District. Naika is of the Gauda caste which has migrated to the Tuluva region in relatively recent times from the Kannada-speaking areas of Karnataka to the East. While many Gaudas are bilingual in Tulu and Kannada, most speak Kannada as their house language. The Siri PaaDdana   as Naika recites it, and the Siri rituals he participates in are probably typical only of the Belthangadi region.

12. I have recorded portions of at least three Siri cult performances. All of these are limited to selected portions of the ritual and to ritual-context recitations by selected individuals. The recording quality is relatively low-level, using less-than-professional quality equipment. In all cases, the recording was made in the midst of the ritual action and consequently there is a high degree of ambient noise. Although these tapes were used to transcribe portions of the recitation which were later translated for publications, the recordings were not made with any attempt at collecting or documenting the full Siri Epic.

At various times, but in particular in 1969, 1975 and 1986, I did record the complete Siri Epic as the singer understood it to be, under artificial conditions, mostly consisting of the house or courtyard of the raconteurs. In all cases, the singers were regarded by many others in the locality as people who know the full Siri PaaDdana  . These have all been transcribed, some of them somewhat loosely translated into English and several of them published in summary form in a prose narrative style. Copies of most of these tapes have been deposited in the Regional Resource Centre for Folk Performing Arts, Udupi or the American Resource Centre for Ethnomusicology, Delhi.

13. I deal with the complex issues of variability and authenticity of the Siri Epic in a separate paper (Claus 1993b).

14. I have described this "mental text" variously, under several different contexts and in relation to different situations. In one sense, a PaaDdana   and its hero are a model for an individual in distress to reintegrate themselves or their group into a moral order (Claus 1984). In another, the PaaDdana   is a reference (text) for societal ideals or ideal-appropriate action (Claus 1978). In yet another, it expresses a group's place in relation to other groups in a larger societal scheme which is represented in the whole of the PaaDdana   corpus (Claus 1989). In yet another paper I use the concept in reference to compositional techniques use by singers to produce the performed texts which I had collected (Claus 1993). This last sense, elaborating upon the work of Lord and Perry, is very close to that also used by Honko (1998: 92-116).

15. After collecting many PaaDdanas in many different contexts, by different types of singers, and after translating, analyzing and commenting upon them through a number of published and unpublished papers, I have finally come to the conclusion that most of the PaaDdana   oral literature is a genre of formal testimonial. This is not the place to argue this point, but I believe that it can be best understood as evidencing the power of the Tuluva spirits (bhuta  , daiva  ) in a relatively fixed, traditional, historical narrative form. As such, they come to resemble legends in some cases, myths in others. I am now convinced, though, that they differ significantly from both ballads and epics, in that the focus of the story is not actually the "hero," but the deity, even though the deity's acts normally constitute only a small portion of the content and the narrative dimension usually follows a human actor ("hero") and the deity only appears in a few scenes. This shift in the generic classification and the focus of our perspective has significant implications in how we interpret the text and its performance contexts. For example, while both ballads and epics may contain, evidence, or express a morality, the Tuluva pana   literature by and large collapses the distinction between morality and deity, equates the two, or perhaps even ignores "morality" in favor of the nature of the spiritual (the divinity) itself. When tragedy (the main mood of virtually all PaaDdanas)   befalls heroes, it is not necessarily because they acted immorally - indeed, morality may not even be an issue - but simply because they ignored the deity. Deities are facts, the paaDdanas attest, and it is the deity who ensures that the singer portrays their nature correctly by troubling (uppadra  ) the singer if they don't.

16. In a recent conversation with Dr. Elisabeth Shoembucher, a German Anthropologist who is similarly poring over a huge corpus of discourse text recorded during possession cult sessions in another part of southern India, put it this way: "I think maybe our fascination with possession is that we are looking for the Other's Other." While she may have meant the Other's Divinity, for me, this could only mean Ourselves.

References

Claus, Peter J.

----- 1975. "The Siri Myth and Ritual: A Mass Possession Cult of South India," Ethnology XIV: 1:47-58.

----- l978. "Heroes and Heroines in the Conceptual Framework of Tulu Culture: Journal of Indian Folkloristics, 1, (2):28-42.

----- l979. "Spirit Possession and Mediumship from the Perspective of Tulu Oral Traditions" Culture, Medicine and Psychiatry, 3:29-52.

----- l984. "Medical Anthropology and the Ethnography of Spirit Possession" Contributions to Asian Studies edited by Judy Pugh and Valentine Daniel.

----- 1986. Playing cenne: The meanings of a folk game" in Another Harmony: New Essays on the Folklore of India, Stuart Blackburn and A.K. Ramanujan, eds.,. pp. 265-293. Berkeley: University of California Press.

----- 1988. Future Research in Tulu Culture. Academy Silver Jubilee Lecture 4. Manipal (India): Academy of General Education, Manipal.

----- 1989. "Behind the Text: Performance and Ideology in a Tulu Oral Tradition". Indian Oral Epics edited by Peter J. Claus, S. Blackburn, S. Wadley and J. Flueckiger.

----- 1991a. "Kinsongs." In Arjun Appadurai, Frank J. Korom and Margaret Mills, eds., Gender, Genre, and Power in South Asian Expressive Traditions. Philadelphia: University of Pennsylvania Press.

----- 1991b. "Tulu PaaDdanas: Text and Performance." In Perspectives on Dakshina Kannada and Kodagu, edited by the Editorial Committee, Mangalore University Decennial Volume. Mangalagangotri (Karnataka, India): Mangalore University. pp. 1-10.

----- 1993a. "Variability in the Tulu PaaDdanas." Paper presented at the Conference on Dravidian Language and Literature Stuttgart, Germany. On the Internet at http://class.csueastbay.edu/anthropology/claus/gundert/gundert.htm

----- 1993b. "Text Variability and Authenticity in the Siri Cult." In Bruckner et. al. Flags of Fame (Heidelberg University, South Asian Studies No. XXVII). New Delhi: Manohar. Pp. 335-74.

----- 1997. "Ritual Transforms a Myth." Journal of South Asian Folklore 1 (1): pp. 37-57. (A version of this paper was presented under the same title at the South Asia Conference, Madison, WI, 1991 and published on the Internet in 1994 at http:\\class.csueastbay.edu/anthropology/claus/index.htm)

Dorson, Richard M.

----- 1968. The British Folklorists, A History. Chicago: University of Chicago Press.

Honko, Lauri.

----- 1998. Textualizing the Siri Epic (FF Communications No. 264.) Helsinki: Suomalainen Tiedeakatemia, Academia Scientiarum Fennica.

Rai, B. A. Viveka.

----- 1993. The Genres of TuLu Folk-Poetry: An Introduction. In Bruckner et. al. Flags of Fame (Heidelberg University, South Asian Studies No. XXVII). New Delhi: Manohar. Pp. 269-82.

Rao, N. V.

----- 1984. TuLunaaDu Siri   (in Kannada.) Bangalore: Pragathi Publications.

Someshwara, Amrita.

----- 1962. PaaaDdanagaLu   (in Kannada.)

Upadhyaya, U. P.

----- 1986. Folk Epics of Tulunadu. Udupi: Regional Resource Centre for Folk Performing Arts, Udupi.