SECTION THREE—
THE BOUNDARIES OF EPIC PERFORMANCE

8—
Problematic Performances:
Overlapping Genres and Levels of Participation in Arabic Oral Epic-Singing

Dwight Reynolds

Dwight Reynolds focuses on the Egyptian oral epic of Sirat Bani Hilal, which relates the heroic exploits and westward migration of the Bedouin Arabic Bani Hilal tribe during the tenth and eleventh centuries and belongs to an age-old narrative cycle that includes both poetic and prose genres. Though the epic circulated widely for centuries, the twentieth-century tradition has survived more narrowly. It is still performed in northern Egypt, however, where it is perpetuated by hereditary, professional singers whose spirited and multifarious renditions resonate deeply within the community. Reynolds suggests that this tradition derives much of its strength and appeal from the artistry of these oral poets, who skillfully manipulate generic and performance boundaries, constantly generating and sustaining levels of dynamic interplay between themselves and their audiences.

One of the central ideas of performance studies is that performance is a distinct, identifiable type of human communication that in turn allows for or even requires special types of interpretation. Key to this idea is the understanding that performances are bounded or framed behaviors that are marked for the purpose of recognition.^[1] The implication is thus that within the context of performance events at least two types of communicative behavior will be found: "performance" and "nonperformance." By expanding our theoretical focus, however, from the internal analysis of a single genre of performance, such as oral epic-singing, to the larger domain of entire performance events, we encounter problematic areas where genres appear to "overlap" or "interpenetrate," where the "breakthrough" into and out of performance occurs over and over again, or where different "levels" of performance seem to be present all at the same time.^[2] These phenomena directly challenge any simple definition of performance as an object of study.

In focusing upon this central idea of complex performances, I hope to suggest a line of thought that might lead us to examine the problem of complex units, such as whole performance events, composed either of multiple

genres or multiple movements into and out of performance such that the problem of boundaries and their interpretation arises, and further, to speculate on the ramifications of such problematic boundaries for performance studies.

Thus in this essay I would like to examine this metaphorical conceptualization of boundaries, frames, margins, and edges within the question of where performance and nonperformance do (or do not) meet in the northern Egyptian Arabic oral epic tradition of Sirat Bani Hilal (The epic of the Bani Hilal Bedouin tribe). Specifically, what I want to offer here is a portrait of a performance tradition as a whole, rather than the analysis of the style of a specific singer or the close analysis of an individual performance event. In other words, I am making the claim that just as individual singers can be observed to deploy a certain standard repertory of techniques of composition and interactions in their performances, we can also trace the rough parameters of an entire regional tradition through the observation of many performances in order to provide an image of the typical, recurring elements in the larger tradition. I hope that this broader portrait may provide both the material and impetus for comparative analysis with other epic traditions around the world.

In brief, I will examine three types of problematic boundaries here. By problematic I mean aspects of this northern Egyptian oral epic performance tradition that appear to challenge our commonly received notions of the "boundedness" of epic as a textual genre and as a performance genre. These three types are: (1) the boundary between the epic itself and the auxiliary genres of song with which it is most commonly performed; (2) narrative strategies within the epic that transgress or bring into question the boundaries of the epic story world; and finally, (3) audience interaction and participation in the epic performance event that create "overlapping performances."

The Sirat Bani Hilal Tradition

The gist of the story of Sirat Bani Hilal is historical: the Bani Hilal Bedouin tribe did indeed exist, and they did indeed embark upon a major migration from their homeland in the Arabian peninsula, across Egypt and Libya, and into Tunisia and Algeria. The Bani Hilal then controlled much of the hinterlands of North Africa for more than a century. These events took place between the tenth and eleventh centuries and form the core narrative of the epic tradition. Later, in the mid-twelfth century, the tribal confederation suffered a series of military defeats at the hands of an eastward-expanding Moroccan dynasty, after which the tribe ceased to exist as a political and social entity. Interestingly enough, though it appears that historically the Bani Hilal were destroyed by outside forces (that is, the Moroccan Almohad dynasty), in the folk epic tradition the tribe defeats itself.

Wracked by a series of internecine conflicts, the various clans go to war against one another until the tribe is annihilated, providing an even more tragic close to the tribe's history.

In all probability it is the disappearance of the tribe itself that has allowed its history to become appropriated as a pan-Arab epic tradition. If the tribe had survived, it is quite probable that the exploits of the Hilali heroes would have remained a tribal history restricted to poets of the Bani Hilal tribe. Instead, the tales of the Hilali migration, their battles and conquests, their romances, their winning of brides, and so forth have been woven into a narrative of extraordinary length that has been documented in Arabic oral tradition in virtually every corner of the Arab world, from Morocco in the west to the Sultanate of Oman in the east. We possess historical evidence documenting the performance of this poetic tradition from the fourteenth century onward; in addition, a rather impressive manuscript tradition emerges from the eighteenth century onward, and a tradition of cheap chapbook publications dates from the late nineteenth and early twentieth centuries. Finally, in the late twentieth century, we have a number of audio and video recordings of live performances from several different regions.^[3]

A primary characteristic of this tradition, however, is that it exists in many different types of performance. In some areas it is narrated as a prose tale cycle, in some areas in prose embellished with brief passages in verse, in some areas as chante-fable, where the narration of events is in spoken prose, and the speeches of the heroes are in sung or cantillated poetry, and finally, in some regions the entire story (with the exception of brief introductory passages) is performed in sung, rhymed verse to the accompaniment of one or more musical instruments. It is this latter mode of performance that I will address here. This sung, versified mode of performance was much more widespread in previous centuries, but in the latter half of the twentieth century it is found in only two main regions: northern and southern Egypt.

I will focus specifically on the northern Egyptian performance tradition as performed by the hereditary, professional epic-singers of the Nile Delta region. By hereditary and professional I mean that the occupation is almost always inherited by descent and that this is the sole livelihood of these performers. These epic-singers constitute a defined social group within northern Egyptian rural society and are understood to be one of the several groups commonly referred to as Gypsies [Ghajar]. This group in particular, however, the Halaba, are associated with two occupations: blacksmithing/tinkering and epic-singing. Over a number of years I have studied a single village that is home to fourteen households of professional epic-singers. These fourteen households are, in addition, related by blood and/or marriage to another seventeen households of epic-singers scattered across northern Egypt. This population almost certainly constitutes the entire extant body of hereditary epic-singers in northern Egypt.

With the full understanding that there can be no such thing as a complete version of the epic, in order to communicate some sense of the proportions of this tradition, I can cite three versions of the epic that I recorded from three different poets in the village, each of which can be said to cover the general narrative scope of the tradition as it exists in this region. One version was thirty-four hours long, approximately 5,500 verses (and a verse here is an end-rhymed unit 24 to 30 syllables long—in contrast, for example, to the southern Egyptian quatrain style of performance in which a verse is 8 to 12 syllables long). A second version is fifty-three hours long—that is, approximately 12,000 verses. And a third rendition had reached seventy hours in length when we had to abandon the effort due to the poet's health, and represented approximately two-thirds of his repertory. A rough estimate based upon the poet's listing of the episodes he had yet to sing would place his version at well over a hundred hours in length. Incidentally, when these "full-version" recordings (field experiments I purposefully conducted to explore the narrative breadth of the tradition) are contrasted with recordings I made of naturally occurring performances, the poet who sang the thirty-four hour version beginning to end typically takes nearly twice that time to sing each episode when performing for an enthusiastic audience.

As for the topic of problematic boundaries of performance, the first area I will explore is that of the interaction between the epic as genre and the auxiliary genres with which it is typically performed. Different performance situations in the village of al-Bakatush draw different types of audiences and create rather differently structured events. Until the very recent past, Sirat Bani Hilal was typically performed at weddings, saints' festivals, in coffeehouses, at circumcision ceremonies, at harvest festivals, or in a private evening gathering, that is, a sahra or a lela. This latter is currently by far the most common epic-singing event. During my various stays in the village in 1983, 19861987, and 1988, I was able to attend well over a hundred such gatherings and was able to record seventy-six full performances and fragments of an additional twenty. It is from this specific type of event that I draw most of the following examples.

Auxiliary Genres: Madih , Mawwal , And Sira

The sahra as a performance context for epic-singing in al-Bakatush is defined by the sahra's patron or host, who will have negotiated with a poet to perform that evening. When the poet arrives, a round of tea and cigarettes will probably be offered before the performance is set in motion. When the poet deems fit, he begins by readying his instrument, putting rosin on the bow and on the rabab's two strings, and finally tuning the instrument and testing it briefly against his voice. After first pronouncing the basmalah ("In the name of God, the All-Merciful, the Compassionate"), he plays a brief musical

introduction. This instrumental interlude serves to alert guests who may still be waiting outside that the performance is about to begin. The first genre sung is a song of praise to the prophet Muhammad (madih or madh al-nabi ) This may be accomplished in a few short verses by citing some of the Prophet's best known attributes (his compassion, his beauty, etc.), by allusion to well-known tales from the Prophet's life (he who ransomed the gazelle . . . , he to whom the camel spoke . . .), or by the poet's choosing to recount one or more such tales at length. Unlike the monorhymed odes of the epic, these songs of praise are constructed in quatrains, usually rhymed abab/cdcd/efef. . .

This opening praise song to the prophet Muhammad unifies the audience and casts them immediately into an interactive mode through its repeated references by name to the prophet Muhammad, to which listeners respond with one of the nearly obligatory traditional blessings, "May God Bless and Preserve him [salla llahu 'alayhi wa-sallam]" or "Upon him Be God's Blessings and Peace ['alayhi al-salat wa-l-salam]."

In most performances of epic-singing in al-Bakatush, the praise song to the prophet Muhammad will be following directly by a mawwal, a short lyric poem of five, seven, or occasionally nine verses that is performed in a rhythmically free and often very melismatic style in which a single syllable may be lengthened and sung over several notes. During the mawwal there is a marked shift in presentation style and in the role played by audience members. Here the focus is the quality of the singer's voice, his ability to ornament the melodic patterns, and rather than responding phrase by phrase, the audience typically waits until the end of the brief song and responds to the entire text as a whole with approbation or criticism. The themes of the mawwal are also quite distinct from the earlier praise of the Prophet, for this is a song form that focuses on the trials and tribulations of everyday life. Its texts are often couched as advice to the listeners in a proverbial rhetoric using stock images of Egyptian folklore. The forces of Fate are addressed directly as Time, Destiny, the Nights, or the World; the lion functions as a symbol of the ruler, the mosquito as the interloper or sycophant, the camel as the stalwart man, the doctor as an image of the Beloved, since she or he alone can cure the disease from which the Lover suffers. But the mawwal is a complaint addressed in the end to one's fellow human beings and not to God; it is not a plea for intercession, but rather a plea for understanding.

Mawwal  (Shaykh Biyali Abu Fahmi, 2/11/87; rhyme pattern aabaa )

1. is-sabr 'uqbuh farag li-lli nshaghil baluh

Patience-its result is release for him whose mind is occupied (with cares and troubles),

2. ahsan min illi yifadfad yihutt il-fikr fi T baluh

(Which is) better than he who grumbles and fills his mind with thoughts.

3. ma fish ahsan min illi yisbur

There is nothing better than he who is patient

4. li-hikam iz-zaman wi-awanuh

(and endures) the judgments of Fate and his Era;

5. min husn 'aql il-gad' biyi 'dil ahmaluh

From the good sense of the stalwart fellow he is able to balance his loads (in life)!

This aphoristic mode of thought is of course reminiscent of many different epic-singing traditions, such as the proverbial mode used in Mandé hunter songs or the brief aphorisms used by Muslim singers in Yugoslavia or even the improvised terme songs of Central Asia traditions.^[4] In all of these traditions, the proverbial or aphoristic song occurs at the point of transition into the more heroic epic genre. In the case of the Arabic tradition, it promotes a fundamental shift away from the ideal as represented in the life of the prophet Muhammad toward the level of identification with the characters in the epic who display at one and the same time both superhuman characteristics and all-too-human weaknesses and failings.

Finally, the poet directs us into the epic world, but in order to do so he moves carefully through a complex series of steps that are repeated whenever he commences the epic or returns to the epic after an interruption. The poet begins with a brief scene-setting that informs the listeners where he is in the overall story. If this is the very beginning of a sahra, he will include the phrase "The narrator of these words said [qal ar-rawi]" or "The author of these words said [qal mu 'allif ik-kalam] ." The following brief prose introduction is narrated in an ordinary, though occasionally rhythmic, voice. While narrating, he begins to ornament the prose with rhymes [Arabic saj' ] which occur at progressively regular intervals. Then a character within the epic must be emotionally moved to speak—that is, the situation a character is in must produce such fear, joy, or sadness, for example, that he or she is compelled to rise and compose a poem that addresses the situation. When the character begins to recite the poem, the poet moves into full song. It is this movement into an "authoritative speech-act" that brings us into the world of epic speech. Once we have arrived in that world, many different voices can be deployed, including that of third-person narrator, but the movement from speech to song, from the outer world to the world of epic verse is always accomplished by the shift into an authoritative first-person utterance.^[5] I have elsewhere analyzed this process in much closer detail as a series of

very intricate shifts of voice, melody, and poetic form.^[6] The basic steps, however, are as follows:

instrumental music —> rhymed prose —> emotional crisis —> sung epic verse

When entering this world of authoritative heroic speech—that is, the epic verse itself—the performing poet briefly reiterates all of the framing devices of the sahra performance. He stops and performs a short musical interlude. Then the opening verses of the poem will be in praise of the Prophet (a miniature madih ) The poet then calls upon the audience to listen to the words of the epic character in a verbalized quotation marker parallel to the opening formula "The narrator said" and usually describes the character as being in emotional turmoil:

Quotation Marker  (Shaykh Taha Abu Zayd, 6/I/87 )

O listeners to these words, wish God's blessings on the Prophet, 
Taha, fortunate is he who visits him [i.e., on pilgrimage].^[7]

The emir Barakat said-and his heart was in pain; 
The fire of his heart in his soul did sear him.

When the epic character actually begins to sing, he or she sings either a full mawwal or several lines of aphoristic poetry in the style of the mawwal but in the form of epic verse:

Aphoristic Verses  (Shaykh Taha Abu Zayd, 6/1/87 )

Said King Fadl from what had befallen him: 
"It is a wretched world, and Fate is a tyrant!

Ah! From the world and the reversals of all that happens in it, 
In my opinion Fate spins in circles.

After the good things of the world and the cup of sweetness, 
There is no escape from drinking also from the cup of bitterness.

Each stage of the evening performance to that point is thus reproduced in miniature as a frame for the "language of heroes" (to borrow a phrase from Richard Martin):^[8] instrumental introduction/praise of the prophet Muhammad/quotation marker/aphoristic verse/narrative material. This frame is repeated each time the poet begins to sing, even after brief interruptions, providing a constant mirroring between the performance of the living epic-singer and the performances of the heroes of the epic.

In the following short example, Khadra the Noble (mother of the hero Abu Zayd) has had her livestock stolen by a marauding band claiming to be collecting a "tax" from her. The dialogue between Khadra and the head of the raiding party has been recounted succinctly in prose, and the poet now leads us back into epic verse with Khadra's lament:

Transition into Epic Verse (Shaykh Taha Abu Zayd, 6/2/87 )

Spoken:

"So Khadra wept in anguish; she could not have wept harder. See now what Khadra will say, and we shall now cause the listeners all to hear it. He who loves the Prophet wishes God's blessings upon him!

[Audience: May God Bless and Preserve him!]

[Musical Interlude]

Sung:

My first words are in praise of the Chosen One, the Hashemite [Muhammad], 
We have no intercession [with God] save him.

Said Khadra when Fate leaned upon her, 
"By the life of my Lord [God], there is no god but He.

The Creator of creation . . . Ah! . . . and He knows their reckoning, 
In His power He knows, yes, all matters indeed.

I shall complain of my sufferings to [King] Fadl, yes, the hero, 
Under whose protection these long years I have remained.

If he is able to defend my rights, 
I shall stay the coming years here with him.

[But] if he is not able to defend my rights, 
I shall go seek a people who defend their refuge seekers!

Where shall I go? Whence am I coming? 
My lowly state pains me; woe is he whose strength has been crushed by Fate."

In terms of frames and boundaries, the sahra performance as conducted in northern Egypt thus challenges our ideas of where the beginning of the epic lies. The poets themselves answer that the epic begins after the formula "The narrator said." Yet the intense replication of the musical interlude, the praise song to the prophet Muhammad, the verbalized quotation marker, and the aphoristic rhetoric of the mawwal sequence within the epic to frame the speech of heroes suggest that the "auxiliary" genres are in fact critical parts of the epic itself. Richard Martin, who has studied the speech-acts of characters within the Iliad in such rich detail, has also pondered the framing of the Iliad as a whole in order to ask, "What speech-act does the poem make?"^[9] My response to Martin's observations and question, in part, is that despite the highly significant relationship between the madih, the mawwal, and the sira, which I have summarized here, when asked to provide a translation or summary of the epic, I instinctively find myself beginning simply with the

sira, that is, the "epic" itself, and discarding the accompanying genres, perhaps unwittingly repeating a decision possibly taken by the scribe of the IIiad long ago. In northern Egyptian performances of the Hilali epic, the speech-act of the poem is indeed parallel to the speech-acts of the heroes within the poem. Thus without the constant dynamic of the interaction between the frame of the larger epic-singing event and the frames of the heroes' performances within the epic, a highly significant part of the epic tradition as a whole would indeed be lost. In addition, as I have explored in some detail in other publications, this constant equation of the speech of poets and the speech of heroes is part of a larger social dynamic involving the proper roles of poets and heroes that places the epic at the very center of symbolic conflict within the context of the village.^[10]

Narrative Strategies

The second set of observations I would like to make concerns narrative moves within the epic performance. The first example is of a process found in many traditions—that is, the process of glossing the text in a series of asides or excursive remarks that overtly depart from the narrative and are openly addressed to the audience. In the following example, a poet comments to his listeners about the rank and respect accorded to the poets of old (with obvious reference to the contrasting status of poets in modern village society):

Shaykh Biyali Abu Fahmi, 2/14/87

Spoken Narrative

So [King Hasan of the Bani Hilal] said to him, "O poet Jamil." He answered him, "Yes, O father of 'Ali [=Hasan]."

He said to him, "You have journeyed among many peoples, among the great ones of the noble Arabs—has anyone given you gifts and treated you as generously as have I and my Arabs?"

Poet's Excursus

Now the poet was of great politesse [Arabic adub, emphatic form of "polite"]. The poet was of great politesse, for every poet who picks up the rabab is of great politesse. Why? Because he sits with good people. Because a poet never possesses bad manners. He travels with his rabab . . . . I do not laud poets merely because I am a poet! [laughter from audience]. . . . It is because the histories tell us so! The poet was of great politesse. Were he not of great politesse, he would never pick up the rabab and sit with good people. How could he be a poet of kings and Arabs and be impolite? He was of great politesse. And the audience, as well, when they listened to a poet, they were the pinnacle of respectfulness.

There is another less obvious but equally effective means of moving back and forth through the boundaries of the story world: a poet may make a request or an observation concerning his host or audience by manipulating a character within the epic. If an audience member falls asleep, for example, the poet may cause a hero within the story to fall asleep, and begin singing in a softer and softer voice. Knowledgeable listeners will immediately begin looking about to see who is going to be the butt of the poet's joke. The poet sings more and more softly, and then another character arrives at the scene to shout "Wake up!" The shout usually rouses the unsuspecting napper and provokes a round of laughter among the other listeners. Or, when seeking a glass of water, the poet may cause the hero to ride out into the desert, and then describe in great detail the sun and the heat and how the hero's throat is getting scratchy and sore: "Yes, the sun is beating down and the heat is intense, . . . his sword is so hot he cannot touch it, . . . and the sweat is pouring off his stallion . . . . He is getting thirsty . .. oh so thirsty." [In fact, when this paper was first presented, Margaret Beissinger, who was sitting in the front row, leapt up to pour the speaker a glass of water. Reynolds responded, "Well, it is clear that not all boundaries are problematic, since certain rhetorical devices function quite well cross-culturally!" Ed.]

A particularly fascinating use of blurred boundaries occurs in this epic tradition when heroes within the epic are portrayed as also being epic poets in their own right, and we are then treated to the description of a performance of praise songs to the prophet Muhammad or bits of epic poetry as sung by a hero within the epic. The result is thus a complete reduplication of performance event, host, poet, audience, and song. And it is exactly at this moment that northern Egyptian epic-singers often choose to express their social or political commentary about life in the village. The hero in the epic who is singing as a poet may criticize his patron or his audience, and of course should any offense be taken by our host or listeners, the performing poet denies any intent on his part and distances himself from the offending words as part of the epic—after all, it was the hero Abu Zayd who sang them!^[11] In these three short examples—a poet's aside to his audience in an attempt to strengthen his own social position by reference to the epic story world, waking a sleeping listener or asking for a glass of water by manipulating the story line, and burying social commentary inside a performance within a performance—we can see in miniature the powerful dialogue that exists between the worlds inside and outside the epic as mediated by the poet who has primary control of the boundaries between them.

Audience Participation

Audience members listening to a performance of Sirat Bani Hilal in northern Egypt, as in many traditions, may take a very active role in shaping the

performance by critiquing or encouraging the poet, by requesting particular scenes, or by pointing out that a particular detail has been overlooked. But in this tradition, audience members may also find themselves suddenly part of the actual stuff of the performance, rather than in the somewhat exterior role of supporters, encouragers, or critics. This may occur in one of several different ways. First, as we have seen, the poet may direct commentary or observation toward an audience member by manipulating the plot or characters of the epic as demonstrated above. In addition, there is a genre of performance that is often performed along with the epic in which audience members themselves are the material used by the poet. This genre is called hitat baladi, which can be loosely translated as "bits of local color," and consists primarily of a comic routine in which the poet attempts to weave the names of his listeners, their characteristics, references to families, friends, land, and so on, into a loose narrative recounted in rhyme. This genre also constitutes a form of social Russian roulette, since one never knows whether one is going to be mentioned in a positive or negative fashion, complimented or made the butt of a joke. In this sense the situation is similar to attending performances in certain comedy clubs in the United States, where the performer is likely to abandon pre-prepared material and make jokes with or at audience members at any moment. This is a performance drawn directly from the performance context.

But there is still another means by which audience members become part of the very fabric of the performance. During tea breaks and other interruptions, listeners often engage in discussions of the epic and its characters; this in turn often leads to discussion of events external to the epic that are parallel or relevant to the discussion. For example, the motif of a hero in the epic sleeping at prodigious length (the Heldenschlaffe of European folktales) one evening provoked a long series of first-person narratives recounted by listeners. One man told of a time when he had been returning in the evening from a trip outside the village and was so tired that he fell asleep on his donkey. When he finally arrived home in the middle of the night he simply continued to sleep on his mount at the door of the house till the following morning when his wife finally discovered him, still sound asleep! Another listener jumped into the discussion with the tale of how he had one time had a similar experience but had actually fallen off his donkey on the way home and remained asleep, his donkey standing by peacefully all night long, and had awoken in the morning alone, in the middle of a strange field, lying alongside the path!

The final story in this series was told by two listeners who had served in the army together and had received leave to go home for one of the holidays. They caught the train with several friends from the same village and, since they were stationed many hours from home, eventually all fell asleep, only to wake up in the middle of the night in the next province having trav-

eled quite a distance past the village in the other direction. But when they piled off the train in confusion at two in the morning in a strange city, who should they find standing in front of the station (through God's mercy!) but a man from their own village with his taxicab who had just delivered someone to the station and was now planning on heading home to the village for the holiday, too. Everyone arrived safely and in time for the feast!

When we finally returned to the epic, the poet took his cue from his listener's tales and wove references to their stories into his narration of the epic. In this tradition, this process of weaving elements from the epic into discussions that then provoke performances by audience members of first-person narratives, proverbs, folktales, and other conversational genres is quite common, as is the poet's response of making reference to, or even citing, these audience performances in subsequent portions of his rendition of Sirat Bani Hilal. The thread linking different elements of the performance event as a whole often runs through an entire series of different genres performed by a series of different performers.

Conclusion

In our first set of examples we saw that the auxiliary genres commonly performed with epic in this northern Egyptian performance tradition provide a highly significant frame not only for the epic as genre but for the performances that occur within the epic itself. These frames or cues are essentially formal devices and are only partially rooted in narrative developments. By contrast, the second set of examples we examined were ones in which the poets manipulate narrative elements so as to affect or comment upon situations outside the narrative story world. They constitute narrative strategies much as Kenneth Burke conceived or, strategies that purposefully appear to blur any possible divisions between the inner and outer world of the performance.^[12] And finally, we have seen examples of how chains of performances can be sparked, creating interactions in which the role of performer constantly shifts, and a dialogue of performances exists alongside the more usual dialogue of comments, criticisms, and compliments.

From these examples I conclude that the clarity and rigidity with which performance boundaries and genre boundaries are maintained, or the lack thereof, are qualities specific to individual traditions and must therefore be carefully observed, documented, and compared along with other more commonly recognized characteristics of performance, text, and context. It seems clear that some traditions may in fact draw a great deal of their strength and potency from forcefully maintaining strict boundaries of genre or performance. I think it also clear that this particular oral epic tradition, Sirat Bani Hilal as performed in northern Egypt, thrives on the dynamic interplay be-

tween the creation of performance boundaries and the willful obscuring and even violation of those boundaries. Here then is an example of an oral epic that draws a portion of its popularity and vitality from the constant negotiation of its margins by playing with the space in which performance and nonperformance meet. There is surely a spectrum of possibilities of which this is but one, thus this closing question: Where do the other epics of the world lie on this same spectrum?

Works Cited

Bauman, Richard.

1977.
Verbal Art as Performance. Prospect Heights, Ill.: Waveland Press.

Bird, Charles.

1972.
"Heroic Songs of the Mande Hunters." In African Folklore, edited by Richard Dorson, 275-293. Bloomington: Indiana University Press.

Briggs, Charles.

1988.
Competence in Performance: The Creativity of Tradition in Mexicano Verbal Art. Philadelphia: University of Pennsylvania Press.

Burke, Kenneth.

1931.
Counter-Statement. New York: Harcourt Brace.

Goffman, Erving.

1974.
Frame Analysis: An Essay on the Organization of Experience. Cambridge, Mass.: Harvard University Press.

Hymes, Dell.

1981.
"In Vain I Tried to Tell You": Essays in Native American Ethopoetics. Philadelphia: University of Pennsylvania Press.

1971.
"The Contribution of Folklore to Sociolinguistic Research."Journal of American Folklore 84: 42-50.

Lord, Albert B., trans.

1974.
Serbo-Croatian Heroic Songs Collected by Milman Parry. Vol 3, The Wedding of Smailagic Meho. Cambridge, Mass.: Harvard University Press.

Martin, Richard P.

1989.
The Language of Heroes: Speech and Performance in the "Iliad. "Ithaca: Cornell University Press.

Reichl, Karl.

1992.
Turkic Oral Epic Poetry: Traditions, Forms, Poetic Structure. New York: Garland Publishing.

Reynolds, Dwight Fletcher.

1995.
Heroic Poets, Poetic Heroes: The Ethnography of Performance in an Arabic Oral Tradition. Ithaca, N.Y.: Cornell University Press.

1991.
"Orality and Veracity: The Construction of Voice in Early Arabic Literature." Paper presented at the meeting of the Middle East Studies Association, Washington, D.C., 23 November.

1989.
"Sirat Bani Hilal: Introduction and Notes to an Arab Oral Epic Tradition." Oral Tradition: Special Issue on Arabic Oral Traditions 4 (1-2): 80-100.

Slyomovics, Susan.

1987.
The Merchant of Art: An Egyptian Hilali Oral Epic Poet in Performance. Berkeley: University of California Press.

9—
Worshiping Epic Villains:
A Kaurava Cult in the Central Himalayas^[1]

William S. Sax

It is perhaps no surprise that India's ancient epic Mahabharata continues to thrive in the Central Himalayas of North India as a form of ritual drama. More astonishing is the fact that in one remote valley the villains of Mahabharata are major cult figures, and subjects of religious veneration. In this essay, William Sax describes that cult and analyzes current processes of social change that are rapidly transforming it.

In the Central Himalayas of North India, India's great epic Mahabharata lives on. Indeed, it is the greatest single source of folklore in this predominantly Hindu region,^[2] constantly invoked to explain everything from the nature of south Indian cross-cousin marriage to the origin of warts. While the vast majority of local persons identify with the heroes of the epic, there is one valley, high and isolated, in which the villains rather than the protagonists of the story are worshiped as divine kings. The situation, however, is complex and ambiguous: as the valley is integrated ever more closely into the contemporary world, its epic traditions are transformed. In what follows, I hope to show how Mahabharata lives on, and also how it changes in response to wider social processes. But before doing so, I must say something about the epic itself.

Mahabharata is the longest epic poem in the world, more than eight times the length of the Iliad and the Odyssey combined, a veritable encyclopedia of Indian culture.^[3] In one of its recensions, it boasts: "Whatever is here may be found elsewhere; what is not here cannot be found anywhere else" ("yadihasti tadanyatra, yannehasti na tatkvacit" ) The central narrative of Mahabharata involves a dynastic rivalry between two sets of cousins, the five Pandava brothers versus the hundred Kaurava brothers. The Pandavas are the protagonists, sons of gods, born in order to rescue Earth from her oppression by demons who have taken birth in other royal lineages. Their social father, Pandu, was cursed to die if he made love to a woman, and so the five are actually half-brothers, begotten on Pandu's wife Kunti by various deities. There is also a sixth half-brother named Karna, born before Kunti was married. Soon after birth, she set him adrift in a river, and later he was found and raised by a lower-caste couple, so that neither he nor his five half-brothers

know they are related. The irony of the situation only increases when Karna becomes a firm ally of Duryodhana, eldest of the Kauravas. Karna is one of the two "villains" who is worshiped in western Garhwal; Duryodhana is the other.

Although the Kauravas' father Dhritarashtra was eldest, his blindness made him ineligible to rule, and so the throne went to his younger brother Pandu. The two sets of cousins grew up in an atmosphere of rivalry, which intensified after Pandu died and the administration of the kingdom was taken over by the Kauravas' blind father Dhritarashtra. For a time the kingdom was divided in two, but when the Pandavas performed an elaborate royal sacrifice to lay claim to universal lordship, Duryodhana felt threatened, and he responded by inviting the Pandavas to a dice match in which the eldest brother, Yudhisthira, lost everything: his wealth, his kingdom, himself and his brothers, and finally their common wife Draupadi. Duryodhana ordered his vicious brother Duhshasana to drag her into the assembly hall and strip her naked. But this proved to be impossible because in answer to her prayer, Krishna provided Draupadi with an endless sari. Enraged by the Kauravas' cruelty, Bhima, the second Pandava brother, vowed that he would take revenge by drinking Duhshasana's blood and breaking Duryodhana's thigh.

The blind king Dhritarashtra restored to the Pandavas their weapons and their freedom, but in a subsequent gambling match Yudhisthira again lost everything, and they were exiled for thirteen years. They spent their final year in exile without being recognized, and therefore, by the terms of the wager, Duryodhana should have restored half the kingdom to them. But this he refused to do, and his intransigence led to the great Mahabharata war. It lasted for eighteen days—a vast holocaust, a bloody sacrifice that consumed the earth's great kings and champions, including Karna (dishonorably slain by Arjuna at the urging of Krishna, while attempting to free his mired chariot) and Duryodhana (also treacherously slain at Krishna's urging, by Bhima who violated the rules of fair play by breaking Duryodhana's thigh in a club fight and thus fulfilling his vow). This great slaughter brought to a close the third age of the world and ushered in kaliyuga, the final and most decadent age, in which we now live.

Both the Kaurvas and the Pandavas belonged to the Bharata lineage, from which the word Mahabharata is conventionally said to be derived, namely, "the great maha [story of the] Bharatas." But according to both popular and learned etymologies, the proper derivation is "the Great (maha ) War (bharata )," and the eighteen-day battle is undeniably the central event in the epic. This should come as no surprise since Mahabharata, like India's other great epic, Ramayana, was originally composed and performed for the warrior class of Kshatriyas in ancient India, and thus reflects their central values and concerns.^[4] Mahabharata is in fact a kind of extended meditation on the moral dilemmas arising from the fact that war is both terrible and necessary. On

the one hand, it is wrong to slay anyone, especially one's friends and kin, as Arjuna explains to Krishna immediately before the great battle (their dialogue is famous as the Bhagavadgita ) It is also wrong to gain victory through deception and trickery, but this is precisely what the Pandavas did at Krishna's urging, a fact that has generated much debate and discussion amongst Hindus. On the other hand, Earth must be defended against evil, and this responsibility lies squarely with the warrior class, men like the Pandavas, whose dharma or code for conduct may include violence. Mahabharata can of course be interpreted in many ways: many scholars in India and elsewhere have tried to discern a historical kernel beneath the sprawling epic, while some currently influential interpretations see it as a vast, transformed sacrifice,^[5] or as the first and greatest example of devotional Hinduism.^[6] But in a general sense—especially in its contemporary representations in comic books, films, television, and performance genres Mahabharata remains a powerful Indian metaphor for the struggle between good and evil.

Mahabharata in Garhwal

Traveling in India, one is continually struck by the many ways in which both Ramayana and Mahabharata are popularly invoked. This boulder is said to have been placed by Bhima; that cave is said to have been a resting place for Rama and Lakshmana, heroes of the Ramayana, as they searched the forest for the kidnapped Sita; here is the site where Arjuna the Pandava performed asceticism in order to gain his magical weapons; there is the place where Rama left his footprint in the rock. Through such localizations, the epics are made immediate on both the local and the national scale: many of India's holiest places derive their importance from their association with the epics. Rama, Lakshmana, and Sita spent the first and most idyllic part of their forest exile in a place called Chitrakut, which is today identified as a pilgrimage place in the North Indian state of Madhya Pradesh. Before attacking his enemy Ravana's island fortress of Lanka, Rama prayed to the god Shiva, and the place where he did so is known today as Rameshvaram, one of the most important Hindu pilgrimage places in the world. Rama's supposed birthplace of Ayodhya is not only a famous pilgrimage place but also a politically contested site.^[7]

Places throughout peninsular India are mentioned in the Mahabharata, but the main action of the story takes place in the north, roughly between present-day Delhi and the Himalayan foothills. It is widely believed that Indraprastha, the capital of the Pandavas, was located on the present site of Delhi's Old Fort, while Kurukshetra, scene of the Mahabharata's climactic battle, is still a famous place of pilgrimage. But nowhere does Mahabharata have such tremendous social and religious importance as in the former Hi-

malayan kingdom of Garhwal, in the Himalayan districts of the state of Uttar Pradesh, where it is invoked to account for the origins, not only of social customs and features of the landscape, but also of the local people, who consider themselves descendants of the Pandavas. Sanskrit versions of the epic mention Badarinath, the foremost local pilgrimage place, and one that was intimately associated with the kings of Garhwal.^[8] Garhwalis believe that the Pandavas were born at the local temple town of Pandukeshvara below Badarinath, that the attempted assassination of the Pandavas by burning them in the palace made of lac took place at Lakha Mandal in the Yamuna River valley, that the kingdom of Virat where the Pandavas spent their final year of exile was in the Jaunsar region of Garhwal, and that the Pandavas made their final ascent to heaven through the Garhwal Himalayas.

The major vehicle by which Garhwalis express and explore their understandings of Mahabharata is pandav lila, a local tradition of ritual performance. The word pandav refers of course to the five Pandava brothers, but what about lila? The word means "play" in both the nominal and verbal senses of the English word.^[9] In the first place it means free, spontaneous play, like the play of a child, and this concept is, in effect, the Hindu answer to certain theological problems. If god is self-fulfilled and complete, then what could motivate him (or her or it) to create the world? In some strands of Hindu philosophy this question is answered in terms of lila, so that God's creative activity is likened to the playfulness of a child. The term is also used in this sense to address questions of theodicy, so that when confronted with death and suffering, Hindus are more apt to say "It is God's lila" than "It is God's will."

But lila- also means "play" in the sense of drama and theater. A lila is a play about religious themes, so that pandav lila is the "play of/about the Pandavas." At this point, the reader might suppose that pandav lila is a kind of folk drama, but it is more precisely described as a "ritual drama." It is a ritual because, like religious ceremonies but unlike secular theater, it is believed to be efficacious; that is, it is performed for specific ends, such as helping the crops grow, keeping disease and misfortune at bay, and promoting health and well-being by pleasing the gods, who are physically present upon the stage. During performances the five Pandava brothers, their common wife Draupadi, and other gods and goddesses enter the bodies of the men and women playing their parts. There is no "fourth wall" here: participants are constantly slipping back and forth between the roles of spectator, bard, character, and dancer, and as character or dancer they are normally considered to be the incarnations of those epic characters whom they represent. Such divine incarnation is frequent in Hinduism, where there is no ontological gulf separating humans and gods.^[10]

In performing pandav lila, Garhwali society collectively represents itself to itself by means of a public ritual performance. But this representation

clearly favors the Rajputs, who identify themselves as members of the class of Kshatriyas or warriors. Performances of pandav lila make implicit claims about who and what these Rajputs are.^[11] There are several reasons for saying this, the most important having to do with blood and ancestry. Most local Rajputs claim to be descended from the Pandavas, and thus pandav lila is understood as a form of ancestor worship. Such an understanding is also revealed by the generic term for pandav lila in western Garhwal: saraddh, from Sanskrit sraddha, the annual rite of ancestor worship that is obligatory for all Hindus. Rajputs are thus doubly linked to the Pandavas: not only do they perform a metaphorical sraddha for them by sponsoring a pandav lila performance, but in addition a central part of that lila involves a dramatic search by the Pandavas for the necessary materials to perform a sraddha for their dead father, king Pandu.

Pandav lila is also self-definitional for Garhwalis because Garhwal is prominently mentioned in Sanskrit versions of the epic, lending plausibility to the local belief that much of the story happened there. After all, the Indian literary tradition refers to the Mahabharata as "history" (itihasa ) as opposed to the more purely literary "poetry" (kavya ) exemplified by India's other great "epic," the Ramayana.

A third reason that pandav lila is self-referential has to do with the importance of the weapons that are carried by the dancers: a scythe for Draupadi, a staff for Yudhisthira, clubs for Bhima and his son Babarik, bows and arrows for Arjuna and his son Nagarjuna, a scythe for Nakula, and a slate for Sahadeva. The iron arrowheads of Arjuna and Nagarjuna are clearly the most powerful objects in a performance.^[12] They must be held in the right and not in the unclean left hand. They must never be allowed to touch the ground lest their "power" or "energy" (sakti ) escape. Large penalties (usually a goat or its cash equivalent) are levied on those who accidentally drop them, and women, low-caste people, and those in a state of pollution are not allowed to handle them. In Jakh village in 1991 my assistant Dabar Singh and I were invited to dance, then later told that I would have to do so with a headless shaft. We politely declined. In the next lila we attended, in Kaphalori village, I was allowed to dance with an "empowered" shaft complete with arrowhead. I found the experience exhilarating.^[13] At the beginning of a pandav lila, the old weapons that were used by the fathers of the present generation in the previous pandav lila, perhaps twenty or thirty years ago, are taken down from a secret and honored place (often under the eaves of a house) where they have been kept free from pollution and negative influences. They are placed on the altar in the dancing square, and kept there for the remainder of the performance. Meanwhile a new set of weapons is fashioned—sometimes on the spot by the local blacksmith—and used in the current performance. After the lila is over, these new weapons are placed under the eaves, and the old weapons of the previous generation are ritually disposed of at some holy

place, often a water source. In this way the Rajputs' martial energy—the ksatra of the Kshatriyas—is embodied in the weapons of the pandav lila and passed from generation to generation.

Finally, pandav lila may be considered a form of self-representation, because of the kinds of episodes that are selected for dramatic representation. Because of Mahabharata's great length, some principles of selection must be employed in choosing these, and martial episodes are clearly favored. This is in keeping with Garhwali Rajputs' self-image as warriors, an image that was strengthened when they were defined by the British as one of the "martial races" of India and were actively recruited for the colonial army. Even today, military service is a highly prestigious career, sought after by many local youths. The theme of martial prowess runs right through pandav lila, and the fight to the death between Arjuna and his son Nagarjuna is of particular importance. East of the Alakananda River, this event is represented far more often than any other. Not only is it sung of by bards and enacted by possessed players, but it is danced literally hundreds of times over the course of a nine-day lila. Every night, each surrounding village is given the opportunity for its two best dancers to display their skill, which they do by dancing the Arjuna-Nagarjuna dance. This lasts until the wee hours of the night, with anywhere from fifteen to thirty pairs of dancers performing the same episode. In terms of sheer quantity, this father-son battle is represented more than any other episode. But it is also enacted (in a less complex fashion) in the Mandakini River basin and in the far west in the Tons River basin, making it one of the few features that is common to pandav lilas throughout Garhwal. The details of the episode are much too complex to be recited here; ^[14] suffice it to say that one of its messages seems to be that a Kshatriya must always accept a challenge, even when it comes from a kinsman, a brother or even a father. To refuse a challenge is to lose one's most precious possession—one's honor—and so the fight between father and son once again points to war's dual nature, its necessity as well as its tragedy.

It should be clear by now that for the people of Garhwal, Mahabharata is much more than just a book. This seems an obvious enough statement, but I find that it needs to be made—repeatedly—to my academic friends and colleagues. As members of a bibliocentric profession, many us have bound copies of Mahabharata on our shelves, and it is only natural for us to think of it as a book. But it would be well to remember that even the critical edition of Mahabharata is only one version among many; that the story has been translated into many languages of many nations; that it traveled throughout Asia as far as Vietnam and Indonesia;^[15] that to begin with it was not a book at all, but rather an oral epic; and that the problems associated with establishing a critical edition proved intractable. Over thirty years ago the editor of the critical edition wrote that the "essential fact in Mahabharata textual criticism is that the Mahabharata is not and never was a fixed rigid text, but

is fluctuating epic tradition. . . . To put it in other words, the Mahabharata is the whole of the epic tradition: the entire Critical Apparatus."^[16] There are many Mahabharatas, not just one, and that is why I refer to "Mahabharata" and not to "the Mahabharata." I want to insist that books are just one part of this tradition; that they have no ontological or epistemological precedence; that Mahabharata is not only a book but also a political model, a bedtime story, a tradition of dance, a dramatic spectacle, and much more. In Garhwal, it is an ancestral tale, one that is periodically enacted as a form of ancestor worship, and one that provides a basic—probably the basic—paradigm for what it is to be a Rajput.

A Cult of the Kauravas in Western Garhwal

In the far west of Garhwal, near the Himachal Pradesh border, is the Tons River basin, an area that has long been associated with unorthodox social and religious customs such as bride-price, intercaste marriage, and especially fraternal polyandry of the sort exemplified by the marriage between Draupadi and the five Pandava brothers. When my village friends in district Chamoli, in the eastern part of Garhwal, heard that I was planning to travel to this area, they did their best to dissuade me. I was warned that men did not return from the valleys of the Rupin and Supin rivers (tributaries of the Tons): the women there enslaved them, turning them into goats or frogs by day, and back into men at night "for their pleasure." Rumors of poison cults abound: the women of the area are said to worship supernatural beings who demand one human sacrifice per year. A friend of mine, a traditional healer, went so far as to empower some salt with special magical spells, telling me that if any local woman were to offer me food, I should sprinkle this salt on it, and if it turned blood red, then I shouldn't eat it. A retired government officer said that several decades ago when he toured through the area, whenever he was offered food by the local people it would first be tasted by a prepubescent girl, in order to demonstrate that it was not poisoned.

I had long been aware of rumors of a Kaurava cult in this region,^[17] but had dismissed them as fantasies. I reasoned that because some of the customs of the area were very unusual, people in other parts of Garhwal assumed that the entire culture was inverted and strange, and this was why Garhwalis elsewhere were so willing to believe the most outlandish tales about the local people. The idea that there might actually be a group of people who worshipped the Kauravas struck me the way it strikes most most Hindus—as utterly implausible.

But I was wrong. Soon after arriving in the area for the first time in the winter of 1991, I was directed to the large and imposing temple of Raja Karan ("King Karna"), the presiding deity of Singtur, a traditional land division or patti whose social and religious life centers on the god. Karna's main tem-

ple is in Dyora village, but there are others scattered throughout Singtur, including a temple of his putative son Vikhasan.^[18] Raja Karan is regarded by the population of Singtur as their king: he is served by numerous lineages of priests, musicians, carpenters, and watchmen; he is often called upon to settle local disputes, which he does through his oracle (mali ) he is the subject of a rich devotional folklore; and like other kings he travels frequently, sometimes in royal processions to the villages under his rule, sometimes to drive away other gods encroaching on his domains, and sometimes as a pilgrim to local sacred places.

In many respects, Raja Karan and his priests are social and religious reformers. For example, they are fierce prohibitionists, opposed to drinking in general and especially on religious occasions. They are opposed to animal sacrifice within the temple compound, even though it is practiced in other nearby temples, and they seek to reform local marriage customs. Despite the stereotypes noted earlier, polyandry is not generally practiced in Singtur or the neighboring areas.^[19] However, the prevalent form of marriage, by bride-price, is still unorthodox, and Raja Karan's priests refuse to accept money in exchange for their daughters.^[20] The reason, they say, is that Raja Karan was famed for his generosity (one of their most popular epithets for him is dani raja or "the giving king"), and so they, like him, are keen to give their daughters as a pure gift. In doing so, they conform to the normative type of marriage among North Indian Hindus, the kanyadana or "gift of a virgin." In these and other respects they self-consciously distinguish themselves from Duryodhana and his subjects higher up the valley, where such reforms are only just beginning. How appropriate that Raja Karan's divine kingdom should lie midway between the more conventional cult of the Pandavas in the lower end of the valley and the cult of Duryodhana in the high mountains. The tragic figure of Karna, unknown brother of the Pandavas and devoted ally of the Kauravas, still mediates between the two sides, just as he did in the epic.

The people of Singtur assured me that there were indeed several temples of Duryodhana higher up the valley, the chief ones being located in the villages of Jakhol and Osla. Some of them also told me that there was a Duhshasana temple nearby.^[21] They said that the oracle of Duryodhana, whenever he was possessed, had to lean on a crutch for support, since Duryodhana's thighs were broken in his final combat with Bhima.^[22] I was told that Duryodhana used to go on royal tours throughout the region, and that wherever he halted, the people of that village were obliged to offer their finest animals, grain, milk, and butter to the god and his priests. So frightened were they of his curse that they would do so without demur. The people of Singtur asssured me that they had suffered a great deal at the hands of Duryodhana and his minions. For years their high-altitude herdsmen were forced to offer the finest of their flocks as annual tribute, until one year five brothers

defeated Duryodhana's followers, even though they were vastly outnumbered, thereby ending the custom.^[23] When I asked why it was that Duryodhana and Karna were now rivals when once they had been allies, people usually answered that the quarrel was not between the gods, but between their human devotees.

Why would anyone wish to worship the Kauravas? Duryodhana and his brothers are symbols of evil, of adharma, and Karna is a tragic figure strongly associated with death and defeat. Duryodhana makes only the briefest appearance in pandav lilas elsewhere in Garhwal, and Karna is considered as so inauspicious that people in that area do not even read the karnaparva or "Book of Karna" from the Mahabharata without first performing a goat sacrifice: they believe that if they do so, catastrophe will strike.

The short answer is that people do not worship Duryodhana because he is evil, but rather because he is powerful. Specifically, he is thought to have the power to bring rain or to withhold it, something that is obviously of crucial importance to local farmers. There are other reasons too for worshipping the Kauravas, and these are really no different from the reasons people elsewhere in Garhwal give for worshipping the Pandavas: in both cases they have to do with people's ideas about who they are, and how they became that way. The belligerence of the Kauravas, their eagerness to do battle, fit in well with local Rajputs' images of themselves as courageous and warlike. People sometimes say that Duryodhayana did not really die, but rather fled to the Tons Basin after he was injured in the Mahabharata war. It has been claimed that people from this area fought on the side of the Kauravas, and that the present inhabitants are their descendants^[24] —in other words, that Duryodhana is king of his subjects in the Rupin and Supin valleys, just as he was king of their ancestors. In worshipping him they are remaining loyal to their ancestral traditions; to do otherwise would be dishonorable.

Neither the kings of Garhwal nor the British ever had much authority in this area, and its people were notoriously "turbulent and refractory,"^[25] a reputation that persists to the present day. Political and religious authority is vested in deities like Duryodhana and King Karna, who are regarded as divine kings. Gods in this region have for a long time been among the most prominent political actors, and the boundaries of their domains are the subjects of continual and lively dispute.^[26] Devotion to the ancestral god is, therefore, a kind of protonationalism, in which loyalty to one's lineage, caste, and region are all mutually reinforced in the cult of the deity. To its neighbors, Duryodhana's domain might look like an "evil empire," but to those within it, loyalty to the cult is an appropriate and honorable attitude.

So perhaps my earlier skepticism was wrong, and this exotic Himalayan valley harbored an equally exotic religious cult, in which the most notorious villains of Hindu mythology were adored by the local population. This was, after all, explicitly stated by Garhwali scholars and journalists, and by peo-

ple living in the area. There have always been those who insist that only indigenous interpretations are valid, so perhaps this interpretation, clearly indigenous, was correct. The problem, of course, is that there is always more than one interpretation: "natives" have different points of view just as anthropologists do, and neither perspective is monolithic. This was implied by my friends in Singtur, who told me that in the upper valleys, the old ways were dying. Villages outside of Duryodhana's core territory were reluctant to invite him to visit, and so his annual tours were growing fewer, his domain shrinking, his influence diminishing. Inspired by Raja Karan's example, people were beginning to question the old ways, and now there were two factions in Duryodhana's main village, divided over animal sacrifice and the use of intoxicants. Young men had returned with university degrees from as far away as Delhi, and Duryodhana's followers were beginning to claim that he was actually Someshvara, a form of the great Hindu god Shiva. But I was told that despite these changes, there could be no doubt that that the god worshipped in the upper valleys was Duryodhana.

I pondered this information while trying to arrange a visit to the god's chief cult center, in the village of Jakhol. This was not easy to do, as it was winter and the mountain trails were buried in snow and often impassable. One evening I was explaining my interest in Duryodhana to a fellow patron of the tiny inn where I took my meals, when a small group of men sitting next to us fell silent. One of them angrily called out: "Who says our god is Duryodhana? His name is Someshvara!" I realized that my opportunity had arrived. The situation was very delicate, so I asked the man, Kula Singh from Jakhol village, to come outside with me where we would not be overheard. I apologized to him for causing offense, and said that I was only repeating what I had been told by the local people. Perhaps he would be willing to take me to Jakhol and show me the truth. He said that he would indeed be willing to take me there, but not now—it was midwinter, and the paths were too dangerous. But if I would come for the god's spring festival, I could stay with him, and he would help me with my research.

Later that spring I spent several days in Kula Singh's house, observing and participating in the god's festival, and three years later, in 1994, I returned again. During these two trips I discovered that, just as my friends had said, there were two factions in Jakhol. The "traditionalists" supported old customs such as demanding sheep and goats as tribute and sacrificing them in front of the deity, while the reformers regarded this as a form of theft inappropriate for a religious institution. There was also a dispute over the way in which the god makes his appearance during festivals. Traditionally, he is carried outside on a "chariot," actually a kind of palanquin, made of freshly cut pine saplings, and the young men of the village have great fun leaping on the saplings, jumping up and down on them and trying to break them. (The practice hearkens back to an incident in the god's biography when, according

to local folk songs, he had been forgotten by everyone except a group of children, who began a festival in which his image was worshipped in childish ways.) The reformers felt that this custom demeaned the deity by showing him disrespect; moreover, people were allowed to wear shoes when they approached the "chariot," and it was quite possible for low-caste persons to pollute the deity by coming in contact with him during the melee. Raja Karan's temple had abolished similar customs, animals were no longer sacrificed in front of the god, and when he left the inner sanctum he was placed on a tiger skin, symbol of royalty, rather than on a wooden palanquin. Reformers in Jakhol wished to follow suit, and to that end they convinced the rival faction to take the god on a pilgrimage to the famous nearby temple of Kedarnath. They felt that after completing such a virtuous act, he would certainly adopt a vegetarian diet and be more circumspect about ritual pollution. Sometimes the rivalry between these two factions from the same village got rather nasty. During my visit for the spring festival in 1991, there was an altercation between the reformers and the traditionalists, and that evening the reformers sought me out and earnestly requested that I bring my camera the next day. They were certain that their rivals were going to shoot them, and they wanted me to document the massacre for posterity.

But despite the depth of feeling on these issues, the fact remains that both sides are committed, at least publicly, to the view that the god is Someshvara and not Duryodhana. In my two visits to Jakhol I have recorded several of the god's songs, including his "history" as sung by temple musicians; spoken at length with two of his priests; recorded a number of brief, quasi-historical songs from lay villagers; and conducted a lengthy interview with Sundar Singh, who, as both village elder and chief administrative officer of the deity, probably knows more about Someshvar and his history than any other person. Everyone I talked to in Jakhol-priests, Rajputs, and musicians—was unanimous that the god was not Duryodhana. The most common argument in support of this position was based on iconography: the image was held to be that of Shiva, with the river Ganges emerging from his head, a half-moon in his hair, a garland of serpents on his throat, and earrings, though this was impossible to confirm, as outsiders (including local officials) are not allowed to view the image. Others claimed that the traditional form of worship was of the sort that is directed only to Shiva. Probably the most articulate spokesperson for this point of view was Sundar Singh, who told me in 1994 that the reason the old men used to call the god "Duryodhana" is that in their day no one had taught them any better. But, he said, when educated persons began coming from outside and looked closely at the image, they realized that it was not Duryodhana. I have also been told that it is not the main image that is called Duryodhana, but rather a smaller image that is visible on the upper part of the palanquin when the god emerges during his festivals. According to this story, the image was made by a famous blacksmith

named Bisar Katha, from across the border in Himachal Pradesh. He made this second image in order to add to the splendor of the main one, but when he made a third image, the local people saw that it was a replica of the first, and they went to kill him. With his dying breath he said: "I served you, and yet you're killing me, just like the Kauravas and the Pandavas, so I name this image Duryodhana."^[27] Although some of these denials are inconsistent or demonstrably inaccurate, nevertheless the god's subjects agreed that he was Someshvara and not Duryodhana.

Yet it still seemed to me (and to Raja Karan's subjects from Singtur) that the temple up the valley belonged to Duryodhana. Perhaps I accepted the Singtur people's version because they were my friends. Maybe I believed them because my anthropological fascination with cultural difference encouraged me to see a Duryodhana cult that wasn't really there. In the welter of inconsistent and mutually exclusive interpretations, one thing that emerged quite clearly was that each interpretation was related to some specific interest, not least my own. But self-interest often requires self-deception: just because there are multiple perspectives on some issue does not mean that they are all equally true. In the end, I persisted in believing that there was a Duryodhana cult because of the evidence. There is, for example, the matter of the god's songs. In 1991, one of the village musicians had referred to the deity both as Duryodhana and as Someshvara,^[28] but by 1994 all references to Duryodhana had disappeared from his song. Similarly, while attending a pandav lila in Singtur in 1991, I recorded some folk songs from two men from Jakhol in which they referred to the god both as Duryodhana and as Someshvara. In informal conversation in Jakhol, many people called the god "Duryodhana," which the reformers found rather embarrassing. One of the god's priests admitted to me that they used to worship the god as Duryodhana, and only recently began worshiping him as Shiva. Along the top of the exterior walls of the temple in Jakhol are a number of carvings, prominent among which is one of two men facing each other, each flanked by their supporters: underneath are carved the names "Duryodhana" and "Arjuna."

But the most persuasive evidence of all was provided by the god himself. Surely his testimony is authoritative, since villagers in the upper valleys respect and fear him as their ruler. In 1994, he returned from a brief tour to some nearby villages and was ritually welcomed home to Jakhol, at which time he possessed his oracle, as is customary. It was the beginning of winter, and there had been a prolonged dry spell, but as we stood in the flagstone square before the temple, dark storm clouds whirled around us, and we were briefly pelted with hail, while the slopes above the village were blanketed with fresh snow. The god's speech exemplifies many of the issues already discussed, such as the tension between the two village factions, the god as a divine king, and the difficulties that arise when ancestral traditions undergo change and transformation:

Oracle: Education (sikhi ) is happening! I will accept whatever you do and say. Keep your old traditions, keep them! But times are changing, and I'll change with them.

Sundar Singh: We are only children. Our ancestors' ancestor—Dhangkato and the others—came with you, but when they challenged you, you destroyed them. We will do as you say.

Oracle: It would be best for you to unite with the others.

Sundar Singh: I also want this. But things happen.

Oracle: You should unite, yes, but the two sides are always quarreling; you won't listen to each other. This new generation has also dishonored me. Everyone has his own opinion.

Sundar Singh: We are not acting for ourselves; we do whatever you tell us. Please protect my family.

Oracle: I'll protect all my subjects (janata ) not just your family. Everyone knows how my rituals are conducted, and they should act accordingly. They say that even the musicians have their freedom, but we'll see how free they are!

Sundar Singh: Musicians, messengers, watchmen, priests, subjects—if we are separate from you, then where is our place? We follow you.

Oracle: Stay within your limits, follow my orders, and all will be well. Maintain your dignity (maryada ) don't slip from your old traditions. Let the others slip instead.

Sundar Singh: You have the power and the virtue (gun ), not us. The rain has fallen.

Oracle: Yes. I'll take care of the rain.

Sundar Singh: Let the domain (mulk ) be united.

Oracle: Wait four days, then see. I'll provide rain; don't worry.

During this speech, the oracle supported himself by leaning on what looked rather like a long metal sword with an unusually short handle. Was it not the crutch about which I had already been told, required by Duryodhana because his thighs had been dishonorably broken by Bhima in combat? His speech was punctuated by frequent interjections—"Ak! Ak!"—was this not, as one man from Singtur suggested when he heard my tape recording, an expression of pain from his broken thighs? After the possession was over, the oracle leaped up to a standing position, obviously relieved that the trance was complete—was he also relieved of Duryodhana's pain?

Conclusions

Who is the god worshipped in the upper valleys of the Tons River basin? It seems likely that formerly he was widely recognized as Duryodhana, but nowadays his identity is less certain. Perhaps, by the time this essay is published, he will have completely metamorphosed into Someshvara. Duryodhana's followers are clearly aware that other Hindus regard them as perverse and back-

ward because they worship a notorious villain; therefore they are in the process of "reinventing" their deity. That this process should now be gathering momentum is hardly surprising, given the fact that these formerly isolated valleys are being integrated with the rest of India with unprecedented speed. A generation ago, there were no roads into the area, few visitors from outside, and virtually no public education. Today the transportation network is burgeoning,^[29] domestic and international tourism is rapidly expanding, and government or private schools are found everywhere. One local youth obtained a law degree in Delhi several years ago and returned to become a very influential proponent of the view that the god is Someshvara and not Duryodhana.

The problem with trying to determine, once and for all, such an abstract matter as the identity of a god is that it is a matter of beliefs and meanings, which are notoriously unstable. The god means different things to different people and even to the same people—in different contexts. Moreover, collective ideas about his identity change over time. There are relatively stable periods when such ideas are fairly consistent, as seems now to be the case in the cult of Raja Karan, and there are times of rapid change, as seems now to be the case in the cult of Someshvara. To assume that the god has a permanent, stable personality is as mistaken as assuming that a person has one. Cultural representations of gods, like representations of selves, are neither monolithic nor unchanging, but rather are related to the ever-changing contexts of culture and history.

Works Cited

Bharati, Agehananda.

1970.
"Pilgrimage Sites and Indian Civilization." In Chapters in Indian Civilization, edited by Joe Elder, 1 83-126. Dubuque: Kendall-Hunt.

Bhardwaj, Surinder Mohan.

1973.
Hindu Places of Pilgrimage in India: A Study in Cultural Geography. Berkeley: University of California Press.

Biardeau, Madeleine.

1978.
"Études de mythologie hindoue: 5. Bhakti et avatara. " Bulletin de l'École Française d'Extrême Orient 65: 111-263.

1976.
"Études de mythologie hindoue: 4. Bhakti et avatara. " Bulletin de l'École Française d'Extrême Orient 63: 87-237.

Blackburn, Stuart H., and Joyce B. Flueckiger.

1989.
"Introduction" to Oral Epics in India, edited by Stuart H. Blackburn, Peter J. Claus, Joyce B. Flueckiger, and Susan S. Wadley, 1-11. Berkeley: University of California Press.

Chandola, Anoop.

1977
Folk Drumming in the Himalayas: A Linguistic Approach to Music. New York: AMS Press.

Desai, Santosh N.

1970.
"Ramayana-An Instrument of Historical Contact and Cultural Transmission between India and Asia. "Journal of Asian Studies 30 (1): 5-20.

Eck, Diana.

1981.
"India's Tirthas: 'Crossings' in Sacred Geography." History of Religions 20 (4): 323-344.

Egnor, Margaret.

1980.
"On the Meaning of akti to Women in Tamil Nadu." In The Powers of Tamil Women, edited by Susan S. Wadley, 1-34. Syracuse: Maxwell School of Citizenship and Public Affairs.

Embree, Ainslie Thomas.

1990.
Utopias in Conflict: Religion and Nationalism in Modern India. Berkeley: University of California Press.

Fanger, Alan C.

1987.
"Brideprice, Dowry, and Diverted Bridewealth among the Rajputs of Kumaon." In The Himalayan Heritage, edited by Manis Kumar Raha, 139-153. Delhi: Gian Publishing House.

Fuller, Christopher J.

1992.
The Camphor Flame: Popular Hinduism and Society in India. Princeton: Princeton University Press.

Galey, Jean-Claude.

1986.
"Totalité et hierarchie dans les sanctuaires royaux du Tehri-Garhwal." Purusartha 10: 55-95.

Ganguli, K. M.

1981-1982.
The Mahabharata of Krishna-Dwaipayana Vyasa. 4th ed. 12 vols. Originally published by P. C. Roy (Calcutta, 1883-1896).

Haberman, David L.

1988.
Acting as a Way of Salvation: A Study of Raganuga Bhakti Sadhana. New York: Oxford University Press.

Hein, Norvin.

1987.
"Lila." In The Encyclopedia of Religion, 8: 550-554. New York: Macmillan.

1972.
The Miracle Plays of Mathura. New Haven: Yale University Press.

Hiltebeitel, Alf.

1991.
The Cult of Draupadi. Vol. 2, On Hindu Ritual and the Goddess. Chicago: University of Chicago Press.

1988.
The Cult of Draupadi. Vol. 1, Mythologies: From Gingee to Kuruksetra. Chicago: University of Chicago Press.

1976.
The Ritual of Battle: Krishna in the Mahabharata. Symbol, Myth and Ritual Series, edited by Victor Turner. Ithaca, N.Y: Cornell University Press.

Ibbetson, Denzil, and Edward MacLagan.

1919.
A Glossary of the Tribes and Castes of the Punjab and Northwest Frontier Province. Lahore: Government Printing, Punjab.

Majumdar, D. N.

1963.
Himalayan Polyandry: Structure Functioning and Culture Change: A Case Study of Jaunsar Bawar . New York: Asia Publishing House.

Morinis, E. Alan.

1984.
Pilgrimage in the Hindu Tradition: A Case Study of West Bengal. Delhi: Oxford University Press.

Narasimhan, Chakravarthi V.

1965.
The Mahabharata: An English Version Based on Selected Verses. New York: Columbia University Press.

Nautiyal, Shivanand.

1971.
Gadhval ke lokanrtya git (Folkdance songs of Garhwal). Prayag: Hindi [1902 saka]. Sahitya Sammelan.

1986.
The Ramayana of Valmiki: An Epic of Ancient India. Vol. 2, Ayodhyakanda. Introduction, translation, and annotation by Sheldon I. Pollock. Edited by Robert P. Goldman. Princeton: Princeton University Press.

Raha, Manis Kumar.

1979
"Stratification and Religion in a Himalayan Society." In Himalayan Anthropology, edited by James F. Fisher. The Hague and Paris: Mouton.

Rosser, Colin.

1955.
"A 'Hermit' Village in Kulu." In India's Villages, edited by M. N. Srinivas. Bombay: Asia Publishing House.

Saklani, Atul.

1987.
The History of a Himalayan Princely State: Change, Conflicts, and Awakening. Delhi: Durga Publications.

Sax, William S.

1997.
"Fathers, Sons, and Rhinoceroses: Masculinity and Violence in the Pandav Lila." Journal of the American Oriental Society 117 (1): 278-294.

1994a.
The Gods at Play: Lila in South Asia. New York: Oxford University Press.

1994b.
"Who's Who in the Pandav Lila?" In The Gods at Play: Lila in South Asia. New York: Oxford University Press.

1991a.
Mountain Goddess: Gender and Politics in a Himalayan Pilgrimage. New York: Oxford University Press.

1991b.
"Ritual and Performance in the Pandavalila of Uttarakhand." In Essays on the "Mahabharata," edited by Arvind Sharma, 274-295. Leiden: E. J. Brill.

Sharma, Man Mohan.

1977.
Through the Valley of Gods: Travels in the Central Himalayas. New Delhi: Vision.

Sukthankar, Vishnu S.

1933-1966.
The Mahabharata. 19 vols. Poona: Bhandarkar Oriental Research Institute.

Sutherland, Peter.

1988
"The Travelling Deities of the Western Himalaya." Unpublished manuscript.

Tambiah, Stanley J.

1973.
"Dowry and Bridewealth and the Property Rights of Women in South Asia." In Bridewealth and Dowry, edited by Stanley J. Tambiah and Jack Goody, 59-169. Cambridge Papers in Social Anthropology, no. 7. Cambridge: Cambridge University Press.

Tarabout, Gilles.

1986.
Sacrifier et donner à voir en pays Malabar. Paris: École Française d'Extrême Orient.

Thukral, Gurmeet, and Elizabeth Thukral.

1987.
Garhwal Himalaya. New Delhi: Frank Bros & Co.

van Buitenen, J. A. B.

1973-.
The Mahabharata. Chicago: University of Chicago Press.

van der Veer, Peter.

1987.
"'God Must be Liberated!' A Hindu Liberation Movement in Ayodhya." Modern Asian Studies 21.2: 283-301.

Waghorne, Joanne Punzo, and Norman Cutler, eds.

1985.
Gods of Flesh, Gods of Stone: The Embodiment of Divinity in India. Chambersburg, Pa.: Anima.