Preferred Citation: George, Kenneth M. Showing Signs of Violence: The Cultural Politics of a Twentieth-Century Headhunting Ritual. Berkeley:  University of California Press,  c1996 1996. http://ark.cdlib.org/ark:/13030/ft2m3nb13r/

Showing Signs of Violence The Cultural Politics of a Twentieth-Century Headhunting Ritual Kenneth M. George UNIVERSITY OF CALIFORNIA PRESS Berkeley · Los Angeles · Oxford © 1996 The Regents of the University of California

Preferred Citation: George, Kenneth M. Showing Signs of Violence: The Cultural Politics of a Twentieth-Century Headhunting Ritual. Berkeley:  University of California Press,  c1996 1996. http://ark.cdlib.org/ark:/13030/ft2m3nb13r/

for the people of the headwaters

Preface

This book has to do with the language and cultural politics of ritual violence in a minority religious community in highland Sulawesi, Indonesia. More particularly, it is about headhunting ceremonies and their stubborn presence in the contemporary social life of a marginal, upland enclave, an enclave that has suffered the wounds of social and cultural dislocation just by staying in place. No longer so remotely situated from the centers of state and post-colonial order, this minority community has had to look for social terrain in which its legitimacy and autonomy can be asserted. In their search, they use what is at hand as well as that which has been brought from over the horizon. The discourse and violence of headhunting ritual have been tangled up in the social life and struggles of this enclave for a long time. Becoming familiar with the discursive construction of violence in the theater of head-hunting ritual is, for me, a way to acknowledge and make plain what this community has at stake as it tries to shape its history and its fate.

So as to be clear from the outset, this work is not a comparative study of headhunting traditions. Although I will refer to practices in other societies from time to time, I have no interest in putting forward here a general theory of headhunting. Neither am I interested in developing general theories of violence, ritual language, or social marginality. This is not a refusal of theory. Rather, it is a rejection of transcendant constructs and vantage points that would harness ethnography to the manufacture of a certain and fixed human universe. Gerald L. Bruns (1988) has encouraged us to think of philosophy as revisionary work, a means for keeping our ideas and accounts of the world plural, loose, and open to correction. Ethnography, I would

argue, is a revisionary task as well. In taking ideas and theories and placing them in tension with a lived-in world, ethnography offers us a measure of our limits and our historicity. It is a means for acknowledging those who remain resistant to our sense of things, an entanglement with that recalcitrant other who demands nothing less—and perhaps nothing more—than recognition. It calls for readjustments to our moral and conceptual horizons.

Begun as a study of ritual speech and music, this work has led me to see violence as a crucial scene of textuality, a discursive bridge to other social, political, and experiential dimensions of human life—in this case, in an out-of-the-way community. To call attention to the textuality of ritual violence is not to undermine the seriousness of its consequences for that small community or for us. Rather, I see it as a way of throwing light on the complexity and power of violence and on its reach into other sites of symbolic exchange. Though there may always be slippage between violence and its representation in talk, song, and oratory, the language of violence always has pragmatic entailments. Dictating what the story of violence will be is fundamental to its exercise, and being in a position to do so puts someone or some group very close to the matrices of order and disorder.

Working with the discourse of headhunting ceremonies has obliged me to go poaching in other disciplines in an effort to relate music, literacy, Southeast Asian political history, gender hierarchy, topographies of grief and envy, and the dynamics of commemorative tradition to questions concerning ritual violence. I very much hope that readers in disciplines other than anthropology—in, say, ethnomusicology, Asian studies, religion and ritual studies, postcolonial history, folklore, psychology, and literary and cultural studies—will find something worthwhile in the following pages and forgive any shortcomings they may find. By the same token, I look forward to a multidisciplinary readership that can enlarge and revise the work undertaken here.

This book reflects the influence and assistance of many teachers and friends. My first words of thanks go to Jeff Titon, Daniel W. Patterson, James L. Peacock, Kenneth Irby, Jonathan Strong, and Dennis Tedlock for showing me some of the paths I might follow when going off to stalk poems, songs, and stories in other places. While at the University of Michigan, I had the good luck to study with Aram Yengoyan, Sherry Ortner, A. L. Becker, and Judith Becker; they have my warm and enduring thanks for their friendship, insight, and patient advice. Much in this book takes inspiration from the work of these ten scholars and writers, and I hope it measures up to their standards of care, creativity, and critical reflection.

Colleagues in the Departments of Anthropology at Tulane, the University of South Carolina, and Harvard have been a constant source of encouragement and intellectual challenge, and I hope they will take pleasure in reading this book. I am grateful in particular to Stanley Tambiah, Arthur Kleinman, and Byron Good for their critical response as I worked through ideas for Chapter 4.

Two other Harvard colleagues deserve special acknowledgment. First, I thank Michael Herzfeld for his endless puns and for his boundless critical energy and generosity. I joined his NEH Seminar on "Poetics and Social Life" at Indiana University during the summer of 1990, a time when I was brooding over how to turn a sprawling Ph.D. thesis into a book worth reading. Michael volunteered to read the thesis and gave me some enormously productive comments and suggestions. That kind, mad act was instrumental in giving me direction and resolve. He has my lasting gratitude.

Mary Steedly has been a steadfast colleague, friend, and confidante for many years. Circumstances have found us together in Chapel Hill, Ann Arbor, and Cambridge, and perhaps that is why it is difficult for me to imagine not being able to turn to her for personal and intellectual support. Her critical approach to ethnography and Karo social history has pushed me to reexamine my own work, and her being around has deepened my understanding of friendship and collegiality. Mary's interest and concern, not to mention her comments on several chapters, have made this a better book, and I thank her for that.

I am indebted to many others who have helped this book along by commenting on early versions of individual chapters or on the manuscript as a whole. I would like to mention in particular Donald Brenneis, Johannes Fabian, Douglas Hollan, Janet Hoskins, Webb Keane, Rita Kipp, Joel Kuipers, Toby Volkman, and Jane Wellenkamp. Susan Ferguson, Lindsay French, and Kate Hoshour also made comments that helped bring clarity or reason to the manuscript in some way. Last, but hardly least, Kirin Narayan and John and Karen Campbell-Nelson remained faithful listeners and eager readers, gently chiding me toward finer ethnographic work.

I want to express my very deep appreciation to Renato Rosaldo, Peter Metcalf, Bob McKinley, and Janet Hoskins (once more) for their encouraging letters or thoughtful remarks as I put this book together. I cannot imagine doing work on Southeast Asian headhunting traditions without building on their accomplishments. A number of friends and scholars also helped me fathom the music of headhunting ritual: particular thanks go to René Lysloff; his expertise and patient help were fundamental to my understanding of sumengo song structure and performance styles. I note, too, Greg Nagy,

Kay Kaufman Shelemay, and Deborah Wong; all offered key insights on choral singing.

Countless friends and acquaintances in Sulawesi made this a rewarding project. A. Makmur Makka helped in very crucial ways, as did Syamsul Arief, H. A. Oddang, S. Mengga, M. Saiyadi K., Mukhtar Saleh, and their respective families. The people of Mambi and Bambang showed me extraordinary kindness and hospitality for two and a half years. I cannot possibly name all of them here, but let me mention with heartfelt thanks the generosity and understanding of Abdul Rahman Enang, the late Johannes and Anton Puatipanna, and their families. I owe a special debt to the mappurondo communities in Bambang and Mambi, not only for their boundless hospitality, but for privileges seldom given to strangers. Their trust and patience, their kindness and knowledge, lie at the heart of this book. Finally, I can only hint at the immeasurable debt I owe Bombeng Boaz Rendeng (Papa Ati). He worked faithfully with me on this project, taking me to the farthest reaches of the Salu Mambi headwaters to teach me how to listen for the breathing of rivers and the felling of songs. His voice is in nearly all the lyrics that follow.

Vida Mazulis worked valiantly to keep the manuscript from smothering other kinds of intellectual interests in our home. Still, she took the time to go through portions of the book closely and critically, taking care to remind me of the need to historicize wherever possible, and making reasoned complaint whenever my prose got too clever or tangled up in itself. Her incisive ideas and patient companionship have helped me see my way through and beyond this work, and on to new projects.

Funding for the research that led to this book came from several sources. I gratefully acknowledge support from the Social Science Research Council; the Fulbright-Hays Doctoral Dissertation Abroad Program (Project No. G00-82-0543); the Wenner-Gren Foundation for Anthropological Research (Grant No. 4144); the University of Michigan Institute for the Humanities; and the National Endowment for the Humanities. My thanks go also to my Indonesian sponsors at Lembaga Ilmu Pengetahuan Indonesia (LIPI , the Indonesian Institute of Science) and Pusat Latihan Penelitian Ilmu-Ilmu Sosial (PLPIIS , the Center for Training in Social Science Research) at Universitas Hasanuddin (UNHAS ).

Some passages and sections of this book have been published elsewhere. I thank the editors and publishers of the following journals for permission to reprint materials from:

Lyric, History and Allegory, or the End of Headhunting Ritual in Highland Sulawesi. American Ethnologist 20(4):697-717.

Dark Trembling: Ethnographic Remarks on Secrecy and Concealment in Highland Sulawesi. Anthropological Quarterly 66(4):230-239.

Violence, Solace, and Ritual: A Case Study from Island Southeast Asia. Culture, Medicine, and Psychiatry 19(2):225-260.

Music-Making, Ritual, and Gender in a Southeast Asian Hill Society. Ethnomusicology 37(1):1-26.

Felling a Song with a New Ax: Writing and the Reshaping of Ritual Song Performance in Upland Sulawesi. Journal of American Folklore 103(407):3-23.

Headhunting, History, and Exchange in Upland Sulawesi. Journal of Asian Studies 50(3):536-564.

All of the photographs are mine.

In preparing this book I had a good deal of technical assistance. I would like to mention Vivian Montgomery, who did a marvelous job transcribing the sumengo melodies into Western notation, and John P. Mallia, who carefully put the notations into computerized form. I also thank Daniel Glazer for redrawing my research maps and developing a simple sketch of Southeast Asia. Zack Whyte helped with the index.

Last, I want to thank Stan Holwitz, Rebecca Frazier, and others at the Los Angeles offices of the University of California Press, and the copy editor, Bill Carver, for taking interest in this book, and giving it their outstanding care. They managed to get the book reviewed, copyedited, and produced during the earthquake of 1994, the floods of 1995, and the several other trials that have come their way. I take that as a good sign.

K. M. G.

rel-ic (rel' ik), n. I. a surviving memorial of something past. 2. an object having interest by reason of its age or its association with the past: a museum of historic relics. 3. a surviving trace of something: a custom that is a relic of paganism. 4. relics, a. remaining parts or fragments. b. the remains of a deceased person. 5. something kept in remembrance; souvenir; memento. 6. Eccles. (esp. in the Roman Catholic and Greek churches) the body, a part of the body, or some personal memorial of a saint, martyr, or other sacred person, preserved as worthy of veneration. 7. a once widespread linguistic form that survives in a limited area but is otherwise obsolete.
—The Random House Dictionary of the English Language, 2nd Edition Unabridged , 1987

1
Relics from Alien Parts An Introduction

It has been said that there are only two plots that really matter to stories and storytelling: "You go on a journey or a stranger comes to town."^[1] If departures, returns, and unexpected arrivals make great material, then it may be that those we call headhunters have a real tale to tell. Not only do they go off on a journey, but they bring back a mute, disfigured stranger with them. And they tell stories about it. I know this because I have listened to them. I know, too, that there is something disquieting about finding wonder and grace in their songs of terror and blood and noise and death.^[2] It is not a matter of surprise. After all, violence is no stranger to art: the bodies heaped or strewn across paintings, epics, and theater tell us that. I think the disquiet may have to do with finding oneself seduced by the spectacle and adventure of violence, by its clarity. It is the disquiet of thinking that you could be its victim or the one telling its tale. It is the disquiet that comes from assenting to a story of violence and acknowledging the passion and revulsion that quickened within you.

The force and seductiveness of headhunting stories remain real even when violence is left behind. In a sense, the headhunter's story always leaves violence behind—the bodies of the fallen are absent, elsewhere, just over the horizon of the senses. There is nothing "here" except the narrative and the dread trophy of the violent feats that happened "then and there." Consider, then, the oddness of a dumb, stolen head listening to the story that celebrates the death and dismemberment of its own person.^[3] For many

traditions of headhunting, though, it is a narrative of some kind that remains as the only presentable or recoverable trace of violence. Some head-hunters—like the Ilongot men who have figured prominently in the work of Michelle and Renato Rosaldo—simply abandon the severed head of their victim, and return home to take up boasting and song. "We came home and sang and sang" goes an Ilongot headhunting story—a remark that prompted Michelle Rosaldo to say that celebratory song itself was the source of the "anger" that led men to kill (1980:56-57). Yet there are those, too, who have put the violence of headhunting behind them in a different way. Here I have in mind those headhunters who make their predatory raids in the past. Although they have long put away the headhunter's weapons, they conjure relics of violence from the past in order to animate the present. Their violence happens only in the commemorative work of ritual narrative and song. These are the kind of headhunters who people this book.

Where headhunters have put down their weapons and taken up song, the mimetic discourse of ritual offers an especially revealing look at violence. We see violence as a narrated form of symbolic exchange, something "figured" and "read," so to speak, through the discursive tensions linking dominant and oppositional social formations (Armstrong and Tennenhouse 1989; Chambers 1991; Feldman 1991; Medick 1987; White 1978), and shaped through the social poetics that frame experience (Herzfeld 1985; cf. Brenneis 1987; R. Rosaldo 1986). Nevertheless, the intrusions of history, along with the ambiguities of commemoration, complicate our readings of the headhunter's violence. Slippage, disjuncture, and irony begin to characterize the way in which a circumstanced and recalcitrant social world stands apart from the idealized vision of the headhunt. The edifice of ritual and all of its pragmatic concerns begin to tremble.

The problematic gaps between the discursive projects of ritual and the instabilities of the social world beg a story, too. What stories need to be told about headhunters who do not take heads, but who nonetheless stage rituals about headhunting? How should they be told? Consider the following narrative:

This simple narrative sketch posits an originary violence that has been displaced over time. It gives the terrors of the past an illusory firmness and clar-

ity. But each of the four "editions" or "versions" invites a different reading of the headhunters' story. If I close the tale with "They take effigy heads," a sense of violence lingers to haunt the present; the headhunter seems to be exercising choice. Closing the story with "They take head-shaped surrogates," I imply something else—perhaps that the headhunters' violence and capacity to act have been thwarted or contained. If I say, "Now they take coconuts," I am almost sure to raise a laugh, for the only sign of violence is a harmless, pathetic theft—one of the weapons of the weak. And last, should I close with "They use coconuts bought in a market," I risk giving the impression that contemporary headhunting is a masquerade or a relic amusement for those tamed or humiliated by historical change and the expansion of a commoditized world.

Other stories can be read in this narrative as well. It might be read as a fall from authenticity into fakery, a story in which "really real" violence and "really real" heads devolve into gesture and stage prop. Read from another prospect altogether, the narrative may look like the story of all simulacra, a story in which there is no distinction between originals and copies, between reality and representation. Yet still another reading may be the most common and troubling one: a story of progress, a story in which pagan violence is subdued and in which human characters ascend from a state of primitivism to one just shy of civility and modernity. This modernist story of progress has been a powerful one in the popular and intellectual traditions of the West. Those traditions have been instrumental in making a fetish of headhunters, headhunting, and the severed head: turned into extravagantly magical figures, trophy heads and headhunters have served a rhetoric of control in the West's encounter with "others" (cf. Fabian 1983; R. Rosaldo 1978; Trouillot 1991; White 1978). In short, these representations of other people's violence have played into the discursive violence emanating from colonial and postcolonial centers.

Among other things, this book is about the language, music, and violence of headhunting ritual, and about the cultural politics that have shaped them. Yet I hope it may help dislodge the cathected figure of the headhunter from that rhetoric which peoples our world with "savages," "pagans," and "monsters." My focus will be on the song and chant of a ritual headhunt called pangngae , a ceremony observed by a minority religious community on the island of Sulawesi, in Indonesia (see Maps 1 and 2). What makes this ritual so striking, these days, is its artifice. No one is killed, no actual head is taken. But its rhetoric of violence is unmistakable. By exploring the discourse of pangngae—especially its themes, its projects, and its situatedness—I hope to recognize what is at stake for the community when it convenes in the theater of ritual headhunting.

Map 1. Island Southeast Asia

Map 2. South Sulawesi, Indonesia, showing study area

Objects and Encounters: October 1982 and July 1983

"It's a kind of harvest festival." The Kepala Desa —the head of the village district—was leading me over the paths to the hamlet of Lasodehata, where he planned to show me something about local ritual tradition. Though his remark was quite clear, I didn't know what to listen for. I was a stranger to Bambang (the name of the village district). I didn't know the people, I didn't know the place, and I certainly didn't know a euphemism when I heard one. Guided by wonder and literality, I simply went along with him.

Some months later, when I began to get a hold on the local language, I learned that Lasodehata means "the phallus of the debata (or spirit)." The story goes that when some settlers sank the first hole for a housepost on that site sixteen or seventeen generations ago, they hit a large red rock. The moment they did, the angry bellowing voice of a mountain spirit complained of the wound. The settlers made amends to the spirit and named the hamlet in commemoration of the event. Years later, civil and military authorities objected to the hamlet's name and changed it to Rantepalado, "the field of jackfruit"—although no jackfruit trees grow there. About the same time, I also learned that the "harvest festival" was called pangngae, and that pangngae means "to take a head" in raid or ambush. But on that day in October 1982, walking with the Kepala Desa, my working premise was that I was about to see something that had to do with harvest ceremonies.

Lasodehata was very still when we arrived. People must have been away in their gardens and coffee groves. We stood below the shuttered door of a large house, the comb from a nest of wasps or bees hanging beside it, set out to thwart malevolent spirit-beings from intruding into the house.^[4] No one was at home, so the Kepala Desa found a neighbor, stationed him at the foot of the house ladder, and then let himself in, beckoning me to follow. Fine shafts of sunlight reached down from the thatch roof into the dim, smokeless room. A large drum hung not far from the door, cinched up near a rafter. I took a picture of it, and a few weeks later carefully labeled the slide: DRUM used in FERTILITY RITUAL, BAMBANG . Moments after telling me about the drum, the Kepala Desa pointed up toward the loft, saying, "Those instruments (Indonesian, alat-alat ) are blown to make a sound after prayers are made to the dewi padi (Ind. rice goddesses).'

I looked up, craning my neck to make out something I might recognize resting in the loft. I saw several large bamboo tubes—they looked like they might be flutes—and a snarl of plaited leaves. A small offering rack hung

next to the instruments. I fussed with my camera trying to find the angle, the right light, a sharp focus, and I vaguely recall lying on the floor peering upward through the lens. I took another shot. That one is labeled: RITUAL PARAPHERNALIA for Fertility Ritual, Bambang. Kept on rafters .

That was my first look at the relics from a headhunt. They had been put away after the previous harvest (March 1982), and with the exception of the drum have since been replaced. I take out those two photographs from time to time, usually to wind up remembering how on that day I had hopefully imagined sounds coming from a silent drum and sleeping flutes. I also remember Ambe Lusa, the Kepala Desa and a Christian of abundant good will and humor, who would die before I left the region with much-changed ideas about the harvest festivals he had described to me. His showing me that loft persuaded me to live and work in the area for the next three years, and I hold on to those photographs so that I can think back on the moment. But to erase my ethnographic naïveté and mistakenness, I have recaptioned them "Violence: Still-Life I," "Violence: Still-Life II " (see Figure 1).

Circumstances kept me out of the mountains for the next five months, and I would learn little about pangngae until the following October. Shortly after moving to the region, I found out that those who had turned to the church and the mosque had more or less forsaken the local ritual tradition that interested me, a tradition called ada' mappurondo . The mappurondo communities—that is, the groups of people who adhere to the traditional ritual order—were quiet and remote. During this time, it was largely Christians and Muslims who interpreted local culture for me.

Two months or so after settling in the area, I joined Christian acquaintances in wedding festivities at Salutabang, a village in Bambang with a significant number of mappurondo households. One of the mappurondo elders there, Ambe Teppu, confronted the host of the wedding. Mappurondo terraces had already been "wounded" by the till, putting into effect a village-wide tabu on noise, laughter, music, weddings, and other festive rituals. His anger went even deeper upon learning that my Christian friends had put on a surprise "culture-show" (Ind., pameran budaya ) for me as light entertainment and education, a show that included some old headhunting songs, a chant invoking the spirits of women's household rites, and a staged version of a dance that in ritual circumstances would involve trance. Ambe Teppu saw it not only as poking fun at ada' mappurondo, but as sacred things out of place. Bitter about what had happened, he would later scold the newlywed husband:

I have coffee beans. I will give them to you and you can roast them up and make coffee. I have sheaves of rice. Go ahead and take them, and cook up some rice. But this is my religion. I will give you my religion. But don't, don't turn my religion into culture.

Fig. 1.
Violence: Still Life II. Bamboo flutes resting in the gable-end loft of a home.
These large, decorated flutes—called tambolâ—are unique to local headhunting cere-
monies. 1982.

For me the words stung. This was a version of "the raw and the cooked" quite unlike the one I had digested in preparation for research.^[5] Ambe Teppu was taking aim against the arrogation of local tradition by intruding ideologies and social groups. Powerful institutions were competing with the mappurondo communities for the ideological control and social production of words, meanings, and practice (cf. Vološinov 1973; Williams 1977). Resisting a civil discourse that denied him a religion and treated his sacred tradition as "culture" and "art"—matters I take up in this book—Ambe Teppu was also protesting conduct that abused and concealed the real nature of mappurondo practices.

Though they were aimed at his nephew, I took Ambe Teppu's words as an admonishment for me as well. After all, I was part of a "culture industry," albeit of a different sort. Although I had no intention of turning traditional sacred practices into ethnic song and dance (Ambe Teppu's most immediate fear), I wanted to be sure that I did not slight the mappurondo communities' authoritative claims to their own traditions. Ambe Teppu's words also reminded me that though he and I had been situated in unique ways by the varied histories and politics that worked through us, both of us were caught up inextricably in the same world. I met him at a time when the discipline of anthropology was subjecting itself to a critique of its representational authority and its complicity in the politics of domination.^[6] Ambe Teppu's reprimand was forged in a similar spirit: the politics of culture do matter.

I had not gone to Sulawesi with the idea of working on the problem of headhunting or violence, but with plans to explore the biographical and autobiographical strains of ritual discourse. And indeed, ethnographic reports on Sulawesi had given me the impression that headhunting ritual was a thing of the past. As in other parts of Indonesia, headhunting rumors were common forms of terror, linked now, as in the colonial past, to the presence of police, military personnel, workers from large construction projects, and other alien figures (Drake 1989; Erb 1991; Pannell 1992; Tsing 1993, forth.). When I learned, some months after meeting Ambe Teppu, that mappurondo "harvest festivals" are headhunting rites, I was drawn to them, for several reasons. For one thing, it is very clear to me that pangngae is the central ritual apparatus within the mappurondo communities for asserting their cultural autonomy: the community that no longer holds pangngae is in effect moribund. Just as ritual theater was the basic cultural idiom for the nineteenth-century Balinese state (Geertz 1980), the ceremonial headhunt of pangngae forms a political drama through which the mappurondo communities try to control their past and their present. By the

same token, the rituals are a means for these relatively egalitarian communities to display and read signs of spiritual and material potency (cf. Atkinson 1989; Errington 1989; M. Rosaldo 1980). But I do not of course mean to exorcise conflict from pangngae, for as Dirks (1994) reminds us, ritual can as easily exhibit or bring about conflict and instability, as put authority and order on display. For the moment I want to stress that the mappurondo communities' hold on autonomy, potency, continuity, and stability has been thrown into question by a civil order that contests or censures the very means for grasping them—the rhetoric of ritual violence. At stake is how pangngae will be claimed, reclaimed, or forgotten, and by whom. Conflict over ritual practice and representation is a fact of life for the mappurondo enclave in Bambang, and Ambe Teppu's protest was just one more skirmish in the ongoing struggle of a minority religious community to shape its own fate.

It would be misleading to describe this struggle as a struggle for identity. The conflict has to do with ritual practice and legitimacy, with authoritative claims to tradition, and with ways of coping with an uncertain world. For people who have remained faithful to ada' mappurondo, rituals are the events of time, memory, and tradition itself. Authoring and authorizing ritual "texts" are at issue in this rivalry with Christians, Muslims, and the civil administration. This brings me to another reason I was drawn to the ritual. The texts and textuality of pangngae not only were contested resources, but also promised a valuable point of entry into understanding the discursive construction of violence and ritual tradition. A study that dwells upon the discourse of pangngae, it seems to me, might take us far in coming to terms with ritual violence. At the same time, I think it will give us a glimpse of what the mappurondo communities have at stake as they negotiate their place in the world through a tradition of headhunting ritual.

Recent Commentaries on Regional Headhunting Traditions

Some of the more insightful commentaries on Southeast Asian headhunting traditions have shown a deep regard for local discourse and history. In particular, I have in mind here the work of the Rosaldos on Ilongot headhunting in the Northern Luzon region of the Philippines and, indeed, this book is in part an effort to resume the conversations they began. Let me sketch what I take to be their accomplishments, noting, too,

the works of others that have shaped my interest in the discourse of head-hunting ritual.

Putting headhunting into historical perspective has been Renato Rosaldo's key contribution (1980). In very general terms, he was able to show how headhunting works as a central moving force in the improvisation of Ilongot social life and in the shaping of local memory and historical thought. For example, headhunting, as part of a broader pattern of feuding, often motivated marriages and residential moves in the Ilongot communities. At the same time, it served as a focal episode in personal and collaborative recollections of the past. What is clear from Rosaldo's treatment of violence is that social improvisation and memory are mutually shaping: headhunting (as discourse and practice) was at once a way to apprehend the past and a way to respond to circumstances and contingencies of an ever-changing social world.^[7] The historical turn taken by Rosaldo also helped debunk the image of the timeless primitive. Not only did Ilongot headhunting cease many times in the last century or so, but it was fundamental to local history as process and thought—that is, it was fundamental to measuring change and the passage of time.

More recently, Janet Hoskins has pushed the analysis of history and head-hunting in another, fresh and illuminating, direction (1987). Hoskins argues that sejarah , or "history," has become a new genre of authoritative discourse at local and national levels in Indonesia. National history turns Wona Kaka, a Sumbanese headhunter of the early twentieth century, into a heroic figure in the national resistance to the Dutch colonial order. Yet in the historical view of some Sumbanese, he symbolizes local resistance to encroachment and absorption by any outsiders, whether Dutch or Indonesian. As a result, the two histories compete with one another, each trying to claim this headhunter as its own heroic figure by "reinventing" his past and his heroism. The history of local headhunting is thus a contested cultural resource, rather than a formative episode in the reproduction of the social order. It is the proving ground for heroic figures crucial to the ideological control of the past. In a subsequent study, Hoskins (1989) revisits the problem of Sumbanese headhunting, this time with an interest in the historical transformation of "things." By examining the life history of a severed head, and tracing changes in its identity and value as it moves through eras of exchange, alliance, and trade between rival groups, Hoskins shows how the relics of violence can be historied and reappropriated in the changing currents of social life.

Michelle Rosaldo followed a different path in her work on Ilongot head-hunting (M. Rosaldo 1977, 1980, 1983). Unhappy with the shortcomings

of structuralist approaches, she proposed an alternative framework based on explorations into the "tones of thought" through which particular cultures associate severed heads with passion, mourning, fertility, and envy. Ilongot headhunting, in her analysis, was caught up in local discourses on emotions, in age- and gender-based social hierarchies, and in notions of personhood. Her understanding of Ilongot emotions found tragic reprise in Renato Rosaldo's "Grief and a Headhunter's Rage" (1984), written after her unforeseeable death in 1981. An important statement on emotion, ritual analysis, and the positioned subject, this essay showed once more that an exploration of emotion and meaning could render headhunting as a plausible and compelling cultural practice.

I think I can distinguish my own interests in headhunting by quoting a brief passage from the final remarks in Michelle Rosaldo's monograph (1980:231): she wrote that she was "involved in showing how particular modes of speaking are illuminated by the social actions and relationships such speech describes." In this book, I want to show how ritual modes of speaking are themselves modes of social action and relationship and, as such, are practices that shape the social and historical world in which headhunters find themselves.^[8] This approach has the virtue of making central to analysis a body of discourse that Clifford Geertz (1973:448) might call a "story people tell themselves about themselves" (see also Ortner 1978; R. Rosaldo 1986). Ritual songs, narratives, and liturgies do not merely reflect or represent social life, but are of practical consequence. The politics of representation consist not only of struggles over signifiers and signifieds, but also of efforts to control the pragmatic acts and social relationships emanating from representational practice.

Although Michelle Rosaldo was able to show the rich conceptual realm in which Ilongot headhunting made sense, she did lead our attention away from the kinds of discourse that stirred the hearts of Ilongot men and gave their violence purpose: the ritual songs called buayat . Interestingly, the buayat is prominent in an opening vignette in her book (M. Rosaldo 1980). She reports about how, on her return to Ilongot country in 1974, she pulled out an old tape of a headhunting ceremony she and her husband had attended during their first stay—a sacrifice and songfest that celebrated the murder of a lowlander by an Ilongot youth named Burur.^[9] Hearing the buayat, her friends demanded that she stop the tape: it was too wrenching and painful for their hearts. "[T]he song itself . . . made their breath twist and turn inside them; it pained them because it made them want to kill" (M. Rosaldo 1980:34). It seems to me that this ritual song was a criti-

cal form of discourse that stirred men to hunt heads, one that rendered headhunting morally intelligible and valuable to the Ilongot themselves (cf. R. Rosaldo 1984:190). As Rosaldo herself notes, the song celebrated and renewed the angry passions of the headhunter (1980:56); and she reports that Burur had been sullen and withdrawn over the failure of his kin to celebrate his accomplishment promptly (1980:68). To be fair, the buayat probably thwart in-depth analysis: Michelle Rosaldo explains that the songs have "no linguistic 'sense' of which Ilongots are aware," adding indirectly that they are distinct from the boasts shouted out in them (1980:54-55). It was also the case that by 1974 the buayat had gone the way of Ilongot head-hunting; they could not be retrieved for study. The same appears true of the boasting narrative song called tarapandet , which recounts a headhunter's journey (M. Rosaldo 1980:156-157).

For all of its ethnographic wealth and insight, the corpus of work by the Rosaldos does play down the significance of ritual in the analysis of head-hunting. This has prompted complaint from Peter Metcalf (forth.), who has explored Berawan headhunting practices in Sarawak, Malaysia (Metcalf 1982). He remarks that the Ilongot case may be atypical when measured against the elaborate headhunting ceremonies commonly found elsewhere in the island region (cf. McKinley 1976; Hoskins forth. [a] and [b]). I share with Metcalf a fascination for the place of ritual in social life, but I am not inclined to worry over whether Ilongot headhunting is "typical." The Rosaldos may have had purpose in writing against the privileged position of ritual in symbolic anthropology (e.g. R. Rosaldo 1984)—indeed, I have a hunch that they felt it necessary to work around ritual in order to get at the complexities of experience and history—but their ethnographic work bears sufficient trace of ceremonial discourse to convince me that ritual speech and song may afford important insights into headhunting throughout island Southeast Asia.

Violence and Ritual Discourse: Themes, Projects, and Situations

There is no one way to get at ritual discourse. For example, explorations into the poetics and politics of ritual speech in island Southeast Asia have included detailed accounts of: canonical parallelism (J. Fox 1988); entextualization and authority (Kuipers 1990); the problems of voice and

agency (Keane 1991); stylistic and thematic analyses of prayer (Metcalf 1989); the politics and history of genres (Bowen 1991); assertions of ethnicity and identity in a contested social terrain (Hefner 1985); oratory and political poetry (Atkinson 1984, M. Rosaldo 1984b); the efficacy of chant and its relation to local notions of power (Atkinson 1987, 1989); the eccentricities of trance accounts (Steedly 1993, Tsing 1993); the production of cultural coherence (Becker and Yengoyan, eds., 1979; Traube 1986); and the production of gender difference (Siegel 1978, Kuipers 1986, Rodgers 1990). Yet there is a common thread running through much of this literature: the authors recognize, albeit in different ways, that the history of society and the history of language are bound up with one another, and consequently that the study of ritual speech requires us to see how poetics and performance are situated in social life (cf. Fabian 1974; Bakhtin 1986; Bauman and Briggs 1990, 1992).

Much of the ethnographic literature prior to the historical turn in anthropology represented ritual not only as an instrument of tradition and traditional authority, but as a rigid, unchanging structure as well (cf. Kelly and Kaplan 1990). The same goes for ritual discourse, which commonly is portrayed as repetitive, formalized, conventionalized, and fixed (for example, see Bloch 1974, Rappaport 1979, and Tambiah 1985). In my experience, religious language is more supple and improvised than these descriptions would allow. Yet I would agree that ritual and ritual language show a certain stability through time. What troubles me is that we too easily forget that the fixity and formality we observe in ritual (and in tradition, for that matter) are not the properties of ritual language per se, but, rather, the outcome of a community trying to stabilize a body of discourse for continuous interpretive work. Whether that "stabilized" discourse serves the pragmatic interests of traditional authority or factors into the reformation of the social order depends largely on historical conditions, as Stanley Tambiah (1985) has pointed out. At issue for me is how communities produce, reproduce, and tactically alter ritual discourse in ever-changing social and historical contexts. In my view, there is no ur-text, no abstract cultural schema, no basically basic story or structure behind ritual.^[10] There is instead an ongoing history of prior ritual events and texts being recalled and put in productive tension with the present.

As Jerome McGann has reminded us in his recent exploration of materialist hermeneutics (1991), texts are made and remade under specific social, historical, and institutional conditions, conditions that mark the horizon within which texts are interpreted and transformed (cf. Williams 1977). Dis-

cursive practices are always in service of one or more situated projects, and it is such projects that come to dominate (or set the grounds of struggle for) the pragmatic and hermeneutic dimensions of the textual acts in question (cf. Fish 1980, 1989). Different "versions" of a text, I would argue, trace its situatedness and throw light on the conflicts that determine its multiaccentuality. Indeed, remaking texts—in McGann's terms, "producing editions" (1991:33)—is a way to shape meaning (cf. Tedlock 1983, Fabian 1990). How this process took place in one specific Indonesian "situation" comes through clearly in the work of James L. Peacock (1968), who was able to show that differences and deformations of plot in performances by Surabaya's ludruk theater-troupes in the 1960s made sense with respect to the goals and ideological programs of opposed political camps. For traditional ceremonial discourse, I believe it is possible to demonstrate as well that rituals and their multiple stagings are made "for particular purposes by particular people and institutions, and [that] they may be used (and reused) in multiple ways, many of which run counter to uses otherwise or elsewhere imagined" (McGann 1991:47). Staging a ritual places it in historical relation to prior stagings and yet aligns it with immediate social goals, tensions, and anxieties.

For the mappurondo communities in Bambang, staging pangngae is part of a broader struggle to assert authority and control over their ritual tradition and their past. Yet there is no fixed way to do this. The communities instead bring out different "editions" of the headhunt in an effort to repeat the past and to respond to the contingencies of the present. Each performance narrativizes violence in keeping with living memory, and with the interests and problems of the moment. Because I have no record of ritual materials from the colonial or precolonial past except those recalled in the discourse of the present, my understanding of how the mappurondo communities fashion their headhunting ceremonies is limited largely to the horizons of a two-year period in the mid-1980s. I look, of course, to various liturgical genres in an effort to see what it is the communities are trying to accomplish. But central to my analysis is a genre of choral song called sumengo , which may be translated as "the singing." The songs are short: the typical sumengo consists of but three octosyllabic phrases sung by a song-leader and chorus. Yet the songs are so prominent in pangngae that it would not be off the mark to describe the ritual as a kind of sumengo-fest. In fact, the sumengo arc the principle textual and performative vehicle for narrativizing the violence of pangngae, and in that sense they are analogues to the Ilongot buayat and tarapandet discussed by Michelle Rosaldo. For this

reason, my exploration of different stagings or editions of pangngae rests largely on differences I noted in sumengo performances. That is, differences in sumengo performances suggested to me different ways to narrativize or aestheticize violence, and different ways to reproduce and make claims to ritual tradition in the theater of social life.

Much in this book is in keeping with text- and performance-centered ethnography (e.g. Becker 1979; Bauman 1977; Fabian 1990; Hymes 1981; Tedlock 1983), and with recent trends in practice theory (see: Ortner 1984, 1989; Dirks, Eley, and Ortner 1994). So as to elucidate a local narrative of violence and relate it to the flux of social history in the mappurondo communities, I will organize my exploration of pangngae largely in terms of its themes, its projects, and its situatedness. By "themes" I mean broadly and simply the subject matter of specific lyrics, liturgical texts, or canonical acts as they are generated in specific contexts of production and interpretation.^[11] To give but a few examples, adornment, noise, and pity are themes that one commonly encounters in the representational and performative dimensions of pangngae. The challenge will be to relate them to the theme of ritual violence. By "projects" I mean the recurrent goals of the ritual. Although the search for prosperity might be considered the overriding point of pangngae, it is clear that the domination of an ethnic rival, the lifting of mourning prohibitions, the celebration of manhood, and the act of commemoration itself also factor into the ritual as purposive tasks. The themes and projects of ritual discourse, I should stress, are intertwined and lend each other intelligibility. It is through them that the mappurondo communities are able to conjure the headhunter's violence and give it significance. A look at the situatedness of pangngae can take us further still. In what kinds of communities do these rituals take place? What social conflicts and contingencies surround the headhunt? What events and forces lie behind and ahead of the ritual tradition we encounter today?

Let me outline, then, the structure of this book. Immediately after this introductory chapter, there follows a general social and historical sketch of the mappurondo enclave in Bambang. It should provide readers with a rough sense of the political culture and cultural politics surrounding pangngae. The next four chapters (Chapters 3 through 6) dwell largely on the themes and projects of mappurondo headhunting ritual, relating them very broadly to the social and cultural reproduction of the mappurondo communities. Chapter 3 explores the headhunting of the past and the way it is recalled in ritual discourse and oral history today. Here I try to make sense of headhunting and the polemics of ethnicity and highland-lowland exchange in the precolonial and colonial periods. I also use this discussion of history

to critique the excesses of structuralist and other symbolic approaches to ritual headhunting, and to recuperate the political import of violence and the trophy head.

Chapters 4 and 5 trace the processual features of a contemporary head-hunt from start to finish. Grief, violence, and solace are the key themes in Chapter 4; I relate them to the projects of mourning and making vows. As many readers will recognize, this chapter takes inspiration from and responds to Renato Rosaldo's "Grief and a Headhunter's Rage" (1984). Mappurondo practices suggest that the resolution of communal mourning is more significant than personal catharsis in motivating a headhunter's violence; that ritual refigures individual affect as "political affect" (consonant with the political import of violence); and that varied discursive forms such as vows, songs, and noise mediate the ways in which people (and the community as a whole) put grief behind them and resume their lives. I argue, then, that personal motives for taking part in contemporary headhunting spring not from powerful emotions but from obligations and acts made in discourse.

Chapter 4 follows pangngae from the time the headhunters leave their village to the time they return home. Chapter 5, by way of contrast, deals with the week of ceremonial activity that takes place after their triumphant return. Here I relate the politics of adornment and envy to the discourse of manhood, and to the local social hierarchies relating men to men and men to women. In this chapter we see males figured anew: they not only are valorous and violent as headhunters, but also show their virtuosity and authority as rhetoricians, as speechmakers. At the same time, the rivalry that linked the headhunter with his victim is now supplanted by a rivalry between husbands and wives.

Chapter 6 addresses the final project of pangngae: commemoration. In this chapter I try to locate the violence of pangngae in commemorative discourse itself, arguing that violence is both subject and generative force for its own reenactment in chant and lyric. I point out as well that commemoration is a mode of sociality. I look in particular at the social organization of choruses and relate such organization to everyday social hierarchy. But I also raise the issue of sociality to suggest that the politics of representation not only involve struggles for authority over signs, but also imply a struggle to inhabit distinct forms of social relationship.

Chapter 7 takes a look at differences in the commemorative work of four villages as they hold pangngae. Taking inspiration from McGann, I try to see how the discourse of pangngae has been situated and circumstanced—to scan, in other words, the performative and interpretive horizons for this ritual. Against a background of a tradition in crisis, I explore how agonistic

play, irony, trance critiques, and written texts have become commonplace in the headhunt. The violence and memory of headhunting, and the interpretive work linked to them, are now inseparable from writing, reading, and the excited play of voice.

Chapter 8 extends further my concern for the situatedness of pangngae. I finally come around to the Christian and Muslim efforts to claim or forget local headhunting practices, efforts that are worked out largely in terms set down by the Indonesian state. My "texts" in this chapter are not sumengo, but a tourist pamphlet, analyses of local tradition by two theologians, and the quips and remarks of friends. Against these portraits of headhunting, I return to the figure of Ambe Teppu, the man who reminded me that the politics of culture do matter, most especially for minority communities too small to count in state and national orders. I close with a brief look at the mappurondo communities in 1994, and with some reflections on this work as a whole.

A Note on Research in Sulawesi

Most materials for this study come from roughly 30 months of ethnographic research in the mappurondo communities of Bambang between October 1982 and August 1985.^[12] have made no effort to disguise the places where I worked. Friends and hosts in the mappurondo communities, however, appear in these pages with their names changed. During the time I lived and worked in Bambang, I was able to witness and take part in pangngae eight times: twice in the village of Salutabang, three times in Minanga, twice in Rantepalado, and once in Saludengen. I recorded ritual songs and liturgical performances in each of those places, and all the lyrics that appear in this book are quoted from specific ritual performances. Day-to-day observations, casual conversation, household surveys, and village histories were the source of other materials that came to shape my understanding of ritual song and violence.

Much in this work hinges on the translation and interpretation of ritual speech and song, especially the sumengo. In an important sense, the project of translation and interpretation is an unfinished one. I keep going back to these songs, sometimes finding something I overlooked, other times finding new ways to connect the materials meaningfully to other features of mappurondo life and language. Readers will further extend and revise the project of interpretation. That said, it may be helpful if I provide a quick

sketch of my encounter with these songs, and of my tactics and methods of translation.

As may already be clear, the genres of ritual song peculiar to the head-hunt are subject to a rigid set of local tabus. This holds for ethnographic dialogue as well. Villagers may not sing, rehearse, or even discuss these materials unless it is "headhunting season," so to speak—a period lasting between three days and two weeks. These tabus necessarily shaped my work with singers, lyrics, and performance. In particular, they framed the contexts in which performers and I could comfortably engage in a give-and-take with one another. By placing my inquiry within the sphere of traditional practices, I put a partial check to my impulse to control ethnographic dialogue. At the same time, the sanctioned period for ritual performance gave me a chance to raise questions about the sumengo and other genres and to ask villagers to reflect on the meaning and significance of the material.

Listening in and listening to mappurondo headhunting ritual is complicated work. Gatherings are exuberant; conversations, songs, and drumming all take place at the same time. Liturgical acts will silence the talk and instrumental music, but these acts are often whispered or muttered, or demand choral responses that are difficult to make out. My understanding of ritual emerges from being there, and listening to tapes and reading transcriptions prepared by my Christian language assistant from Salutabang, Bombeng Rendeng (Papa Ati).

During the postharvest season of 1984, I moved from one village to the next as the headhunting rituals closed in one and began in another. Papa Ati joined me in Salutabang and shortly afterward we repaired to the mountain market town of Mambi—downriver and out of the orbit of mappurondo social space—where we would be free to work on transcribing and translating materials into Indonesian and English. Once basic transcriptions were complete, we would together explore the songs' lyrics. We then took our understanding of the songs back to the different mappurondo communities for comment in 1985, when ritual conditions next allowed. Most of those to whom we turned were men, but at Salutabang and Minanga we were fortunate to have the help of women who were skilled sumengo soloists. Interviews with soloists, ritual specialists, and other villagers supplied me with information on singing, the origin and history of headhunting, and on people's impressions of sumengo performances and their encompassing tradition. We also gathered new materials in 1985, and rushed to bring them into our conversations about pangngae.

Translating the songs was hardly straightforward. Translations emerged from talk between a native speaker/singer (raised mappurondo but now

professing to be Christian) and a stranger trying to become familiar with local tongues and with the life-worlds and political anxieties of the mappurondo communities. Talk was a mix of Indonesian and languages peculiar to Bambang and Mambi. Whole stories inform word glosses and the details of interlinear worksheets. The interpretive commentaries of villagers, too, were hardly straightforward, leading off, as they often did, into local history, personal memories, and other songs. But by and large the meaning of the songs and liturgies did not occasion much debate or disagreement. Interpreting the songs was not a pressing cultural problem; remembering them was. Sometimes people found songs enigmatic and beyond understanding. Other times singers and listeners acknowledged a plurality of meaning in the lyrics—and I have tried to capture some of that surplus of meaning in the chapters that follow.

The final months of fieldwork brought a fresh problem. Late in April 1985, I was able to attend pangngae at Saludengen, a village in which I yet had no friends or acquaintances. Aside from 45 new song texts, the ceremony yielded a performance style unlike those I had previously encountered. Further, it was in Saludengen that I discovered villagers who had begun to write down sumengo lyrics (though not the tune). Confronted with all of that, I had to rapidly reformulate my ideas regarding sumengo tradition and, indeed, headhunting tradition. I had opportunity to pursue some of my most important questions. Unfortunately, my visit was brief and many more questions had to go unasked.

Once again, I should stress that the work of translation and interpretation has not come to a halt. For example, working with ethnomusicologists after I returned home from Sulawesi gave me an opportunity to grapple with textual and melodic structures in ways I never imagined during my initial study (George 1989). Finally, in writing this book I have tried to place songs so as to help identify and make intelligible some of the key strains and themes in mappurondo headhunting ritual. That effort should not preclude further interpretive work—further "editions" in McGann's terms—by me or by others.

2
The Mappurondo Enclave at Bambang

The Salu Mambi runs shallow and cold over a twisted, boulder-strewn course, its headwaters tangled in the forests of the Messila and Mambuliling massif. The name "Mambi" comes from the word mambie , meaning "hidden from view"—as by trees or tall grasses. A glimpse from the flanking trails is enough to confirm the name. The river winds into ravines and behind mountain spurs, and, like the hamlets nestled above its banks, can simply disappear beneath shrouds of mist and rain. With the exception of a few visits elsewhere in the mountain region, I lived and worked for roughly 30 months (between 1982 and 1985) along the part of the Salu Mambi watershed that stretches from the upper reaches in Bambang to the small but fertile basin at the town of Mambi.

Mambi is a full-fledged, if largely thatch-roofed, town, with market, schools, mosque and church, shops, kecamatan (Ind. regional subdistrict) offices, and a significant number of outsiders who work there as teachers, traders, or civil administrators. Most of those born and raised in Mambi are Muslim, and most nurture hopes of finding a place in a wider, more complex world—somewhat stalled beyond the horizon—be it through their religion, the Indonesian bureaucracy, the coffee trade, video cassettes and radio, or the promised roadway that in 1985 had yet to reach town. As one townsman remarked to me, their interests run with the river waters: to the coast, the sea, and Mecca.

East and upriver from Mambi, the settlements of Bambang form a rural precinct—or desa (Ind.)—of terraces, swiddens, and coffee gardens (see Map 3). When leaving town to walk the riverside trails up toward the

Map 3. The Salu Mambi headwaters, the study area circled on Map 2

headwaters, one speaks of "going inside Bambang" (le'ba' illalam Bambang ) or even just "going inside." The phrase is apt in several senses. It suggests tracing the course of the river back behind the looming escarpment of Buntu Seppon into unseen places among the forests and folds of the mountain range. For the Muslim townsperson, Bambang is foreboding, jinn- ridden, and unclean, a place to be avoided and left to the imagination. Movement, real or imagined, into the terrain of Bambang captures something of a cultural outlook that associates upriver regions with ancestral tradition and authority. "Going inside" involves not only a journey into a remote and

Fig. 2.
A hamlet in Bambang, at the headwaters of the Salu Mambi. 1984.

brooding terrain, but also an acquaintance with a past and some of its ungovernable memories.

"It's like Texas up there," wisecracked Tapuli, a civil servant—and a Christian-always ready to beat me to the draw in a showdown over our knowledge of each other's cultural icons and images, "The people are cowboys." I don't think he meant to paint an image of lawlessness and violence so much as to hint that the people living in Bambang were a bunch of stubborn and occasionally hot-headed rubes who didn't have the sense to give up tradition for the civility and progress of the Indonesian order. Yet it is precisely the images of the violent, the lawless, and the pagan that come into use when the contemporary state wishes to define—and so oppose itself toga suspect citizenry or a suspect era of history.^[1]

For a century and more, Bambang has been a pocket for the ritual traditions once practiced throughout the Salu Mambi and Salu Hau watersheds. Although most of Bambang's residents have converted to Christianity since 1970, a significant minority has retreated deeply into ancestral tradition. Depending on time, place, and the persons to whom one is speaking, the ritual tradition can be called alu' (sacred knowledge and doings), pemali appa' randanna (literally, "tabus of the four strands"), or ada' mappurondo (the term that I favor and will explore below). Followers of this ancestral religion do not maintain or hold claim to an autonomous territory, nor have they joined together in formal political union. Considering the fracture of upland society along religious lines, and given the region's incorporation into the Indonesian state, I should emphasize that those who still adhere to this ritual order do not constitute an ethnic group, but, rather, a religious minority with a distinct ideological focus and identity as well as a distinct set of claims about headwater religious traditions. Although it is useful to think of them as forming a coherent religious enclave or collectivity, their communities and households are scattered throughout a dozen settlements along the upper reaches of the Salu Mambi. In a loose sense, these communities are autonomous, village-based, ritual polities. Living uneasily beside their Christian or Muslim neighbors and kin, and largely excluded from desa administration, members of each community oversee their own household and religious affairs, and usually convene as a village-wide ceremonial polity only during the annual headhunting rites of pangngae.

I took residence with a Muslim family in Mambi in April 1983, a few weeks prior to a spectacular solar eclipse that had everyone worrying about what prayers and ancestral observances would help ward off sinar merah (Ind. "red rays," i.e. infrared radiation). Two months would pass before I was to meet someone from the mappurondo communities upriver in Barn-

bang. During that time, I stayed in Mambi and used the eclipse as a start-up topic in lengthy talks with Muslims, Christians, and members of the civil service about local ritual traditions. The last of these usually talked about the dangers of animisme (Ind. "animism"), of "praying to rocks and trees," giving proof once more that anthropological discourse can race ahead of ethnographers and greet them in the field. For these civil servants, the communities that adhered to local ritual tradition were terbelakang and terasing (Ind. "backward" and "estranged"), groups in exile that had yet to embrace a "true" religion (Ind. belum beragama ).^[2] Christians, and an occasional Muslim, had another way of naming those who followed such ritual practices: tomalillin , or "people of the dark," a term I took as a kind of slur on pagan belief, reckoning that it had roots in the scriptural and evangelical discourse of enlightenment that accompanies religions of "The Book." Most Muslims, meanwhile, spoke of these persons as orang kafir (Ind. "unbelievers") and called the pagan ritual order ada' mappurondo , "mappurondo adat ."

The word ada ' had clear roots in Malay and Arabic terminology for "custom" and "tradition"—adat . The word mappurondo , however, troubled me in my first months in Mambi. When I asked for glosses of the word in Indonesian, people usually responded with sudah dipesankan , roughly "already given [or passed on] as a message." But my early work with word roots in the Salu Mambi area suggested that this gloss was missing something. I tried a Torajan-Indonesian dictionary for clues. Ma'parondo , for the Sa'dan Toraja, means to shake with malarial chill or to tremble in fright ("like when you see a mad dog," someone was later to write me, quoting a Torajan acquaintance). After working later in the year with old manuscripts from Muslim lowland communities, I realized that mappurondo derived from the Bugis and Mandar word ma-pura-onro , meaning "already in place."^[3]

From the vantage point of a social poetics—with its interest in the semantic and pragmatic history of signs and the contests waged over them—the term ada' mappurondo stands out as a response to the Muslim lowland cultures that had been extending their influence into the highlands for over three centuries. The term, I would argue, reflects the effort of highlanders to translate their ancestral practices into the interpretive and discursive framework of an intruding culture. As such, it captures something of the ambivalence that uplanders may have felt with respect to lowland culture. On one hand, it portrays (and thus constructs) ancestral ritual as prior au-tochthonous practice. All practices to which ada' mappurondo may be opposed thus appear as derived or alien. On the other hand, the term's derivation from non-autochthonous language suggests a concession—willing or forced—to a long history of engagement with coastal principalities.

Of the several terms that can be used to name local ritual tradition, I have chosen to use ada' mappurondo. When mentioned, alu' (sacred knowledge and doings) refers rather narrowly to liturgical acts and speech, most often in connection with women's household rites, rather than with men's head-hunting ceremonies. Alu' also smacks of the Sa'dan Toraja ritual tradition aluk to dolo , and, to be honest, I want to deflect the claims and opinions of those who too closely associate the Salu Mambi and Salu Sa'dan peoples, and who are apt to view Salu Mambi ritual practice as a derivative or marginal form of traditions born in a Sa'dan cultural hearth. Pemali appa' randanna designates much more effectively a unique, local order of ritual activity. Yet in the cultural politics of the headwater region, pemali appa' randanna is subject to the claims of Christians and Muslims, claims that would transform it from lived ritual order to past cultural heritage. For its general use among Muslims, Christians, and those faithful to ritual tradition, and for its social historical resonance, mappurondo strikes me as the more apt term.

In writing, then, about ada' mappurondo, mappurondo tradition, mappurondo ritual, mappurondo history, and the mappurondo enclave, I see myself caught up inextricably in the shifting cultural politics of upland tradition. At the very least, I want to acknowledge the authority of the mappurondo communities and the claims they make to local tradition through ritual performance. While recognizing the counterclaims of Christians and Muslims, and some of the more significant cultural ties to communities and societies elsewhere in Sulawesi, I think it critical that we listen for mappurondo voices beneath the din of a pejorative discourse that would define and then erase a contemporary minority religion. Ritual tradition—saying what it is and holding claim to it through practice—has become a cultural problem for the mappurondo enclave, for what is at stake is the making of social relations and social meanings, and the capacity of several small communities to shape their own fate.

The People of the Headwaters

Writing a patrol report in 1924, an anonymous military figure described villagers from Bambang as "coastal Toraja," an incongruous, if telling, remark meaning "coastal highlander" (anonymous [Militaire Nota] 1924). For most outsiders, the people of the Salu Mambi region are of

marginal and ambiguous ethnic stock. Of course, the shifting ethnic affiliations and boundaries of the past may once have afforded them greater autonomy and visibility. But for the last 50 years or more, they have disappeared from the ethnic map of South Sulawesi, a map constructed, in part, through a colonial and national interest in identifying a governable plurality of societies and cultures. That map includes only four ethnic groups—Bugis, Makassar, Mandar, and Toraja—each presumed to have a distinct language, culture, and hearth territory. Depending on their social, political, and religious interests, most outsiders would recognize the remote area of the Salu Mambi as Mandar or Torajan. Still others—like the Indonesian Department of Social Affairs and my sponsors at LIPI (Lembaga Ilmu Pengetahuan Indonesia , the Indonesian Institute of Science)—would write off the Salu Mambi settlements as suku terasing , "estranged," "remote," "people in exile," thus placing them at the margins of the civil and cultural order of contemporary Indonesia (correspondence from LIPI 30 August, 1983).

For their part, the people of Bambang and Mambi call themselves Todiulunnasalu , "people of the headwaters." Although economic and religious factors can serve as points of contrast and comparison, the people of the headwaters usually distinguish their communities from the Mandar and Toraja on the basis of language, custom, and history. Though Christian villagers occasionally stress their linguistic and cultural affinities with the Sa'dan and Mamasa Toraja, the people of the headwaters more widely acknowledge a long history of involvement with the Mandar, the seafaring people who live on the neighboring coasts to the south and west. While recognizing such ties, uplanders are careful to point out the social and cultural differences that mark and shape their own ethnic order.

The speech communities of the mountain region are diverse, polylinguai, and in flux (see Appendix I). Yet by and large these communities form dialect chains throughout the region, a pattern in which the languages of neighboring villages or territories share a significant level of lexical, grammatical, and pragmatic features. In a sense, then, the languages of the Salu Mambi form a linguistic bridge between Toraja and Mandar. The chaining of "dialects" (i.e. speech communities with real social, political, and cultural interests) has kept a single headwater language or linguistic standard from emerging. In fact, ambiguity, polylingualism, and the "marriage" of exogenous tongues are at the heart of Todiulunnasalu ethnic identity.

If linguistic differences have promoted highly localized identities associated with villages or territories, common adat tradition and political union have helped to construct a regional headwater identity. There are, of course,

the shared ritual traditions of the pre-Islamic and pre-Christian past—an issue treated throughout this book. Here, it is worth noting headwater juridical tradition, called ada' tuo , or "adat of life." Under ada' tuo, no human life may be sacrificed in redress for a crime or offense. How well persons and communities abided by this tradition is open to question. Actual practices notwithstanding, adat tuo has served as an acknowledged ideal, one that was introduced, according to upland oral histories, to put a check to rebounding violence. The jural customs it replaced—ada' mate , or "adat of death"—linger, it is said, as the vengeful juridical tradition of the coastal communities in Mandar territory. For the people of the headwaters, then, ada' tuo stands out as an emblem of their ethnic differences vis-à-vis the Mandar coast. Note, too, that this way of declaring ethnic differences underscores the fearsome violence of lowland society and the need to manage upland feuding.

Contemporary headwater ethnicity relies particularly on recollections of a now-vanished upland political order.^[4] Sometime in the sixteenth century, the scattered hill settlements in the Salu Mambi, Salu Hau, and Salu Maloso watersheds came together under covenant of mutual interests to form a relatively egalitarian political league called Pitu Ulunna Salu , the "Seven Headwaters." The members of the league were not individual villages (lembâ ), but socially homogenous, multivillage territories (pa'lembângan ) whose inhabitants acknowledged common ancestry and common adat traditions. The seven founding "adat territories" included Tabulahan, Aralle, Mambi, Matangnga, Tu'bi, Rantebulahan, and Bambang (see Map 4). A key factor that appears to have brought these adat territories together as Pitu Ulunna Salu was a common upland concern over trade and conflict with outsiders, and in particular with the Mandar coast. Indeed, oral accounts speak of the upland league as a response to Mandar political organization, wherein the territory of Balanipa assumed lordship over six other coastal communities by founding Pitu Ba'bana Binanga , roughly, the "Seven Rivermouths."

Over the course of the next three centuries, the league would experience turmoil and change as regional and islandwide power relations shifted.^[5] But the rubric of Pitu Ulunna Salu endured into the late nineteenth century as means and metaphor for upland political cooperation. Along with ada' tuo, the covenant of Pitu Ulunna Salu no doubt promoted the ethnic affiliation of the highland territories. And less than 50 years later, the ethnic and political boundaries of Pitu Ulunna Salu would be resurrected under the Dutch colonial order to construct and legitimate a mountain administrative district (1917, Staatsblad No. 43). Yet with the coming of the Indonesian order,

Map 4. The regional adat territories: "The Seven Headwaters" and "The Seven
Rivermouths"

Pitu Ulunna Salu sank once more into obscurity, and with it faded the ethnic visibility of the headwater peoples.

The ethnic picture today is no less ambiguous than the one offered in the anonymous patrol report of 1924. Neither Mandar nor Toraja, but something in between: The coastal Toraja? The hinterland Mandar? As Pitu Ulunna Salu recedes further into the past, "Todiulunnasalu" grows more anachronistic as an ethnic designation. Some people have turned back to the mosaic of territorial identifies—thus ToMambi, ToAralle, ToBambang , a person from Mambi, from Aralle, from Bambang. Others have already begun to adopt identities from contemporary Indonesian districts. For example, it would not be unusual to hear someone from the Salu Mambi area claim to be an orang Polmas , "a person from the regency (Ind. Kabupaten ) of Polewali-Mamasa," and let it go at that. Still, a majority of people along the Salu Mambi consider themselves Todiulunnasalu, or people of the headwaters, and nowhere more so than in the mappurondo enclave in Bambang.

The History of Bambang in the Precolonial Era

Headwater political history begins with the marriage of Pongka Padang and Torije'ne' at Tabulahan sometime in the fifteenth century. Songs and narratives recall Pongka Padang as an elite Torajan son who wandered from the Sa'dan headwaters that lie in the mountains to the east.^[6] Torije'ne', meanwhile, is something of a mystery to the uplanders. Her name is in the Makassan language and means "person from water" or "one from the sea." Who she was, why she carried a Makassan name, and why she was living at Tabulahan at the time of Pongka Padang's arrival is unknown, though written sources from the Mandar coast suggest she might have been I Sanre Bone, an elite daughter from the Makassar region. This "coastal-highlander" couple would give birth to four incestuous offspring (two brothers and two sisters) whose children would go on to settle Pitu Ulunna Salu and the Mandar coast. Accounts differ on the names, birth order, and settlements of Pongka Padang's descendants, yet all with which I am familiar agree that a son named Tammi' formed the adat territory of Bambang by opening hamlets at Lasodehata and Rantekaneei (which today make up part of the village of Rantepalado). His descendants subsequently settled both Bambang and the Salu Mokanan, the watershed that lies to the south in Rantebulahan.

The foundational discourse to which these stories of social and political origin belong consistently emphasizes themes of siblingship, seniority, and gender. For example, the stories of Pongka Padang and Torije'ne' assert a common parentage for the scattered mountain settlements, and thereby relate the upland adat territories to one another as siblings. As such, the mountain league used an idiom of siblingship as a basis of organization and cooperation. The same idiom, however, also captured some of the rivalry and divisiveness that periodically disturbed the upland alliance, throughout its history. In short, the stories of ancestral parents and siblings serve as a basis of cooperation and authority, even while masking or making intelligible the tensions implicit in upland politics.

Like many other Southeast Asian hill communities, the people of the headwaters reckon prestige and authority on the basis of age. Those more senior by and large wield more influence and authority, and enjoy higher status, than their juniors. But if ideas about seniority help order mountain social life, they also introduce the seeds of conflict. For instance, the most heated and feared disputes occur between a descendant's senior son and junior brother in their claims to be "next-in-line" with respect to inheritable wealth, office, and prestige. It is not surprising, then, that the clash and competition of oral histories and genealogies linking Pongka Padang, Torije'ne', and their son Tammi' to contemporary Bambang place a strong emphasis on seniority. The purpose is to weave stories that legitimize, clarify, or (in other contexts) challenge existing claims to political authority and status in the settlements. For example, the settlement at Rantepalado (and one of its male residents) has remained the only community that may hold claim to the title Indona Bambang ("The Mother of Bambang"), and with it, political leadership over Bambang, by tracing descent along senior lines to Tammi' and Tammi"s senior son. Other villages, meanwhile, carry other adat rifles by virtue of descent from Tammi"s other children and grandchildren.

The stories of origin, descent, and authority have an unmistakably masculine character. They are plotted to foreground fathers, senior sons, and the political offices they hold. Mothers and daughters may figure in genealogical reckonings as "juniors," as "outsiders," or as convenient steppingstones to more senior male lines. In these cases, women are bypassed as focal authorities. To take but one example, genealogical recitations and asides in the Bambang communities of Rantepalado and Minangs claim authority through Tammi"s sons while arguing that the settlements of the Salu Mokanan can trace relationship to the region's founder only through a daughter, said to be his youngest child. From the viewpoint of narrators in

Rantepalado and Minanga, feminized origins here undermine Salu Mokanan claims to higher political authority.

It is worth mentioning, too, that uplanders use stories about Pongka Padang and Torije'ne' to situate differences in language and ritual practice in the mountain communities. Villagers in the western hill-country of Tabulahan, Aralle, and Mambi speak "language of the mother" (basa indo ), while those living upriver and to the east in the rugged reaches of Bambang, Rantebulahan, and Matangnga speak "language of the father" (basa ambe ). The distinction corresponds with contemporary linguistic patterns, for indeed, communities with "paternal" dialects have more in common with the language spoken by the Sa'dan Toraja (Pongka Padang's ethnic background, it will be recalled), whereas communities with "maternal" dialects show greater commonalities with the dialects of the Mandar coast. Father/mountain, mother/sea associations at one time held as well for differences in ritual practices. Prior to their embrace of Islam or Christianity, communities at Tabulahan, Aralle, and Mambi staged postharvest household rituals known as pa'todilopian , "the time of the ones in the sailing vessel." Corresponding rites for those speaking "father tongues" were called, collectively, pa'bisuan , "the time of quickening spirits," with individual ceremonies bearing names often applied to Sa'dan rituals. Commonalities in the regional ceremonial order, meanwhile, are understood as the wedding of different ritual traditions brought respectively by Torije'ne' and Pongka Padang.

From its beginnings in the sixteenth century to its final collapse around 1870, stability and cohesiveness always threatened to elude the upland league. Villages, for the most part, were free to run their own affairs under the leadership of the tomatuatonda' , or village head. Matters of interest to, or involving, neighboring settlements were resolved by convening village leaders under the authority of the Indona Lita' ("Mother of the Ground") of each adat territory. These adat communities, such as Bambang, were themselves fairly autonomous polities, except when matters commanded the attention of the entire headwater league. In this hierarchic political structure, power and authority had no center. Rather, authority lodged throughout the hierarchy and rested on the consensus of male elders and those leaders selected from an elite to be first among equals. To sum up, then, the league could neither overcome internal dissension nor compel participation and compliance of its members, and so lacked cohesivehess.

As we have seen, a common concern over trade and conflict with the Mandar coast helped bring the upland communities together as Pitu Ulunna Salu. Once again, the foundational idiom of siblingship structured the mountain people's understanding of regional political and economic relations. In

their reckoning, it was the junior descendants of Pongka Padang and Torije'ne' who settled the Mandar coast. So not only did this discourse bind mountain and coastal settlements together as siblings, but it also introduced a hierarchy of authority that encouraged the mountain settlements to look upon the coast as a dependent junior. This claim for greater authority and prestige also had a basis in the "natural facts" of mountain cosmography, in which the upstream world sits in an "elder" position vis-à-vis regions downstream. I will have more to say about the political, economic, and ideological dimensions of this relationship in later chapters. For now, it is enough to mention that Pitu Ulunna Salu political organization had its roots in the interdependent and oppositional patterns linking mountain and coast. It arose in response to potentially hostile groups, and with the idea of stabilizing trade with the coast.

As I already have noted, the internal stability of Pitu Ulunna Salu was another matter. Remarks and remembered stories about the 300 years stretching from the late sixteenth through the late nineteenth century paint an image of treachery and discord. Persistent rivalries and feuds between settlements suggest that "feud management" may have been a key issue for the upland league. The vigor of the slave trade further disrupted the highlands. Bands of warriors attacked other villages, seized captives, and then sold them to the Mandar for cloth, porcelain, and weapons. Uplanders, too, took some interest in building retinues of slaves for the purpose of expanding power and prestige, although slavery continued to be viewed as a coastal or lowland institution. Yet it was the arrival of Islam in the uplands during the mid-eighteenth century that seriously fractured the mountain league, as Muslim settlements moved closer culturally, socially, and politically to the Mandar, who had begun to embrace this religion over a century before.

The Enclave in the Headwaters

The coming of Islam to the mountains is remembered in different ways. Contemporary narratives from the Muslim settlements that comprise Mambi and Aralle mention wandering teachers (Ind. ulama ) and saints (Ind. wali ) as the figures who first brought Islam to the headwaters more than 200 years ago. Such stories put emphasis on evangelical effort, sacrifice, and dedication, and on the growing understanding and enlightenment of the mountain converts. Mappurondo villagers in Bambang have a different story to tell, however.

The story takes place at Mambi. The coastal principality at Balanipa has sent a gift to Mambi—a copy of the Koran—igniting a quarrel with Mambi's upstream neighbor, Bambang. The recitation goes:

Says Indona Bambang to Indona Mambi:
Do not eat away at the house
by bringing a newcomer to the adat of Pitu Ulunna Salu.
Says Mambi:
Don't come to complain, Indona Bambang,
about the gift in my sarong from my lords at Balanipa.
Says Bambang:
Don't complain, then, about the smoke at the door of Salukona
who burns incense
and sacrifices a chicken at the border of the land.
I will mark the borders of the land at Bambang with thread
so that the newcomer does not eat away at the house.

The story, so far, tells of the heated words of political and ideological conflict over ritual practices. In headwater political tradition, neither territory may intrude into the affairs of the other, although Bambang, in its regionally acknowledged role as "watchpost of adat" (su'buan ada' ), is obliged to sound alarm when it senses threat to the integrity of adat traditions. So as to contain the dangerous intrusion of Islam, here embodied in the Koran, Bambang secures its borders with thread; that is, it simply declares itself sealed off from Islamic discourse and its consequences.

Soon after, the debata (spirits) took offense at the alien text and abandoned the rice fields at Mambi, the ritual hub of upland agriculture. As a result, the harvest failed and famine set in. The recitation continues with the Indona Mambi pleading to the Indona Bambang for help:

Indona Mambi calls out, calls to Indona Bambang.
Says Indona Mambo:
Come here, Indona Bambang.
Pick up the fallen land and the fallen people
these your lands and your people.
Indona Bambang goes downriver saying:
Mambi, move it downriver to Panompa that Koran of yours.
Indona Mambi moves the Koran downriver to Panompa.

Once the offending Koran had been moved to a village downstream, the debata resumed their guardianship of the rice crop and prosperity returned to Mambi.

Although the immediate source of misfortune had been carried off to a

remote village where it would do no harm, Mambi nevertheless held to its new faith. A dilemma remained: how could the moral and ritual tradition be kept safe from the advance of Islam and the ToSalam , as the Muslims were called? So as to prevent Islam from encroaching further on upland ritual tradition, the league met and declared Bambang a retreat for mappurondo tradition:

The newcomer eats away at the house.
The watchpost of adat
is a room covered by a dark shroud,
not to be seen
not to be heard.

With these words the highlanders concealed local ritual tradition from the gaze of Islam, and placed it out of earshot in the upstream reaches of Bambang. At the same time, these words transformed ada' mappurondo into the religion of a social enclave.

This version of upland political and religious history suggests that as long as eight or nine generations ago the upland league was susceptible to social and ideological fracture. It is also a defining and determining episode in "local mappurondo history"—i.e. a vision of social time through which the contemporary mappurondo communities at Bambang maintain ideological control of the past. This story acknowledges the religious plurality of the mountain region, and so makes concession to the growing political and cultural influence of the Mandar coast. I suspect, too, that the episode recounted here marks the time when "mappurondo"—the coastal term for the already said and the already in place—began to slip into the discourse of highland ritual tradition. If the story is one of concession and retreat, it also gives foundation to the proud exclusionary discourse that distinguishes Bambang from neighboring territories. Villagers from Bambang have shouldered the role of tradition-keeper as their peculiar historical task and destiny. That moral and historical vision remains sharp among mappurondo and Christian factions in present-day Bambang. For them, the history of Bambang has been the history of guarding the living authenticity of adat for all of Pitu Ulunna Salu.

The fervor and consistency of adat tradition beneath the "dark shroud" was not enough, however, to put a check to social turmoil and discord. Around 1870, raids from the Salu Mokanan watershed in Rantebulahan devastated Bambang.^[7] Villagers fled north to find sanctuary in Aralle as warriors spilled into Bambang to take captives and raze hamlets. The intruders,

called ToSalu ("people of the river") began to settle Bambang. The original inhabitants of Bambang, the ToIssilita ' (roughly, "people of the ground"), meanwhile opened settlements in the Buntumalangka' region of Aralle. During the course of the next 30 years, significant numbers of the ToIssilita' made their way back to Bambang to reclaim their abandoned hamlets and terraces. After resolving land disputes, the ToIssilita' rebuilt settlements at Lasodehata, Minanga, Saludengen, Masoso, and Ulumambi, leaving the ToSalu firmly lodged in Bambang's upriver region, having settled at Salubulo, Karakean, Limba, Rantelemo, Salukadi, Salutabang, and a different part of Ulumambi.

The ToSalu intruders powerfully altered the social, political, and cultural landscape of Bambang. Speaking a different dialect, practicing "irregularities" during ritual, and having usurped land and belongings, the ToSalu raised enduring suspicions among the ToIssilita'. Only the fact that the Salu Mokanan people, too, were able to trace descent from Tammi' allowed minimal political unity to be restored throughout Bambang. ToSalu recognition of the Indona Bambang at Lasodehata did little to overcome mutual suspicions or sore relations deriving from the raids. Nearly a third of the earliest Dutch commentary concerning Bambang (Smit 1937) deals with the ToSalu intrusion and the subsequent political calamity. More than 120 years after the fact, events are talked about still, smoldering in ToIssilita' minds, glowing in the ToSalu. Rarely do the ToSalu and ToIssilita' marry one another, and ToSalu representatives have asked the Indonesian civil administration for provisional status as a separate village-precinct, or desa, so as not to share authority with, or subject their interests to, the ToIssilita'.^[8]

The Dutch and Indonesian Periods

Dutch patrols passed through Bambang for the first time during the 1906-7 mountain campaigns, and by 1912 civil administration was in place, the Dutch having set up small posts at Mambi and Aralle. The next twenty years brought many changes: the abrupt cessation of slavery, the introduction of taxes, the arrival of Bugis traders and the opening of the first upland marketplaces (at Mambi and Aralle, c. 1925), the considerable loss of life to famine in 1913 and to the epidemic of 1918, and the arrival of Christianity. Resistance to the Dutch was sporadic and of little consequence, showing none of the anti-colonial millenarianism noted in Central Sulawesi

(Adriani & Kruyt 1950 [1912]; Atkinson 1989; van der Kroef 1970). Indeed, Bambang remained something of an administrative backwater, never subject to the prolonged and intense colonial intrusion that occurred in Sa'dan or Bugis territory. After administrative reorganization of the region in 1924, the nearest civil and military posts of any importance were at Mamasa and Mamuju (40 and 60 kilometers distant, respectively), and colonial officers seldom interfered with the tomatuatonda' (village head) and indona lira' (territory leader) who were already in place, except to enforce collection of taxes and to make sure villagers adhered to Dutch policy and directives (cf. Smit 1937).

When the Dutch arrived in the highlands, the social and ideological enclave at Bambang was intact but still recovering from the ToSalu raids 30 years earlier. The upstream district was a hearth of mappurondo activity, whereas Mambi was fully Muslim, if nominally so.^[9] It was no accident, then, that the Dutch missions ignored Mambi and devoted their efforts to converting the tomalillin ("people of the dark") in Bambang. In contrast to the social and historical forces that brought Islam to the mountains, Christianity arrived in the person of colonial and mission figures who were interested in direct administrative rule. There probably was little the mappurondo community could do to keep Christianity out of Bambang. A story is told about the Indona Bambang in the first years of Dutch rule. Faced with the dilemma of either forsaking local tradition or defying Dutch authority, he decided to subject Christianity to trial and curse:

They say he said, "lf the adat of the ToSarani [Christians] is good, Bambang will prosper. If it brings ruin, it will be gored on the horns of this water buffalo:"

The water buffalo—by tradition the grandest sacrificial animal—is an emblem of ada' mappurondo. Resonating within the declaration, then, was the threat that Christianity risked being gored on the horns of ada' mappurondo. Things did not turn out well. Before a year had passed, the Indona Bambang was dead—gored, it is said, by the very water buffalo to which he had pointed when making his declaration about Christianity.

The message was not lost on the mappurondo community. Yet there remained enormous flux and ambiguity in the reception of Christianity at Bambang. Impressive baptismal figures (Atlas 1925) and glowing reports about the conversion of the entire district (Veen 1933) at best reflect a transient and misunderstood engagement with the church. For example, initial mission work commenced in 1912 under the aegis of the Indische Protestantsche Kerk (the Indies Protestant Church) but amounted to little

more than mass baptisms by an itinerant missionary (Krüger 1966; Rauws 1930; Veen 1933).^[10] Schools were built with government assistance at Karakean and Rantelemo, and the Ambonese teachers installed there baptized nearly 2,500 persons (Atlas 1925:129; Krüger 1966:127). Yet elders today still remember these zealous Ambonese and insist that baptism was misconstrued at the time as a directive from the colonial administration. Indeed, a report on Bambang's local leadership written in the mid-1930s suggests that the church was yet to gain wide and lasting influence (Smit 1937).

The Protestantsche Kerk was unable to sustain its mission in the uplands, and so, in 1927, surrendered the evangelical field to the more conservative and pietistic Christliche Gereformeerd Kerk (Christian Reformed Church), the church that was to oversee the area until 1942 (Krüger 1966). The missionaries Bikker and Geleynse surveyed Bambang, and responsibility for Christian life there passed to Geleynse, who took brief residence at Lasodehata (Rantepalado) in 1930. His impact was comparatively modest, for later that year Geleynse opted to oversee this district from Mela'bo, a Mamasa Toraja village 30 kilometers from the settlements at Bambang. Few conversions took place, and many of those baptized earlier simply continued mappurondo practices without any pretense of being Christian. Locals mention that those who did convert often did so with the idea of winning Dutch support for their political ambitions; in other cases, sons or daughters of the elite were encouraged to convert in order to advance through the schools.

Missionary activity ceased with the Japanese occupation, and by 1950, the year Sulawesi joined the Republic of Indonesia, Christianity had only a limited foothold in Bambang, claiming but a few households in any village, save at Rantelemo and Karakean, where Christian missions and schools had succeeded in forming stable communities and a cadre of leaders. The fledgling Gereja Toraja Mamasa (the Mamasa Toraja Church) had by this time assumed control of the churches in Bambang and began to intensify its efforts in converting the district. Yet it was not until the early 1970s that large numbers abandoned ada' mappurondo for the GTM, a time when those supposedly lacking religion became suspect in the eyes of an Indonesian bureaucracy prone to associating lack of faith in God with subversive, procommunist leanings.

Mission figures were not without influence in cultural and ethnic politics. Throughout the colonial and early independence periods there were efforts to incorporate Bambang (and all of Pitu Ulunna Salu, for that matter) politically and ethnically within the Toraja sphere. Several civil officers who oversaw Pitu Ulunna Salu acknowledged the area's cultural and economic ties to

the Mandar coast (e.g. Hoorweg 1911; Maurenbrecher 1947), however, and consistently argued against incorporation. The people of Bambang, for their part, viewed the Mamasa Toraja as political and cultural subordinates, and protested the political realignment of the mountain region, to little avail. Although Bambang and the rest of Pitu Ulunna Salu remained within Afdeeling Mandar , they were merged with the Mamasa Toraja to form a single administrative subdivision, or onderafdeeling , called Boven Binoeang en Pitoe Oeloenna Saloe —Upper Binuang and Pitu Ulunna Salu (1924, Staatsblad No. 467). Thereafter, Mamasa acted as a political, cultural, and economic center for the onderafdeeling, with the consequence of bringing about the political and cultural marginalization of Bambang and Pitu Ulunna Salu in the hinterland region.

From 1950 through 1965, banditry and rebellion troubled the area. The devoutly Muslim rebels led by Kahar Muzakkar took control of Mambi in 1958 and seemed poised to assault Bambang. Bambang organized rapidly under the ToIssilita' leaders of OPR , or Organisasi Pertahanan Rakyat (Ind. People's Defense Organization), who operated secretly from Saludengen. With aid from nationalist Battalion 710, Bambang attacked Mambi, driving the rebels and most of Mambi's townsfolk back to the coast near Mamuju. Not long after, treachery and rapacity on the part of the Bugis-led 710 caused the OPR to break off the alliance. OPR forced 710 to retreat to Mamasa, and then sealed off all trails into Bambang until civil order was restored fully in 1964.

The Political Stature of the Mappurondo Enclave at Bambang

Just as Bambang was an administrative backwater during the Dutch period, so it has been a backwater of development since 1965. Forming a part of Kecamatan Mambi , and organized and administered during the time of my work as a single desa from the ToIssilita' village of Rantepalado, it has no middle school, no health clinic, no electrification, no roads. Although a marketplace was successfully opened east and upriver, in the ToSalu village of Rantelemo, government-sponsored development programs usually falter at Bambang's door. Indeed, the desa required a visit by the camat (head of the subdistrict) and his entourage in 1984 to spur compliance on the issues of taxes and photo ID cards. The development picture in

Bambang was perhaps best summed up by a 1981 land-use map of Kecamatan Mambi prepared by the provincial offices of the Department of Agriculture and placed in the Kecamatan office. Desa Bambang was missing from the map, left undrawn. In its place were symbols indicating forest and mountains, and lines showing the land as part of other desas.

Since 1970, the mappurondo following at Bambang has suffered a rapid and continued erosion of its community. The heart of the problem is that local civil authorities do not view ada' mappurondo as a religion. Owing to a narrow interpretation of Pancasila , the creed of the Indonesian state, mappurondo practices go unsanctioned and fall under the rubric of adat (i.e. custom, not religion) or animism (i.e. a mistaken pagan enchantment with nature, also not religion). Thus putatively lacking religion (Ind. agama ), mappurondo villagers are thus seen as lacking something in the way of good citizenship. The villagers became suspect in the eyes of civil and religious authorities, who from time to time have charged the mappurondo communities with harboring pro-communist sentiments or anti-development attitudes. Clear efforts were made to pressure mappurondo children to convert to Christianity (or Islam) in order to advance scholastically, and, where possible, to exclude mappurondo adults from holding office in desa or lingkungan (village) administration.

The result has been a sharp rise in conversions to Christianity since 1970, especially among the young, such that ada' mappurondo is increasingly the ritual tradition for elders and the disenfranchised. Today, it is impossible to find a village oriented wholly to ada' mappurondo. In fact, nearly three-fourths of Bambang's 5,000 residents have converted to Christianity, and look condescendingly upon their mappurondo neighbors and kin. Worse, their censorious attitudes have gained ideological and institutional support from the Indonesian government, whose state policies continue to champion monotheistic religion as the keystone of solid, progress-oriented citizenship. In many cases, villages have seen bitter tensions drive Christian and mappurondo camps apart, perhaps nowhere so dramatically as in the ToIssilita' village of Saludengen, where the sides actually exchanged rice terraces and moved. houses so as to create an upstream mappurondo settlement and a downstream Christian settlement. And under strictures issued by Gereja Toraja Mamasa (the Mamasa Toraja Church), no marriages may take place between Christians and those who adhere to ada' mappurondo.

As should be clear, the mappurondo minority has little voice in modern desa political life. Although the Dutch at first drew from the mappurondo leadership to legitimate colonial directives (Hoorweg 1911; Smit 1937),

subsequent administrations long ago supplanted Pitu Ulunna Salu and the Indona Lita', bringing an end to the character and scope of precolonial political authority. The mappurondo leadership of today rests with local groups of elders and "ritual specialists," whose authority extends no further than (village) lingkungan boundaries and revolves around the ritual and moral life of the mappurondo households located in a given settlement. Through 1985, efforts to forge a districtwide organization to shield mappurondo interests consistently met with failure, thwarted by the deep mistrust between ToIssilita' and ToSalu, lingering intervillage disputes, and a fear of relinquishing authority to persons belonging to other villages and other kin groups. Seeking recourse and state sanction through the auspices of Parisada Hindu Dharma—the Hindu leadership through whom the Sa'dan To-raja and the Dayak gained recognition for their minority religions (respectively, aluk to dolo and kaharingan )—has also met with failure, owing to the fears and disinterest of mappurondo elders and to the inability of Hindu representatives to work directly at desa levels without first establishing kabupaten (regional district) and kecamatan (regional subdistrict) offices.^[11]

Mappurondo villagers see no way out of their predicament. Bewildered, marginalized, and made vulnerable by the changes that have swept through Bambang, they have lost ground in the struggle for control of social meanings and social relations. Still, they find ways to understand the new political order, to come to terms with its bureaucratic institutions and its concern for citizenship, through the vehicle of traditional cosmology. As one elder put it, voicing his complaint about the attitudes of schoolteachers and officials, "Animism? What is it? I don't know this animism. I only know that I am afraid, afraid of the spirits and afraid of the government:"

The Contours of Mappurondo Social Life

What is social life like today in this minority religious enclave? As in the past, most mappurondo villagers are farmers and gardeners whose lives center on the household, the hamlet, and the relatives that make up a person's bilateral kindred. Owing to marriage patterns and ways of reckoning kinship, the mappurondo households in each village form a relatively close-knit group of kin with shared interests and anxieties. In my experience, these persons show a deep sense of belonging to their birthplace and homestead, of being tied not only to other people born in the village, but

Fig. 3.
A single-hearth house and rice barn in Bambang. A woman works in the yard
between them. 1984.

to its paths, the shadows of the surrounding hills, and to the sound of the rivers running along the valley floor. The village and its lands thus promote a comforting image through which people recall a common history and a common way of life. To offer but a glimpse of that life, let me here sketch a few basic "facts" about the mappurondo communities in Bambang.

Settlement Pattern

Organization of the settlement begins with the house, or banua . In contrast to the multiple-hearth houses of the past, present-day banua are usually single-hearth homes consisting of a front room, a partitioned sleeping area, and a smaller back room that encloses the kitchen area and hearth.^[12] A married couple and their offspring commonly reside at each hearth and maintain their own rice barn in the yard. A cluster of houses—rarely more than 30—makes up a hamlet, or tonda' (also called botto ), and the terraces and gardens resting on the slopes beneath a tonda' make up its bamba . All the hamlets, their respective bamba, and their outlying gardens and terraces together make up the lembâ , or village, which is administered by desa civil authorities as a lingkungan .

The villages in Bambang vary significantly with respect to the number of households and hamlets, and to the amount of land in cultivation (see Table 1, Appendix II). During the 1980s there were nine lingkungan in Desa Bambang, two of which incorporated settlements that at one time were organized as villages. They included Rantepalado (formerly Lasodehata, and incorporating Salubulo), Minanga, Saludengen, Masoso, Karakean, Limba, Rantelemo (incorporating Salukadi), Salutabang, and Ulumambi. As of 1984, these settlements together comprised 5,742 persons and 1,120 households.

It is worth mentioning once more that the ToIssilita' and the ToSalu make up distinct social and territorial groups in Bambang. Salutabang, Limba, Rantelemo, Salubulo, and the hamlet of Seppang at Ulumambi are the principal ToSalu communities. The ToSalu have also opened homesteads in Masoso, Karakean, and Rantepalado. The ToIssilita', meanwhile, have clustered in the settlements at Saludengen, Minanga, and Rantepalado, as well as in the older hamlets at Ulumambi. Although founded as ToIssilita' settlements, Masoso and Karakean are moving further within the orbit of the ToSalu.

The religious composition of the villages varies dramatically (see Table 2, Appendix II). As should be clear, there is a strong Christian presence throughout Bambang. Limba, Rantelemo, and Masoso, for example, have embraced the church. The mappurondo households remaining in those villages are so few that they can no longer stage rituals on their own.^[13] Christians, of course, have power and influence in other settlements as well. But these same places also have significant mappurondo communities, albeit of varying size and activity. As for Islam, there is a single Muslim household in all of Bambang, located in Minanga, the village lying closest to Mambi.

Economy

The local economy is largely one of subsistence farming, supplemented by income from cash crops, labor, and small-scale trading ventures. Hunting-and-gathering activities are in decline and contribute minimally to subsistence. Although the household is the basic unit of production and consumption, a great deal of cooperation takes place among kin with respect to labor and distribution. There are a few comparatively wealthy mappurondo households in each village, and a few very poor ones as well, but differences seem marginal in light of resource-sharing and the struggle of even the most prosperous households to carry on from one harvest to the next.

Hillside gardens are a mainstay of subsistence. The typical household, for

example, maintains roughly half a hectare of land in garden crops. Emphasis is on semipermanent rather than shifting swidden gardens, with virtually no crop rotation and only brief fallows. Some plots, especially those planted with cassava, are kept in continuous cultivation until exhausted soils force a farmer to move on. Key garden staples are cassava, sweet potatoes, corn, and taro; these foodstuffs, rather than rice, make up the bulk of the diet. Gardeners also plant bananas, cane, beans, and different kinds of greens.

The prominence of the garden notwithstanding, locals say that rice is their most important crop and chief index to prosperity. Unlike garden cultigens, rice is a special, prestigious food, an absolute necessity for expressions of ritual hospitality. Mappurondo villagers consider it sacred and medicinal, believing that it grows under the supernatural guardianship of the girlish and easily startled Debata Totibojongam . The people of Bambang cultivate rice in flooded terraces, and make no use whatsoever of dry swiddens. Keeping with tradition, villagers generally plant one crop of rice annually. Without exception, mappurondo households are careful to plant and harvest in accordance with ritual practices and tabus, and tend to avoid new strains of rice introduced by the government (George 1989). Some families may own as much as 2 hectares of terrace land (see Table 3, Appendix II), but with rare exception, no household is able to raise a crop sufficient to make daily, year-round consumption of rice possible. Households report that they normally deplete their rice stores in anywhere from three to six months, and virtually all buy additional rice in the markets at Mambi and Rantelemo.

As a marker of prosperity and prestige, rice overshadows the contribution of garden crops to subsistence. At the same time, the importance attached to rice somewhat hides the crucial role of coffee in the local economy. Coffee cultivation antedates the arrival of the Dutch (Hoorweg 1911).^[14] Colonial policy and market penetration of the highlands after 1920, however, did much to spur interest in this crop. Local farmers were at first ambivalent about coffee. Elders recall disputes over whether beans should be dried alongside of rice sheaves, some persons fearing that Debata Totibojongam would take offense and abandon her guardianship of the rice. Yet others prophesied that coffee would be a "tree of gold." At issue was whether a sacred or a mundane crop would dominate subsistence. The growth of markets and an encroaching cash economy ultimately worked in favor of coffee, although it has yet to dislodge rice as the basic measure of subsistence.

Since 1980, farmers have become more confident in committing land and labor to coffee (see Table 4, Appendix II), the value of which is subject to fluctuations in world prices and the strength of the Indonesian rupiah.

Growing coffee as a market crop has made it possible for households to buy their way out of subsistence shortfalls (as in the case of the 1983-84 rice harvest) and to gain the cash needed to pay taxes and school fees; to purchase clothing, kerosene, zinc roofing, and tools; and to buy rice terraces. Coffee has also allowed young, landless households to start up small trading ventures.

At the same time, villagers have grown dependent on this cash crop. For example, the 1985 coffee harvest was a striking failure, and despite an excellent rice harvest, farmers expressed worries and reservations about making it through the year. Mappurondo ritual activity during 1985 was a particularly revealing measure of the impact of the coffee failure. Ordinarily, an abundant rice harvest would presage a burst of prosperity rites. No such large-scale ceremonies were held, however, as villagers saw that they would have no coffee income to refill rice barns depleted by ritual.

Mappurondo villagers still play a subordinate role in the market network, in general lacking sufficient capital and social networks beyond the region to broaden or strengthen their ventures, and to challenge Bugis control of trade at Mambi and Melabo. Their attitudes toward debt, risk, and the social obligations that enter into trade take shape from a background of subsistence farming, a communalism informed by kinship, and a history of limited, if well-defined, exchange networks with the Muslim coast.

Kinship, Family, And Household

Villagers reckon their kin relations bilaterally and can usually find persons whom they consider relatives settled throughout Bambang. Personal kindreds, for example, extend to third and fourth cousins. But in particular, people see themselves related through overlapping, ancestor-focused groups called hapu . Hapu organization is especially important to mappurondo society, for it permits the "networking" of persons, families, households, and territorial groupings. In principle, the hapu can be extended indefinitely, such that hundreds of persons, or in fact all of Bambang, may be considered sahapu , "of one cluster," that is, "of one origin." But genealogical reckoning remains shallow in most cases, going back no further than three or four generations, and even then in imperfect fashion. As ancestors fade from the memory of the living, so too does their hapu. Commonality of interest withers as memories blur or fade, and descendants take up then own special concerns. Thus, the hapu are fundamentally kaleidoscopic—shifting patterns of relationship that take shape as prior ones dissolve, overlap, or break apart.

The hapu offers a network of kin in which task groups take shape and lifelong friendships are born. Persons who are sahapu, and especially those who live in the same village, casually visit each other, help one another with labor or a loan of some sort, and often become inseparable companions. Relations within the hapu, then, are usually cooperative, harmonious, and reciprocal, and for hapu members to act otherwise amounts to a breach of proper conduct, causing serious offense and considerable social tension. Conflict between persons or households who are sahapu is hardly absent in daily life, to be sure, but rather than being torn apart by problems, the hapu membership attempts to soothe and settle tensions.

Though people necessarily look to ascendant generations in reckoning the bonds and obligations of kinship among peers and elders, they also look downward to descendants. The interest in and emphasis on descendants is clearly seen, for example, in the use of teknonyms, where persons assume names designating their relationship to firstborn children and grandchildren. Thus, the father and mother of a child named Ta'bu become Ambe Ta'bu and Indo Ta'bu, respectively, and all four grandparents (providing Ta'bu is their first grandchild) will take on the name Nene' Ta'bu. These teknonyms of course signal one's movement into different stages of adulthood, but they also imply a concern for descendants that runs beyond the matter of identity alone. That concern is for founding a prosperous hapu made up of all of one's descendants. As focal ancestor of an ever-growing hapu, a person gains prominence and respect, and at death is buried and honored by the hapu members.

Before moving on to other features of social organization, it is worth remarking that kinship in the mappurondo enclave is also predicated with respect to place of birth, or origin, in a manner akin to the Balinese concept of kawitan, as described by Geertz and Geertz (1975). For the villagers living along the Salu Mambi, the key term is turunam (and also, katurunam ). The best, albeit unwieldy, gloss for the term might be "[the] place at which someone [or something] comes forth and descends to touch the earth." As such, turunam implies a "first appearance," and in its most common usage refers to one's village of birth. In other contexts it may refer to one's natal hamlet, house, or hearth.^[15] What is clear is that persons experience a keen sense of belonging to their birthplace. Evidence for this strikingly deep attachment appears not only in song and poem, but comes across in the surprising energy and resolve shown by the gravely ill in their attempts to return to their place of birth before dying.

For the moment, the point to be made is that the villagers' own concep-

tualization of kinship rests not only on bilateral paths of descent, but upon a sense of place as well. In fact, the customary way of imagining kinship is to think of a hapu in the botanical sense: a cluster of bamboo, the jointed stalks of which represent lines of descent emerging from the common base of village or territory. Where one is born and to whom one is born together link a person not only to a specific set of relatives, but also to a whole community, its history, and its tradition.

The nuclear family, comprising husband, wife, and (unwed) offspring, is the most elementary social group in the villages, but it takes on a significant social, religious, and economic identity only when it tends a hearth and thereby acts as the core for a household. The importance of the hearth cannot go unmentioned. Until a family keeps a hearth of its own, it remains an adjunct to other families (generally that of the wife's parents), unable to conduct religious ceremonies, to have a voice in village deliberations, or to manage economic resources or exchanges autonomously. Once a family has its own hearth, it takes charge of its own religious, political, and economic interests, and effectively becomes a household of its own, even though it may share a dwelling with another hearth-owning group. Household membership, however, always stays somewhat fluid, as relatives are usually welcome to take up temporary or semipermanent residence with the hearth-owners and their children. In time, a household often encompasses an adjunct family, too, as a daughter and her in-marrying spouse begin a family of their own at the hearth of her parents. Or as parents become aged and infirm, the central or founding hearth is passed on to a son or daughter whose family becomes the core of the household. The new hearthkeepers will care for their elders and watch over their interests.

Marriage And Inheritance

Marriages, today, are monogamous. Historically, the case may have been otherwise, for people report that polygyny was a common and acceptable alternative for their elders. In fact, to judge from the way elder men spoke, polygyny appears to have been a marker of male prestige, connoting prosperity, sexual potency, charisma, and power.^[16] On the other hand, accounts of the polygynous past link the practice to an era of wife-stealing, the passing of which men mention with a clear sense of relief. Divorce, meanwhile, is comparatively easy, though not without socioeconomic risk. It is something both men and women can and do initiate.

Incest tabus prohibit marriage with siblings, with parental siblings, with

first cousins or their children, or with persons who are considered to have been raised at the same hearth.^[17] A person may take a second or third cousin as a spouse, but marriages of either kind, called sisahapuam ("both of one hapu"), require that special fines, exchanges, and offerings be made. Villagers are ambivalent about sisahapuam marriages. For one thing, they frankly acknowledge that such marriages permit a consolidation of land and a restoration of office that once belonged to the hapu founder. For another, people observe that second and third cousins already feel a fondness for, and an obligation toward, one another, and thus make excellent husbands and wives. Yet villagers also feel that sisahapuam marriages can also introduce or exacerbate strains within a hapu, and many people would just as soon avoid such conflict. That ambivalence notwithstanding, roughly 50 to 60 percent of all marriages within the mappurondo communities are sisahapuam.

Uxorilocal residence strategies are prevalent and have promoted the formation of mother-sister-daughter "cores" within the household and hamlet. Indeed, a new husband is said to "live beneath the ends of the floorbeams" of his wife's natal home, a saying that captures his marginal status vis-à-vis his in-laws. Owing to hamlet (tonda ') exogamy, a preference for village (lembâ) endogamy (see Table 5, Appendix II), and the perceived advantages of sisahapuam marriages, the village settlement tends to be a strongly bounded, close-knit community.

Bridewealth, inheritance, prestige, and social realignments make preparations for marriage a tinderbox, especially in sisahapuam cases. Error and oversight can lead to serious affronts and bitter quarrels. In order to marry, a youth must give somba (roughly, "honor") to his prospective bride's kin. A form of bridewealth, somba customarily is made up of cloth and tools, goods of exogenous origin. But seldom does a youth collect these goods entirely on his own. He more often enlists the help of his kin in gathering enough materials for somba. During the wedding ceremony, representatives of the husband's side will make a gift of these valuables to the bride's kin, who divide the cloth and tools among themselves in proportion to the honor and respect due each kin member. Ostensibly flattered and honored by this bridewealth, the bride's kin give her to the groom.^[18] They also reciprocate with meats and bundles of rice that are divided among the husband's kin so as to reflect the honor due each kinsperson and to acknowledge help in amassing the valuables used as somba.

Following their wedding, husband and wife each take their mana ' inheritance from their elders: rice terraces are the customary form of mana', particularly for women, and are given in part with the idea of providing the

couple with a base of subsistence. Secondary inheritance takes place after a parent or relative dies. The remaining inheritable possessions of the deceased, called kambi ', are distributed to relatives, who are required, in turn, to make a sacrifice to the soul of the deceased. Kambi' usually consists of household goods and small parcels of land, for more substantial holdings are normally transferred beforehand to descendants as mana'. Ideal distribution of kambi', according to villagers, should level out disparities in gifts of mana' to siblings.

Ritual titles and paraphernalia are not included in mana' or kambi'. It often happens that one of the ritual elite grows old and wishes to relinquish title and obligations. Unless village elders voice disapproval, ritual paraphernalia and duties are simply passed on to the handpicked successor in the family or hapu. The death of an officeholder creates a more problematic succession. Here, succession is entirely in the hands of village elders, who must pick one of the deceased's relatives for the vacant position. Fierce competition can break out among the deceased's siblings and children as they try to sway the elders in their choice.

Inheritance and somba figure critically into decisions about marriage. Never absent in the minds of husband, wife, and their respective kin is an interest in advancing their prosperity and status, the bases of which lie in inheritance. Meanwhile, somba exchange not only has the practical purpose of forging alliances, but also may reflect intra-hapu strategies and interests. Not surprisingly, the snare of village politics, too, is deeply planted in, and consumed with, matters of marriage, inheritance, and status.

Prestige And Social Hierarchy

Prestige and social status in the mappurondo communities rest largely on age and seniority. Persons show deference to parents, grandparents, parents-in-law, and elder siblings, and expect the same from their juniors and offspring. Seniority in descent also plays a crucial part in making successful claims to inheritable and prestigious positions of ritual leadership. At the same time, significant terrace and coffee holdings, ritual displays of wealth and prosperity, and the ability to convene a retinue of followers bestow prestige on a person or household. For the community, then, age and seniority tend to create an unalterable set of status positions, while wealth and achievement offer a means for elevating prestige and influence. Given these two axes of prestige, a person's social status is inevitably and incessantly negotiated, contested, and kept in flux.

Disparity in wealth within the mappurondo community is not that great, and few households enjoy marked prosperity.^[19] The villagers reckon wealth in terms of rice: those who grow and eat more rice are rich, and those who grow and eat little are poor. The more prosperous become tomakaka ' ("one who is like an elder sibling"). The tomakaka' act as patrons to less prosperous relatives and villagers, permitting the latter to sharecrop or to borrow generously. Thus, relations of dependency and prestige find definition in idioms of siblingship and seniority.

Social hierarchy clearly has undergone change over the last century with the cessation of slavery and the erosion of mappurondo ritual life. Villagers gave several different descriptions of former ranking systems, but all accounts included mention of social dependents and inferiors who had been enslaved through capture or debt bondage. Yet idioms of siblingship and seniority once more entered the picture, here mitigating against the sharp demarcation of a slave class. Slaveowners incorporated slaves into their families as adi ', or junior siblings, such that relations of debt bondage and coerced dependency took definition from the reciprocal obligations that bound siblings to one another.

Villagers are emphatic that persons holding ritual or political titles form a social elite. Ritual cannot properly take place without the men and women who hold the appropriate ceremonial rifles presiding. For their participation and the "burdens" they undertake, the ritual elite take home gifts of special meats. Thus, the ritual elite gain prestige through their specialized knowledge and their customary rights to sacrificial meats. The prestige of ritual rifles remains strong, and claims to such positions continue to be a source of competition, conflict, and envy among mappurondo households. At the same time, the burdens of ritual office—living under special sets of tabu and hosting certain gatherings—make the rifles less attractive to some villagers.

A Glimpse At Gender Relations

Sexual politics and gender hierarchy make up an important dimension of social and ritual life in the mappurondo communities. Later chapters will treat the issue of gender in some depth. For now, a thumbnail sketch of gender relations will be helpful.

Generally speaking, the relations between mappurondo men (muane ) and women (baine ) steer toward balance and collaboration, and in many contexts appear to have the relatively egalitarian and complementary character one might expect in a society that uses age as the most basic criterion of

prestige (Ortner 1981, 1990). Similarities aside, gender differences do exist and do order village life, especially in the division of labor and ritual practice. Yet most persons tend to view these differences as complementary in nature, and it is in part this perceived complementarity that keeps gender distinctions from becoming a problem in everyday life. It needs to be asked, of course, whether this complementarity has a basically egalitarian or hierarchical character. Indeed, the issue of control troubles and haunts the talk of equality, balance, sameness, and complementarity. But as a starting point for discussion, it is important to note that both men and women in the mappurondo communities use a discourse of complementarity to level gender differences and to portray them as fitting, natural, and reasonable.

This egalitarian discourse appears to be the common or predominant one. To my knowledge, no one claims men or women to be fundamentally superior to the other. Along with some basic ideas about human sameness, there exists a basic moral view about what makes a "good person" (mapia penabanna , literally, "with good breath"), irrespective of gender. Generosity, patience, and industriousness, for example, are valued in both men and women. Yet ideas about manhood and womanhood also need to be brought into the picture. In the mappurondo community, a man should have the valor and skills of a headhunter (cf. Atkinson 1990, and Rosaldo and Atkinson 1975). He has a reputation to keep and must learn to bluff or threaten others with a "don't mess with me" attitude. At the same time, a man should be able to persuade with high oratory and a sound knowledge of tradition. A woman, on the other hand, should be a skilled gardener and, when possessed by the debata in household ritual, a graceful dancer. Villagers do not stress a woman's ability to bear children but, rather, her domestic discipline. The favor shown a woman's capacity for domestic discipline does point to a troubling current in the seemingly egalitarian relations between men and women. Ex-airing women for their domestic discipline is perhaps a way to celebrate their controllability. In this sense, a good woman is one who monitors her own domestic behavior and thus spares another person—perhaps a husband or father—the need to rein her in.

Unlike the division of ritual activity, the sexual division of labor in the mappurondo community is flexible and relatively balanced. Women customarily plant, weed, harvest, and gather, while men fill rice terraces, burn off garden plots, fish, and make homes, huts, and rice barns. Yet men and women are free to assist one another in these tasks. Childtending and cooking preoccupy women more often than men, but are quickly passed on as chores for children aged roughly seven and up. The division of labor is far more

rigid, however, when it comes to hunting game or weaving: without exception, hunting falls to men and weaving to women. I have no evidence that villagers place differential value or worth on these various tasks.

It is marriage that brings men and women into complete adult status and binds them into a complementary relationship. As husband and wife, they become bela , a term unmarked for gender denoting a companion in a mutual task.^[20] Through cooperation and companionship a couple aim at increasing the prosperity and status of the household, something usually measured by harvest surpluses, landholdings, sound health, number of children, and displays of hospitality and generosity. Although the bela relationship is a culturally acknowledged ideal, it is not unusual for it to be troubled with tension and asymmetry. First, and without exception, \ men marry women several years their junior. Thus, a nongendered axis of prestige and authority based on age enters into the bela relationship, compelling junior wives to show deference to senior husbands. Owing to their seniority, husbands may be said to enjoy an edge of control over their wives. But it needs to be remembered, too, that uxorilocal residence strategies put a husband in a subordinate relationship vis-à-vis his wife's family. In fact, the most common cause of household discord during my time in Bambang had to do with soured status relations between sons-in-law and parents-in-law.

Traditionally, men have had a nearly exclusive hold on village political positions. The pattern continues today, at a time when the mappurondo community has to make concessions to the civil administration, and to those who have embraced Islam or Christianity. Still, women are not kept from bringing an issue before village men, nor are they prevented from holding village political office. That they seldom do so points rather directly to the asymmetries cloaked by the discourse of gender equality and complementarity.

Without question, village endogamy, hamlet exogamy, and uxorilocality are themselves political acts and constructions: they shape the way women and men participate in village politics. By virtue of their role in local social structure, married men have to become village politicians. Their interests and obligations span natal and affinal households as well as natal and affinal hamlets, and indeed, men show themselves to be consumed by local prestige politics. Women have more narrow political interests, ones that focus on their natal hamlet and that show deep commitment to the social bonds of mother-sister-daughter cohorts. It is not surprising, then, that locals think of the settlement as a male sphere, and the household as a female sphere. Positions of village leadership customarily fall to men, and the voices heard in gatherings to discuss village affairs are those of men. In principle, women may take part in these discussions or sit as village head. But for a woman

to do so would strike villagers as peculiar, and would certainly be a case of "beating the odds" (cf. Atkinson 1989, 1990, from whom I borrow the phrase). Beyond the rule that "village" rituals performed under the authority of men must take place before the household prosperity rites overseen by women, there is little evidence that the male-dominated village sphere is consistently valued above the female-dominated sphere of the household. I would suggest, then, that gender asymmetry in village-level political life may not be so much a matter of hierarchical value (male over female, public over domestic) as a matter of political scope. The village politics of men encompass the socially and residentially limited politics of women, but are not consistently accorded higher value.^[21]

Both men and women take part in ritual life, and both have opportunities to become recognized specialists vested with the authority to perform certain ceremonial roles. More generally, men take center stage in ukusam botto, the rites of the settlement, while women assume authority during ukusam banua , the rites of the household (about which, more below). Although both husband and wife share a concern for the material prosperity of the hearth and family, it is the wife who is felt to be the guardian of household welfare. Her capacity to bring a providential presence into the home by being possessed or "taken by the debata" (diala debata ) during ritual accords a wife and mother a significant degree of religious dominance, and places her at the threshold of the sacred. Men, meanwhile, play a more peripheral role in the religious management of household prosperity. Theirs is a world oriented to village welfare, prestige, and male peership.

To sum up, the gender differences and sexual politics of the mappurondo communities tend to resemble those described by Collier and Rosaldo (1981) in their study of hunter-gatherers and simple horticultural societies. Furthermore, they fit reasonably well with patterns described for other bilateral "hill tribes" in island Southeast Asia. Overall, there is in mappurondo culture a fairly egalitarian gender ideology that stresses the complementarity of women and men. If that ideological system posits differences between men and women, it nonetheless fails to organize the distinctions into a stable hierarchy of value wherein one gender could be said to be consistently subordinate to the other in terms of prestige.^[22] With the exception of certain ritual practices, very little effort is expended on policing the boundaries of gender difference in daily activity. Asymmetries in authority do exist, but for the most part reflect a relational structure that puts emphasis on men in some contexts, and on women in others. Men emerge consistently as community leaders, while women assume privileged roles in the sacred business of the household. On the other hand, clear vectors of asymmetry appear in

marriage, idealized as the egalitarian and complementary bela relationship. Here, a husband's anxieties about male prestige and peership lead him to find ways to control his wife and her domestic and sexual services.

Mappurondo Ritual Life

I have saved for last a very simple sketch of traditional religious practices, the practices that define the mappurondo enclave as an ideological minority. The ritual tradition at Bambang, like that throughout the Pitu Ulunna Salu hinterlands, goes by the name pemali appa' randanna , the "tabus of the four strands." With this metaphoric term, villagers think of ritual life as a four-stranded necklace, the loops marking four kinds of ritual time, and the beads representing prohibitions and admonitions regarding practice and conduct. The looped strands of the necklace suggest that ritual life not only adorns the village—a theme articulated in headhunting ceremonies—but also assumes a circular, cyclical, and encompassing form. In a sense, the ritual order of pemali appa' randanna adorns both the village and time itself.

Briefly, pemali appa' randanna divides the harvest year into four periods, each with a characteristic focus, mood, and purpose. Patomatean , "the rime of the dead," pertains to death and mortuary practices, and for that reason, lacks some of the seasonality of the other periods. Patotibojongan is in effect from the time the rice crop is planted to the time it is put into storage. The men's headhunting rite of pangngae then opens up pealloan ("the place of the sun"), a period that subsumes the remaining two "strands." The first of those strands is pa'bisuan ; it concerns the exuberant rites centered on the village and the household, including pangngae. Once pa'bisuan comes to an end, pa'bannetauan , the season of marriages and divorces, takes place. After wedding festivities have come to a close, villagers await the return of patotibojongan.

A closer look at this ritual calendar, beginning with patotibojongan, should help frame a discussion of pangngae in later chapters. Patotibojongan refers to the timid and girlish rice spirits who descend from the sky-world to occupy each terrace. Upon the plowing of terraces and the casting of seed or the transplanting of seedlings, every mappurondo household will give offerings and prayer to these dabata. Subsequently, the community is absorbed in labor and enters into a somber tabu period that lasts until harvested rice is placed in the granaries—very roughly, an eight-month period stretching from July through February. All of that which might disturb the

easily startled rice spirits—noisiness, laughter, storytelling, singing, drumming, and ceremony—is strictly prohibited. (Two important exceptions are funerary ritual and, as the rice matures and ripens, the manufacture of tops, swings, and windmills, the sounds of which delight the rice spirits.) At harvest, final offerings are given to the rice spirits, who ascend to the skyworld. Sheaves of rice are then dried on vertical racks. When the last of the households have collapsed their drying racks, patotibojongan comes to an end.

Not long after the drying racks have been put away, a group of men and youths will slip out of the village and go on the headhunting journey associated with pangngae. Their boisterous return in the dead of night releases the mappurondo community as a whole from public mourning for the deceased and shifts ritual practices to pa'bisuan, "the time of the quickening spirit." Pa'bisuan divides in two: pa'bisuam muane , those village-level rituals held under the authority of men (also called ukusam botto ) and the ensuing pa'bisuam baine , a series of sometimes elaborate rituals held under the authority of women for the benefit of individual households (also called ukusam banua ).

Men's ritual is limited to the annual, obligatory performances of pangngae and the prestigious and comparatively rare morara ("the bleeding" [i.e. of a sacrificial animal]), both of which are staged by entire villages.^[23] It is in these rituals that the debata restore and enliven men's sumanga ' (elan vital, or soul).^[24] Both rituals exalt the valor and cunning of headhunters, and extol the prosperity and prestige of the village and village tradition. But before morara may be staged, pangngae must be held. Pangngae involves as well, in most cases, only village residents, with occasional guests (most often kin or distinguished elders) from mappurondo communities in other settlements. Morara, on the other hand, requires a host village to invite guests from all corners of Pitu Ulunna Salu.^[25]

Rituals belonging to women (pa'bisuam baine) commence once the rites of men have come to a close. Because morara is seldom run, pa'bisuam baine usually begins when the headhunting celebrations of pangngae end. Thus, the focus of ritual shifts from male to female, and from village to household. Run under the authority of women who hold ritual office, the ceremonies of pa'bisuam baine (five according to the ToIssilita' and six according to the ToSalu) aim at bringing prosperity and health to the presiding household and its hapu, by way of sacrifice and entrancing visitations from the spirits. The rituals of the female domain are sequential, moving from the small and simple to the grand and elaborate (called parri ', "heavy," or "burdened"]).^[26] Households must adhere to the set sequence and may

not stage more than one rite per year. At the same time, a household is under no obligation to run one of these rituals unless husband and wife have made a vow to do so.

Following the season's final pa'bisuam baine ceremony, activity shifts to the fourth strand of ritual injunctions and admonitions—pa'bannetauan, "the time of human seed." Like the rituals associated with women, the weddings of pa'bannetauan could be called rice-sensitive, in the sense that they take place only when a household has enough rice to put on a marriage feast for a daughter. In practice, households will not stage an elaborate rite during pa'bisuam baine if they anticipate proposals from prospective sons-in-law, preferring instead to reserve the necessary rice supplies for a marriage feast. As might be expected, harvest failures dramatically limit the number of weddings staged in a given year. With the close of pa'bannetauan, the exuberance of pealloan subsides. Although no major ceremonies may be staged, villagers continue to labor or relax in relative freedom from tabus.

There exist a number of other rituals that are more contingent in nature, though some of them have a connection with pealloan. These stray beads, so to speak, have to do with epidemic, natural disaster, early stages of a person's life-cycle, and construction of a house. The crises of epidemic and natural disaster can befall a village at any time. Prolonged or serious illness is usually dealt with privately, after enlisting the aid of recognized curers (usually female) who possess knowledge of the incantations and offerings used to heal.^[27] In the case of epidemic, however, village elders try to drive away the supernatural beings thought to bring illness upon the hamlets. A public sacrifice is held on a path downriver from the village (the cosmological direction of death and illness), and a frightening effigy is installed to ward off the beings' return. Similarly, disturbances within a hamlet—landslides, the toppling of trees, or the collapse of a house—are followed with calming sacrifices to the startled debata who may have unleashed the calamity.

Life-cycle rituals associated with pealloan include the now-defunct practices of male tooth-filing and superincision that were halted earlier in this century by Dutch and Indonesian authorities. Among extant practices, the ritualized presentation of clothes to children in their fourth year (peculiar to ToIssilita' communities) takes place during pealloan. Minor rites seeking a long life for a year-old infant, marking a child's contact with the ground (i.e. taking first steps), or beautifying an infant girl by ear-piercing happen by parents' preference and discretion during pealloan, too. But the earliest life-cycle rites—naming, and introduction of the child to the sleeping sling—happen within a month of birth and are not coordinated with pealloan. Finally, the very visible ritual event of installing a hearth in a newly con-

structed house (by taking ashes from the wife's mother's fireplace) happens without exception during pealloan.

This quick sketch hardly does justice to the complexity and texture of mappurondo ritual practice. Nonetheless, the basic character of the ritual order comes across in this outline of practice. At root, the ceremonial order—in its regulation of time, sociality, and discourse—establishes a powerful set of constraints on labor, consumption, and domestic and communal reproduction. Prosperity is perhaps its key theme—whether of person, household, hapu, or settlement. That discourse of prosperity, however, also includes, and makes intelligible, an anxious desire for prestige. Local religion thus sacralizes, and even fuels, the prestige politics that shape social life in the mappurondo communities. The rich symbolism of pemali appa' randanna also forms the context in which local gender differences reach their most radical formulation and display.^[28] Such shifting asymmetries not only play a part in shaping human relationships but also inform the making of household and village as the key sites for the sacred politics of prosperity, authority, and power.

I should not leave the impression that the ritual order is free of pressure and change. Judging by the recollections of older men and women, pa'bisuam baine—in contrast to other portions of the ritual cycle—is in some communities on its way to becoming badly stalled with respect to the staging of parri' ceremonies. These elaborate rituals have always been difficult and costly to perform, and the current economy and the vast number of conversions to Christianity since 1970 have made them more burdensome and problematic for a household. As a result, the ritual cycle of today is one geared especially to weddings, funerals, and the headhunting rites of pangngae. The consequences for women should be clear—their place in ritual life has diminished somewhat and taken new focus.

Having sketched the basic social and historical outlines of the mappurondo enclave at Bambang, I can resume exploration of local headhunting practices. I should stress that Christians dominate the religious and political landscape in Bambang, just as Muslims do downriver in Mambi. Followers of these world religions have rejected pemali appa' randanna in everyday moral practice. The mappurondo enclave in Bambang persists, meanwhile, as a number of village-based ritual polities that have yet to coalesce into a territory-wide religious body. Living uneasily with their Christian neighbors and kin, reluctant to put aside the tensions that separate ToSalu and ToIssilita' camps, and often bewildered by the intrusions of the Indonesian state, mappurondo households seldom look beyond village boundaries and show

little interest in joining a larger cooperative union. Nevertheless, all the mappurondo communities in this enclave have a consistent, historically informed vision of Bambang as repository and guardian of adat tradition. The tenacity of pangngae relative to other traditional ceremonies shows that the village remains the pivotal site for reproducing and giving institutional form to the broader ideological order of ada' mappurondo.

3
Defaced Images from the Past On the Disfigured Histories and Disfiguring Violence of Pangngae

At first glance, the mappurondo communities may seem an unlikely place for us to make sense of headhunting ritual. After all, during pangngae no enemy actually is felled, no human head is taken. What takes place is not a "real" headhunt, but something staged to look like one. The drums, songs, and loud cries, the feasting and speechmaking, the bloodless trophies come into play as villagers set loose their imagination in the ritualized mock-slaying of an outsider. It is easy—as it is reasonable—to suppose that the social, cultural, and political forces of the last 100 years have brought an end to "real" headhunting, leaving today's villagers with the ghost of former practices. From this vantage point, pangngae lingers as a troubling or perhaps amusing relic of more primitive and violent times. Yet I think what makes the annual headhunt of pangngae so striking and instructive is its artifice. The lack of actual violence, more broadly the frank recognition and exposure of incongruities between ideology, act, and circumstance, make explicit the desire of mappurondo villagers to recreate their ritual practices—and meanings for them—from the givens of the past and the contingencies of the present.

From the start, let me acknowledge that there are difficulties in writing about the headhunts of the past. The problem goes beyond knowing "what happened" along the banks of the Salu Mambi. In fact, it goes right to the issue of how ethnology has theorized and explained headhunting. I will

suppose—rightly, I believe, and with some evidence—that the upland villagers once upon a time took heads. At some later date, they stopped, or almost stopped. Forsaking real victims and real heads, villagers began (or resumed) their traffic in surrogate skulls. Yet, there has been at least one instance in the last 50 years when a cohort of upland men have taken the head of a human victim. The historical movements back and forth between actual heads and surrogate heads, however, should not mislead us regarding the ontological status of the latter.^[1] We need not assume that the use of a surrogate head has to follow upon the use and abandonment of human heads. To the contrary, a human head does not enjoy a privileged and prior position in the scheme of the real. A human head and a surrogate are equally real, powerfully vested signs of violence. Seen from another angle, though, the head and the surrogate are disparate versions of the "same" sign—each with critically different entailments and consequences.^[2] Whether a head or a surrogate comes into use depends, of course, on social and historical conditions. The problem for analysis, then, is to write the biography of the sign—here, the image of a severed head—noting its movement through time and the way it falls subject to varying social and historical forces.

A second difficulty in writing about the past is to come to terms with the "textuality of history" (Montrose 1989; cf. White 1978).^[3] I can know about the headhunting of the past only by way of the discordant traces, silences, and erasures that make up local history. For their part, the mappurondo villagers of today claim that their tradition of headhunting ritual is unbroken, stretching back sixteen generations or more. Two alien histories—that of a colonial power and that of the headhunters' victims—are silent. Dutch administrative and missionary records, for example, are mute about headhunting in and around the Salu Mambi region, save for a remark about a supposed "headbarn" to the north in Tabulahan (Kruyt 1942:550, citing correspondence with Bikker). That it goes unmentioned suggests that the ritual of pangngae never seriously troubled the colonial authorities. Similarly, the ritual does not appear to have troubled the headhunters' usual victims—the Mandar. They have no idea that they have lost heads, past or present, actual and surrogate, to the uplanders. If the exuberant claims of the uplanders are true, why do the victims and colonial authorities fail to mention Salu Mambi headhunting practices? Coming to terms with the textuality of local history means coming to terms with this mystery, and I will offer a reading of the past that will reconcile these discordant, competing views.

Texts about the past have their own historical character, of course. Written or recited, remembered or forgotten, these texts owe their life to particular times and places and take shape in accord with the culturally specific

tensions and constraints of a lived-in social order. The same is true of the critical commentaries and ethnographic analyses heaped upon headhunting practices in island Southeast Asia. How anthropology has encountered and represented the headhunter is an issue that deserves the close scrutiny of a cultural or intellectual historian. Here I want to note only a few very general points about that literature.

Much of the early theorizing and "fact-finding" about Southeast Asian headhunting took place just after hill and island communities were "pacified" and brought within the orbit of colonial jurisdiction. I think it fair to say that headhunting ritual stands out in this early literature as a sign of an unruly and pagan otherness.^[4] For some, the headhunter may have figured as a human grotesque, sparking curiosity, contemplation, and awe. More significant, however, was a desire on the part of the social order producing this literature to tame the headhunter's violence. That impulse or desire to tame demanded subjects that would acquiesce under threat, reprisal, and enticement. No surprise, then, that headhunting became warrant for reconnaissance and pacification campaigns, while its decline or cessation served as a measure of colonial and missionary success. The dark romanticism, the fear, and the will to control are all present, for example, in a brief passage from a report by J. F. W. L. Goslings (1933 [1924]), a Dutch civil officer assigned to the Galumpang region of western Sulawesi. He writes that, "Before 1905 Galumpang was 'terra incognita' where no European had ever set foot. The Galumpang people lived in a time of underlying feuds. . . The hunting of heads until our regulation was in unbridled practice. . ." (1933:57-58). As for curbing this entrenched ritual violence, Goslings showed little faith in the native tribunals and courts set up by the Dutch. Rather, he left it to "time, the Civil Administration, and the Mission" to bring about an end to head-hunting (1933:68).

Seldom do these early accounts look at headhunting practices with respect to questions of domination and subordination.^[5] Rather, the tendency is to fathom headhunting in terms of belief , an approach decidedly more theological than political or social. Once they had exacted control, colonial and mission authorities needed only to comprehend their pagan wards. With the collapse of precolonial political systems and relations of exchange, observers were less likely to witness the taking of human heads than to see a body of customs neutralized and distorted by the new social order. The violence of headhunting, in most cases, belonged to the past, where it could be interrogated safely for the intellectual projects of ethnology.

The early ethnographies and travel accounts that dealt with headhunting created a legacy for later work. The consuming question—"Why a head?"—

organized inquiry around problems of belief and causality (cf. Needham 1976). At the same time, it enshrined the trophy skull as a fetish for ethnographic analysis. Solving the mystery of the severed head, finding in it an intelligible form of native rationality—a belief that heads contained magical soul-stuff, for example—would surely solve the problem of headhunting. Subsequent eras of structuralism, Freudianism, and related modes of symbolic anthropology more or less revealed the same fascination with trophy skulls. The aim was to identify a perduring symbolic logic that necessitated the ritualized taking of heads. Contingencies of circumstance, purpose, and meaning, of course, had little place in this determinative and oftentimes univocal structural logic. Meanwhile, theories that emphasized the resonant multivocalism of ritual symbols did little to dampen the fascination with the head. To the contrary, theories that emphasized the semantic abundance of ritual objects, that trumpeted their rich polysemy, still construed the severed head as the locus of meaning. If ethnologists could no longer accept the view that the skull contained soul-stuff or potent seed, they nonetheless remained convinced that the head was full of meaning. The theoretical danger, here, was to overlook the intertextually contingent and constructed character of meaning, treating it instead as a decodable property of the symbolic object. In practical terms, this kept the severed head at the center of ethnographic interest, holding it in place as the proper object for interpretive work and play.

The mystery of the head lingers still, holding and being held by the anthropological imagination. Without question, it has shaped nay work on pangngae and the mappurondo community. Yet I am interested less in finding an unequivocal answer to the problem, "Why a head?" What I hope to establish instead in the present work is a vantage point, or series of vantage points, from which the shifting ambiguities and ironies of the stolen head can be glimpsed.

Containers Contained, or Theory Turned on Its Head

A cohort of youthful headhunters came home to the village of Rantepalado on a moonless night in late April 1985. Let in on the secret of their approach, I had joined several mappurondo elders in their darkened homes, waiting for the shouts and flutes that would open the hamlets. Once the

youths were back and the hamlet fires burned again, we sat up until dawn celebrating their return, gathered under a makeshift lean-to at the edge of the hamlet of Lasodehata. I talked a lot that night with Ambe Ka'du, one of the more brash and vocal young leaders of the mappurondo community in Rantepalado. Telling me about pangngae, he looked over at the boy-warriors and remarked, "The point is for the headhunters to bring back sumanga', got it?"

Sumanga ' is a common term throughout southern Sulawesi for what might be called ardor, vitality, spiritedness, drive, potency, or soul (cf. Errington 1983, 1989). Sumanga' can rise, and it can be pulled by a woman's lips as she draws in her breath over the crown of someone's head during a ritual act known as meala sumanga ', "gathering sumanga'." In a sense, sumanga' is a kind of animating "soul-substance." Ambe Ka'du's remark seemed to echo support for the long-abandoned "soul-stuff" hypothesis of A. C. Kruyt (1906). Like other theories put forward around the turn of the century, the soul-stuff theory tended to be somewhat instrumental in its view, and wound up not only making conjectures about the mechanisms that made headtaking so efficacious, but confusing them for the principal cause or motivation behind the practice as well. In Kruyt's view, the headhunter planned to make off with the beneficial soul-substance contained in enemy heads. Taken back to the warriors' village, the victim's skull would exude a life-energy that would restore health and replenish crops and livestock. Although convincing enough to take in a skilled ethnographer like Raymond Kennedy (1942), the theory did not hold up very well on several counts, and was even abandoned by Kruyt himself (Downs 1955). For example, it fails to explain those cases, like that of the Ilongot, where the head was not taken home but was discarded instead (M. Rosaldo 1980; R. Rosaldo 1980, 1984). Second, headhunting practices vary with respect to the number of heads brought home. The majority of groups were like the Berawan, who sought just a single head (Hose & McDougall 1912; Metcalf 1982). Others trafficked only in bits of skull (Goslings 1933 [1924]; McKinley 1976). Even the Bare'e Toraja, whose headhunting practices provided the ethnographic context for Kruyt's formulation of soul-stuff theory, were content to take a single enemy head (Downs 1956:64). The number of heads customarily taken in a raid offered no uniform evidence, then, that headhunters were trying to amass or stockpile soul-stuff. More damagingly, few if any headhunters in the ethnographic literature made use of a soul-stuff concept in explaining the how and why of headhunting. In fact, headhunters characteristically had little to say to ethnographers about why they took heads, except to remark that it was a matter of tradition.

Ambe Ka'du's words gave me eyebrow-raising pause. I pushed him: "The sumanga' is in the head that's carried back."

"No. It's like this, carrying back a head brings back sumanga'."

"So they carry sumanga' back from the coast."

"No. Bringing back something brings back sumanga'."

"Oh. Got it. So the head doesn't contain sumanga'."

"I don't know. Probably."

What I think Ambe Ka'du was trying to get across to me was that it is the ritual adventure of going headhunting that restores or "brings back" sumanga', not the traces of sumanga' substance that may rest in the trophy skull. His remarks suggest that by exercising and displaying their potency, the headhunters, like the ones who had just returned home in the middle of the night, do not lose their sumanga', but rather, add to it. I have no doubt that the young headhunters of Rantepalado were exhausted from their adventure. But their faces and laughter showed exuberant pride and lively humor. And they brought no trophy head with them.

As Needham makes clear in his critique of Kruyt and others (1976), soul-stuff theorists needed "a mysterious factor X" that enabled them to bring native thought on headhunting into line with their own sense of what causal logic must demand. Nonetheless, the line of inquiry that looks for the logic behind headtaking practices holds some appeal for Needham, inasmuch as he argues that an understanding of headhunting and its association with prosperity must involve an understanding of the "modes of causality as actually conceived in alien traditions" (1976:81). Others, like Derek Freeman (1979), place less faith in local discourse. But the familiar logic is presumed once more: taking heads leads to prosperity and promotes fertility and well-being.

Freeman's thinking is instructive, and typical. Sharply rejecting sociological approaches, he argues that the phallic and procreative symbolism of the head—a symbolism that he sees as psychologically grounded rather than culturally constructed—is the necessary basis for Iban headhunting. For him, the head equals the phallus and is a source of seed. Thus, it is "germinating heads" that provide the mysterious factor (se)x that causally links violence to fertility and prosperity. Enthralled with this kind of sexual symbolism, the poet-ethnographer Carol Rubenstein (1985:231) cannot contain what I read as a scatterbrained parody of Freeman's approach:

. . . the act of emptying out the brains of the taken head to nourish the earth of the group is the metaphor of the ancient practice. This creativity is the ultimate conclusion of the headhunting endeavor. The throwing of the brains onto the earth is the

spermatic movement so that manhood may thrive. The warrior is seen in his godlike attribute of both taker and maker of life. Taking the head with its arcane powers and implanting the stuff of the head into the earth are the elements of a mythology for the creation of man and the world itself.

It is easy to poke fun at Freeman's and Rubenstein's fertile imagination as they work at understanding Bornean headhunting traditions. To their credit, they are making an effort to come to terms with the abundant reproductive imagery of Southeast Asian headhunting practices. But their work amounts to little more than a seed-and-sex version of soul-stuff theory. It still clings to the idea that there is something in human heads that brings about the prosperity of person, community, and field.^[6] If the chants and songs of pangngae are any indication, it may be the discourse of disfiguring violence that in part connects heads and prosperity. A good example comes from the ma'paisun , the chant delivered by the babalako (or, in ToSalu villages, the topuppu )—the ritual specialist in charge of pangngae—when presenting the surrogate head to the debata. Facing east into the sun of mid-morning, he calls to the debata met on the journey downriver, chanting their names with the steady rhythm and grammar of formulaic repetition. Once the chant "arrives" at the place of ambush, it begins to dismember the victim. But moments later the throng of people at the ceremony bursts with noise and keeps the babalako from chanting further:

Here the gathering erupts with exuberant whoops and shouts as the chant disfigures the victim's head. At the same time that the crowd bursts with noise, girls and young women let loose a shower of areca nuts, being careful to pelt their favorite headhunters and lovers. Meanwhile, the babalako flings water from a small bamboo tube over his shoulder and onto the gathering. It is said that whoever and whatever is hit by the rain of water will meet with blessings from the debata.

Freeman or Rubenstein might argue that the shower of areca nuts and

water symbolizes and makes palpable the fertile substance of the head. I think not. Rather, I am convinced that the words and performance of the ma'paisun form an imaginative context in which the disfigured human head and the ritual throng are made oppositional doubles. First, the chant brings the gathering along on the journey downstream, the approach into enemy lands, the ambush, and the murder. Suspense grows, and in the end the crowd lets loose a burst of noise, areca nuts, and water. The throng cannot contain itself . And they erupt just as the chant disfigures the severed head, taking its eyes, the brains, the ears—the moment when the head cannot contain itself . For a moment, then, the communal throng is like the disfigured head of an ambushed victim. With at least one important difference: the throng survives while the victim joins the dead. The doubling of throng and head requires a discursive context—here, the ma'paisun chant—that locates a "narrating" social body upstream and a "narrated" one downstream. The chanted discourse ends in a crescendo of violence and noise, and allows the narrating social body to unleash its will in murderous negation of the narrated double downstream.^[7] The uplanders' sense and sensation of a violent, enlivening energy, and of their survivorhood, become signs of an enduring and unfaltering prosperity.

My critique, then, takes aim against a "causal logic" that puts the source (or cause) of social and material prosperity inside the head. There is nothing in a trophy skull that promotes well-being. Rather, it is doing something to a head, or with a head, that helps bring about prosperity.^[8] But mappurondo headhunters complicate the picture with another telling consideration. More often than not, villagers at Bambang speak of headhunting as the outcome of prosperity, rather than its cause. Take, for example, the lyrics to one of the sumengo:

The lyrics paint an image of increase. The uplanders prosper and the floor to their house sprouts runners, showing signs of life and growth. It is then that the headhunters look for a victim in regions downstream. No one goes on the journey of pangngae when ill or burdened with grief for deceased kin, and, indeed, the scale and enthusiasm with which this ritual adventure is run suffer when a poor harvest or epidemic takes place.

In saying that headhunting is the outcome of prosperity, no claim is being made that prosperity causes this kind of violence. Rather, the mappurondo understanding of headhunting suggests that this kind of ritual vio-

lence comprises a part and a measure of what prosperity is all about. To put it somewhat differently, headtaking is a sign of prosperity, one that indexes and pragmatically shapes well-being. Like Michelle Rosaldo (1977: 169, 1980), then, I find it more fruitful to abandon the search for causal logics and to start looking, instead, for the forms of discourse and thought that connect headtaking and prosperity.

Having listened to the commentaries and explanations of villagers, and to the sacred speech and song of pangngae, I am inclined to think that a severed head and prosperity are brought together largely through a series of speech acts and discursive exchanges. I have already suggested that the ma'paisun counterposes the disfigured human head and the celebrating social body. Yet heads have a distinct and more significant place in the vows and entreaties that men make in discourse with or about the debata. In return for providential gifts—a good harvest, unflagging health, the resources to build a new home—men and youths make silent promise to the debata to join in the ritual ambush and beheading of an enemy. Related to the moral entailments of a speech act as such, the headhunt is a way to make good on a promise with the spirits.^[9] I will underscore the significance of the headhunter's vow later in this book. For now, it is enough to recognize the trophy head as an object in the moral traffic of exchange between human and spirit worlds. After it has been obtained in fulfillment of sacred vows, the head is made to hear the desires and entreaties of the village:

Thus, the head is a reciprocating sign. With it, the headhunter fulfills a vow to the spirits. As such, the severed head is proof of the headhunter's piety and virtue. The presentation of the head to the debata not only frees the headhunter from the burden of the vow, but also marks the trophy as a gift in return for providential blessings, thereby earning the future favor and good will of the spirits.

In light of these discursive exchanges, we will do well to examine the

transit and transformation of the head, bearing in mind, of course, that even if the mappurondo communities of today do not take real heads, they nonetheless talk and sing as if they do. Imagine, then, a world in which spirits dwell far beyond the horizon "upstream" and enemies live beside the horizon "downriver." The violent acts and discursive exchanges that take place in headhunting move the head from one horizon to another. Called toibirin , "ones on the horizon," before their ambush and dismemberment, the enemies in death become sibirin debata , "sharing one horizon with the spirits." This "cosmological move" along the river is also a journey that brings about changes in the social and phenomenological status of the head. Murdering, dismembering, and disfiguring a stranger in a remote land are willful acts to dehumanize a person-victim. In felling and beheading a victim, the headhunter turns a person into a lifeless object. The ritual discourse of pangngae, however, restores to the trophy skull its humanity. Brought back to an upland village, the head becomes a guest and friend, albeit one who must suffer humiliation along with the customary gestures of hospitality and honor.^[10] What is significant here is that the language of hospitality, humor, and entreaty give back to the victim-head its subjectivity, its person-hood. The head is once more a "you."

Robert McKinley's (1976) deft comparative account of Southeast Asian headtaking practices puts together a strong case that similar notions of cosmology and personhood not only are widespread in the region, but also play the key role in giving headhunting its rationale. Using a structuralist approach and a set of ideas that recall Georg Simmel's essay on "The Stranger" (1950), McKinley argues that the head is the preeminent sign of personhood, and that the taking of a head is a way to make enemy faces and names "friendly." Taking heads so as to incorporate enemies, in his view, traces back to a need to ritually (and logically) overcome existential contradictions (and threats) posed by the social categories and the cosmological or cosmographic notions that make up local ideology.

McKinley's insights are compelling, and he has gone further than any other commentator in unraveling the mystery of "Why heads?" Yet in claiming that headhunters are busy "winning souls for humanity" through the procurement of enemy names and faces, McKinley comes very close to building theory around sociological soul-stuff. Although sympathetic to any account that allows for the existential problems of others and that can tie head-hunting to "reality maintenance," I think more has to be done to ground McKinley's work in the social and historical contingencies of a lived-in world, and in the practical effects of ritual discourse. For example, care needs to be taken to read, and thus theorize, "existential" predicaments so as not to

overlook the political and social strains that give birth to them. In particular, those identified as strangers, enemies, or other outsiders never fall under those rubrics in a timeless or transparent way. Headhunting may be not only a response to the strange humanity of the outsider, but also a practical and historically contingent way to formulate (and perhaps control) outsiderhood. Similarly, a semiotic analysis of headhunting that ignores local discourse runs the risk of missing, or even suppressing, the ambiguities and strategized understandings of ritual violence.

Headhunting, History, and Exchange

Mappurondo cosmography fits very well with the topologies of violence and personhood described by McKinley. Recognizing, however, that cosmographic ideas can shape and take shape from a lived-in ethnic and political terrain, I want to examine how pangngae has related over time to the polemics, strains, and predicaments of regional order. I will put off for later treatment the problem of how contemporary headhunts (and a contemporary history of headhunts) mediate relations between the mappurondo community and the outside world. For now, my focus will be on the headhunting rituals and the highland-lowland ethnic relations of the pre-colonial and colonial eras. My purpose is to show that headhunting practices amounted to a violent declaration of social difference, a set of acts that not only fabricated the problem of outsiderhood, but also partially resolved it. In this connection, I will suggest—if only speculatively—how trade and labor exchanges of the past provided an important context for the headhunters' violent practices and thereby lent them shape, purpose, and intelligibility. At the very least, I hope to capture a sense of how headhunting mediated shifting patterns of dominance and subordination in regional politics.

In opening this chapter, I mentioned difficulties in writing confidently about the headhunts of the past. The problems have mostly to do with the textuality of history. On one hand, any account I give is mediated by and dependent upon textual sources, be they written or oral, local or alien, that already are deeply textualized and historicized. On the other, to tell the history of mappurondo headhunting requires a plot, a story of some kind. For the case at hand, I think I can draw the two problems together. To be more specific, the story of the surrogate head, of headhunting that is not head-hunting, relates directly to the discordant claims and odd silences of competing histories.

Recall that villagers insist that their tradition of headhunting ritual is unbroken, dating back almost sixteen generations. Meanwhile, published records of the colonial administration and mission for the Salu Mambi backwater are sketchy and few, and make no mention of headhunting. The Mandar, too, are silent about headhunting raids from the uplands. Why do victims and colonial authorities fail to mention pangngae? Clearing up this mystery, it seems to me, should help clear up the dynamics of Salu Mambi headhunting, past and present.

Without wishing to exclude the possibility of colonial intervention in this headhunting tradition, I believe it useful and fitting to privilege local accounts in reconstructing the history of headhunting. Not to do so would be to deny the mappurondo community their history—yet another act of pacification. Judging from the chanted or sung discourse of pangngae itself, and from the informal commentaries of elders, the headhunts of the past played into exchange relations between this highland region and the coastal settlements of the Mandar and Bugis.^[11] At the time, the highland and lowland communities were mutually interdependent, with the mountain settlements markedly subordinate in terms of prestige, material wealth, and political power. Pangngae momentarily reversed the hierarchical relations of interdependence: headhunters would ambush a person in a coastal hamlet, take the victim's head back to the mountains, and ritually present it to the debata so as to assure the prosperity of the upstream region. Headhunting practices thus served as a dramatic expression of upland resistance toward a powerful and oftentimes threatening coast. By the late nineteenth century, however, uplanders were already using a skull-shaped surrogate in place of a severed head. If the use of a surrogate marked a crucial shift in the politics and polemics of regional exchange, it also ushered in an era of bloodless resistance to forces that threatened the highlands.

Did They Really Take Heads?

Ethnographic reports from Sulawesi make it clear that head-hunting practices were common throughout the island's mountainous interior prior to Dutch administration.^[12] In most cases, headtaking had its basis in ritual rather than in warfare or feuding as such, a pattern that suggests that claims linking headhunting to warfare and expansionism (e.g. Vayda 1969) are not applicable throughout all of insular Southeast Asia. To be sure, regional tensions and intercultural polemics played a part in shaping head-hunting traditions. Yet the reports indicate that headhunters found more powerful motivations in ritual themes and obligations that linked such vio-

lence to grief, prosperity, and the reproduction of the social body.^[13] In particular, the death of an important political figure appears to have been a key reason for holding a headtaking ceremony. In the context of ritual, these concerns made headhunting purposeful, intelligible, and right. They spelled out the headhunter's obligations and fueled a desire to kill. That headhunting was a form of sacred violence may also account for why the practice was so entrenched in the central mountain region. The post-suppression trade in skull fragments, for example, is good evidence that the rituals associated with headtaking remained crucial in local ceremonial life. Even then, actual headtaking occurred from time to time. De Jongh's report (1923) on the Karama River area, for instance, makes note of two headhunting cases adjudicated by the Native Tribunal at Mamuju in 1920.

Given what is known about the island as a whole, and about the central mountain region in particular, it would be exceptional if the territories making up Pitu Ulunna Salu did not have a tradition of ritualized headtaking. Headhunting customs were widespread and strikingly similar in character. Indeed, throughout the mountain region the term mangae , or a cognate form, meant "to go out in search of a head" (Kruyt 1942:544).^[14] Insofar as the neighboring polities in Galumpang, Seko, Rongkong, Sa'dan, and Mamasa all practiced some form of headhunting, it is very difficult to imagine that Pitu Ulunna Salu did not traffic in heads, too.

The silence of the colonial record on Salu Mambi headhunting may simply reflect a lack of familiarity with the remote hinterland. Although a post was temporarily set up and occupied at Mambi within a few years of Dutch contact, civil and military authorities by 1920 had shifted their offices 40 kilometers away, to the larger posts of Mamuju and Mamasa. These officials were chiefly interested in security and economic development, and subsequently recorded events and conditions having a bearing on those issues. Pangngae goes unnoticed in their reports. I think it safe to argue, then, that the annual rituals did not disturb the enforced calm of the period or pose any special problems for the authorities. This comes as no surprise, for elders today insist that when peaceful conditions prevail, pangngae calls for surrogate heads.

We can be sure, however, that the missionaries were ready to seek out, confront, and subvert cultural practices that would impede the conversion of local peoples (cf. Bigalke 1981 and Volkman 1985). Given the mission program, it is difficult to imagine that Christians would have refrained from taking steps to halt headtaking activity, real or symbolic. Still, I think it important to ask whether the Dutch missions largely overlooked or ignored the annual rites of pangngae.

In contrast to efforts in the Sa'dan and northern mountain regions, missionary work in the Pitu Ulunna Salu communities was extremely shallow and uneven. As discussed in the preceding chapter, the initial mission work of the Indische Protestantsche Kerk amounted to little more than mass baptisms of uncomprehending villagers. A few Ambonese teachers later opened two Christian schools in the Salu Mambi district (Atlas 1925). Some ToSalu elders told me that in one village bordering on the Mamasa Toraja region these teachers gathered up and crushed trophy skulls kept from earlier times. Whether they did so with the idea of putting a halt to ongoing ritual practices or, more generally, of erasing "pagan signs" needs to be resolved. Yet no one recalls a mission or administrative campaign to end pangngae. It was perhaps enough for these Ambonese teachers to destroy the outward signs and traces of a violent local past, and leave ritual practice undisturbed. Whatever the case, the principal mission efforts were centered on the Toraja settlements around Mamasa, the site of the new administrative and military post, and left Pitu Ulunna Salu largely unshepherded.

The situation did not change dramatically when the Christliche Gereformeered Kerk took over the evangelical field from the Indische Protestantsche Kerk in 1927. The first of the Christian Reformed missionaries, A. Bikker, worked primarily with the Mamasa Toraja. Save for a remark about a Tabulahan "headbarn" in a letter to A. C. Kruyt, I have found no references to headhunting in any of his published reports, including those that do touch upon ritual life at Pitu Ulunna Salu. Geleynse, assigned to Pitu Ulunna Salu in 1930, took very brief residence in the village of Lasodehata (currently Rantepalado), where pangngae is still performed today, but left no published reports. For these missionaries, then, pangngae does not seem to have been a critical or even noteworthy practice. Their silence is especially odd in light of the attention given to headhunting ceremonies in other regions.

It is worth mentioning once more that the headwater villages claim that pangngae has taken place annually and without disruption since the dawn of their own history, and that neither the mission nor the civil administration took direct steps to suppress the ritual. If this is the case, the failure of the Dutch to put a check to pangngae may seem especially strange, given their long-standing policy of eradicating headhunting. However, we need to bear in mind yet again the claims of today's villagers. They insist that Pitu Ulunna Salu had already stopped trafficking in real heads before the arrival of the Dutch, having reverted to a ritual search for surrogate skulls. Perhaps, then, a quieted and bloodless version of pangngae was already in place when the colonial powers began to administer the mountains.

There is thus little to guide us in determining whether "surrogate" head-hunting ever raised administrative or missionary alarm in the Salu Mambi region. But convinced that even surrogate headhunting would earn a remark or two, I think the silence of the colonial record tells us something very important: looking for headhunting, the authorities did not see pangngae . As to why, I want to suggest that pangngae masked itself, that to the few witnesses from the colonial and missionary order it did not, on the surface, appear to be headhunting. In fact, it may have appeared to them as little more than "a kind of harvest ritual," the way contemporary Christians describe this particular mappurondo rite. The custom of using surrogate heads probably did help cloak headhunting ritual from the view of the Dutch. Yet I believe the disguise was more elaborate still.

Here it is helpful to remember that other alien history: not only is pangngae missing from the commentaries of local colonial authorities, but it also has no place in the historic memory of its traditional victims—the lowland Mandar. What disguise could have helped the upland headhunter hide his violence from his victim?

The Mountains, The Coast, And Regional Exchange

The Mandar homeland stretches along the coast between Polewali and Mamuju. Devoutly Muslim, the Mandar are known as Sulawesi's finest sailors and fishermen, and for turning out some strikingly fine silk fabrics. The Mandar devotion to the pursuit of status runs deep. In my experience, siri ', the keeping of dignity and face so often ascribed to the Bugis (Andi Zainal Abidin 1983a, 1983b; Errington 1989), reaches more radical expression among rivalrous and status-conscious Mandar men.^[15] In any event, the Mandar cultural emphasis on a righteous defense of personal or family honor has impressed mountainfolk as a tripcord for violence.

According to oral and written histories, the mountain and coastal communities established regional polities in their respective areas during the late fifteenth or early sixteenth century, the period of Makassan expansion (Abd. Razak Daeng Patunru 1983; Reid 1983c; Samar and Mandadung 1979). As discussed in the preceding chapter, the upstream or "headwater" territories of Tabulahan, Aralle, Mambi, Bambang, Matangnga, Rantebulahan, and Tu'bi formed Pitu Ulunna Salu, while the downstream or "rivermouth" principalities of Balanipa, Binuang, Banggae, Pamboang, Sendana, Tapalang, and Mamuju took the name Pitu Ba'bana Binanga (see Map 4). Trade, armed

incursions by the Bugis, and other regional concerns sometimes led these twin politics to act in concert. But more often than not the uplands and lowlands looked upon one another with suspicion. Above all, these polities were consumed with their own internal rivalries and power struggles (Sutherland 1983a; Yayasan n.d.). By 1872, both the upland league and the lowland federation had collapsed (see: Smit 1937; Sutherland 1983a).

Local and regional interests notwithstanding, the Mandar were very caught up in the political and mercantile dynamics of Sulawesi and the archipelago as a whole, helping Sultan Hasanuddin, for example, resist the Dutch and Arung Palakka at Makassar, and failing subject to the trade restrictions of the Bungaya Treaty of 1667 (Amin 1963; Andi Zainal Abidin 1982; ENI 1918; Abd. Razak Daeng Patunru 1983). Although a common ancestry linked the elite houses of the lowlands and the mountain region, the former were far more interested in gaining and consolidating prestige and power through marriage to Bugis, Makassan, and off-island nobility (see DepDikBud n.d. [a-d], 1981; Abd. Razak Daeng Patunru 1983; Sinrang n.d.; Sutherland 1983a; and Yayasan n.d.). Social networks of this kind not only bolstered the prestige and authority of the Mandar rulers, but also enmeshed them in the political and commercial intrigues of the island. The same networks also promoted social stratification, the proliferation of political offices, and the Islamization of the elite in the early seventeenth century.^[16]

The Mandar economy hinged on the slave trade and on the export of upland or marine commodities. From the uplands came rattan, resin, corn, and fragrant woods, while the coast produced tortoise shell, tripang, dried and salted fish, copra, coconut, kapok, sago, oils, hides, and silk (Galestin 1936; Hoorweg 1911; Rijsdijk 1935; van Goor 1922; Veen 1933; Zeemansgijd, n.d.). Key imports were rice, salt, fabrics, weapons, opium, and ceramic goods. Although bound together as a political family under the leadership of Balanipa, the Mandar principalities were undoubtedly commercial rivals, each of which had access by rivers or mountain passes to hinterland areas that provided a flow of exportable goods.

When the Dutch patrols first entered the highlands in 1906, the communities of Pitu Ulunna Salu were economically interdependent With the Mandar coast (cf. Maurenbrecher 1947).^[17] Although the uplands remained politically autonomous, the coast held far more power and prestige in regional relations. The Mandar states probably had little interest in placing Pitu Ulunna Salu under direct rule, especially if lowland politics and maritime trade were drawing most of their attention. Furthermore, in the latter half of the eighteenth century several communities in the upstream territories of

Tu'bi, Aralle, and Mambi had begun converting to Islam. The conversion of these communities strengthened their social ties to the coast and substantially weakened the mountain league. So long as the flow of upland goods and slaves passed into their ports, and so long as Pitu Ulunna Salu remained a fragmented political body unallied to other powers, the Mandar had little reason to exhaust material and human resources in subjugating the highlands.

For all the uplanders, salt, dried fish, weapons, ceramics, and cloth (which figured as a favorite article in the transfer of somba at marriage) were the key items to be brought back in return for the rattan, corn, and resin taken downstream. Oral accounts by uplanders indicate that the mountain communities kept up a long-standing trade in forest goods and labor with partner coastal settlements. Elders from the northern mountain districts, for example, say that their ancestors traded with partner communities along the northern Mandar coast: older men at Mambi, Aralle, and Tabulahan mention that their fathers made trade journeys to Mamuju, Tapalang, and Sendana. Similarly, settlements from southern mountain districts claim ties to Balanipa and Binuang on the southern Mandar coast.

Taken altogether, the reported pattern of historic trade is enough to suggest that pairings may have existed between the seven coastal polities and the seven upland territories, such that each Mandar realm had a reliable hinterland partner with whom to exchange goods. In reality, trade patterns were probably not always so neat: seasonality in the availability of goods, competition, and feuds, to name but a few factors, could have encouraged uplanders to switch partners on occasion. Still, to take one example, it was hardly a matter of accident that Rantebulahan, the nominal leader of the upland league, should act as hinterland to Balanipa, the dominant polity of the coastal federation. In fact, the elites of the two territories had convergent genealogies.

Slavery played a critical role in exchange throughout the upstream and downstream regions. For example, it is clear that the Mandar ports had a significant role in the export of slaves to Makassar, Pare-Pare, Kalimantan, and Batavia (see: Abeyasekere 1983; Bigalke 1981; Reid 1983a, 1983b; Sutherland 1983b). In addition, the Mandar were interested in obtaining slaves (and other dependents) in order to build substantial retinues of workers (cf. Macknight 1983). The mountain communities, meanwhile, had a different interest in the slave trade. Reciprocal labor exchanges between kin and limited opportunities for agricultural expansion meant that slaves could contribute little to the upland subsistence economy, but slaves were nonetheless of value, because of the price they could fetch on the coast. Captives,

debtors, and even junior relatives were traded to the Mandar for weapons, cloth, or cash. The Mandar, like their Bugis neighbors, did raid their hinterlands for slaves, but Hoorweg (1911) and Goslings (1933 [1924]) indicate that sustained raiding rarely struck deeper than the mountain swidden communities lying just beyond the periphery of Mandar lands.

Nothing is known about the beginnings of migratory labor in the region, but by the late nineteenth century it had become an important factor in the economic interchange between uplands and coast. For example, Hoorweg (1911:103-105) reports large numbers of mountain folk making the trip to Mamuju in order to work as coolies, to assist coastal partners in "vandalism" (i.e. piracy), and to work dry-rice fields. The trek was annual and routine, and enough to provoke Hoorweg to complain that it resulted in the underdevelopment of agriculture in the mountain region (1911:104).^[18] As he described it, these migrant laborers exchanged their labor for cloth, household effects, and small amounts of money. Yet given the food shortages and sociopolitical turmoil that troubled the mountain communities from time to time, migrant labor probably allowed uplanders to get supplies of food as well as prestige goods.^[19]

Hoorweg does not indicate when the migrants usually arrived. The traditional upland agricultural calendar finds men at work on rice terraces from July through September, and October is devoted to clearing and planting gardens. Corn, another of the principal upland crops, usually goes into the ground in January, by which time the rice crop is almost ready for harvest. It would be unlikely, then, for men to journey down to the coast before late February, for to do so would run counter to their most effective subsistence strategy. (According to locals, women did not become migrant workers.) An exception, of course, would be the landless, but the local need for labor probably would be sufficient to keep them home in any case. Coastal fields, according to Hoorweg (1911:103), are planted in November. The next period in the cultivation of crops requiring intensive labor would be during harvest—about five months later, during March. It is in March, too, that the monsoon begins to swing to a westward direction and give favorable winds to Mandar sailors. In my reckoning, it seems likely that upland men would travel to the coast to find work at this time: the upland harvest would have been over, and the agricultural cycle required few labor inputs from males until July. At the same time, the coast could absorb the influx of manpower. March would also be a suitable time to carry forest products to the coast, especially if boats that had brought imported goods were readying to sail away with regional exports.

Upstream, Downstream, And Practical Cosmography

If contemporary historical views are a reliable index to the past, ideas about siblingship, social reciprocity, and cosmography must have colored upland relations with the Mandar. Recall, first, that the cosmographic perspective begins with a reading of moral and historical facts from the natural fact of a river. In the local imagination, river waters flow from their sacred sources in the skyworld, down through earthly terrain, onward to the ocean, and then, finally, to the region of the dead beneath the sea. Whatever lies closer to the source is more sacred and more authoritative than that which lies below. By the same token, upstream regions are "before" and "elder" with respect to downstream areas. Sacred knowledge and tradition, too, have their birth in the headwaters, and flow "down" through time, person, and generation. The mountain settlements of Pitu Ulunna Salu, sitting as they did beside the headwaters that eventually empty into the sea along the Mandar coast, deemed themselves guardians to the prosperity of the region. Just as the rivers flowed from the headwaters to rivermouth, so, too, did authority, tradition, and well-being.

As detailed in the preceding chapter, the genealogies of both the coast and the hinterlands ran back to the children of Pongka Padang and Torije'ne'. According to upland recitations (and consonant with local cosmography), the junior siblings inhabited the lowlands while the senior siblings populated the highlands. Uplanders thus looked upon their coastal trading partners as siblings, but claimed higher status and authority for themselves by virtue of their descent and their place in the "natural facts" of cosmography.

Cosmography and siblingship, taken as metaphor and fact, set the basic terms for reciprocity and exchange between the mountains and coast. The uplanders' claim to higher status anticipated obedience and deference from their coastal kin. By the same token, the uplands—as senior sibling—arried an obligation to support and guard their coastal juniors, by sacrifice of their own interests if necessary. Though the perspective was decidedly hierarchic, it helped frame the complementarity between upstream and downstream siblings, and served as an ideological vehicle for rendering economic exchange as the sharing of gifts and services. Complementarity and hierarchy notwithstanding, the sense of siblingship between mountain and coast also suggested that ineradicable strains and rivalries existed between the two, that envy as much as sharing could propel the relationship.

In a sense, upland ideology contained certain truths regarding exchange relations with the Mandar. Forest and swidden goods, critical to the coast's maritime trade, flowed from the uplands. The relative prosperity of the up-lands—measured by goods brought down from the hills—assured the prosperity of the coast. In short, the elder siblings were taking care of the younger siblings; the headwaters were replenishing the rivermouths. Bearing in mind that uplanders swapped their humble products for prestige goods, the "younger sibling" from the coast was bestowing respect and honor on the upstream "elder sibling." Taking a different tack, one could say that the Mandar probably were getting the better deal, and that the imbalances of economic exchange were in their favor. Yet, if upland ideology worked to mask distortions and asymmetries in exchange relations with the coast, it also served as a useful, if chiefly metaphoric, rendering of mutual interdependency.

Further still, the idiom of siblingship did not so much hide or mystify the realities of trade as act as their moral basis. If exchange is potentially dissociative and explosive in nature, then the imagined sibling bond between mountain and coast may have helped regulate the conflict engendered by trade (cf. Foster 1977). In other words, the mythically predicated idiom of siblingship, rather than the exchange of goods per se, made the trade relationship between upstream and downstream reliable. Without denying that economic interests and needs motivated regional trade, the idiom of sibling-ship became the moral context in which exchange could succeed (or, for that matter, fail).

There is little evidence that the Mandar viewed their exchange relations with the Pitu Ulunna Salu uplands in the same ideological terms; indeed, some claim that the coast is "senior" to the upstream regions (e.g. Syah 1980:18). But as long as regional trade patterns conformed to, or were amenable to, the upland interpretation, contradictory or divergent Mandar attitudes may have posed few challenges to upland thought. It appears that in their several centuries of exchange with an alien and potentially threatening coast, the mountain communities never scrapped the idiom of sibling-ship as their interpretive perspective on trade until the Bugis established upland markets in 1925.

Going To The Sea

The highlanders all agree that their ancestors sought Mandar victims when hunting heads for pangngae. Indeed, the discourse of contem-

porary ritual continues to portray the headhunters' victim as Mandar. For example, the antiphonal singing called the ma'denna taunts the head to call to the Mandar nobility for help:

Other songs tell of the panic among Mandar hamlets, the churning of their rivers, and the spilling of their seawater. And the babalako , as presiding specialist, honors the head as a "grandchild" of the Mandar lord, Daeng Maressa. I will come back to these songs and look at them in more detail later on. For now, I want to point out that pangngae is a time for the uplanders to prey upon their coastal trading partners.

Songs and chants of this kind, of course, do not literally kill. But the images of violence that they call forth are direct and unmistakable. The ambush of a trading partner—actual or imagined—also tips us off to the way in which headhunting was disguised. The masking of pangngae at the time of Dutch contact had not only to do with the fact that headhunters sought heads that were not heads, but also with the fact that the quest was concealed in the ideology and practice of trade. Trade and ritual violence happened within one another's shadow, and headhunting thus dissolved before the eyes of colonial observer and coastal "victim" alike. The masking of pangngae is now deeply entrenched in local ritual tradition, even if the Mandar are no longer the principal trading partners for the uplanders. But contemporary evidence still affords a chance to understand why headhunting may have become concealed in this way.

Some telling data lies in the way villagers mention trade and headhunting. Le'ba' le'bo ', literally, "going to the sea," is the term used for trade journeys and seeking adventure in other lands, and villagers still use the phrase as the most popular euphemism for the headtaking journey of pangngae. In a sense, the term collapses distinctions between headhunting and trade, subsuming them under a broader conceptual category having to do with journeys. Yet, following lames Scott (1990), it is worthwhile, too, to recognize

that the euphemism is a discursive tactic intended to obscure, to disguise. The pragmatic ambiguity of the term lets the headhunter make a threat or disavow violence.

Equally significant is the ma'paisun chant of the babalako discussed earlier. Calling out the dwelling places of the debata who live along the headhunters' trail, the specialist is in effect marking a route to the sea. These chants trace but one trail and name only one coastal settlement where a victim may be killed. Thus, the warriors from the ToSalu communities of Rantebulahan and Bambang travel (in chant) to Tenggelan in the Mandar principality of Balanipa; those from Mambi head off to Tanisi below Tapalang; the ToIssilita' villages of Bambang go to Abo in Sendana; and those from Tabulahan trek to Mamuju. In addition, the returning warriors typically identify the home of their victim when, arriving at the edge of the hamlet lands, they respond to the ritualized greeting of a village elder:

These invocations and greetings bear out the claims of today's elders that cohorts of upland men returned year after year to the same Mandar settlement to take a head. Indeed, these ritual genres imply that the mountain communities tacitly coordinated their raids in such a way as to make sure that headhunters from Rantebulahan, for example, would not strike at Mamuju, the trading partner of Tabulahan.

Contemporary practices also suggest that headhunting "season" may have coincided with the uplanders' annual trade journeys to the coast.^[20] Pangngae takes place once the harvest rice has been put away in the barns, usually between late February and early April. For warriors to leave their villages at this time on the 40-90-kilometer journey to Mandar territories means that they would appear on the coast right when agriculture and commerce (marked respectively by the March harvest and the shifting monsoon) could absorb manpower. Returning from a headhunt, warriors invariably carried what they claimed to be stolen cloth, weapons, and porcelain. Even today, theft has a place in headhunting ritual: along with the surrogate head, a

small pile of goods—said to be stolen from the victim—is presented to the spirits.

Coinciding with a critical moment in the labor and trade calendar on the coast, terminologically fused with trade journeys, and targeted at reliable coastal partners and patrons, pangngae must have figured importantly in regional exchange networks. Claiming to be out on a headhunt, upland men were actually bartering and laboring on the coast. Or perhaps more accurately, while men were headhunting, they were also trading goods and labor.

It should be kept in mind that the uplanders of today insist that there were points in history when Mandar victims were indeed slain for their heads, just as they insist that current tradition prohibits the taking of real heads for the purposes of pangngae. Nearly all claim that the reason for taking heads from the Mandar had to do with feuding and retribution. As one ToIssilita' elder explained it: "It was to strike back. Our grandfathers would go to the sea [to trade or work] and someone would be killed or made slave."

A youth from a different village had a different story: "At first, no one took a head. Then slavery and war appeared. Then people killed one another and taking heads became custom."

However true explanations such as these may be, they divorce head-taking from its customary ritual frame, and link it to patterns of reciprocal violence. In other words, the accounts shift the moral perspective on head-hunting from the vantage point of ritual obligations to the desire for revenge in the endless back-and-forth of rebounding violence.

Papa Ati, my language assistant, gave me a broader historical account of headhunting practices. As he had heard it from his father, pangngae at first called for a surrogate head and stolen goods to be offered up to the spirits. The custom then underwent change in the time of kende' tats asu ("crazy dog rises"), when slavery, wars, and feuding took place; hunting for real heads became the rule. Once peaceful relations were restored between the coast and the uplands, headhunters sought out a kind of tuber, tullu bulam ("egg of the moon"), as a surrogate head. More recently, the coconut, a product associated with the coast, has become the customary surrogate. This account keeps headhunting within the moral frame of ritual, but also indicates that regional strife was a trip mechanism whereby symbolic dramatizations of headtaking were "rescripted" to include very real killings.

I pressed several elders about the matter of actual killings and how the Mandar would respond. According to most, the Mandar knew full well when the raids and ambushes were to take place—more than 80 mountain settlements would be sending out cohorts of headhunters during a six- to

eight-week period—but did not want to disturb a practice upon which regional prosperity depended. The Mandar elite were even said to have designated specific settlements where the uplanders could take heads.^[21] A Christian commentator of upland birth provides a similar explanation (Makatonan 1985:105), saying that the Mandar helped the uplanders seek a head because they, too, needed a human victim for their own ritual sacrifices. According to these upland formulations, the taking of real heads did not disrupt trade, especially in a time of violence, vendetta, and enslavement. In fact, such explanations presume that headhunting was in the interest of the victim's survivors, a viewpoint expressed in the ritual entreaties to the head (see the chant above, p. 67).

It is hard to imagine the Mandar going along with this. In my acquaintance with the coastal dwellers, I saw only a few signs that the Mandar associate upstream regions with greater prestige and authority and with sources of prosperity. In fact, the Mandar of my acquaintance had difficulty imagining their ancestors surrendering a head to the people of the headwaters under any circumstances. One acquaintance, for example, listened to my speculations about the uplanders' headhunting raids with some amusement. "That must have been long ago," he snorted, "and if they had tried that, we would have just cut them dead."

Still, actual headhunting may not have significantly disrupted regional order, an order that by its nature was probably given over more to turmoil than to calm. Judging by the upland accounts that found "reasonable cause" for taking real heads in regional patterns of vengeance and slavery, head-hunting was merely symptomatic of intercultural tension. Still, the same ac-counts—ignoring as they do, the ritual framework of pangngae—do not at the start provide "reasonable cause" for surrogate headtaking.

If headhunting rites gained purposefulness and intelligibility in light of ideas about fertility, grief, masculinity., and village political life, they nonetheless put down roots in the volatile and ambivalent social relations of exchange between upstream and downstream. It follows, then, that the idioms that gave regional trade its moral context may also be the terms that provided a moral basis for headhunting. In other words, ideas about sibling-ship, reciprocity, and cosmography may offer important clues about why uplanders wanted to take the heads of Mandar victims.

Here, a story about the origins of headhunting told to me by Ambe Teppu may throw light upon the moral dynamics of the pangngae. As he told it, there were two brothers, the younger of which went to live on the coast. Some time later, the younger brother on the coast beckoned to his elder brother to come down from the headwaters. The younger one asked

for help because everything had ceased growing in the lowlands. The elder brother pledged to help by appealing to the spirits at the headwaters, but told his sibling that he needed gifts that could make an appropriate offering. Thinking the matter over that night, the younger brother told his wife to clear out the house, leaving only tattered cloth and other damaged goods. When his brother came again the next day, the younger one told him to take everything that was in the house. The elder one carried the meager effects back to the headwaters and offered the prizes up to the spirits. In the meantime, the younger brother made offerings downstream. The spirits were pleased and restored prosperity to the coast. Thereafter, the elder came back yearly, took something from his younger sibling, and in this way guaranteed the prosperity of the coast. But the younger brother remained greedy and hoarded fine belongings to himself. Unwilling to share the fruits of prosperity, he continued to leave coarse or damaged goods for the elder brother. In time, descendants of the younger brother began to enslave the people of the headwaters. Slaves died in sacrifices to the spirits, and people began to kill one another upriver and downriver. It was then that Mandar heads were taken and presented to the spirits at the headwaters.

The narrative no doubt collapses myth and history. Yet it is not so important to separate the elements of myth and history as it is to grasp the moral drama that fuses them together. The key to the story is a violation of the moral relationship that binds siblings to one another. The elder sibling is obliged to help his junior brother and does so. The younger one, meanwhile, owes his senior brother respect and honor. Yet the junior brother deceives his elder sibling, hoarding valuable goods to himself, and thus withholds the signs (and currency) of honor and prestige. Later on, descendants of the junior sibling enslave their "senior brothers" at the headwaters. Duped, humiliated, and subordinated by his junior, the enraged senior brother begins to take heads in reprisal.

Although the story shows how the exchange of goods (and, later, of slaves and heads) promotes prosperity and cosmological balance, the scenario is one of ambivalent, inverted brotherhood. Trade is the correct and desired expression of exchange, but with the inversion of the brothers-partners relationship upsetting the moral order, reciprocity must become manifest as the ritualized killing and looting of pangngae. In this sense, pangngae is a special variation on patterns noted by McKinley in his analysis of Southeast Asian headhunting (1976). He finds headhunting to be the means for making distant enemies friendly within the exchange of souls critical to cosmological balance. Though headhunters in the Pitu Ulunna Salu highlands also seek cosmological balance, they do so through the exchange of goods and

by taking a distant junior "brother-gone-bad" and turning him into a proper junior sibling again by resubordinating him through fatal ambush and placating ceremony.

Two sumengo from the ritual headhunt shed further light on the inverted moral order of the upstream-downstream exchange system. They concern prosperity, labor, and patronage. The first is one I have already mentioned in discussing heads and fertility. It is sung in warning to the Mandar:

The lyrics warn of the headhunters' prosperity and power. But the following sumengo may be sung in reply:

For the uplanders, one works for another under the terms of saro : within the context of saro, a person works for the sake of a relative or friend at the request of the latter, generally receiving a meal and a token share of the harvest (or wage) in the course of the task. The friends or relatives are obliged in turn to come work for the person who lent them a hand, and they, too, will receive a meal and shares. Failure to reciprocate does not indebt the friend or relative, but amounts to the arrogant claim that one's status is too great to permit one to work for another. In certain situations, this would offend the person who once lent aid. But to work for another person without being able to call for labor in return marks a person as a social inferior. When this happens—because of meager landholdings, for example—one is not affronted, but humiliated by one's own circumstances.

The song of reply presents the latter case. The uplanders are made laborers for the coastal powers without hope of calling them to the hills to reciprocate under saro arrangements. The uplanders get handouts of food (tena ) for their work, but have no opportunity to give food to the Mandar in their fields back home in the mountains. These humiliating conditions provoke the stereotypic expression of futility heard in the last line of the song. The uplanders' chance to regain prestige and honor "fails off into the (river) waters" and is washed away. With it goes any real chance of restoring the relationship between senior and junior brothers to the moral ideal.

If the upland villagers resigned themselves to the fact that the Mandar had the greater means of power and prestige, they remained certain of their own place along the moral terrain that stretched across ideas about cosmography, siblingship, and reciprocity. Pangngae was less a facade for trading ventures than a complex moral statement or declaration that temporarily negated a more persistent skewing of correct relations between mountain and coast. The uplanders were willing, perhaps compelled, to deliver their gifts of labor and forest products to the coast, but the humbling realities of regional power and commerce gave them little in return. The symbolic killing and looting of pangngae, then, countered trade and labor exchange, and momentarily righted the conditions of upland interdependency with the Mandar. The people of the headwaters could thus recover their virtue and slake their envy by turning their partner-brothers into victims. Pangngae gave the uplanders a sense of survivorhood, and with it a taste of power and control not ordinarily theirs. The irony is that the uplanders' survival, in a material sense, did not depend on ritualized headtaking, but upon their acquiescence in regional trade and labor.

Seen from the vantage point of some of the contemporary mappurondo narratives and ritual practices, pangngae grapples with the incongruity between ideology and circumstance, much in the way that Jonathan Smith (1978, 1982) claims is characteristic of all ritual. In a sense, pangngae was—and to a degree still is—a way to enact or present an ideal vision of exchange relations that could be recalled and put into conscious tension against the real order of things. If the facts of regional exchange humbled the uplanders, pangngae was a counterfact that momentarily restored their virtue, authority, and status as seniors. The ritualized theft of heads and goods from the coast tilted the social relations of exchange in a direction most favorable to the mountain communities. The seizure of goods was a way to "make them ours"—incorporating the valuables without giving anything in return. At the same time, taking a head was a way to make the victim "ours"—a way to subjugate and incorporate the rival junior who was at once trading partner, sibling, and enemy.

Postscript, 1906-1985

Surrogate headhunting is not the postscript to "the real thing" and to the related terrors of the precolonial past. Listening to elders from the mappurondo communities of today, "taking something else"—a surrogate skull, tattered cloth, coins, porcelain, salt, or goods to be used in

somba wedding exchanges—has been at the heart of pangngae since its origin. When the Dutch stepped into mountain history in 1906, pangngae was already euphemized as le'ba' le'bo ' ("going to the sea"), and thus tied to the ongoing regional exchange of gifts, labor, commodities, and slaves. Cloaked by various kinds of exchange between upland and coast, momentarily "scripted" for tuber or coconut surrogates, and staged in a language unknown to Europeans, pangngae showed few or no signs of violence to the small number of Dutch figures posted in the headwater region. Let me mention once more, then, two paths that the colonial response may have taken: first, the Dutch simply had not seen pangnage, or at least had failed to recognize the rites as headhunting ceremonies; or second, they recognized pangngae as a form of headhunting but were untroubled because of its lack of explicit violence.

The facts remain in doubt, of course. But I think a strong case can be made that the colonial order, for all its power and intrusiveness in exercising dominion over the mountain region, did not bring an end to pangngae. It is worth further underscoring that the colonial and missionary presence did not bring about the use of surrogate heads. Rather, surrogacy was already included as an originary feature of pangngae, and its currency at the time of colonial contact came as the result of regional politics, economics, and social relations. The interdependence of mountain and coast—an interdependence, it will be remembered, that led an observer to call the uplanders "Coastal Toraja"—was one in which the Mandar dominated. Considering that the taking of real heads might invite reprisal from a powerful adversary and partner, it seems reasonable that the uplanders would resort to effigies in order to practice their ritualized arts of resistance.

What the Dutch and subsequent Indonesian administrations did do was eliminate the conditions and circumstances under which villagers might "rescript" pangngae once again for the taking of human heads. Since 1925, the building of mountain roads, the spread of upland markets, the expansion of cash cropping, and the growth of a Bugis-dominated trading class have changed the patterns of regional exchange and, with them, the context for headhunting journeys. By 1975, mappurondo villagers from Bambang had more or less ceased making trade journeys to the coast, relying instead on the markets of Mambi and Mamasa. Today pangngae takes place only among the mappurondo communities at Bambang, along the upper reaches of the Salu Mokanan in Rantebulahan, and in the more northerly district of Bun-tumalangka'. In Mambi proper, the ritual is well within the memory of elders, having come to a halt several years before the Japanese occupation

(1942-45). In the meantime, it lingered until 1970 in Mambi's outlying settlements, usually as a ceremonial meal and (since about 1950) without the effigies and reenactments.

The mappurondo communities continue to face dilemmas posed by the recontextualization of ritual life. The headhunting that belonged to the violence, adventure, and politics of the past cannot be the headhunting of a censorious and socially fragmented present. As the patterns of precolonial upland-lowland exchange recede further into the past—patterns in which headhunting made sense and upon which headhunting made comment—I cannot help thinking of pangngae as a ceremony troubled with contradictions and incongruities, a reflexive metacommentary, a ritual about head-hunting ritual. Men and youths no longer undertake an actual foray to the sea when off on the headhunt of pangngae. In fact, the journeys out of the village, once measured in months,^[22] now take place within a few days' or few hours' time. Meanwhile, an increasing number of young boys have joined the cohorts of warriors over the years as the dangers associated with travel have become less threatening. No one carries any weapons save a work machete. In the modern order of things, these journeys bear little resemblance to those depicted in song and dramatic dance. Not surprisingly, the contradictions and incongruities between what is said to take place and what actually does take place in pangngae have grown, promoting a sense of loss and irony among villagers.

Yet ritualized representations of headtaking are still the crucial and lasting idiom of pangnage. Two years into my work I turned to my language assistant, Papa Ati, and told him of some doubts, "Maybe I am wrong in thinking that taking heads is the important part of pangngae. Maybe it is something else."

He held up his hand and signaled me to stop. "No, you're not wrong. It's like this. If you discuss pangngae, you discuss taking heads. That is the point—taking heads."

Perhaps talking about headhunting is a way to talk about a small community's efforts in resisting social and cultural threats from without. While there is some truth to that, I have to note that as a ritual of resistance, pangngae remains significantly lodged in the volatile and ambivalent relations of the past. An explicitly anti-Dutch, anti-Indonesian, anti-Muslim, or anti-Christian counterposture is next to absent in the discourse of contemporary rites. The ritual still pits the headwaters against the rivermouth, and in so doing makes a violent and local formulation of social differences. But those differences, forged in ritual ambush and murder, already see the

"victim-stranger" as someone belonging to the headhunters' community, yet whose membership involves a being outside and against the group (cf. Simmel 1950:402-403). In this sense, the headhunter's "other" is always a partner, a sibling, a person just downriver. Within the discourse of pangngae, the Dutch and the Indonesian, the Muslim and the Christian, barely exist at all, except, perhaps, as shadowy figures of an allegorized headhunt. Yet in a more complicated and indeterminable way, the absence of these figures may mark either a very strong political challenge or a profound mystification of the world, namely, a (ceremonial) refusal to acknowledge the relations of interdependency that tie the mappurondo community to powerful social and historical formations.

The "making of an other" in pangngae, meanwhile, calls for a ritual object capable of holding the imagination and mooring the ceremonial practices of the community. The traffic of goods in regional trade offered a variety of objects that found their way into ceremonial discourse. But the prize, of course, came by way of murder and theft—a stolen, disfigured head.

Terror, Laughter, and the Grotesque

Myriad themes and issues come together in pangngae: grief and mourning, envy, eroticism, fertility, and ideas about masculinity, for example, routinely surface and intermingle in the discourse of the ritual head-hunt. And they do so in complex ways. In the tangle of ritual discourse, the existential and sociohistorical problem of "making an other"—one of the principal and formative strains in pangngae—links with these other themes by way of the imagery of the severed head. For this to happen, the head must possess the character (and do the work) of the grotesque.^[23] Lake other kinds of grotesque images, the severed head can wed awe, terror, and violence to a ludic fascination with bodies, fertility, and ways of transcending death.^[24]

First, the making of the other. Treating the problem of the other as a structural and existential feature of social life, McKinley (1976) argues that the trophy head is the most apt symbol for ceremonies like pangngae because it contains the rice, the most concrete image of social personhood. The ritual incorporates the face of the stranger into the headhunters' community, and thus "restores the stranger's misplaced humanity" (McKinley 1976:124). I would add that such a "rice-to-rice" encounter restores the

possibility of speaking to others, thus acknowledging "strangers" and making them accessible within the social give-and-take of dialogue (cf. Gusdorf 1965; Natanson 1970:34-36). At the same time, bringing this alien other within the headhunters' community becomes evidence for the stability and rightness of local power and tradition.

As helpful as these arguments may be, they gloss over or ignore the terrifying violence of the headhunter. Not only does the headhunter kill, but he dismembers. The headhunter's violence, in an important sense, ends in the objectification of "the other." But that violence also involves the mortification and debasement of the fallen. The ritualized project of violence that requires the taking of a life and a head is thus one that produces a demonized and degraded other. Further, it surrounds the degradation of the fallen in an atmosphere of festivity and laughter.

Here is where a discussion of the grotesque may shed light on headhunting ritual. By "the grotesque" I mean kinds of imagery and discourse that foreground the monstrous, the fantastic and bizarre, the exaggerated, the incongruous, and the terrifying or awe-provoking. The aesthetic and political aspects of the grotesque are by no means stable and uniform, but appear as particularized cultural and historical modes of discourse (cf. Bakhtin 1984, Stallybrass and White 1986). Nonetheless, the interplay of terror and laughter seems central to any notion of the grotesque, no matter whether it is an alien or comic world that is imagined. And to twist and disfigure a statement by Flannery O'Connor, it is in the language and imagery of grotesque that "man is forced to meet the extremes of his own nature."^[25]

Taking a head in ambush is a marked, rather than unmarked, form of violence.^[26] In the Salu Mambi region, it is recalled not as a feature of warfare and feud, but as a ritual act only. The ritual of pangngae serves, then, to mark the taking of the head as an extraordinary kind of terror, constrained with respect to time, place, purpose, and person. Already marked in this way, the violence takes a grotesque turn with the dismemberment and deformation of the victim. As such, the violence of pangngae is marked not only by its ritual dimensions but also by an anatomizing fascination with the body.

How grotesque violence figured into the ritual discourse of the past is uncertain. The contemporary language of pangngae, however, shows a recurrent interest in the grotesque and lurid. For example, the theme of dismemberment figures importantly in the ma'paisun discussed earlier in this chapter. The babalako chants out the protocol of mutilation—slashing the neck, taking the head, removing the eyes, removing the brain, removing the ears, and (in a 1985 recitation) removing the tongue. The dismembering

and debasing of the victim's body also appear in the sumengo, but often are euphemized as the felling of a tree. In contrast to the precise mutilations enunciated in the ma'paisun, sumengo lyrics are more likely to deflect the imagination toward the "scene" of dismemberment. For instance,

Both songs make a topographic move, placing the scene of dismemberment along a river. The lyrics in the first counterpose place and action through the image of branching. The confluence of two rivers, or, rather, where a river branches into two lesser ones, serves as the spot for dismembering and beheading a victim. The verb baratoi means to cut off limbs, ribs, or branches, things that protrude in number from a vertical "spine." Thus, the spot where the river branches ironically serves as the place for the "de-branching" of the victim. As for the lyrics that follow in the next song, the trampled ferns connote the excited rush to deliver the first slash. The lyrics direct the gaze of the listeners' imagination first to the ground, then upward to the wounded neck, held aloft as a gesture of exuberant dominance.

The discourse of the headhunt reveals, too, fascination or disgust with rot and decay. One song, for example, depicts the body of the victim as a fallen palm left to be consumed by a swarm of maggots and grubs:

Slain and abandoned to rot in a makeshift grave, the headhunter's adversary lies abused and debased. The most explicit gesture of revulsion often occurs, however, when the trophy head first is brought into the village. As the prize is raised up for all to see, the crowd of villagers turn their heads and shout Bossi'! Bossi '! "It stinks! It stinks!"

The grotesque imagery of dismemberment and rot renders the humiliation of the fallen in ways that stir curiosity, awe, and revulsion. The violent deformation of the victim's body also should be read as the disintegration and debasement of an oppositional or adversarial community. The blows delivered to the body of the fallen are blows to the Mandar community. Themes

of noise and panic only underscore the disintegration of the rivermouth communities. Recall the opening lines of the ma'denna :

mountains banter, talk back and forth
peaks taunt, lash out at each other
cockerels frighten the hamlets
whom do they throw into a panic
it's the head that they throw into a panic

Terror and anguish over the destruction of home and settlement echo in the lyrics of many sumengo, too. For example,

Finally, the highland throngs who sing of this terror abuse their fallen adversaries with laughter. Umpaningoi ulunna , "playing with the head," is the penultimate ceremony of pangngae, staged the night before the feast of the cohort and the presentation of the head to the debata. During this ceremony, the community has the chance to heap honor and humiliating abuse on the enemy head. As is typical of headhunting rituals elsewhere in island Southeast Asia, the head sits as an honored guest in the home of the tomatuatonda' (the village head), or in the home of the cohort leader, and is presented with gifts of tobacco and betel quid. The babalako or topuppu will chant soothingly to the head, mixing pity and hospitality (see Figure 4):

For you this grandchild of Daeng Maressa
the one who is seated below at the rivermouth
. . . don't be red of face
. . . don't ooze blood from your throat . . .
savor your betel with a relaxed look
smoke your tobacco with contented breath
soon you will arrive there at your place of gold

Fig. 4.
Ambe Sope, the topuppu (the one who presides at ritual festivities) in one of the
mappurondo communities, offers areca, betel, lime, and tobacco to the surrogate head,
here perched on a makeshift centerpost and swathed in a headcloth. After honoring
their victim as a guest, villagers will taunt the head with song. Ambe Sope's face is
caked with rice paste, intended to lighten and refine his complexion for the final cere-
monies of pangngae. 1984.

But as the evening wears on, the crowd will taunt the head with the caustic verses of the ma'denna. And from time to time someone will still shout out Bossi! Bossi'! and thereby provoke laughter from the gathering. Some ToSalu remarked to me that in the past, village women would occasionally dance in bawdy fashion with the head.^[27] Things are tamer now, and the laughter of pangngae appears to have lost some of the derisiveness it may have had in the past. Yet laughter endures as a mark of survivorhood, power, and the capacity to surmount and control terror.^[28]

As in medieval carnival (Bakhtin 1984) or the feasts of Saturnalia (Nagy 1990), the grotesque imagery of pangngae serves a discourse of regeneration and fertility. The dismemberment and rot of the headhunter's other becomes the source of renewal and rejuvenation for the upland communities. Put somewhat differently, grotesque imagery renders the political reversal of a dominant adversarial community as the earth from which the upland villages prosper. At the same time, the exaggerated and anatomizing violence of the headhunt evokes astonishment and awe among the uplanders. Taking a head is an extreme and marked act of terror, intended not only to frighten and humiliate an adversarial community, but also to stir dread and awe among those in the hills, and thereby allow them troubled recognition of their own violence. In the theater of pangngae, it is not enough to show stolen cloth and weapons as a sign of domination over the coast. It takes a disfigured head. For in gazing at the head, the uplander does face the extremes of his own nature: that alien other downstream and his own terrifying violence.

Coconut Relics

The language and imagery of the grotesque in Bakhtin's (1984) formulation is emphatically political, a form of undominated discourse that resists all regimes in their official exercise of talk and power.^[29] Is this the case in pangngae? On one hand, yes; it presents counterfacts to the regional exercise of Bugis and Mandar hegemony. On the other hand, it is a mistake to forget that the grotesque imagery of pangngae lies at the heart of traditional upland "official" discourse and power. "Played straight," pangngae is not a festive or carnivalesque critique of mappurondo society.

The movement back and forth between human heads and surrogate skulls marked crucial shifts in the political relations between the mountain and coastal communities. Judging from the remarks of mappurondo elders,

using a surrogate skull was already common practice by the time the colonial administration set out to eradicate headhunting in Sulawesi, and has remained so up to the present. As to the specific events and conditions that led uplanders to resume what is now a century-long traffic in surrogate heads, virtually nothing is known or remembered. Still, it is reasonable to assume that the idea of taking tubers or coconuts in place of severed heads was but part of a new, bloodless resistance to forces that threatened to dominate the highlands.

It is no accident that a coconut has become the favored surrogate. The formal similarities between a coconut and a skull—something obvious to both headhunter and ethnographer—speak for themselves. More significant, I think, is the fact that the coconut is a product associated with the coast. The palms do not thrive in the highlands, and as a result the coconut is considered a prestige food. It seldom appears in daily cuisine (save among Muslims—persons, it should be noted, who practice a "downstream" religion), but finds key use as a gift to a household hosting a wedding or prosperity rite. In the context of pangngae, however, the trophy coconut is never consumed or later given away. It remains a trophy-offering from the coast. As such, the coconut continues to do the work of the severed head, signifying and restoring the ideal political and cosmological relations linking upstream and downstream.

No one mistakes the surrogate for anything other than what it is—a stage prop—just as no one misreads its blatant symbolism. Forsaking the ambush of a coastal victim for the purchase of a coconut in an upland market, headhunters have made concession to political realities that have brought them powerlessness and despair. Part of that concession probably involved reformulating their historical understanding of regional exchange. In this context, the stories that portray real beheadings as an unwelcome, but nonetheless just, transformation of innocent surrogacy give the uplanders the moral high ground in the region's past.^[30] Still, many of today's villagers show a considerable uneasiness when it comes to the violent representations of pangngae. On one hand, the ritual lets them discover and reflect upon the heroic virtues and sentiments necessary for the preservation of village polity and tradition. On the other hand, a real beheading would fill most villagers with dread. The sacred obligations of tradition, as presently interpreted, leave them trapped. Headhunting is terribly right and terribly wrong: it is good to think and bad to do.

It is worth mentioning that contemporary ritual practice suggests a symbolic avoidance or masking of this dilemma. When the warriors return to the village, they jubilantly hold aloft the trophy bag that contains the coconut.

Fig. 5.
Ambe Sope in the upstream, gable-end loft of a house, preparing to present the
head to the debata (spirits). The head hangs enclosed in a dark woven pouch (center).
Beside it hangs a rack for offerings. and above it rest the tambolâ flutes frown a prior
headhunting ritual. 1984.

From the time the trophy is brought back to the village to the time it is placed as an offering to the spirits, out of view, in the eastern (or upstream) loft of an elder's house, it never comes out of the bag. The bag affords no glimpse of the prize within. It conceals what the trophy is—a coconut obtained through purchase or barter—as well as what it is not—a disfigured face stolen in ambush. In this way, no one can gaze upon the awful truth of the headhunters' violence or their humiliation in reciprocal relations with distant lands.

The violent and grotesque imagery of ritual headhunting is, among other things, caught up in the problem of outsiderhood and the historical shifts in regional patterns of dominance and subordination. Showing signs of violence has been the headhunter's way of making a practical cultural response to the threatening power of the outsider and the disorderly circumstances of exchange. As rites of reversal, current stagings of pangngae symbolically invert the social order of the region's past. Whether the predatory journey into the past evidences a displacement of contemporary tensions is difficult to confirm. The headhunt could arguably be a commemorative eruption of a sublimated history. These questions notwithstanding, the effort to disguise violent intent—journeying in the past, carrying no weapons, displaying no trophy heads—suggests the infrapolitics typical of subordinate groups (Scott 1990). In this sense, pangngae is a form of undominated discourse—contributing to, rather than substituting for, a practical program of resistance. But as subsequent chapters will show, headhunting celebrations are also very much a form of official discourse, a key to legitimizing mappurondo social order in each mountain settlement.

As Stallybrass and White (1986:26) remind us, "transgressions and the attempt to control them obsessively return to somatic symbols." Dialectical polarities—inside:outside, high:low, life:death, pure:polluting, us:others—are familiar markers on the social and symbolic terrain of pangngae. As the ritual's dominant somatic symbol—begging both reverence and disgust—the severed head works as an icon of the taxonomic or categorical violence done to the enemy's "body politic." In this light, the grotesque image of the severed head serves the social and political purpose of demonizing an adversarial community. Even in its momentary return from the status of "object" to re-humanized subject, the head suffers the humiliating "blows" of hospitality and laughter. Although the head elicits expressions of disgust, it also stirs awe. The relic of a marked and unusual violence, the head has the capacity to astonish, to exalt. In this sense, the severed head is a grotesque guise for the sublime.

The ambiguities of taking a head—real or imagined—have, I suspect, the potential to trouble the headhunter. After all, pangngae calls for the headhunter to meet the extremes of his own nature. In face-to-face encounter with his alien double, he must confront his own capacity for terror, violence, and laughter. Yet it must remind him of his subordinate place in the world as well. Since no later than the close of the nineteenth century, the trophy head has been masked and withheld from public gaze. The masking of the stolen head paradoxically announces and suppresses its own history. Surrogation, euphemism, and head-bag declare the humiliation of the upland communities under Mandar and Bugis hegemony. But they also hide a history of "real" headhunting from view, leaving the past uncertain.^[31]

Ironically, the social changes of the twentieth century have brought the mappurondo communities to a point where they recognize that the head is now a sign of their own outsiderhood, their own oppositional stance vis-à-vis Indonesian and global orders. In the current social and ideological climate, headhunting does not resolve the problem of the outsider but inverts and exacerbates it. Yet local moral tradition requires this ritualized gesture of violence from them.

As for a century of ethnographic speculation on the logic and meaning of a severed head, it appears to be based on a misrecognition of the social and political dimensions of ritual violence. Commentators have been hypnotized by the grotesque, forgetting that it is a mode of political discourse, and have enshrined the head as a fetishized product of pagan depravity and primitive thought. If the mappurondo case provides a corrective lesson, the connection between severed heads and fertility had nothing to do with a logic of seed and soul-stuff, but, rather, with the mortification of the rival, an act that served as the very measure of the headhunters' prosperity. The political turnabout envisioned in pangngae underscored the providential favor shown the headhunters' communities when the world is restored to the cosmographic ideal. As one song puts it:

However fleeting or illusory the turnabout may have been as a political ideal, the headhunter's violence compelled those downriver to subserve upland prosperity.

4
Violence, Solace, Vows, Noise, and Song Ritual Headhunting and the Community in Mourning

Common to nearly all Southeast Asian headhunting traditions is a practical discourse that ties the remorseless felling of a victim to the cessation of personal and collective mourning. In most cases, taking a life in raid or ambush becomes an act through which a headhunter can put aside the emblems of grief carried by himself and others. Pangngae is not very different from these other traditions, for it, too, connects bereavement and violence. Although the scale and enthusiasm with which the ritual is run may vary, villagers conduct the headhunt with the idea of bringing public mourning to an end. For example, a term used throughout Bambang to describe the tenor of pangngae is maringngangi '. The word means to "ease," "lighten a burden," or "relieve," and conveys the way in which headhunting ceremonies dispel grief (kabarataan ). In fact, the ToIssilita' of Minanga, Saludengen, and Rantepalado use maringngangi ' as a common name for their headhunting practices. No less telling are two ToSalu terms for the headhunt, untungka'i botto , "opening the hamlet" (i.e. from prohibitions of mourning) and untauam bahata , "being persons who mourn."^[1]

From my vantage point as an outsider, killing another human being for the purpose of bringing mourning to an end raised complex and disturbing questions about what people in headhunting communities think and feel about death, violence, and loss. Mappurondo villagers, by contrast, seemed less troubled by the moral and interpretive dilemmas that the violence of

headhunting posed for me. Those I knew well showed great conscience and sympathy in everyday affairs, and saw murder as wrongful and alarming. For them, any murder within the headwater communities was a sign of treachery and discord, and threatened rebounding violence. Fortunately, the lessons and constraints of ada' tuo (the "adat of life" discussed in Chapter 2) have worked to keep violence subdued and comparatively rare. In fact, the uplanders in my acquaintance regard violence and murderousness as attributes of the very lowland communities from which they claim to take trophy skulls. The ethos of pacifism that guides upland life also draws strength from a curse said to have been placed upon the mountain territory of Rantebulahan over 250 years ago for the sacrifice of a relative or dependent during malangngi ', a household prosperity rite run by women. This frightening negation of the sacrificial order sparked a social and moral crisis, and is said to have brought an era of poverty and misfortune upon Rantebulahan that has yet to be brought to a close.^[2] Some elders even attribute the collapse of Pitu Ulunna Salu to this ritual murder.

The violence of pangngae did not provoke the same kind of dread among my friends. I did make some of them uneasy by speaking as if the headhunters, past and present, were culpable of real and imagined murders. In doing so I conjured a world in which the headhunters' victims lived not as strangers and enemies, but as distant partners and persons worthy of human concern. My remarks also forced upon my friends a critical and demonized portrait of mappurondo practice. It was in the context of such conversations that mappurondo acquaintances would retreat into the past to find a moral basis for their headhunting practices. Stories about retribution in a time of feuds and enslavement were a way to make mappurondo ancestors and head-hunting practices appear just.

The discomfort my remarks and questions aroused from time to time confirms for me the predicament of the mappurondo communities in the Indonesian national order, and the ambivalence villagers can feel when discussing or holding traditional rites. In the discourse of Indonesian citizenship, the mappurondo headhunter cannot find an appropriate victim, for citizenship insists on a sense of mutual belonging that encompasses and transcends regional and ethnic distinctions.^[3] The social horizons of the past have been widened, shifting the boundaries of the human beyond those called for in pangngae.^[4] In the terrain of state citizenship, taking the head of a lowlander is murder. Yet the state also denies the mappurondo community ideological and institutional support. Local ancestral practices go unrecognized: they have not been accepted as religion nor have they been canon-

ized as kebudayaan , (Ind.) "culture." Given the pressures of the contemporary order, it is hardly surprising that this minority religious community would seek legitimacy in the past. The ceremonies and effigy heads of the present are reminders that call forth tradition and history. Looking back into the past in ritual performance, mappurondo villagers can see their practices as virtuous, intelligible, and right.

Stalking a victim in the past requires, in some sense, a forgetting of the present. But caution and concessionary gestures to the here-and-now are called for. The warriors of today carry no weapons except for their work machetes (used to fashion a tambolâ —a flute made from bamboo and palm), and all are careful not to be seen near mountain settlements in daylight hours. As recently as 1974, some warriors would travel to the coastal markets at Polewali and spend a week or more in and on the town. During my time in Bambang, however, cohorts would journey no more than a dozen kilometers from their home, and then linger in the forest for a short time—a day, three days, a week at most. Certain settlements, like the one at Minanga, have begun to limit their warriors' expeditions to Saturdays, so that the young boys who take part will not be missed in school. And in what may be the most telling concession to the present, some cohorts, like those in Rantepalado, do not even bother to find a head trophy, but simply return to their hamlets with tambolâ flutes.

I wondered aloud to Ambe Sope, the topuppu of Salutabang, whether pangngae made sense in the current scheme of things. As the authority charged with directing pangngae, Ambe Sope could appreciate the purposefulness and contingency of ritual practice. Accepting of the concessions the community might have to make in conducting the ritual, he nonetheless regarded pangngae as an obligation of tradition. Sense or no sense, ancestral teachings demand the songs and liturgies of pangngae. But Ambe Sope also mentioned obligation fostered by need. Something had to be done, for it would be unimaginable to leave a village in mourning.

Ambe Sope's remarks, along with the common formulation of pangngae as an "opening," an "unburdening," or a "relieving," call attention to bereavement as a crisis or force to be put to rest in mappurondo headhunting ritual. Let me be careful to explain, then, that pangngae brings an end to communal mourning . People put away public emblems of their grief, and the community, as a collective, declares itself free from the burdens of mourning. In an important sense, pangngae occasions a discursive space in which individuals can work through personal feelings of grief or loss and get on with life. Taking advantage of the ritual to work through feelings of loss,

however, does not always lead villagers to put grief and anguish irrevocably behind them. The end of communal mourning is hardly coincident with the passing of personal grief, nor should it be confused with the latter. The sentiments of specific participants are diverse, complex, ongoing, and largely indeterminate. Nevertheless, the celebratory mood and performative (or pragmatic) work of pangngae highlight a collective effort at overcoming loss. At the very least, the ritual tries to bring the scope of bereavement within measure. In this way, the headhunt plays an important role in patterning the rhythms of mourning and solace in the mappurondo communities.

Put in broad terms, it may be asked why the problems or crises of mourning have to be worked upon in the precinct of demonizing an other. In tackling this mystery, some ethnographers have suggested that grief or other turbulent emotions motivate and find catharsis in a headhunter's violence. I will discuss their work in the pages that follow. For the moment, I will just note that I see things somewhat differently: my understanding of pangngae convinces me that the resolution of communal mourning is more significant than personal catharsis in motivating the violent act of headhunting. I also place an emphasis on discourse as the generative site of both violence and solace. A variety of discursive forms—such as vows, songs, or noise—mediate the ways in which people and a community put grief behind them. Above all, the discourse of headhunting ritual refigures individual affect in collective terms as political affect, as a resource emanating from and giving meaning to shifting political situations and asymmetries.

Further questions may be framed around the "problem" of bereavement and ritual violence in the mappurondo community. Consider the ironic absence of bereavement in stories about the ritual's past. Although Ambe Sope and many others made it clear the pangngae was run for the purpose of brining mourning to a close, not once did anyone suggest that a need to dispel grief motivated ancestral practice: the genealogy of pangngae was always recounted and justified in light of intercultural tensions and polemics. I cannot offer much on this point except speculation. Perhaps the crisis of loss and anguish "goes without saying," buried deep in the moral foundations of the headhunt. Perhaps it is dangerous for this minority religious community to structure history around grief.^[5] In any case, bereavement was never mentioned in a way to confer virtue or intelligibility upon the headhunts of the past .

Given this tradition that attaches ritual violence to the human project of mourning, I am led to wonder, too, about what has happened within pangngae as the mappurondo communities have adjusted to an unstable

and sometimes menacing word. Their dilemma is not unlike the one encountered by the Ilongot: what can be done about mourning, how can one live with suffering, when headhunting is out of the question (R. Rosaldo 1984:181-182)? Shifting social and political orders have put the violence of the headhunt into flux, sometimes fueling it, other times dampening it. The current suppression of ritual violence implies profound changes in the discourse of anguish and mourning for both the mappurondo and the Ilongot headhunter.^[6] The Ilongot response appears to be pain and despondency. The long-standing "mappurondo solution" is to work with effigy heads so as to bring mourning to a close. But are "effigy emotions" brought into play? If the mappurondo headhunter of today cannot carry a weapon on his journey, can we expect him to carry his anguish and rage? I suggest that e does not, relying instead on "effigy emotions"—that is, representations of emotions made in ritual discourse—to characterize the sentiments that become entangled in violence.

Topographies of Anguish and Pity, and the Problem of Catharsis

Earlier in this book I made mention of the riparian cosmography in which the mappurondo headhunter situates himself and his victim. It will be recalled that regions downstream from the mountain territories lie n the path of souls who follow riverwaters to the sea, where they sink in a lightless whirlpool to the realm of the dead. The warrior cohorts of pangngae always journey in a downstream direction to find their trophies. Their movement downriver and back could be read, following Downs (1955) and McKinley (1976), as an effort to restore social and cosmological balance to a world disturbed by the loss of life and the crisis of mourning. In this structuralist scheme, the headhunter tries to subjugate death; the precise identity or social origins of the stolen head seem beside the point—anyone outside and downstream should do as a victim.^[7] The life lost downriver balances the life lost upriver; the soul that traveled downstream is countered by the trophy head brought upstream.

The problem of cosmological balance does not inform every headhunting tradition in island Southeast Asia. Discussing the Berawan of northern Borneo, Metcalf finds that headhunting is intended to "deflect the malice of hovering souls" (1982:127), to "export" the resentment and jealousy of

the recently deceased (Metcalf forth.). Within Berawan eschatology, "death has a chain quality to it," and until deflected outside the community, "death will follow upon death" within the group (Metcalf 1982:127). The point of Berawan headhunting, then, is to make death someone else's problem, not to even the score of souls lost. Yet Metcalf's findings, like those of de Josselin de long, Downs, and McKinley, remain loyal to exchange theory: the latter three describe the direct, symmetrical exchange of the dead, whereas Metcalf sketches something akin to the indirect and asymmetrical exchange of unquiet souls.

Judging from my experience, mappurondo villagers do not link pangngae in any explicit way to the problem of restoring cosmological balance. Neither do they speak about hovering souls whose anger and jealousy might bring about further deaths.^[8] Yet a rhetoric of "here and there" persists in the management of mourning. In shedding the burden of mourning, the community imagines an act of ritual violence taking place in the home of a rival subaltern from the past. Looking upon this violent scene through the ritual imagination has the potential to distract, dislodge, or possibly negate the anguish of those in mourning. If I am correct, then topography, the emotions, and social hierarchy work as transcodings of one another. In the discourse of pangngae, the journey downriver to fell and dismember a victim also marks a crossing into an emotion-scape of pity, remorselessness, and disgust. As I see it, these emotions are the expressive "masks" of political attitudes. Though they are not to be confused for one another, pity, remorselessness, and disgust do have something in common: they are felt in encounter with the subordinate and the debased, and always involve a "looking down" upon another, however fleetingly. They are political affects felt in the terrain of social hierarchy.^[9] One of the key tasks of mappurondo headhunting, then, is to shift momentarily the look of the community away from "egalitarian" feelings of anguish and empathy that link villagers together toward the "superordinate" feelings of pity and remorselessness that position uplanders with respect to their historical rivals on the coast.

My views derive from listening to the songs and liturgies of pangngae, but also follow upon long consideration of work by Michelle and Renato Rosaldo on the emotional discourse of Ilongot headhunting (M. Rosaldo 1977, 1980, 1983; K. Rosaldo 1980, 1984, 1987, 1988). Though I risk con-flaring and oversimplifying their respective studies of the Ilongot, their understanding runs as follows: the violent ambush and beheading of a victim affords a cathartic release of liget , a term the Ilongot use to describe feeling stirred up by grief, anger, envy, or desire.^[10] Comments from Renato Rosaldo's work are illustrative and emphatic:

Regarded as ritual, headhunting resembles a piacular sacrifice: it involves the taking of a human life with a view toward cleansing the participants of the contaminating burdens of their own lives. (1980:140)^[11]

Perhaps we should begin by viewing headhunting as the ritual enactment of a piacular sacrifice. . . . This ritual process involves cleansing and catharsis. (1984:189-190)

[The Ilongot] saw headhunting as a way of venting their anger. . . . [It's] the ultimate place to carry one's rage. . . . They feel it is the ultimate act of violence and anger. Their sense is that it is cathartic. (1987:252)

Entangled with unfocused liget and burdened with the weight of insults, pain, and grief, the Ilongot headhunter discovers that he can cast off his disturbing emotions by tossing a severed head to the ground.

This portrait of the Ilongot fits rather well with most of the literature on the ethnopsychology of highland societies in island Southeast Asia. By and large, the island cultures favor emotional restraint and balance in everyday life. The effort to rein in the emotions, to place constraints on intense feeling, to avoid upset, and to keep dramatic emotional display offstage presumably causes certain emotions to smolder or to diffuse over time (cf. Hollan 1988). In the view of some writers, ritual events offer a context in which everyday constraints on emotion are suspended or relaxed. As a result, rituals become focal occasions for displays of aggressiveness, anger, grief, or exuberant feeling. If this is the case, it would appear that rituals are doubly marked off from the everyday: first by their formalized practices, and second by their cathartic potential. There are considerable theoretical difficulties in characterizing ritual in this way, and, for that matter, in so characterizing the realm of the everyday. Emphasis on the formality of ritual downplays its improvisatory character, and draws analytic attention away from uncertainty and contingency in ceremonial practice. Privileging ritual as the site of catharsis inclines one to treat ceremony as therapeutic practice or safety valve, and to ignore the capacity of ritual to provoke anxiety and trauma. It also runs the risk of skewing our understanding of catharsis by putting emphasis on those contexts in which the release of intense emotions is most subject to explicit ideological and aesthetic mediation. Similarly, an emphasis on the formality and cathartic potential of ritual makes everyday practice and experience appear unmediated and insignificant. Although I will not pursue the issue in the present work, I think the challenge for ethnographers will be to find ways to avoid pitting the ritual vs. the everyday, or the cathartic vs. the constrained, and instead capture them in the indeterminate and contested arenas of human practice.

Despite these reservations about the way the literature has characterized

the ceremonial, the everyday, and the cathartic, I do agree that something that might be called catharsis commonly takes place in a wide range of ritual contexts, and that it informs more than a few traditions of ritual violence. The Ilongot, for example, made very clear statements to the Rosaldos about venting and managing strong emotions in the heat of violence and in the celebratory song-and-boast fests that followed a beheading. Striking emotional discharge also figures prominently in a number of Sa'dan Toraja rituals, leading Jane Wellenkamp (1988a) to argue that the Toraja possess a "cathartic" theory of emotion. Looking at expressions of intense grief in funerals, and at the violence of trance possession and ritualized kickfighting, Wellenkamp argues that the Toraja see personal emotional discharge as socially disruptive and necessary, and thus demanding of ceremonial release and containment. Somewhat surprisingly, her own analysis of headhunting ritual leads her away from connecting this form of ritual violence to bereavement (Wellenkamp 1988b). But her more general point about the ritual management of intense emotion in Toraja society would seem to apply to other cases as well, including the Ilongot.

Saying that headhunting is "cathartic," however, raises a number of problems. Citing a lively debate about the work of Thomas Scheff (1977) on emotional distancing in ritual, Wellenkamp (1988a:494) points out that "there is little consensus in the literature concerning precisely how 'catharsis' should be defined and identified, and under what circumstances it occurs for which specific emotions."^[12] Nor, as Bruce Kapferer (1979a, 1979b, 1979c) seems to suggest, is it at all clear how and why ritual modulates the display and transformation of feeling. On a different note, our views toward catharsis need to be kept flexible. For example, efforts to portray catharsis as a force that threatens to disrupt community stability leads us to forget that the powerful and perhaps violent release of emotion may be in some contexts the very measure and image of social stability. And given that human experience is not so much a conditional state as an unfolding in time, one does well to inquire into the rhythms of pain and catharsis in the theater of ritual.

Renato Rosaldo's very moving essay "Grief and a Headhunter's Rage" (1984) has had an enormous impact on discussions of emotion, ritual violence, and the particulars of Ilongot society. But the essay, along with some misreadings of it, has obscured the far-reaching explorations that characterized the Rosaldos' earlier monographs on Ilongot headhunting.^[13] For example, in trying to shed light on Ilongot violence, it calls attention to rage and bereavement without sufficiently reminding readers that the distress carried into the kill has heterogeneous sources: not only the enraged anguish of

mourning, but the fires of envy, the sting of insult, the heart-piercing lyrics of a song. A close reading of the text shows that Rosaldo has mentioned some of these as the burdens that lead Ilongot men to take a life (1984: 190-191), but the theme of grief overshadows them. The essay also makes a signal contribution about the positioned subject, taking care to distinguish the driving burdens of older men from the volatile desires of Ilongot youths. Unmentioned, nonetheless, are the dilemmas of women (who cannot carry their grief on a headhunting journey), and the terrible turmoil of an adolescent who not only has lost a parent but is yet to wear the hornbill earrings that signal a headhunter's entry into manhood. And in the end, a reader is left wondering, too, about the political dimensions of a headhunter's catharsis. Although anguish or envy may make an Ilongot man want to take a head, there remains the political question, "Whose head?" An unpoliticized notion of catharsis does not, cannot, offer justification or rationale for the choice of a victim.

If I fault the essay for what it leaves unsaid, I do so in order to underscore the complexity and uncertainty of the cathartic act. The felling of a victim is the work of a swarm, some burdened with rage, some entangled with envy, some perhaps stricken with fright at the "smell of blood" (M. Rosaldo 1983:137). The suffering, yearning, and doubt carried into the raid have their origin in a tangle of lives brought together in a collective task. If there is a striking release of emotion in stalking and killing, it is witnessed in the context of comradeship, adventure, suspense, and the taming knowledge of elders. Then, too, there are the raids that fail, that get put off for a more auspicious time, or that wind up forestalled altogether by changing historical conditions. Indeed, one of the key arguments in Renato Rosaldo's Ilongot Headhunting, 1883-1974 (1980), is that Ilongot raiding has been intermittent for the past 100 years. Cathartic violence is susceptible to the currents of political history. As Michelle Rosaldo (1980:34) relates, "even in the old days, grief rarely led to killing. In the past as in the present, intense and disturbing feelings only occasionally knew a headhunter's catharsis; more often they were lost in the distracting claims of daily life." And in those instances where an Ilongot did find a way to vent his rage in tossing the head of the fallen, did the "cathartic act" effectively end his mourning or his envy? In fact, as Renato Rosaldo (1984, 1988:432) himself has been at pains to point out, this ritual moment, for all its violence and drama, may be but one step in an enduring process of bereavement—a momentary respite from the grieving that will linger beyond the headhunt to reawaken in dreams, memories, and conversations.

To return to the mappurondo case, I find the relationship between ritual

violence, catharsis, and the modulation of grief and mourning just as complicated. Under threat of the law and the moral force of their elders' teachings, the mappurondo headhunters of today may not take a life during pangngae. Where, then, do the headhunters carry their anguish? The ritual largely denies them the acts of violence that (from an Ilongot perspective) might help purge their hearts of troubling emotions. One instead encounters songs of violence and the hostile gestures occasionally directed toward the effigy head. In principle, these songs and gestures of violence may be enough to elicit or constitute a cathartic release for some participants.^[14] My own sense is that they do not. They do more to distract one from feelings of anguish and sorrow than they do to lead to a cleansing purge of turbulent emotion. With the potential for cathartic violence so limited, is catharsis itself in question? If violent catharsis does not put the yoke to the tabus of mourning, what does? As I have suggested, mappurondo headhunting may call forth political affect in such a way as to "move" the community out of mourning. The summoning of this particular form of political affect does not tell the whole story, however. Rather, this summons is just one feature of the transformative project in which the ritual practices of pangngae "rework" social experience and social situation.^[15] In this light, the release from mourning that is said to take place in the violence of the headhunt may appear not so much as a purging of troublesome emotions, but as a process in which people suddenly rid themselves of a burdensome situation.

The Journey Downriver

Having raised a complex set of questions about mourning and the violence of mappurondo headhunting ritual, I turn to pangngae celebrations of the mid-1980s. The long, somber period that begins with the plowing of terraces and the casting of seed lingers on after harvest. For eight months, pemali —prohibitions on noisiness, laughter, storytelling, singing, drumming, and exuberant gatherings—silence the mappurondo households and hamlets. The tabus do not bring about soundlessness, but enforce a quiet that allows farmers and spirits to coax forth the rice crop undisturbed. When the last sheaf of the new crop has gone into the barn, and the last drying rack is collapsed, harvest comes to a close. But the quiet continues.

Thought and concern turn toward the deceased for several days. No one leaves the village and no one may do any work. Called palitomate , "back to the dead ones," or bulan anitu , the "moon of ghosts," this three-day

period is marked by the sacrifice of water buffalos and pigs, whose souls will be of benefit to deceased relatives in the afterworld. Bulan anitu is, in a sense, a resumption of the ritual practices concerned with mourning that make up patomatean , "the time of the dead ones." Deaths, of course, happen throughout the year, and bring temporary halt to other activities as burial ceremonies and their concomitant restrictions on labor and the consumption of rice are observed for as long as four days. During the rites of patomatean, women retreat to the center of the homestead, where they keep watch over the deceased and begin to keen for the lost. It falls to men to bury the deceased away from the hamlet, and to coordinate the accompanying sacrifices. From this time forward, the immediate household of the deceased, and especially its female members, wears emblems of grief—ribbons of white cloth and, for wives, shifts blackened with ash and mud.

In effect, bulan anitu contemporizes all deaths that happened throughout the year—those who died during April, for example, are recalled along with those who passed away in November. Thus, the three days of sacrificial ceremonies renew the public representation of loss and mourning. Immediately after the rites of bulan anitu, time is reckoned as panda pealloan , "the edge before day," a period of uncertain duration and lacking ceremony. Although people resume labor in their gardens, an air of sadness and solemnity continues to hang over the hamlets. It is during panda pealloan that men and youths secretly slip out of the village in pre-dawn darkness to journey off on a headhunt.

The solemnity of panda pealloan is tempered with a sense of anxious, unarticulated waiting. People know or suspect that some of the village men have gone to take a head, but the blanket tabu against any mention of pangngae muffles their anxiousness and anticipation. Specifically, villagers believe that mentioning the headhunt may imperil those who "went to the sea." Not surprisingly, the same tabu complicated my knowing when pangngae was to take place in a village. Resorting to euphemism to bring up the issue seldom succeeded, except to amuse and annoy. With time, I became better at reading the "signs" of the headhunters' likely return. Even so, my efforts to know when ceremonies were to commence were at odds with the secrecy and uncertainty that shrouded the headhunters' journey. Like everyone else, I just had to wait for events I could not control or fully foresee. Looking back on it, the term panda pealloan , "the edge before day," captured these moods and tensions, likening them to the way one might look into pre-dawn darkness for a sign of daybreak—training, expectant, and unsure. From the vantage point of this metaphor, solemnity and mourning are heavy nocturnal moods—the darkest hours that come before dawn.

As for the cohort of headhunters, some critical social exclusions constrain and mark its membership. Foremost among them is the one that closes the headhunt to women, an exclusion that institutes a lasting difference in gender ideology. In the context of the present discussion, however, another exclusionary demand is more striking: men and youths in mourning may not take part. This strict ban applies to any male within the household or immediate family of the deceased, and typically is reckoned to extend to first cousins. The effect of the tabu is to sequester and immobilize grief; the bereaved can do nothing about their anguish. In contrast to the Ilongot, then, mappurondo villagers cannot unburden their anguished rage by taking a head. Solace is not discovered in one's own violence, but is received as a gift from others.^[16]

At first glance, mappurondo headhunting would seem to run counter to the findings of the Rosaldos. Yet if mappurondo tabus sever anguish and cathartic violence in practice, they conjoin them in theory. Sequestering bitterness and rage, in my view, is a way to avoid violence. In my reading, the tabu implies that anguish is dangerous and disturbing and can lead one to kill. Put somewhat differently, the tabu suggests that violence does offer catharsis and solace for the bereaved. Remembering that in the current order of things mappurondo headhunters do not want to fell a real victim, and should not fell a real victim, the tabu makes great sense. Because grief is so volatile and threatening to civil order, ways must be found to keep it off the trails leading to the sea.

Here, then, I want to suggest that the social and historical forces that have led the mappurondo headhunters to forsake human heads and to take up effigies of some kind have also brought about change in the discourse of mourning. The villagers have shelved ritual violence and the cathartic release it might bring to an anguished heart. If I am not wrong, this is not a matter of discarding violence and its cathartic potential, but rather, an effort to put them away. Just as villagers put headbags, tambolâ flutes, water-buffalo horns and pig jaws from sacrifices, and other ritual objects up into the loft when they are through with them, so, too, have they put violence away in a place where it can hang as a keepsake. That anguish finds solace in violence is not forgotten. It is recalled in the negative sanctions of tabu and in ritual song.

A few examples from the sumengo tradition may help bring my point across more convincingly. Although the headhunters who depart from the village do not carry with them turbulent feelings of grief, the songs portray them as if they did. The following lyrics picture the headhunter as a preda-

tory hawk, "blackened" with anguish, that is, darkened with the same soot and stain used to prepare mourning shirts:

Another striking song is one performed by women. In it, women respond with alarm as the hawk hovers over the village, darkening the sky with grief-blackened wings. They remind the headhunter not to turn his anguish toward the village and the women, but to carry it downriver:^[17]

The image of the black hawk, a bird that locals regard as an omen of death, renders grief as undomesticated, predatory rage.

Felling the Victim

The transfiguration of the headhunter as a threatening bird of prey is a way to make visible to the imagination the burdensome rage of a community in mourning. The same image also renders a sense of journey, of a "going there," of the tremendous distances to be crossed by the headhunter. The imagination is about to go to another place in order to witness the scene of violence. From the vantage point of the community, the headhunters must move out of sight and out of hearing to find their victim. Moving beyond the horizons of sight and hearing takes the headhunters beyond the social reach of the group, beyond the territory of its senses. The community cannot look upon the ambush and murder that is said to take place. The only violence the community will witness will be that which is described, euphemized, and idealized in ritual song and celebration.

Even though the cohorts of today no longer venture into other ethnic regions, their departure from the hamlets is shrouded by secrecy and marked

by ritual efforts to safeguard the group. Organizing the cohort falls to the babalako (as ToIssilita' call him) or the topuppu (the ToSalu term). Men and youths who want to go off on a headhunt, or who have made pledges to that effect, quietly talk among themselves in search of companions for the venture. Then one by one or in small groups, they secretly confer with the topuppu or babalako, who arranges for the time and place of their departure.

During my two-year stay in Bambang, the number of cohort members ranged from just a few men and youths to nearly 50. On more than one occasion, the pack consisted of two or three married adults, several unmarried youths, and a few novices below the age of ten. Of the married men, the person who vowed to lead the cohort took the role of too'na ("the [tree] trunk"). Another married adult, meanwhile, assumed the role of lolona ("the [tree] top"), the cohort's junior leader. Whatever importance these two roles may have had in the past, they serve today only to mark the higher status of adult males who journey under the obligation of a vow to lead. As for the male elders who oversee pangngae, ToIssilita' custom dictates that the babalako accompany the cohort on their expedition,^[18] while among the ToSalu, the topuppu stays behind in the village.^[19]

As the time to leave nears, those who have decided to join in the foray begin to distance themselves from women by observing certain tabus on foods, sex, and talk.^[20] Then, by prearrangement with the babalako or topuppu, the headhunters meet together at the edge of a hamlet under the cover of pre-dawn darkness. Here, omens and sacrifices become especially important. A chicken is sacrificed, and its blood and entrails read. Bubbly blood, a hard heart, and a gall that protrudes well out from under the liver portend a successful raid. If the cohort happens to get a bad reading, or has, as the saying goes, sala mane ("the wrong chicken"), further birds are sacrificed until good signs are obtained.^[21] When the topuppu or babalako finishes muttering the final words of supplication, the cohort lunges at the offering of food and grabs wildly for its contents. In former times, it was believed that the warrior who got the chicken's head would also take the victim's. Nowadays, the person to snatch the chicken head will be named as the beheading warrior on the return of the cohort.

Before the rising sun begins to lighten the sky, the cohort steals out of the village in a downstream direction.^[22] The headhunters remain out of the village anywhere from one to ten days. During this time, some of the cohort members get the coconut that will serve as a surrogate head.^[23] While still out of the village, the headhunters also gather the bamboo, rattan, and palm fronds needed to make their tambolâ flutes, flutes that when blown

produce an eerie, low-pitched bellow.^[24] Most materials needed for the tambolâ grow on or next to settled land, and are obtained before the cohort makes its trip home. In keeping with the secret nature of their journey, the warriors must be careful not to reveal their identity as headhunters to other villages lest news of their whereabouts be carried back to their home settlement. Above all, the headhunters must not allow themselves to be seen or heard by women as they make their flutes.^[25]

Back in the village, the sudden absence of husbands, brothers, and sons may lead women to become anxious—or to put on a show of anxiety. Bewilderment at the disappearance of their menfolk recedes as the women realize that they are quite likely off on pangngae. To be sure, wives and mothers know that the men do not risk the dangers of real headhunting. Yet any journey poses the danger of accident or violent encounter and can therefore stir deep concern over a traveler's safety.^[26] Gestures of worry and concern are common. While the headhunters are away, wives and mothers will set out special dishes of rice in the loft of the house as a ritual gesture intended to ease the warriors' hunger. A wife is careful not to let anyone touch her husband's betel and lime holder, and no belongings may be taken from the house, an injunction that applies to fire and coals, as well.^[27] Women also take care not to goad, tease, or scold the men and youths who remain in the village during this time.

Later, during the exuberant ceremonies of pangngae, choruses of singers—usually, but not always women—will describe the anxiety and concern of women. For example, one sumengo depicts a wife who wanders about, unsure if her husband—who left without telling her—is away on a headhunt or off somewhere else:

Another chorus frets that those who undertook the adventure in order to adorn themselves with warrior's bracelets have been gone too long:

If the poetics of anxiety heard here provide some mirror of the distress and concern of contemporary villagers, they show women as the ones who endure worry and the threat of loss. Fear of abandonment, in the discourse of

song lyrics, scars the experience of married women. Given the prominence of women in keeping vigil and wailing over the deceased, the lyrics perhaps reflect a thoroughgoing cultural and ideological understanding in which a married woman's experience includes deep strains of brooding, worry, and loss. At the same time, the fact that headhunters do return home suggests that ritual discourse uses women as the foil for showing that loss and abandonment are not certainties.

The poetics and rhetoric of women's anxiety are partner to a broader discourse that idealizes the imagined violence of the headhunters as an agonistic encounter with a distant foe. Just as sumengo lyrics portray the departing headhunters as predatory hawks, the journey's end downriver finds them transfigured as cockerels. By way of contrast to the wild, predatory hawk of anguish, the cockerel offers the image of an aggressive but domesticated fighter, associated with struggles for dominance.

In depicting the headhunters' attack' some songs show only an oblique fascination with blood and violence. For example, many singings portray the victim as a banyan tree or as a black sugar palm. A banyan is a common metaphor for political strength and tradition, and the sugar palm connotes that which is cared for and nurtured. The felling of the banyan and the palm thus symbolizes the fallen body politic and the severed bonds of human care.

But other sumengo, like those quoted in the preceding chapter, reveal fascination or disgust with rot and decay:

The theme of disgust, I should point out, is not limited to song but also figures prominently in the presentation of the trophy head to the community (as we shall see).

Other lyrics call attention to loss, disorder, and anguish in the home of the victim. Above all, the songs do not so much describe a moment of cathartic violence as draw attention to the pitiable figure of the orphan or widow:

The strain of pity in these songs belongs chiefly to women. Men seldom know, or join in singing, these lyrics, and their "silence," it seems to me, is good measure of the remorselessness thought to accompany the slaying of their lowland rivals. In the discourse of song, it falls to women to extend pity to orphans and widows. Casting women this way contributes further to the cultural and ideological patterning of emotion and gender, a pattern in which women figure as the agents of sympathy and longing.

The lyrics of the sumengo, so far, evade mention of anything like cathartic violence. The songs suggest a displacement of anguish—its being carried over the horizon—even though pemali tabus immobilize it in the headhunters' home village. Even then, singers say nothing about headhunters feeling "light" or "unburdened" in their deadly ambush of another. Rather,

the lyrics mask the terrible horror of killing. On one hand, the songs play up the awesome power of the headhunter-cockerel as he knocks over trees, spills seawater, and tears up sand. On the other hand, they conceal the facts of violent ambush. In short, this poetry of domination euphemizes violence, both obscuring and exalting it. Similarly, expressions of pity or remorselessness seem to me but elements in a poetics of humiliation, especially if we recall the lyrics that voiced revulsion at the rotting, mutilated body of the ambushed, or those that conjured images of the panic and disorder amid the community of the slain.

That the ritual songs of pangngae fail to acknowledge ties between violence and catharsis should not surprise us, especially as we keep in mind that this lyric discourse barely mentions the anguish and rage of bereavement. Yet the silence about rage and catharsis does not mean that they are without consequence in shaping mappurondo headhunting tradition (cf. R. Rosaldo 1984:192). Their volatility and force have the potential to reawaken a threatening social order, and in recognition of that threat, they are "put away" rather than celebrated or unleashed by song lyrics. Seen in this light, the lyrics of the sumengo, like the tabus that keep grief-stricken villagers off the trails to the sea, do not suspend or relax restraints on turbulent emotions, but keep them in place. But as for the history behind this body of lyric, I can venture little. Perhaps the songs of the past did more to voice the pain of anguished hearts and to celebrate cathartic violence. Whatever the case, today's lyrics seem fitting in an era when the ambush of an innocent would bring reprisal from the civil order.

If the poetics of catharsis have been shelved in current discourse, the poetics of domination and humiliation have not. The lyrics of many of the sumengo discussed thus far summon pity and disgust, feelings appropriate to a subject who looks down upon another. The terrain that stretches between the mappurondo headhunter and his coastal rival thus includes vantage points for seeing and feeling. That is, local ideology reckons mountain and coastal political bodies in terms of place (inside:outside, upstream: downstream), siblingship (elder:younger), and violence (headhunter:victim), and, in so doing, locates the coastal rival in a lowered or subordinant sphere. Expressing pity for the anguish of the coastal victim serves as yet another transcoding of the idealized social hierarchy that allows the mappurondo community to look down upon a rival polity.

Expressing pity for the fallen is also turnabout for those mappurondo hamlets that have endured the weeks and months of mourning. If an anguished person was ever led to find solace in taking a head, his move from grief to catharsis was never simply a personal project for purging himself of

troubling emotions but was also a political act consonant with communal ideology and interests. And by the same token, the communal political project of subordinating a rival in ambush no doubt created paths for individuals to cope with their feelings.

Signs and Songs of a Return

Within the discourse of ritual lyric, the headhunter begins his journey as a predatory hawk and slays his rival as a strikingly leathered cockerel. Not surprisingly, the headhunter's journey home finds him transfigured once more, this time as the kulu-kulu , or sarese , nocturnal birds whose calls are read as omens of prosperity:

For those who had been left behind in the village, the fate of the headhunters is a matter of uncertainty and anticipation. Banking their fires at night, the villagers listen for omens of the warrior's victorious return.

But what about the headhunters? After getting the coconut that will fill their leader's head-bag (usually a worn rice pouch), the men and youths of the cohort begin making the other objects always carried home from the journey—the tambolâ. The songs of pangngae celebrate the slashing blows of the headhunters in making the instruments:

Each headhunter will prepare his own tambolâ from bamboo, decorating it with palm-leaf streamers, perhaps inscribing his name on the shaft, or etching it with geometric designs.

By custom, the headhunters are always said to be away from the village for tallu bulam , "three moons," regardless of whether they have been away for three weeks, three days, or three hours. With nightfall, the cohort moves out of its hiding place in the forest. Tambolâ flutes strapped to their backs, the returning headhunters approach their village in the dark. The leader car-ties the head and brings the cohort to the edge of the village lands. Above them the hamlets sleep in darkness, night-long fires having been forbidden by pemali tradition.

It is now time to awaken the village. As they move closer to the hamlets, the headhunters take their tambolâ and let loose a long howling note, and then shout "jelelelele yuhuhu " three times in a ritual gesture called mua'da' . Twice more the cohort members sound the tambolâ and let out the cry of mua'da' .^[28] With that the group moves up the paths and deeper into the village, pausing twice more to deliver blasts on their instruments and shouts of "lelelele. " The suddenness of the bellowing and crying startles the villagers out of sleep and marks the beginning of a'dasan ("the shouting"), the night-long rite welcoming the return of the topangngae , as the headhunters are now called.

The topuppu or tomatuatonda' appears at the edge of his hamlet brandishing a torch of blazing kindling, while other villagers remain inside their darkened homes. Arriving at the foot of his hamlet, the headhunters once more let loose a clamor of flutes and shouts. The elder up above calls out into the darkness, hailing the "strangers." A greeting is given in return, and the topangngae let out another series of jubilant cries, while a company of male elders comes down from the hamlet. Soon men and women, bearing torches and singing somengo, stream down from the hamlets to greet the headhunters, bring with them gifts of tobacco, betel, and food. As the crowd prepares to honor the topangngae, the cohort leader holds aloft the pouch containing the head. People turn their heads or make faces, crying "Bossi, boss !"—"It stinks! It stinks!" and then break into gales of laughter.

The song and feasting of a'dasan last until just before dawn. Songs celebrate not only the feat of the topangngae but the occasion of a'dasan itself:

Fig. 6.
The tomatuatonda' ("the elder of the village") comes out to greet the return-
ing cohort of topangngae (headhunters). He carries two torches: the one blazing in
his right hand is the first fire to be rekindled in the village following the return of the
topangngae. 1985.

Throughout the night, the headhunters are obliged to remain just beyond the edge of the hamlets. With first light, the topuppu or babalako puts aside a single tambolâ in the brush near the place of a'dasan; the flute will be retrieved at another time if another village, unable to run pangngae because of epidemic or an unusual number of deaths, calls for someone to release it from mourning. Taking up their tambolâ, the topangngae then follow the babalako or topuppu to the homes of the villagers who mourn for those who have died during the past year. At each home, the headhunters let loose a prolonged roar from their tambolâ and shout out "lelelele" three

Fig. 7.
While away on his journey, each headhunter prepares a tambolâ flute and
decorates it with plaited areca or sugar-palm leaves. The "voice" of the flute releases the
mappurondo community from mourning. 1984.

times. With that, each home and each hamlet visited is free of the signs and prohibitions of mourning, and a week of unbridled festivity begins.

In the discourse of song, grief is "figured" as impenetrable haze or darkness. No one can see "beyond" or "through" the enveloping cloud of mourning. The headhunters, however, usher home the arriving sun of peal-loan (rituals of light) that will clear the atmosphere of bereavement:

Coming out from mourning, in the local imagination, involves a restoration of sight. It also requires the help of others. The burden of mourning is not something that can be shaken off through personal effort, but must be endured until it is removed by the topangngae:

Kept from finding cathartic relief in taking a head, the bereaved endure the signs of grief until the shouts of the headhunter and the sounds of the tambolâ mark an end to the darkness of mourning.

Ritual and the Rhythm of Bereavement

The emotions that villagers carry into headhunting ritual are complex, heterogeneous, uncertain, and shifting. The commemorative practices of bulan anitu (the "moon of ghosts") that take place just before pangngae no doubt intensify a sense of loss among at least some mappurondo villagers, but in my experience, people suffer their hurt with silence and restraint, and seldom show flashes of the anger that may come from grief. Whether a person's sense of loss beforehand has been given greater pitch, or continues to be diminished by the demands of everyday life, the discourse of pangngae offers the bereaved a moment of solace and distraction. A purge of bitterness through physical violence is out of the question,

however, leaving the festivities to "lighten" the burdens of the villagers. Indeed, the idea of being "lightened" is fundamental to their understanding of what takes place in headhunting ceremonies (cf. M. Rosaldo 1980:27). Although from time to time I did spot someone unable to put away his or her sorrow during the festivities of pangngae, villagers usually threw themselves into the music, song, and mirth of the ritual.

What should be underscored is that pangngae enacts a change in the discourse of mourning. The measure of the ritual's efficaciousness has less to do with soothing private hurt and suffering than with changing the social and discursive situations into which villagers bear their grief. The return of the topangngae does not just mark the end of public mourning, it brings it about . Once the shouts of the headhunters and the voice of the tambolâ are heard (and a "voice" [oninna ] is precisely what tambolâ are said to possess), all mourning prohibitions on foods, conduct, and dress are lifted, along with the pemali tabus that kept the village in silence during the planting and harvesting of rice. Removing the burdensome weight of these prohibitions through a declaration of sound is an effort to lighten grief by stripping it of its outward trappings. Here, then, a vocal cry and a droning instrumental voice have illocutionary (or performative) force.^[29] Lacking the semiotic structure and content of the linguistic sign, and so, lacking a locutionary or message-bearing dimension in the strict sense, the shout and the drone nonetheless announce and index (point to) the headhunters' presence. They also announce and index the return of sound and noise to a long silent hamlet, and thereby thwart and negate the set of tabus that prohibited festivity. As constitutive ritual acts, the shout and the drone, timed to coincide with the rising sun, together bring about the exuberant ritual period of pealloan. In this sense, the sound of the shout and the tambolâ measure ritual time and give it shape.

Once this ritual moment takes place, the burdensome tabus and obligations of mourning are removed. Irrespective of their moods and feelings, the bereaved no longer need subject themselves to collectively imposed constraints on conduct, talk, and consumption. As I have already hinted, local discourse uses the natural fact of the rising sun to capture the social fact of this situational "move." Yet the suddenness of the shouting and the tambolâ's drone seem especially significant in effecting the change of discursive situations. Although the villagers wait anxiously for the headhunters to come home, the precise moment of their return and, indeed, whether they are fated to return home, are uncertain matters, shrouded by secrecy. Thus, the shouting and the blowing of the tambolâ come as a sudden (if welcome)

surprise, erupting from below the somnolent quiet of the hamlets. In everyday contexts, a startling noise might raise a panicked response.^[30] Here, ritual makes productive use of surprise, a "rhythm" and "tactic" commonly avoided in day-to-day conduct.

Yet the significance of surprise, of being overtaken by sound, goes beyond being a convenient way to mark time. Let me elaborate. Of the terms used to name rituals, none but pangngae may be rendered with the circumfix, ka- -an (a grammatical construction similar to ke- -an in Indonesian). For a speaker to say "kapangngaean " implies that pangngae happened without the speaker's intent, control, or deliberation (that is, it was unforeseeable, much like kabarataan , experiencing grief over the death of another). Other rituals always take the pa'- -an circumfix, implying a sense of doing something or bringing about something consciously and willfully (e.g., melambe may be rendered as pa'pelambean ). Though too much can be made of terminological constructions, I think "kapangngaean" captures the sense in which it is the sudden and unforeseeable fate of the village to celebrate pangngae.

Furthermore, surprise is basic to the kinds of violence that take place or are said to take place in pangngae. Having encountered so many symbolic reversals and inversions in the discourse of pangngae, is it wrong to think of a'dasan as an event that mirrors and upends the headhunters' ambush of the coastal rival? In both events, the headhunters take the "other" by surprise. And in both cases, noise plays a prominent role. In the ambush that brings death to the unsuspecting, cries of anguish mix with the terrifying noise of attack. In the ambush that "lightens" the burdens of the bereaved, the shouts and the drone of the flutes emanate from an enlivening exuberance. (In fact, the mua'da' shout of lelelele is said to be both the name and the sound of a man's sumanga' , or "soul.")

Two ToIssilita' terms I heard in Minanga can throw further light on the relationship between the ambush and a'dasan. Although maringngangi' , "relieving a burden," is the most common way to refer to pangngae there, other key terms include dia'da'i barata , "shouting at grief," or, alternatively, dirondongi barata , "taking apart [or felling] grief." The latter term goes a long way in establishing parallels between these noisy events. Just as the headhunters fell and dismember a victim downriver, so, too, do they fell and dismember grief in their home upstream; mourning is an abject "other" to be slain with noise. The former term, meanwhile, suggests that the voice itself is a slashing weapon. Vocal noise and the war knife both cut. Yet in the poetics of violence, a movement is implied: the blows of the knife on the

dumb body of the fallen give way to the vocal blows that will slash away at the grief holding the warriors' community mute. The knife will make its victim silent; the shout will give back to the bereaved a voice. Paradoxically, today's headhunters carry no weapons, and so it falls upon the "voice" given to the headhunters' community in a'dasan to enact violence against the rival downstream. That is, language and the imagination deliver blows to the enemy. In this sense, the poetics of violence actually reverse the temporal movement mentioned above: the vocal blows of a'dasan unleash the headhunter in celebratory song and allow the imagination to conjure a scene of originary violence where the war knife can fell a victim.

The sudden violence of a'dasan bespeaks both the unforeseeable and the inevitable destruction of mourning. Later in the ritual, during a period of impassioned speechmaking, men will allude to the unforeseen and inevitable acts of violence. For now, it is enough to note that local discourse places the situational move into pealloan, and thus the end of mourning, outside of willful human action. Equally important, the sudden and inevitable blows of a'dasan impose a rhythm onto grief and suffering. In saying so, I do not mean to imply that the conventional acts and presented ideas of a'dasan bring about a full release from anguish. In keeping with the views of K. Rosaldo (1984), Tambiah (1985), and Suzanne Langer (1951), I would assert that a'dasan is but a gesture—albeit a dramatic one—related to the ongoing and uncertain experience of loss and suffering. Nevertheless, a'dasan makes such a powerful eruption into the experiential terrain of mourning that I cannot help thinking that it serves as a landmark of some kind. In my view, a'dasan momentarily draws the bereaved together such that they are "shouted" at once, and offered solace in solidarity. That is to say, the inner phenomenal time of the tambolâ's drone and the headhunters shout, and the manner in which they form a single heterophonous voice, permit both headhunter and listener to share the vivid immediacy of sound (cf. Schutz 1977 [1951]). It offers communion as antidote to the memory and feelings of loss and separation that characterize bereavement. Furthermore, a'dasan may serve as a measure of experiential "inner" time, of feeling, promoting a sense of before and after (e.g. blinded by the haze of grief and seeing through it and beyond). Whether feelings of loss linger beyond the headhunt is not the point. But it takes a'dasan for a villager to be able to say that he or she has been grieving too long or just long enough. A'dasan makes it clear that enough time has gone by for mourning to be put away. Without a'dasan, no one knows the proper rhythms of experience; without a'dasan, grief would have no measure.

If pangngae affords someone a sudden release from grief, it does not happen in the violent ambush imagined to take place downstream, but rather in the "ambush" upstream. It is not the violence of the blade that offers solace, but the violence of noise and music. It may be that the unfolding ritual process directs grief to the shout and the tambolâ's drone, where it becomes focused, condensed, and brought to a climax in sound. But quite apart from their potential to purge someone of their inner anguish, the shouts of the topangngae and the sound of their flutes bring about an overwhelming situational move. Thus, the shout and the sound of the tambolâ may supplant or dislodge feelings of grief with a mixture of joy and longing. Laden with familiar memories and meanings, the cries of the headhunter and the deep resonant tones of the tambolâ promise relief and festive communion. After months of silence, song and laughter break loose again. Singing, laughing, and feasting can go a long way toward taking people's minds off their anguish. More pointedly, the ceremonies to follow a'dasan extol vitality, good fortune, eroticism, the pleasure of song, and the splendor of men—in short, themes and topics that run counter to the moods, sentiments, and conceptions associated with death and grief.

Though it is wrongheaded to think that a villager's personal sentiments will always be in accord with the moods and pragmatic workings of pangngae, we should not discount the political force of the ritual in convening a "feeling polity." The tambolâ and the warriors' shouts are a distinctly coercive summons: villagers cannot help hearing the noise and music that announces a shift in communal tenor and practice. Getting caught up in or going along with the "official" moods of the community—moods brought about and legitimated by ritual—puts one in political alignment with the dominant trajectories of social thought and action. Not to yield to the enlivening mirth and noise of mappurondo headhunting festivities is to place oneself outside the polity. A refusal to allow oneself to be lightened by the ritual is an act of resistance, a failure to find solidarity with one's community. In this sense, someone's personal sentiments are always politicized by the discursive work of pangngae.

Emotions and emotional discourse aside, pangngae cannot fail to take place, or to effect a discursive and situational move, unless the mappurondo community in a given settlement has become moribund. In cases where epidemic or famine plunge an entire community into mourning (effectively preventing any male from going on the headhunt), a summons will go out to another village (usually Rantepalado, in its capacity as the Indona Bambang) to send someone with a tambolâ to "open the hamlets." Even death

cannot bring a halt to pangngae once the ceremonial fires have been lit—that is, once the first chicken or pig has been sacrificed to welcome the headhunters In the event that someone dies in the village while pangngae is under way, the village will ordinarily divide itself in half. One half will follow pangngae, the other will observe funerary rites with the hamlet of the deceased. Those stained by grief will bear the emblems of mourning until the next postharvest headhunt, a year later.^[31]

From Vow to Violence

So far I have argued that the resolution of communal mourning and the structuring of political affect are more significant than individual sentiment and personal catharsis in shaping the discourse of pangngae. There is no question that personal emotions are caught up in the ritual, but they do relatively little in molding the pragmatic contours of the rite. In contrast to Ilongot practices, the community, not the individual, is the central concern of mappurondo headhunting ritual. Bereaved individuals do not rely on personal acts of violence to rid themselves of a turbulent sense of loss. As we have seen, villagers may not carry their anguish onto the trails in search of a hapless scapegoat. Instead, the bereaved rely on others to secure them solace. What leads these others to assume the role of headhunter?

Renato Rosaldo has argued that turbulent emotions propel the Ilongot headhunter to seek a victim.^[32] He explains that an

. . . account of youthful anger and older men's rage born of devastating loss lends greater human plausibility to the notion that headhunters can find their cultural practices compelling. Because the discipline correctly refuses to say that by nature headhunters are bloodthirsty, it must construct convincing explanations of how headhunters create an intense desire within themselves to cut off human heads. (1984:191-192)

Rosaldo, however, gives us relatively little in his corpus of work to suggest how the Ilongot sustain their emotions in such high pitch. That is to say, anger, rage, and desire appear sharp and without fluctuation and rhythm. In particular, Rosaldo portrays Ilongot grief as virtually all rage; the torpor and weight of personal loss are missing from this picture. At the same time, the emphasis on inner experience runs the risk of missing the ways in which forces external to the feeling-self may lead someone to kill. As should be clear, the mappurondo case poses real difficulties for an approach that teth-

ers the project of violence to personal distress. The mappurondo headhunter is in some ways selfless—he "kills" for the sake of another. What, then, motivates the mappurondo headhunter if the driving force of grief is sequestered? What stirs him to journey seaward?

I think the answer lies in the pledge, the oath, or the vow, or what in Bambang is called samaja . Put in the most general terms, it is the illocutionary entailments of a speech act that motivate the topangngae, not unmediated desire or rage. The binding power of the oath commits the headhunter to violence. Furthermore, the illocutionary force of making a vow, in my view, not only introduces factors of personal will, intention, and hope, but potentially serves as a way of acknowledging and coping with strong currents of desire, want, or anger. In particular, the vow may work to direct, constrain, and display strong personal emotions by linking them to the communal project of headhunting.^[33]

Vows of the sort I have in mind are not absent in the Rosaldos' work. Renato Rosaldo (1984:191) tells us that "older men often vow to punish themselves until they participate in a successful headhunting raid." Michelle Rosaldo offered still more detail:

Some Ilongot men in youthful "clouds" have taken oaths (binatan ), like those occasionally pronounced by mourners, who—to heighten their own liget and notify their fellows of an intention to take a life—forswear the use of certain foods or weapons, or even, in extreme cases, carry whips that they have friends and neighbors use to beat them on entering any but familiar homes. (1980:140)

The authors make very clear that making a vow displays or signals a person's emotional intensity.^[34] In that sense, the discursive act of making a vow bears an indexical relationship to turbulent rage or desire: the smoke of an oath signals the fires of feeling.

But the promissory oath is also a pledge to act. In the Ilongot examples, an oath connects intense feelings of anguish, rage, and envy to displays of sincerity and personal willpower, and to a course of action. In the context of a vow, emotional intensity can convey personal sincerity, purposefulness, and strength of will. At the same time, the pledge to act offers reciprocal measure of one's heightened emotional state. Yet turbulent emotion obliges no one to act; the utterance does . Once the oath has been made, failure to see it through throws one's sincerity, will, and depth of emotion into doubt. Just as important, the vow brings time into the picture: the oath leashes the impulse to act and defers action until the appropriate moment. Until then, the vow holds the oathmaker in its grasp, an entangling or burdensome state of self-imposed obligation.

It is not unusual for mappurondo men and youths to make oaths that commit them to join in the headhunters' cohort. What differentiates them from the Ilongot male is that their vows are made in silence and withheld from public awareness. And nowhere does anguish provide an emotional foundation for the headhunter's pledge. Rather, personal interests, status concerns, and the envy of other men lie at the root of mappurondo headhunting vows. As in the Ilongot case, young men and boys envy those who have been honored as topangngae. In these instances, mimetic desire (to borrow a phrase from Girard [1977]) motivates the pledge to act. Elders, too, make pledges, and these reveal a diverse range of motivations. For example, a ToSalu man has the prerogative of holding the settlement's annual ritual at his home in order to make good on a personal vow to the spirits. Typical are vows in which a man promises to lead or join up with the topangngae in return for providential favors. Vows connected to the restoration of personal health culminate in a version of pangngae called undantai kalena ("striking [making a thudding noise on] one's image"). Those made in order to seek the strength and resources to build a new house or rice terrace end in ulleppa'i samajanna ("releasing the vow"). But it is also possible for a man to combine sacred vows with a desire for personal aggrandizement. One way is to stage umpokasalle kalena ("enlarging one's image"; also called malampa sali-sali , or "making and loading the offering rack"), a ToSalu version of pangngae in which one displays one's good fortune, devout moral character, and stature in the community by giving substantial offerings to the debata. The most elaborate of such "signature" (or personally sponsored) versions of pangngae is ma'patuhu'i botto ("having the hamlets follow"). Here a villager makes a vow of unusual breadth or sincerity; in return for providential blessings he pledges to lead able-bodied men and youths out of the village on a headhunt and to then host an enormous feast in their honor.

When the time to hold pangngae approaches, men and youths will approach the babalako or topuppu and make their pledges known. One virtue of the secrecy surrounding such pledges is that the oathmaker can put off his commitments for another time. Several of my friends volunteered that from time to time men will use the cover of silence and secrecy to back away from their solemn commitments. But they added that for a man to do this would invite possible reprisal from the debata. The same custom also allows someone to fake a sincere pact with the spirits. Ambe Teppu, for example, told me the story of a To Salu man who happened upon a cohort of headhunters returning to his village. Announcing that he, too, carried a

vow to take part in pangngae, the man joined in with the others on their way back into village lands and shared in the honor heaped upon them. The man's wife, however, was not to be fooled. "What vow? she complainingly asked. Convinced that her husband was interested only in food and public praise, she refused to take part in singings and put out only meager portions of rice and meat for her husband.

It should be clear, then, that the mappurondo headhunter carries a vow, rather than turbulent feelings of anguish, onto the trails leading downriver. I should refrain from suggesting that the vow has replaced the originary rage of headhunters past, even if such a replacement would be consistent with the current of historical forces in the mountain region. More reasonable is the claim that the oath—born from complex, heterogeneous, and uncertain wants and interests—may not (or may no longer) include grief as one of its wellsprings. I think there can be no question that the oath is a discursive incarnation of "emotion-work," and I would hazard that the samaja (or oath) has long been a basic mediating form between intense feeling and the headhunter's violent act. Making a vow can be a way for an individual to cope with strong emotions, giving him or her purpose—and thus virtue—while according that individual a measure of control over the flux and intensity of emotion. Furthermore, the vow shows a person's strength of will and character in working and living with intense feeling. If the samaja serves as a discursive restraint on feeling and action, making good on a vow may afford the headhunter not only a sense of accomplishment, but a sense of release as well.^[35] And I would suggest, too, that the exercise of will implicit in the headhunter's vow demonstrates to the bereaved that people can master their temporal existence. In this sense, the vow is both a demonstration that one can get on with life and a gesture of confidence and faith in the community's fate.

The political dimensions of the vow are no less important for understanding the discursive construction of pangngae. It is by way of the vow that the private wants and interests of males are harnessed to communal ends. For men young and old, the "individual" and the "communal," the "private" and the "public," are not opposed spheres of action and understanding, but rather mutually shaping ones where personal wants and concerns gain intelligibility and consequence, and communal tasks open paths for personal pursuits. By the same token, the samaja vows of men grant them the capacity to act and make tactical choices quite unlike that available to women . . . or to the bereaved, who are in a sense feminized by their exclusion from the imagined arenas of ritual violence.

In bringing this chapter to a close, I am too aware of what I cannot know directly—the course of anguish in a villager's life, the words of a solemn oath murmured in solitude, the secret tasks and adventures of the journeying topangngae. The lyrics of ritual song, the burst of noise at a'dasan, and the cautious remarks of friends made so as not to divulge too much are what I have by which to fathom the experiential currents of mappurondo headhunting. I console myself with a remark Gananath Obeyesekere (1990: 55) makes about dreams: "The dream for all of us was not the dream that was dreamt but the dream that was reported." There is no "other" side to which I can cross in search of the lived experiences of mappurondo headhunters as they suffer anguish and envy, or commit themselves to a journey I will never see.

But lived experience is not my quarry in this study. Rather, I am trying to discern the thematic tenor and pragmatic workings of ritual discourse in pangngae. As a starting-off point, I have emphasized that pangngae is staged to bring an end to communal mourning, a process related to, but incommensurate with, the passing of personal anguish or grief. Nevertheless, the ritual provides a discursive space in which participants can work at overcoming feelings of loss. In a very important sense the ritual maps out an "emotion-scape" in which violence, anguish, pity, and solace, along with allusions to heaviness and lightness, take on a political character. Refiguring individual sentiment in collective terms as political affect, ritual discourse exacts a transcoding of social hierarchy and social solidarity. In this context, the demonizing and debasing of a downstream rival together have the potential to distract participants from the crises of mourning. "Looking down" on a subaltern rival from the past through the lenses of pity and disgust potentially confers upon the seeing/feeling subject a reflexive sense of exalted and superordinate being. Put very simply, to be lightened and relieved of the burdens of mourning is to be heightened vis-à-vis a political rival. Yet the discourse of solace also suggests a polity of sentiment constructed around the community's official moods and trajectories of action. Finding solace and giving in to mirth and festivity become expressive signs of social solidarity.

I am sure that catharsis from time to time has had a significant place in the politicized terrain of mappurondo headhunting ritual. But as a product of social history, cathartic violence is susceptible to changing political conditions. In the case of the mappurondo communities, it has been shelved to keep it from disturbing (and bringing reprisal from) regional and state orders. Yet, it is recalled in the negative sanctions of tabu and in ritual song. The current suppression of cathartic violence persuades me that turbulent

emotions are not—and perhaps have never been—the driving force for local headhunting practices. Catharsis and cathartic theories of emotion may play a part in local ethnopsychology and ritual practice, but they do not legitimate or provide a rationale for the institution of headhunting. The catharsis that may come with violence does not lend social, political, or historical intelligibility to pangngae.

The headhunters' violence, I have been at pains to point out, is not the outcome of powerful emotions, but of powerful obligations made in discourse. Making vows is central to the trajectories of action and violence in pangngae, and to linking the personal interests of men to communal ends. Indeed, violence and vow-making appear to be the prerogative of men in this ritual. The will to act overshadows emotion as a pivotal dimension of mappurondo headhunting. From the vantage point of emotion, mourning and oathmaking involve intense feeling. Yet, from the vantage point of human will, mourning and oathmaking underscore different capacities to act. In mourning, mappurondo villagers are helpless, and subject themselves to burdensome and immobilizing tabus. In oathmaking, the young man or elder leashes himself to action and proves himself master of his own existence.

The sudden return of music and sound signals a situational move within the community, and makes coincident the trajectories of mourning and the oath. The durative aspects of bereavement and being held by a vow are constituted and measured by the noise of the headhunters and their tambolâ flutes. More than just marking an end to mourning and the vow, a'dasan brings them to an end in a musical simulacrum of violence. Made possible in the shared (or coerced) apprehension of musical noise, releasing a vow and shedding the public emblems of grief help mappurondo villagers remake the world, allowing the community to resume its ways (cf. Good 1994:130-131).

More remains to be said, however, about the themes and projects of pangngae. In this chapter I have tried to clarify the way in which the problems and crises of mourning are worked upon by demonizing an other within the ritual frame. Intermingled with the ceremonial representations of violence and solace are themes of envy and adornment. But in contrast to the political affects described thus far, the themes of envy and adornment have less to do with mourning and the hierarchical relations linking mountain and coast than they do with the mimetic desires of men and women, a topic to which I now turn.

5
Envy, Adornment, and Words That Make the Floor Shake Pangngae and the Rhetoric of Manhood

Pua' Soja, my friends said, was already a village leader in Minanga when Dutch expeditionary patrols crossed into the Salu Mambi valley for the first time. That was in 1906. Sitting with him in 1984, reticent with wonder and disbelief, I tried to imagine what someone would have seen, heard, and endured in living through the hundred or more years since the internecine collapse of Pitu Ulunna Salu. I pitied Pua' Soja his aches, his clouded eyes and swollen joints, his being plain worn out by nothing except having to live another day. But weariness of body had yet to defeat his courage and dignity, especially during the season of pangngae and other pa'bisuan rituals when he would put on headcloth and ceremonial gear, and struggle the climb in and out of houses and up and down the village paths, to take his place with the other senior men gathered at the rites.

Among my mappurondo acquaintances, no one had firsthand experience of the precolonial order, save for a few who would have been very young children in 1906. Elder women were reluctant to say much about the past. Although their recollections of local events and family histories were every bit as sharp as men's, women left "official" history to the men. As for elder men, most discussed the ancestral past—the time prior to the fall of Pitu Ulunna Salu in 1870—as a turbulent golden age. Ambe Tibo, a grandson of Pua' Soja, was not alone, for example, in claiming that before misfortune fell upon the headwaters, village men were tall and big-boned and always lived to be 100, their bodies evidencing a greatness now gone. He himself had

Fig. 8.
Pua' Soja, 1983.

been in another village one time when someone unearthed a huge bone, said to be part of a human leg. "A span this long or more," he said, making a cutting gesture on the upper part of his extended arm. "Back then, the people here were powerful and rich. But look at what we have become." Others would give a more mixed picture of the past. Ambe Assi', the tomatuatonda' at Minanga, was proud of headwater history and mappurondo tradition. But I painfully recall a remark he made one day as he listened to several of us speaking about the corrosive effects of the Dutch colonial period. "Listen up," he said curtly. "Before the Dutch came we weren't human yet. We didn't even know how to talk."

The prospect of speaking at length with Pua' Soja about the close of the nineteenth century promised both a less gilded and a less demoralized account of local history. That was never to happen. My conversations with the old man were brief and few, and were always distracted by the company of others. He remained a remote figure. Seeing him again during the pangngae celebrations of 1985, I gently pressed Pua' Soja about the headhunting rites of the past. His fleeting remarks never made it clear whether or not heads were taken during his lifetime. In contrast, his answer to my question about the ritual's central theme was unambiguous: pangngae was all about kelondongan (the daring and audaciousness of a cockerel; cockiness), and kemuanean (manhood).

Putting manhood into action and on display in pangngae calls attention to a social poetics (Herzfeld 1985) or a social aesthetics (Brenneis 1987) in which villagers summon a select vocabulary and image of masculine conduct and being. I share with Herzfeld and Brenneis their interest in performance as the means through which sense-making, self-making, and experience coalesce as a strategized and consequential event. But because the ceremonial formulation (or evocation) of manhood in pangngae is so partial and select, and because it so animates the communal body, its basic rhetorical character deserves emphasis. Lake the Glendiot idioms described by Herzfeld (1985: 16), the rhetoric of manhood in mappurondo headhunting rites involves "being good at being a man." Yet I would argue that recourse to a Jakobsonian concept of poeticity—"the set (Einstellung ) toward the message as such, focus on the message for its own sake" (Jakohson 1960:356)—does not take us very far. "Being good at being a man" in the arena of mappurondo ceremony is a way to set specific social and political projects into motion. For that reason, the "set (Einstellung ) toward the message as such" is not enough. The deployment (Ausstellung ) of the message, and its effective expression in performance (Darstellungsgabe ) are basic to the rhetoric of manhood in the mappurondo community, and must play a part in any considered analysis of pangngae.

In previous chapters, I have tried to throw light on ritual violence from the vantage points of political history, mourning, and vowmaking. Pua' Soja's remarks, however, compel a reading of the headhunters' violence in light of ideas about manhood. Exploring the ritual discourse of manhood should also help us to understand why the demonization of the rival "other" and the cessation of mourning take place in a setting that underscores the social hierarchies predicated on seniority and gender. In the course of this chapter, I will argue that a ceremonial politics of envy and emulation animate both hierarchies. That is to say, the headhunting rites of pangngae are a turbine of social difference and mimetic desire. The rhetoric of manhood in pangngae thrives on these political currents. But as I want to show, that rhetoric is also the discursive mode for conjuring envy and emulation.

It is worth repeating that pangngae takes place under the authority of men and is considered both a village concern and a male spiritual pursuit. In a conventional sense, pangngae is not a rite of passage for males; it does not bring about a permanent change in social status. But youths must follow the trails to the sea, so to speak, if they wish to join the company of mature men. And as we have seen, those already in the ranks of men may undertake the journey for personal ends. Women of course play a fundamental if circumscribed role in the ceremonial activities associated with the headhunt. Yet taken as a group, women are not fully subject to the ritual's focal transformative operations. To use Bourdieu's (1991) terminology, pangngae is a rite of consecration and institution. That is, pangngae separates

those who have undergone it, not from those who have not yet undergone it, but from those who will not undergo it in any sense, and thereby [institutes] a lasting difference between those to whom the rite pertains and those to whom it does not pertain. . . . By treating men and women differently, the rite consecrates the difference, institutes it, while instituting man as man . . . and woman as woman. (Bourdieu 1991:117-118)

Mappurondo headhunting rites thus bring gender differences into being and impose them upon participants as their social destiny. At the same time, the ritual makes hierarchical distinctions within a community of males, distinguishing those "seniors" who have been honored as topangngae, and those who have not. The rhetoric of manhood, I would argue, takes up this double business and ties it to the project of ritual violence.

Though it is possible and perhaps necessary to describe these hierarchies in terms of "structural" differences, I think it is important to capture something of the anxiety, want, and restlessness common to the ongoing foment of social categories. Contemporary mappurondo communities are relatively egalitarian, neither producing nor tolerating dramatic differences in prestige

or power.^[1] That said, differences having to do with gender and seniority are basic to local thought and conduct; they help to motivate social life and to make it intelligible. The rhetoric of pangngae, in my view, summons and plays with these differences, juxtaposing them in such a way as to spark social movement.

Although villagers show a deep practical and ideological regard for human sameness, the simple fact is that in mappurondo ritual life, not everyone can be a headhunter. That women are excluded from becoming the central actors and focal subjects of pangngae fractures human sociality and marks part of the faultline that distinguishes manhood and womanhood. Yet because the rituals of pa'bisuam baine put women centerstage, I caution against a rushed analysis that would see pangngae as an occasion for men to gain a position of lasting dominance and privilege vis-à-vis women. By instituting and consecrating a difference between men and women, pangngae declares a limit to human sameness. But it does not relax or suspend the egalitarian ideal. In fact, I would argue that the social distinctions delimited in the ritual rest upon or are carved into the bedrock of sameness. Stated more strongly still, ritual discourse puts ideas about gender difference and human sameness into productive tension and conflict. As a result, the will to sameness, confronting difference, experiences envy, and a desire to emulate. The will to differ, confronting sameness, strives for excess and the "talismans of supremacy" (to borrow a phrase from René Girard 1977). Girard (1977, 1987) would argue that the tension between sameness and difference is a source of generative violence, a violence so terrible and contagious that it must be directed out of the community. And he would likely see mappurondo headhunting ritual as an attempt to solve the problem of rebounding violence (cf. Bloch 1992). I seriously doubt that mimetic desire plays such a decided part in the political violence of pangngae. Yet it is far from absent in this ritual. For the purposes of the present discussion, I merely wish to identify mimetic desire as one of the principal rhetorical strains in pangngae and to call attention to its relevance to local ideas about manhood and womanhood.

In the rhetoric of headhunting ritual, men strive for excess and superiority, and women look on with envy. (a situation that is reversed in the grand rites of pa'bisuam baine). I should stress that women's envy does not derive from lack or absence, but from a wish for still more. That "still more" is denied to women, withheld from them as the defining secret of manhood. Here it is worth noting that the tasks deemed most significant for making men men—slaying a victim, making a tambolâ—happen outside the village and out of the gaze and earshot of women. The excess that belongs to men

is discovered on seaward trails and in the forest during the headhunters' journey downriver. Surrounding these tasks, too, are a set of tabus that put a woman at risk should she happen upon men in their secret work. Violating these tabus is no less than a transgression of the boundaries marking male and female. Thus, the breach of category erupts into somatic symbolism as physical dissolution and death. Seeing men make their tambolâ flutes, for example, is said to bring illness or death to a woman. By concealing some of their practices the way they do, mappurondo men essentially hide themselves, and something of themselves, from public view. It should be evident, then, that masculinity and femininity depend upon differences made and kept in secret as much as upon differences made visible and audible in the domains of ritual and everyday life.^[2] For this reason, women cannot slake their envy during pangngae through emulation. It must remain hungry and restless until the rites of pa'bisuam baine, when women can discover excess in their own secret sources. At the same time, women are complicit—indeed, instrumental—m heaping praise and admiration upon the topangngae. Once more, then, we see the work of others as crucial to the management and representation of self. lust as the bereaved cannot remove the burdens and emblems of grief through self-effort, neither can the headhunter acquire an excess of manhood, a superabundance of being, without the help of women and the collective at large. To sum up, it falls to women to witness, to gaze at, to envy, and to desire the headhunter.

The discursive force of envy and emulation also colors the hierarchy of senior and junior males, of men and not-yet-men. Once again, the desire for difference becomes entangled with the desire for sameness. Young males see presented to them an opportunity to join the ranks of adult men, and in this sense are very much unlike women, who are precluded from achieving sameness with men through the headhunt. The figure of the headhunter is, of course, the model image of the adult male. If the mappurondo youth is to achieve peership in the company of men, and thereby make his passage into adulthood, he must make the headhunting journey of pangngae. Yet there is no need to drag the young across the threshold of maturity. The mimetic aspirations of the young are clear: they want their chance to wear the headhunter's amulets, to have their names celebrated in chant and song, to be honored with gifts of rice and meat, and to sound their tambolâ and deliver a skilled and impassioned oration to the ritual gathering. "It was great," a ToSalu friend recalled, thinking back to his first time among the topangngae in the 1950s, "Everyone was looking at me and my chest was just bursting. I was important. Besides that, I had never seen so much rice, and it was all for me."

Once a youth has proved himself a peer in the company of adult men, there are no further statuses to be pursued. The game becomes one of flexing and holding onto one's status as an adult man by marrying, raising a family, and taking part in village affairs. Yet the sameness of peership, linked as it may be to ideas of harmony and social accord, shows real instability. For example, in the sphere of a domestic politics grounded in uxorilocality, a young husband is said to "live beneath the ends of the floorbeams" of his wife's natal home. As such, a husband must subordinate himself to his parents-in-law. Trysts with married women bring shame to cuckolded husbands and threaten to steal from them both spouse and home. And a man's desire for greater wealth, influence, and reputation—the desire to be a tomakaka' (the prosperous one, who is "like an elder sibling" )—also stirs beneath or behind the politics of peership. Whether to defend himself against a loss of esteem or to enlarge his reputation and influence, a man will often return to the ceremonies of pangngae as a headhunter and once more assert his manhood and difference. These accomplished ones who seek to "enlarge their self-image" (umpokasalle kalena ) in turn can lead novices into the mystery of the headhunt.

What I describe for the mappurondo communities—tensions between ideas about social difference and ideas about sameness, tensions theorized and animated by envy and emulation—bear a decided resemblance to the social dynamics informing Ilongot headhunting (M. Rosaldo 1980, 1983; R. Rosaldo 1980, 1984, 1987). Among Ilongot youth, there is a strong desire to emulate the violent feats of their "fathers" (M. Rosaldo 1980:139), and to wear the coveted red hornbill earrings that will mark them as accomplished men (R. Rosaldo 1980, 1984, 1986). Especially striking is the intensity of the youth's envy. Ilongot youth "coming of age undergo a protracted period of personal turmoil during which they desire nothing so much as to take a head. . . . Young men weep, sing, and burst out in anger because of their intense, fierce desire to take a head" (R. Rosaldo 1984: 191). This adolescent turmoil involves a "tremendous sense of envy or rivalry," a deep "want to become like older men" (R. Rosaldo 1987:242). The emotional intensity of Ilongot men also distinguishes them from their wives and sisters. The hearts of men, according to the Ilongot, are "higher" than women's, and "exceed" them in the intensity of anger, passion, and "knowing" (M. Rosaldo 1980). But curiously, women's envy of the headhunter is missing or muted. As suggested by my discussion of grief and violence in the preceding chapter, the intensity of emotional expression among the Ilongot throws into relief the emotional restraint of the mappurondo

villager. As with anguish, so too with want. Envy and desire may not burn as brightly in the hearts of mappurondo youth as they do in the Ilongot, but their presence is immanent in the discourse of pangngae.

My immediate purpose in this chapter on manhood, then, is to explore the discourse of headhunting ritual for signs of envy and desire, and to say something about their rhetorical force and formulation. As a first step, let me turn to everyday terms for these forms of wanting. Perhaps the most common word meaning to want, like, or wish for something is maelo , which derives from the root elo and the stative verb prefix ma -. It implies a willful state of desiring something. Envy, like grief, is rendered as something that happens to a person without their being able to foresee it or control it: kaindai (from the root inda , the prefix ka -, and the transitive verb suffix -i ). Falling somewhere between the most negative form of wanting (kaindai) and the more positively valued expressions of want (like maelo), there is the term kailui (from the root ilu ), "loving" or "coveting" something. These terms, as such, do not enter the ritual discourse of pangngae as the most prevalent expression of want and envy. But they offer a vantage point from which to understand the celebration of manhood.

The Week of Adornment

A'dasan ends with the light of dawn and the cohort's noisy entrance into the hamlets. As the topangngae move from house to house letting out their shouts and blowing their tambolâ, drumming erupts throughout the village, following the warriors as they release home after home from the prohibitions of mourning. Drums (ganda , also gânda ) have long been cinched up to the rafters of any homestead that owns one, silent since the pa'bisuan ceremonies of the previous year, or since the death of an important villager. Men and youths now lower the drums, and in teams of two or three find the correct beat for the distinctive, stirring rhythms of pabuno (literally "one that kills with a spear"), the percussive emblem of headhunting ceremony. As pabuno swells through the hamlets, males break out with loud whoops, laughter, and the mua'da' shouts of "lelelele." The drumming erupts throughout the day at the whim of the village males, and it is very clear that most youths take real pleasure and pride in beating out the rhythms of pabuno with exuberance and finesse. It is only with the onset of evening and night-long tabus that the drums fall silent.

Fig. 9.
As the sun rises each morning during pangngae, village men lower drums from
the rafters of their households and unleash the exuberant rhythms of pabuno, "one who
kills with a spear." 1984.

The hamlets now belong to men, so to speak, and above all to the headhunters, who are referred to as tobarani , "the courageous," in recognition of their warriorlike bravery and feats. The tobarani carefully hang their tambolâ on the door frames or interior houseposts of their home. For the next seven days the instruments remain silent and on display, transposed into eye-catching icons of a warrior's knowledge and power. Although men and women may from time to time sing sumengo in praise of the tobarani, the drumming of pabuno dominates the musical "space" of the village, its masculine signature holding the hamlets in its grip. Indeed, in a significant sense the drumming is an enacting sign of masculine power, of irresistible social control: no one can not listen to the drumming. It is everywhere and thus consolidates the community through the senses as a listening polity, legitimating its order under masculine authority. Kept by pemali tabus from handling or playing the drums, women are constrained from subjecting themselves and others to these compelling and enlivening rhythms; women cannot find political authority over the senses through drumming. In this light, drumming is part of that masculine "excess" that provokes the envy and desire of women.

The severed head, cloaked in a rice pouch, rests in the home of the tomatuatonda' or the babalako (or topuppu). The headhunters are now free to relax at home or wander about the hamlets during the day. Throughout the community, villagers honor, even spoil, the tobarani, inviting them to share a meal or to take a gift of tobacco and betel. For the time being, the headhunters are forbidden to undertake any kind of labor, and may eat only rice, meat, fish, and eggs, the point being to avoid the "feminine" greens from women's gardens. Evenings are spent relaxing in mirthful gatherings. Youths—and especially the unmarried topangngae—find opportunities in these gatherings to flirt a bit with unmarried girls. At night, however, it is customary for the tobarani to sleep at the household of the tomatuatonda' or at the house of the cohort leader rather than at their own homes. Seven days and six nights pass in this fashion, during which time villagers ready the hamlets in anticipation of the grand closing festivities of pangngae.

A new set of objects, signs, and gestures is brought out of safekeeping. So far, the trophy head and the tambolâ have been the principal working objects of pangngae, signifying its violent destruction of mourning and the "other" downstream. Villagers now bring out rings, bracelets, beaded necklaces and sashes, headcloths, shining lime and betel-quid containers, and fine clothes. These adornments (belo-belo ) figure importantly in the principal discursive project of the ritual until its closing: glorifying the tobarani and, more generally, exalting the image of men and manhood. In spite of their

Fig. 10.
The jewelry of the headhunt, and the basket in which it is kept. The to-
barani—the courageous ones who went on the headhunting journey—wear rings
and bracelets like those to the left. Women wear necklaces similar to the one on the
right. Headhunters will sport a different kind of necklace, one usually made of
silver. 1985.

superfluousness, or rather, because of it, the adornments of the topangngae are crucial in projecting the headhunters' "excess," their possession of "something more." Worn on and around the contours and features of the body, adornment opens up a radiant and stylized surface where the strivings of self and community become fused. No mere aesthetic frill, adornment is key to the politics of envy and emulation.

The penetrating remarks of Georg Simmel's (1950:338-344) essay, "Note on Adornment," require mention here, and most of what follows below reflects his brilliant understanding of the politics of ornamentation.^[3] Adornments are worn to catch and please the eye. It is by way of that pleasing that a person "desires to distinguish himself before others," seeking their envy (338). "Pleasing may thus become a means of the will to power: some . . . need those above whom they elevate themselves by life and deed, for they build their own self-feeling upon the subordinates' realization that they are subordinate" (338). Thus adornment is both self-interested and altruistic, singling out the wearer with objects designed for the pleasure of others, but signifying, too, the wearer's paradoxical dependence on those he wishes to

surpass. The style and radiance of bodily ornamentation absorb the person's individuality, enlarge it. They sensuously announce the wearer's social power and draw the gaze of others. If adornment creates a space for envy ("I have something which you do not have" [342]), it also foments desire by saying "I am here for you . . . take me." As Simmel writes, "adornment creates a highly specific synthesis of the great convergent and divergent forces of the individual and society, namely, the elevation of the ego through existing for others, and the elevation of existing for others through the emphasis and extension of the ego" (344).

I have already introduced key terms for the project of adornment in the mappurondo headhunting rites: umpokasalle kalena , "enlarging the self." Some further comments may be helpful, however Kale denotes neither an interior self (which is usually rendered as penaba , or breath), nor a person per se, but rather the image of the self, the surfaces, features, and reputation a person presents to the world. Adornments and being adorned, meanwhile, are rendered, respectively, as belo-belo and dibeloi (from the root belo ). In local tongue one could speak of decorating the self-image by using the phrase dibeloi kalena , a grammatical form that puts focus on the self-image and assigns agency to someone other than the wearer. This construction captures nicely the political dynamics Simmel so carefully ascribes to adornment.

Like the red hornbill earrings coveted by the Ilongot (M. Rosaldo 1980; R. Rosaldo 1984, 1986), the white beaded bracelets of the tobarani indicate that the wearer has been on a headhunt. In this sense, the bracelets are also the amulets of someone who knows the "secret" of a man's journey to other regions. The tobarani thus are doubly adorned, for the secret also serves as an adornment. As Simmel (1950:337) explains, "what recedes before the consciousness of others and is hidden from them, is to be emphasized in their consciousness; . . . one should appear as a particularly noteworthy person precisely through what one conceals." Further still, the bracelets and finery of the tobarani mark them as persons who have undertaken a heroic task for the collective good. Their deed, a flashing ornament itself, is both altruistic and self-interested.

The week between a'dasan and the closing ceremonies of pangngae finds the tobarani decorating their tambolâ with cascades of plaited streamers made from palm leaves of pale green and gold. On the seventh day after a'dasan, guests arrive from other villages, and hearths become heavy with smoke and noisy talk as people prepare for the evening's ceremonial songfest, called umpaningoi ulunna , "playing with the head" (in the dual sense of putting it on display and teasing it). All during the week, men have been busy preparing the house that will host the gathering. Needing room for

Fig. 11.
Two tobarani, their faces caked with rice paste, their bangs trimmed back.
They are being honored with an evening of song during the ceremony known as
umpaningoi ulunna, "playing with the head." 1984.

roughly 250 to 400 people, they have added floor supports, moved walls, made makeshift porches, and put in place extra ladders. Having enlarged the house, they put in final supplies of firewood, rice, and coffee. As the afternoon passes, the drumming of pabuno grows more fervent and continues without break. Sporting their bracelets and necklaces, the tobarani have their bangs cut by their mothers or sisters, and cake their faces with rice paste in order to lighten and refine their complexion (see Figure 11).

With the setting sun, people throng to the site of the ritual. As the crowd swells, women and girls from the village pack themselves in between the hearth and the downriver side of the house centerpost, which protrudes up through the floor. The beaded and white-faced headhunters begin to arrive, and after hanging their tambolâ in display, find places to sit on the floor upriver from the centerpost. Once it is dark, the drumming comes to an abrupt halt and the entourage of village elders—all men—climb up into the house and solemnly take seats beside the centerpost.

Choruses of men or women begin singing sumengo as the elders help themselves to ample amounts of betel and tobacco from plates being passed around. Some verses extol the radiance of the tobarani:

Another set of metaphoric lyrics likens the tobarani to a kind of sugarcane known for its colorful striping. As the cane ripens and matures, its splendor grows:

The lyrics portray women and girls as parrots—songbirds, I should note—that "dance" back and forth along their perch straining to see the growing sugarcane.

Songs like these not only extol the radiant virtues of the tobarani but also confirm the admiring gaze of others, a gaze that looks up to the headhunters. At the same time, the song performances function as adornments in and of themselves. In a significant way, these lyrics of praise gild the headhunter's image and magnify his presence. Songs can also glorify the name of the elder who led the headhunters on their journey and made it possible for the village to hold the ritual:

The splendor of the topangngae and the praise lavished on them can stir envy among the men and youths who for one reason or another did not join the headhunters on the trek downstream. Some of this envy finds its way into song, and thus to the cars of the headhunters:

Coveting the headhunters' heroic glory, other men and youths vow to travel the seaward trails. The time and place are uncertain, but they will be the to-barani's equal:

The turbulent longing for self-distinction, for being above all others, must be tempered with selfless aspiration to help (or subordinate oneself) to a community of others. If a man thirsts with envy, his heirloom knife (connoting moral tradition) must ache with purposeful desire:

As for who will be fated to make good on their pledge and then "bathe" in communal adulation:

As these examples show, the songs of pangngae celebrate and foster an envy that is productive, rather than destructive, of social order.^[4] An emulative rivalry lies at the heart of male peership and animates it, constantly cycling men through alternating positions of "the gazing" and "the gazed upon." This rebounding mimetic thrives on the bracelets and beads and the songs of praise, adornments that become the focus and incarnating objects of desire.

Women, too, may covet the headhunters' glory, but are denied it under the exclusionary tabus of pangnage. Lyric variations in two sumengo performances I recorded on separate occasions (and in different villages) suggest that women experience a mixture of anxiety and envy during the ritual; or perhaps they reveal an episodic shift in women's experience, a shift in which envy supplants anxiety as the ritual unfolds:

Forestalled from taking part in the mimetic rivalries that conjoin men and youths, women look ahead to the emulative politics that separate and distinguish them from males and the masculine order. Women, too, will have their measure, vowing that no one (no grain of rice) shall be excluded from the community (metaphorized as a cup made from a coconut shell):

Songs such as these refer in particular to the women's ritual known as malangngi '. Difficult to conduct and seldom held, malangngi' not only has a significant place in local ideas about womanhood but also figures as an op-positional or inverting complement to men's headhunting rites. I never had an opportunity to witness malangngi', but in describing the ritual to me, an elder volunteered, "Malangngi', it's the women's pangngae." With a little probing from me he continued, "In pangngae we exalt men. In malangngi' we exalt women." Others, particularly women, were careful to point out that pangngae concerns the village, whereas malangngi' concerns the home and offspring. The ritual promises such real and obvious relevance to the social relationships and practices I am trying to explore that a quick sketch would seem appropriate here.

I am told that malangngi' is set in motion when a group of women and girls gather together in secret under darkness and slip out of the hamlets, moving in an upstream direction.^[5] At the edge of the village lands, they climb up next to a banyan tree and onto a platform built especially for this ritual. While beside the tree, they become possessed by debata. Said to turn into birds as they dance, the women utter the sound of female sumanga', and issue a laughing wit-wit-wit-wit . Panicked by the absence of their wives, sisters, and daughters, village men launch a search and discover them around the banyan. A few men are sent to fetch drums. Upon their return, ritual specialists play set dance rhythms for the possessed women and, after a time, "drum them down" from the banyan. Climbing onto the backs of their

husbands or fathers, the women ride back into the village and ascend into a house where they dance and display amulets and curatives obtained "up in the sky" (the meaning of the word malangngi ). A ceremony extolling the women, the debata, and the prosperity of the household lasts into the afternoon of the following day.

Little wonder, then, that women seek to journey to the sources of their own order so as to momentarily subordinate village males. As one song puts it, men will know the weight of women:

So that there is no mistaking malangngi' (and other rituals of pa'bisuam baine) for a fleeting and insignificant inversion of social hierarchy, I note that during the course of pangngae men express their envy of women. For example, there is a song in which the headhunters see women as more accomplished than themselves. As expressed in the lyrics, women come into their own (metaphorized as the task of weaving a blanket) long before men do; the headhunt is a man's first crawling effort to come into manhood:

If these songs are any indication, mappurondo social order thrives on reciprocal envy within and across gender categories. The desire to distinguish oneself, to surpass others, is something relevant to the experience and interests of both men and women. The power to attract the gaze of others, to stir envy within them, then, is perhaps the most sought-for talisman of superiority and excess being. Yet the exhilaration of being powerful and desired also gets tangled up in ideas of sexual potency. An adoring lover or wife can become a talisman of the headhunter's power. That is, from the vantage point of male peerage and its internal rivalries, women and women's adoration undergo alchemical change to become gilded signs and amulets of the headhunter's elevation above others.

There is no question in my mind that the return of the headhunters marks a festive renewal of erotic discourse, something very much welcome after the long period dominated by the projects of mourning and field labor. Both men and women take part in the mirthful sexual banter. Youths come back from the headhunt feeling their most attractive—vigorous, mas-

Fig. 12.
A young woman about to give betel and areca to two tobarani during the
evening ceremony of "playing with the head." 1984.

culine, and socially important—and are ready to impress and seduce young women. A young woman, too, may show a flirtatious interest in someone, perhaps making attempts to catch the eye of her favorite tobarani. Meanwhile, it is not unusual for older men to boast of their continued potency.^[6] Explicit or demonstrative expressions of sexual interest toward specific persons, however, are out of place. Accordingly, villagers resort to covert or indirect gestures to signal their interest in one another. Indeed, the songs are a safe place to voice their interest in one another.

Consider, then, the cleverness of a song such as the following. On one hand, it is about the shy flirtations that might go on during ceremonies. On the other hand, it can be directed toward someone in particular; sung while giving a revealing glance to someone in the crowd turns the performance itself into a flirtation.

By way of contrast, some lyrics can be fairly graphic in the way they depict lovemaking. For example, one song I heard sung by women captures the sweat and tangle of sex, and makes a rather obvious pun about the singing cockerel, metaphor for both headhunter and penis:

Like the previous example, the song can be performed in such a way as to serve as a come-on to a favorite lover sitting in the crowd.

Seduction and courtship can require stealth on the man's part and, in that sense, are seen as activities that have something in common with head-hunting. A lover must be lured toward a man with flattery and sweet-talk (ussinggi '), not unlike the way the headhunter may lure a victim with magical incantations or baited charms.^[7] Songs like the two that follow candidly depict the whispered sweet-talk of a man who has gone to his lover's house under the cover of darkness in order to sleep with her.

Metaphoric associations that connect headhunting and courtship also prompt different "readings" of sumengo lyrics. For example, both can be likened to the felling of a tree, as in the following sumengo:

From the vantage point of intercultural relations, the song is about felling a Mandar victim. It expresses hope that the victim's head will "fall into the village" and bring prosperity to the upland community. But several ToSalu Mends—all men—also suggested that these lyrics alluded to a marriage proposal. In their view, a woman does not pass into full maturity until she is "felled" by a marriage proposal. A tree that falls into the village becomes timber; a woman who falls into the village becomes a wife.

Other song-flirtations suggest a masculine anxiety that needs to be soothed with a woman's expression of desire. In an exchange of ToIssilita' songs I recorded at Saludengen, men sang the following to a chorus of women:

The lyrics pose a choice of objects, a wooden dish from the mountains, or a porcelain bowl from the coast. As icons of upland poverty and coastal wealth, the dish and the bowl are also metaphoric displacements of rival masculinities. The chorus of women responded, reassuring their husbands and lovers that they like things that come from the mountains, from "home:"

Moments later, the men and women exchanged another question and reply:

In these sumengo, the men probe the feelings and attitudes of village women with the purpose of adorning themselves with words of flattery and reassurance. As one might expect in a ritual that celebrates and enlarges the image of mountain men, the choices of the women reveal a preference for things and persons less prestigious, and a faith in virtues hidden beneath humble appearances.

The oscillating rivalry between men and women is mirrored, too, in the agonistic and antiphonal play of songs and song riddles.^[8] Part of the challenge can be to defeat other choruses simply by outsinging them. Some of the challenge also has to do with knowing an appropriate sequel or rejoinder to the preceding song. For example, a chorus may pose enigmatic questions and riddles in their sumengo, as when men ask:

Women respond with a conventionally spare answer that conceals more than it reveals:

The same holds true of the following sumengo-exchange, again between men and women:

The women are not being evasive in these examples. By concealing or withholding answers to the riddles, the women are able to possess something secret and thereby confer upon themselves a radiant noteworthiness. Although they do not divulge it, they know that the first song refers to a braid, and

that the stringless top in the second song is a headhunter. (Just as a top may spin away for a moment and then return to hover in front of the person who cast it in motion, the headhunter goes off from the village and then comes back to celebrate for the seven nights of ceremony that make up pangngae.) Sumengo exchanges such as these make up a game of knowledge. But the game of knowledge is also a game of power: to pose questions is to exact a hold on another, and to give answers is to escape that grasp (Canetti 1978:284-290). Seen in this light, men and women are playing for power and control over one another during sumengo performance. Yet neither achieves dominance except for a few moments. The conventions of sumengo performance are such that men and women strike a balance with one another within the game of question and answer, image and counterimage, secret and countersecret.

What is perhaps already obvious is that the song-fest of umpaningoi ulunna conjures an image of women as men would like them to be, an image that obscures (and possibly denies) possibilities for a woman's own subjective presence in the social order. These are women who desire a headhunter and envy his excess of being, who offer reassuring words to their husbands, and who like playful sexual banter. If sumengo performance throws the yoke of complementarity upon men and women, it nonetheless appropriates the feminine so as to exalt and construct a manhood reflecting the prestige anxieties of men. In short, a woman's songs still trace the reach of masculine discourse during pangngae. Her voice does not speak with the free authority of her own experience but, rather, with the purpose of turning man's gaze back upon himself. In no small sense, it is man's self-absorbed gaze, displaced and mirrored back to him through women, that adorns him with such radiant effect.

Yet everyone's gaze must turn, too, to the trophy head, the troubled sign of the tobarani's astonishing violence. Led by the elders, the gathering takes part in a liturgical performance that unites the social and moral hierarchies of the headhunt. The topuppu and the entourage of men sitting beside the centerpost take up the sacred strains of the sumengo tomatua , the "sumengo of the elders":

The topuppu then calls out to the leader of the tobisu ("the ones of quickened spirit"), as the warriors are now called, for the head to be brought. One of the tobisu fetches the head and presents it, still wrapped in a rice

pouch, to the topuppu as the gathering shouts "It stinks! It stinks!" He puts the shrouded head on the centerpost and places a headcloth on it. Calls for betel and tobacco go up, and gifts of the same go to the topuppu, the entourage of village elders, and the head. As more betel and tobacco pass through the crowd, men cry out Sanda! Sanda! ("Enough! Enough!"), in a stylized plaint of communal overindulgence and excess: the abundant hospitality (an index of hamlets' and homesteads' material prosperity) has sated and overwhelmed any desire to consume.

The gathering then quiets as the topuppu speaks to the head in ritual address (mualu uln ). Soothing the victim with gifts of tobacco and betel, the topuppu explains the headhunters' act of beneficial violence:

The topuppu and his entourage then erupt with the mua'da' shout and are answered by the men in the crowd. Once more, the elders lead the gathering in singing the sumengo tomatua:

As soon as this very solemn song comes to a close, the topuppu and the elders begin the antiphonal ma'denna, taunting the head, but naming, too, each tobisu who journeyed downstream:

(A complete example of the ma'denna is given in Chapter 3.)

In mocking the head that sits dumb and nameless on the centerpost, the ma'denna dissects and counterposes the male political bodies from upstream and downstream regions. Lyric blows of the ma'denna dismember Mandar political anatomy, while celebrating the names (not the titles) of village men. As the elders and the gathered throng sing out the names of the tobisu, they also connect their names to metaphors of moral virtue. The name of Ta'bu, for example, may be sung out in association with "the straight path"—a metaphor for dedication; that of Sa'du, with "thick iron"—the symbol of endurance. Thus the song also articulates a moral anatomy specific to the men's political order in the village.

If the ma'denna distinguishes the tobisu as virtuous individuals, spreading their names and placing them above others, the ensuing kelonoson celebrates them as a distinguished group. Once again, the entourage of elders leads a mua'da' shout and the sumengo tomatua. Then the kelonoson begins, a performance that combines song and dance to chronicle the assault on the community downriver.^[9] As they sing each verse, the elders sitting beside the head move their hands and arms in a dancing gesture. After each verse, the headhunters stand up around the elders, and while repeating the verse, perform the dance of warriors (slowly skipping, and waving their arms from side to side). One performance at Salutabang went as follows:

The kelonoson portrays the bravery, comradery, and cooperation of men. But the final three verses are especially revealing of the ways in which the violent adventures downriver prefigure a game of dominance in the peer politics of village males. The closing lyrics of the kelonoson entwine threat with unsurpassed sexual bravado. Wife-stealing is a dreaded and explosive issue, for the theft of a spouse undermines a man's prestige and status as a male peer. The lyrics remind those males who did not journey on the headhunt that the sexual license and stealth of the tobarani comes at the expense of other men. At the same time, the cultish and self-admiring boasts of the headhunters persuade them that their irresistible will to dominate can awaken desire in women.

Dipandebarani: The Feast for the Valorous

The sumengo-fest of umpaningoi ulunna can last well past midnight. So long as voices remain strong and plates of betel and tobacco keep circulating, singers are happy to stay up and sing praises to the to-barani. As people drift off to their homes, the topuppu makes his rounds of

Fig. 13.
A small chorus of women singing the somengo during dipandebarani, "
feasting the valorous." 1984.

the hamlets, stopping at the home of each headhunter to make offerings to the debata bisu muane —the spirits who quicken the hearts of men. Standing in the firelight of each household's hearth, the topuppu softly chants over the gift of a chicken and seeks providential blessings for those fulfilling their vows:

With this sacrificial gift, the headhunter not only makes good on his oath, but exceeds it. Returning sacrificial meats and rice to the debata is a sign of the abundance that flows to and adorns the virtuous man.

At dawn, the rhythms of pabuno once more awaken the hamlets. Morning will see the mappurondo villagers gather for the final ceremony of pangngae, dipandebarani ("feeding the valorous"), the rite that honors the headhunters. Celebration reaches its peak in dipandebarani: drumming and song are at their loudest and most exuberant; decorations and finery are at their most elaborate; and food and drink are at their most abundant. Getting ready to appear before the village, the tobarani rub off the last flecks of paste from their faces, leaving a fine powder on their skin. Garbed in their finest shirts and headcloths, and adorned with bracelets, ceremonial knives, and beaded bandoliers, they take up their tambolâ and go to the ceremonial house.

If the harvest has been a good one, the gathering is sure to put up a tremendous clamor of song, drumming, talk, and laughter. The swirling sound of choruses and drums delights and overwhelms the senses. But as Plessner (1970:68) reminds us "We must not forget that sound includes the power of self-verification: one hears oneself." In this sense, the swirl of sound is an icon of the community in its vanity, exhibiting itself in unceasing festivity. Noise and music are adornments, simultaneously enlarging and engulfing the crowd that wishes to hear itself:

Fig. 14.
Radiant in his finery, the head of the cohort listens as villagers celebrate his
name and deeds in song. 1984.

Fig. 15.
Women and girls cluster near the sons, brothers, and husbands who have
gone on the headhunt. The young headhunter pictured here has adorned himself with
headcloth, necklace, bracelets, beaded sash, and his best shirt. 1984.

The music and noise only subside when the topuppu and the entourage of elders begin to conduct the liturgical acts intended to sacralize the headhunters as tobisu and to present the debata with the head and gifts of meat and rice. The liturgical sequence for dipandebarani repeats much of that performed during a'dasan and the previous night: the sumengo tomatua, the ma'denna, and the kelonoson. To these is added mararra topangngae anna tambolâ ("anointing the headhunters and the tambolâ") and the ma'paisun (when the babalako or topuppu presents the surrogate head to the de-

Fig. 16.
A villager brings a gift for one of the tobarani: a bundle of rice, fish, and
meat, decorated with betel leaves, clove-scented cigarettes, and small packs of sugar
cookies. 1984.

bata; see Chapter 3 and below). To start the marrara, the tomatuatonda', in his role as the political head of the community, will chant words to the bird or pig to be sacrificed:

The tomatuatonda', assisted by the topuppu or babalako, then moves among the tobisu and their tambolâ, dabbing them with sacrificial blood. In contrast to the dismembering textual blows of the ma'paisun (see Chapter 3 and below), the gestural discourse of the marrara conjoins, underscoring anatomical movement and strength as the tomatuatonda' anoints each warrior on the ankle, knee, palm, elbow, throat, jaw, and forehead.

Not long after the marrara, the elders ready themselves to present the head and sacrificial foods to the debata through the liturgical act called ma'paisun ("making an offering"). The topuppu or babalako places the head, packets of meat and rice, and a bamboo vessel filled with water in the loft that faces the headwaters of the river and the rising sun, and then begins his chant to the debata:

The chant gives way to a violent crescendo of shouts and noise from the crowd. Women and girls let loose a rain of areca nuts, pelting their favorite men, while up in the loft the topuppu flings a rain of sacral water onto the gathering below. With that, the topuppu closes the ma'paisun:

After the ma'paisun, the village head and the topuppu oversee the division and presentation of the tarakan , broad trays heaped with rice, fish, meat, and eggs for the headhunters.^[10] These prestige foods alone display the community's esteem for the tobarani, but their presentation also includes a gestural demonstration of excessive regard, an enacted rhetoric of over-abundance. As the two elders point out each warrior, a man enlisted to work as a carrier brings forward a large winnowing tray of food hoisted on his shoulder. Acting as if he were struggling with a load of enormous weight, he lets out a mua'da'-like shout ("yuhuhuhu "), buckles at the knees and carefully drops the tarakan tray before the headhunter. This grossly exaggerated portrayal of someone straining and collapsing under a heavy load is itself an enlargement, a stylistic icon of overindulgent excess: too much acting in presenting too much food. Whether the harvest has been meager or abundant, the tarakan always occasions this demonstrative excess. Though it never surprises, the tarakan always impresses.

As the warriors get their tarakan, a team of men pass out smaller packets of food to everyone else in the gathering. No one touches his or her food

Fig. 17.
  The tobarani sound their tambolâ for the final time before flinging them up
into the loft. 1985.

until all has been distributed. As the warriors sit solemnly, if a bit wide-eyed, with their trays before them, the village head designates each tarakan by calling out the name of the headhunter to whom it is being given. Referring to all the foods passed out, but calling forth, too, a hierarchy of social difference, he closes with, "This is for the men. This is for the women. This is for the guests." With that, the men in the gathering make a sudden diving leap across the room at the tarakan of the warriors, digging and tearing at the food, and shoveling a gulp or two of rice and meat into their mouths as the tobarani beneficently look on, making no effort to keep the men from their food. This violent gesture of commensality wrests away some of the headhunters' excess: it is a self-conscious and stylized effort to level hierarchical differences among village males. With it, a community of men place themselves on equal footing, all sharing in the headhunters' glory.

After taking a few mouthfuls of meat and rice, headhunters and guests alike bundle up their portions of food. Later, they will take the packets home and eat there. Singing and drumming resume until about noon. As the sun reaches its zenith, the headhunters take their flutes (which have been silent since the closing moments of a'dasan) and gather near the centerpost. Stand-

ing below the edge of the loft, they begin umbua' tambolâ "raising (or displaying) tambolâ" Urged on by the crowd, the tobarani blow a long note in unison on their instruments for roughly twenty seconds or more (the longer, the better, it is said), and then deliver their mua'da' shout three times: "Jelelelele huuui! " Twice more they sound a long blast on the tambolâ and follow it with three shouts. With their last shout, the headhunters heave their tambolâ up into the loft, where a waiting male will stack them next to the head and the paisun offerings. At the same time, drumming begins again and singers launch into the sumengo tomatua.

Words that Make the Floor Shake

As the rhythms of pabuno pour from the drum, the cohort leader, the tobisu, and any other man so moved take turns leaping up to deliver impassioned speeches at the center of the house. Villagers call this speechmaking mamose ("stomping in fury"). As a man rises, the drumming comes to an abrupt stop. Stepping forward, the speechmaker brings his foot down with a pronounced thud and then circles the floor counterclockwise as he speaks, his right hand raised and sometimes holding a kerchief-sized cloth (the symbol of the Pitu Ulunna Salu league). Delivery is especially striking at Saludengen, where the headhunter-speechmakers take bounding, floor-shaking jumps to the center of the house, and circle about with hardened faces and fiery eyes as they talk. When a speaker finishes, pabuno bursts from the drum. To the cheers and shouts of the gathering, the speechmaker circles the floor once more (sometimes carried atop the shoulders of men, sometimes simply by leaping about). Another speaker takes his place until all those who wish to give a speech have done so. In principle, any male may speak during mamose. But in practice, it usually falls to but a few elders and a headhunter or two to make these impassioned speeches.

Mamose offers a way for a man to display his political maturity and readiness, and thus his peership with other men: a skilled and daring headhunter, he is willing to risk danger out of dedication to village tradition. Capable of taking a life and defending his own with force, he is also a wise and talented speaker who knows how to stir the hearts of others. A youthful warrior, he is self-effacing and humble. Advancing in age, he is ready to counsel younger men. In my experience, young men look up to those elders who speak with wisdom and passion. In this sense, mamose speeches are coveted verbal

forms that inspire others to oratorical heights. With mamose, a man does not just desire to be like other men, but strives to be a leader, a first among equals.

Unlike the agonistic play of sumengo performances, no one tries to outdo the other in delivering a speech.^[11] In fact, a good speech requires some gesture of self-effacement, of humbling oneself before others. On the surface of things, such a gesture serves as a way for a speechmaker to disclaim any interest in distinguishing himself or enlarging his reputation and image. But because this rhetorical tactic of self-effacement draws the admiration of listeners, it only increases the speechmaker's radiance and greatness. At the same time, a good speech must convey the speaker's regard for upholding tradition. So in addition to humbling oneself before others, a speaker tries to commit himself to ritual tradition and the common good.

Youthful headhunters who lack experience in speaking before the community are quick to grasp the importance of playing down the self and playing up tradition. In fact, in most cases their speeches do not go beyond well-rehearsed or memorized expressions of humility and self-sacrifice. However sincere, their efforts tend toward the formulaic and the grandiloquent. For instance, one young headhunter leapt onto center floor and opened his speech with two familiar if well-worn allusions to fate:

With these words, the speechmaker tells the gathering that his becoming a headhunter was unplanned. It was his unforeseen fate to take part in the ceremonial journey downstream. But here he is, a mere infant, so to speak, ready to show what manhood is all about:

If the village elders will hold fast to tradition, he will do his part for the communal good:

The speech is a good one, as first speeches go, for the speaker has shown modesty (to paraphrase: "fate brought me here," "I'm just a child"), and pledged himself to the social and moral order of the community ("I will follow the elders," "I will serve the village and its tradition").

After calling for the attention of the crowd, another young headhunter protested that he did not know very much or have much experience holding forth in speech:

He went on to say that it is hard to make the floor shake, that is, to summon enough spirit and purpose to come before others as a headhunter. It is hard, too, to be an orphan—hard to know what to say and do when one's elders pass away.

His words may reflect a personal loss in the recent past. They might allude, too, to some of the pain and directionlessness a community may feel at having lost an elder leader. Yet the young speaker has been moved to become a headhunter and to stand before the gathering:

Perhaps it was the pleasure of celebration and the attention showered on men that drew him forward. Or perhaps it was a moral calling, for the speech ended with an uplifting if clichéd slogan:^[12]

Combining formulaic expressions of modesty and fate, along with well-worn mottos, some speakers might deliver a more exhortatory speech, like the following:

The speeches of older men tend to be longer, to show more polish, and to be more reflective than the speeches given by the young headhunters. Having less need to prove their dedication to ancestral teachings, older speakers, like the one who gave the following speech, commonly urge their listeners to stay upon the path of mappurondo tradition. Like many other elders, he begins his speech with a self-effacing remark about being crippled or put together wrong:

He shows his modesty and humility, describing himself as a small thread on something greater (i.e. the woven blanket of tradition, ada' mappurondo). He continues with self-effacing comments as he exhorts others to be pious and strong-willed:

And he closes with words of praise for the radiance of mappurondo ceremonial tradition:

Yet orations like these pale when compared to the improvisatory rhetoric of the most skilled speakers. The following is a mamose speech given by Tasomba (seen in Figure 18), a descendant of Pua' Soja from Minanga. It is mamose rhetoric at its finest and most elegant: Tasomba is self-effacing, inspiring, and very knowing and deft in the way he takes hold of the moment to extol local tradition.

Fig 18.
Assuming a traditional oratorical pose, Tasomba struts across the floor and delivers
his mamose speech extolling pemali appa' randanna and mappurondo tradi-tion. 1983.

Then Tasomba invokes the memory of Ambe Ma'dose, whose recent death not only saddened the community, but also took away one of its most prominent leaders. In this village, his death was a blow to mappurondo tradition itself.

Next, Tasomba began to address my being in Minanga for pangngae. Elders in the past told of a stranger who would come, but who had given it any thought?

As Tasomba saw it, I was drawn to Minanga and ada' mappurondo by the ghosts of their ancestors. My being there was also reason to be glad and to hold fast to tradition:

Tasomba now reached the point where he would urge the gathering to hold firm to the virtues of manhood and womanhood and to tradition itself.

Ready to close his speech, Tasomba says that his are not the words of someone who is past his prime, but rather, of someone who is still young and virile, of someone who is still a man:

As may be clear from the speeches of Tasomba and the others, the principal idea behind mamose is to animate the community such that it will follow the moral path of mappurondo tradition. Put somewhat differently, the speechmakers not only call for moral and ideological solidarity on the part of the gathering, but try to spark emulative action as well. Those making speeches thus put themselves forward as moral actors who are to be emulated, as daring and audacious headhunters, as self-effacing speechmakers, but above all, as ones who are animated by tradition. But being animated by tradition is to have given oneself over to the mimetic interests of reenactment. In bringing himself before the gathering, the speechmaker presents the image of someone who follows the village ancestors and elders, who accepts their authority and their authoritative understanding of tradition, and who is ready to perform once more the sacred acts and ceremonies of ada' mappurondo.

The rhetoric of manhood, then, includes a portrayal of man as a rhetorician . That is to say, being good at being a man requires both the valor and violence of the headhunter and the virtuosity and authority of a mamose orator. With respect to pangngae, the violent and the verbal mark the extremes of ceremonial performance, to be reckoned not only as polarities of masculine discourse but as the beginning and end of the ritual's drama as well. In light of these extremes, the rite may be seen as a ceremonial transfiguration of manhood in which speech displaces violence. Successive modulations of

vocal sound trace a path across the social and moral boundaries that distinguish upstream and downstream regions. From the noise of violent ambush downriver, and up through the mua'da' shout at hamlet's edge, emerges a stirring and eloquent speaker. Yet the displacement of violence is never complete. Virtuosity in taking a victim and in giving a speech is likened to felling a tree—a task that falls to men. Cut correctly, the tree "falls into the village" (loe tama ri bamba ) and so can be used and enjoyed. Cut incorrectly, the tree "falls into the river" (loe tama ri uai ) and is washed away. Within the metaphoric discourse of masculine virtuosity, felling a speech and felling a victim are potential transformations and transcodings of one another, and recalling the sumengo performances that connect headtaking and courtship, so too is felling a wife.

Women's Coda

When the last of the mamose speeches comes to an end, drummers beat out the rhythms of pabuno for a final time. As the drum plays, a chorus of women may begin to sing the sumengo tomatua, a song that throughout pangngae has belonged to choruses headed by men. At this time, it is common for a married woman to go into trance possession and begin a lively but silent dance under the influence of the debata bisu baine —the quickening spirit of women. Her brief dance marks the close of pangngae and the onset of a different ritual order, pa'bisuam baine, an order dominated by possessed women and entrancing spirits. In a sense, the currents of envy in pangngae culminate here in the counter-emulative discourse of women. Under the consecrating exclusions of pangngae, women desired to be like the tobisu muane—the enraptured warrior-speakers—but were prevented from sharing their splendor. With the end of pangngae behind them, women are free to seek splendor and to explore the boundaries of otherness on their own terms. They will not be like men; they will be like themselves.

The head and the tambolâ already rest in the attic loft. Men now hoist up the ceremonial drum, cinching it to the rafters. With that, the festivity of pangngae is put away. In the discourse of song, the sense of ending brings with it sadness and regret. For example, women might have sung the following sumengo a little bit earlier in the day:

Fig. 19.
A woman's coda to pangngae: a lively but silent dance under the influence of
debata bisu baine—the quickening spirit of women. 1983.

The song looks ahead to the ritual's closing, when women will put aside their festivity, their songs put away for safekeeping in places associated with mothers, daughters, and sisters. Equally bittersweet is the song exchange of male and female choruses facing the close of headhunting festivities:

And so, the coda to pangngae belongs to women—perhaps it is more accurate to say that the coda offers an image or verbal icon of women—who

give themselves over to weeping and trance, their crying and dancing a somatic capitulation to unanswerable forces.

Felling a coastal rival from the past and shouting away grief are projects that give moral warrant to celebrations of manhood and the ceremonial politics of envy and emulation. Yet in my view, it is the rhetoric of manhood and the interplay of envy and emulation, rather than a will to violence and catharsis, that do the most to motivate mappurondo men and youths to pledge themselves to the headhunt. In this fairly egalitarian community of men, pangngae serves as a ladder by which youths ascend upward toward male adulthood. By no local account did anyone have to take a head if they wished to marry or to hold forth in deliberations with other men. But pangngae is a time for youthful males and elders alike to put their manhood on display and to assert ceremonially their peership with other men. In no small sense, then, being on display, being the focus for the envied gaze of others, is a part of what mappurondo manhood is all about.

In local thinking, ukusam muane (the rites of men) are synonymous with ukusam botto (the rites of the village). The liturgical chants of pangngae occasion the exercise of male authority over collective interests; they are authorized performances in which men hold a monopoly on the establishment of community and the maintenance of communal prosperity. But the authority of men, and the authoritativeness of the elders' liturgical speech (especially that of the topuppu and the tomatuatonda'), in part are grounded in and indeed surrounded by the aggrandizing discourse of praise and song. It is within this exuberant singing that the collaborative assent of the community, including its women, is manifest and recognized.

I do not know if Pua' Soja was ever a "real" headhunter. Maybe he, too, was a weaponless youth when lionized as warrior and orator. The idealized images of manhood that the songs of pangngae conjure up perhaps belong less to the present than to a time Pua' Soja glimpsed in adolescence. Unearthed from memory each year, the sumengo are the remnant bones of men who are remembered as powerful and large, who continue to stir awe, but whose memory stirs nostalgia as well. Although mappurondo understandings of masculinity and tradition still hinge on these images, anxiety and uncertainty color ceremonial practice. The civil order in Sulawesi today holds no place for mappurondo practice, and Christians and administrators more often than not show a lack of sympathy, and even contempt, for ancestral religion. As increasing numbers are lured away from the sacred path of the ancestors, the mappurondo community faces the grave problem of sustain-

ing itself and the practices it claims as tradition. In the past, there were no doubt fewer and less powerful ideological challenges to the mappurondo community. Still, I have little sense of the impassioned eloquence of mamose speeches from the precolonial past—a past, it will be remembered, that Ambe Assi' once despondently described as not fully human and intelligible. Yet listening to the exhortations of contemporary speeches, I cannot help hearing speakers trying to keep others from abandoning the mappurondo fold, from giving up on mappurondo tradition altogether.

6
From Violence to Memory Singing about Singing as a Headhunt Ends in Song

Pangngae, one might say, begins in violence and ends in commemorative representation.^[1] The realities of the headhunters' predatory journey are, and long have been, obscure and in flux. They happen downriver and over the horizon. It is a violence that is absent, something kept out of the community and away from the senses. That violence returns home, and is made present and known, only in the idealized representations of ritual discourse. In the present order of things, the headhunters' violence is doubly distant. After all, the purported victims are the figures of a bygone political order. The headhunters' violence is always marked by its absence and anteriority relative to its commemorative retelling in ritual chant and lyric.

The predations of the mappurondo headhunters can be storied in several ways. Thus far, I have raised the apparition of violence in exploring the polemics of domination and subordination that have colored regional history; in mentioning the untranslatable anguish of the bereaved; and in discussing the celebrated figure of masculinity. I do not consider these reflections to be wrong, but, rather, to be partial. They do not tell the whole story. In the pages that follow, I offer yet another perspective on the violence of pangngae, seeing it instead as an artifact of commemorative discourse itself. I will suggest that the violent subjection of a rival downstream is key to a kind of remembering that keeps the mappurondo polity in each village from passing

into oblivion. Villagers represent the scene of violence again and again, not so much for ieets mythic value, but to occasion the sociality and activity of remembering. Along with Eugene Vance (1979:383), then, I will argue that violent events can be both subject of, and generative force for, their reenactment in ritual chant and lyric. In short, the discursive practices of pangngae stage themselves around violence. Seen in this light, the commemorative work of pangngae keeps an act of exemplary violence moving through time (cf. Yengoyan 1985:171-172). But it should be remarked, too, that this very act of exemplary violence keeps collective memory moving through time.

I turn now to the problem of commemoration, and the construction of commemorated violence, because the difficulties facing the mappurondo community today have less to do with a crisis in meaning than they do with a crisis in remembering. Powerful social groups and institutions are asking villagers to forget or forgo mappurondo practices, to forsake what was already put in place by ancestral tradition. Yet I would venture that remembering and mimetic enactments long have been central to the reproduction of the mappurondo social order. "All beginnings," writes Paul Connerton (1989:6), "contain an element of recollection." The effort to renew and animate the community demanded a mimetic gesture of some kind that would recollect a privileged event from the past. Commemoration was a political end in its own right, and in the moment of its activity summoned or brought into being its own subjects and objects.

There are several aspects of reenactment that will concern me in this chapter. Viewed as practice, mimetic enactments include the expressive and instrumental dimensions we associate with representation and performativity. Yet ritual commemoration is also a mode of sociality, a means of shaping social hierarchy and social experience. As previous chapters clearly show, commemoration can serve other spheres of human concern, but it can also recruit such projects for its own ends. These projects have a great deal to do with shaping the social horizon within which the commemorative rite is staged, and for that reason may be an important source of variability or instability in commemorative activity. Over and against that variability, commemoration reveals yet another aspect, that of imitating its prior incarnations. In a sense, it is the purpose of commemorative tradition to produce a kind of simulacrum, "a copy of a copy for which there is no original" (Rosenau 1992:xiv). But at another level altogether, we can discover commemoration becoming a key theme in its representational surface (cf. Vance 1979). Commemoration will displace violence, and the community will sing about singing as the headhunt ends in song.

Commemoration as Adornment

In the last chapter, I introduced the theme of adornment in exploring the discourse of masculinity. I do so again, but now with the purpose of understanding commemorative song and commemorative tradition. I begin by reminding readers that the mappurondo ritual tradition called pemali appa' randanna evokes the image of a four-stranded beaded necklace. Villagers thus portray local ritual practice as an adorning object, a source of radiance that enlarges the community, elevating it and exalting it. The festive music and singing of headhunting ritual, in particular, adorns the village, the hamlet, and the homestead. The choruses sing about their commemorative festivity:

As the sumengo suggest, this radiant festivity is a mark of prosperity and virtue. In the mappurondo scheme of things there is virtue in the vanity of celebratory extravagance.

After months of labor, sorrow, and musical silence, the chance to sing brings villagers pleasure and affords them a way, momentarily, to divert their attention from the poverty, pain, and ephemerality of mountain life. Singing

for them is an exuberant human act, one to be seized. One song of longing, usually sung late at night after pangngae has come to a close, expresses this sentiment well:

Yet commemorative song also offers the chance to express magnificent and inspiring ideas and attitudes. Within the language of pangngae, the sumengo are an opportunity for villagers to ma'kada senga ', to "speak in another way," that is, to speak in a sublime manner. The urgency with which the community rushes against time to announce and extol itself in celebration comes across in a favorite ToSalu song:

"Speaking in another way" is also foretold by the language of prophetic dream-signs:

The otherness of sublime speech has to do not only with its exalted themes and sentiments, but with its formal features as well. The language of ritual speech and song is itself marked and adorned.^[2] Relative to everyday language, the liturgical speech and ritual songs of pangngae are richly metaphoric, subject to metrical constraints, and full of syllabic play. Many features we find in other oral poetries—the use of conceits and epithets, parallelisms, repetitions—figure prominently in the sumengo.^[3] The idea of adorning or adorned speech, and its otherness, no doubt relates to poetic and rhetorical

markedness, and to the figures, tropes, and patterns that constitute such marking. At the same time, speech that is other and adorned (ma'kada senga ' and kada-kada dibeloi ) potentially masks the situatedness and contingency of its production. Its very markedness helps objectify adorned speech as a seemingly autonomous and authoritative remembered text.^[4]

Thus the themes and metaphors of adornment not only play a significant role in the ceremonial politics of envy and manhood, but also enter into the metalanguage and ideology of commemorative discourse. Language, tradition, community, and the "courageous ones" of the headhunt are all radiant and enlarged in the festivity of pangngae.

The Sociality of Commemorative Song

Singing, in pangngae, is an act of memory, a recollection of lyrics and singings from the past (cf. Halbwachs 1980, Schutz 1977 [1951]). Here, however, the project of recollection is not so much the work of the individual as it is the work of the community. Virtually all the songs of pangngae are choral, and so, before all else, the songs are the commentary of choruses that make up the community. There is, for example, the ma'denna, in which antiphonal exchanges take place between the entourage of male elders and the gathering at large, exchanges that effectively corroborate and communalize the representations of the ma'denna lyrics. Apart from the something remembered, then, there is in the act of recollection a sociality of remembering. From one vantage point, the sociality of the chorus may approximate the mutual tuning-in relationship discussed by Alfred Schutz (1977): a face-to-face relationship grounded in the experience and immediacy of performance.^[5] Yet the ritual chorus—its ephemerality not-withstanding—may also make up a significant if idealized form of social hierarchy (cf. Nagy 1990). By taking a look at sumengo performances, I want to suggest that the fleeting social organization of the sumengo chorus resembles, or is grounded in, the organization of the community at large.

Sumengo are performed by groups of men or women, with each chorus guided by a songleader called a tomantokko . Although men and women occasionally come together to form a chorus, the predominating form is a group consisting solely of men or solely of women. As a rule, the groups are quite informal, and lack sharp boundaries between singers and listeners. People routinely drop in and out of the groups during ceremonies. Those song-leaders with a strong voice, a sharp memory for lyrics, and a practiced feel

for sumengo melody understandably draw more singers and hold them together as a group.

As I noted in chapter I, the sumengo lyric commonly consists of three octosyllabic phrases, sung to a set tune. The tomantokko is usually the one who recalls different sumengo lyrics, and in any event is the singer who delivers the opening solo phrase. The following two phrases are choral, sung by both the chorus and the tomantokko. In principle, anyone can become a songleader; in practice, it is usually older persons who assume this role. When a tomantokko grows tired, his or her chorus may fall quiet for a while. Or the tomantokko may let someone else assume leadership of the chorus, though perhaps not without some resistance.

It should be obvious that the sumengo choruses are social bodies organized around, and instantiating, gender difference. In ceremonial gatherings most everyone takes part in the singing. This means that the celebrating community is divided and constituted largely by choruses and by the gender differences that shape the organization of singing groups. More precisely, the community divides itself around groups consisting of wives, mothers, sisters, and daughters and countergroups made up of husbands, fathers, brothers, and sons. Although the social makeup of the chorus can raise problems in the interpretation of lyrics, it is commonly the case that the "subject's voice" residing in the song text is consistent with the gendered identity of the chorus. Groups of women usually sing songs that represent certain points of view, and men sing lyrics that reveal a different outlook. Though there are some suggestive or playful exceptions to this pattern, they strike me as being quite limited in terms of tactical or social consequence. At root, then, the song choruses communalize the lyric representations of the head-hunt, and ground them in gender difference.

In light of these differences, it is worth observing that the commemorative practices of pangngae include women as political actors and subjects. Normally excluded from everyday participation in public political discourse, women here have a voice. Central as they are to mappurondo social life and to the reproduction of mappurondo political order, the ceremonies of pangngae make up the most important platform for women's involvement in village politics. Thus women's participation in village politics takes place under ritual constraints.^[6] In this sense, women are not free subjects but are subjected by the masculine discourse of pangngae: women appear as celebrants and singers who glorify their husbands, sons, lovers, and brothers. Men, of course, appear as celebrants themselves. Yet they also enter into the ritual as the celebrated and the commemorated, and as those who exercise liturgical authority, roles (i.e. subject positions) not available to women.

Like all ritual crowds, these choral and liturgical gatherings are ephemeral social bodies. Yet even in their transitoriness we can catch a glimpse of the community's enduring forms of organization. The sumengo choruses in particular embody an idealized view of the community. Emergent in ceremony, they form an egalitarian aggregate, a body that thrives on agonistic exchanges between the choruses that divide and constitute it. No single chorus dominates the aggregate. The individual choral groups, however, require a degree of hierarchy. The tomantokko exercise leadership, authority, and privilege over their supporting singers. Younger and less experienced singers subordinate themselves to those who have spent more time in, or have gained more expertise in, remembering and leading songs. Age and experience define the terms by which an individual emerges from the egalitarianism of the whole to assume the status of tomantokko in a constituent chorus hierarchy (cf. Nagy 1990:411), only to relinquish that leadership and drop back into the group. Like the tobarani who enjoys his moment in the sun, the tomantokko sits before others as one to be followed and imitated. As villagers themselves explain it, one fells a victim, and another fells a song (George 1990).

Violence and Commemoration

At the beginning of this chapter, I suggested that for the mappurondo communities of Bambang, commemorative ritual has been a political end in its own right. This is unquestionably the case in the present order of things. In light of the region's religious pluralism and the state's civil policies, ritual tradition is the very basis of mappurondo identity. That is to say, the mappurondo communities are able to assert their autonomy and difference only by keeping ancestral tradition as their practical and ideological focus. Ritual performance, and commemorative rites in particular, are arguably the most crucial political acts the mappurondo communities can undertake, especially in their effort to retain ideological control of the past (cf. Hoskins 1987). Being mappurondo entails a commitment not only to an ancestral past but to ritual performance as such. Ritual performance, in this light, is the only way to reproduce the community.

Yet any tradition of commemorative ritual necessarily has reenactment as its central principle and problem. We know that traditions can be thoroughly reworked, invented, and reinvented; for that reason we see commemorative reenactment as a fundamental means for conjuring the past, for es-

tablishing authentic or authoritative links with the past, and for regulating its debatability (Appadurai 1981; Handler and Linnekin 1984; Hobsbawm and Ranger 1983; Hymes 1981). Commemorative tradition can in this way provide a community with a comforting sense of its own continuity and place in the world.^[7] But it also fosters a sense of obligation: reenactment is one of its moral entailments. Ritual performance puts the ongoingness of tradition, however illusory, on show.

At the same time, reenactment works back upon tradition, not only in the obvious sense of reproducing it (however imperfectly), but in organizing its discourse. If a tradition of commemorative ritual entails reenactment, then ceremonial discourse must be edited, so to speak, in such a way that it can be effectively remembered, reenacted, and commemorated (cf. Havelock 1963, 1986; Lord 1960; Ong 1977; Yates 1966). Such discourse must awaken memory and celebrate it. Further, the scene of commemorative re-enactment can be shown to occupy a place in the narrative and lyric configurations of ritual discourse (cf. Vance 1979). With these issues in mind, I want to take a brief look at how the commemorative discourse and sociality of pangngae constitute themselves through violence and song. Indeed, the ritual itself is a narrative construction that traces a movement from violence to commemoration and the regeneration of the performing community.

Pangngae, we have seen, constructs the world around polarities of Us and Them, so to speak. The self-defining and self-fashioning festivity of pangngae, I would argue, must conjure a demonized figure whose subjection and mortification are essential to the manufacture and hierarchical ordering of social difference. My point is that the commemorative discourse of pangngae requires a demonic other whose function is to help bring into being a textual or representational code suitable for reproducing the mappurondo polity (cf. Folena 1989). The taxonomic violence of creating social others is intrinsic to the commemorative process at issue here. Taking an enemy head or its surrogate, in my view, is an enacted allegory of this representational violence. This physical violence provides a palpability that discursive violence cannot achieve, yet it does so at the cost of concealing its origins. Heads and coconuts are iconic deformations and displacements of an originary and unacknowledgeable representational violence.

McKinley (1976), of course, does not see it this way. In his analysis, the headhunters' violence is the solution, or rather a resolution of existential contradictions emanating from the structures of social life. The "Other" is a problem, a threat to human order, a reminder of the group's social limits. Subjection and incorporation of the Other erase the social difference McKinley believes to be so troubling. In the scheme I am proposing, the

headhunters' social order thrives on the production of difference, not its obliteration. Bringing the trophied remnant of violence back into the settlement does not strike me as an effort to make enemies friendly. To the contrary, it appears to be a way to cast enmity in terms favorable to the headhunters' community.

Pangngae, as I have noted, appears to have had but circumstantial connections to feuding and warfare. The predatory violence of headhunting is not a consequence of historical realpolitik. Rather, political realities form the theater in which ritual headhunting takes place. Those realities have left a trace on pangngae, to be sure. And I think that there can be no question that ritual headhunting made its mark in the practical relations between upland and lowland communities. Yet given its ritual and commemorative form, the violence of pangngae is violence of another order, something driven by and constituted through its own ideals and functions.

The taking of heads, I should emphasize once more, was not a solution to regional political dilemmas. The purpose of bringing a trophy head back to the community was to astonish and awe, to mark the violence as exemplary and ritually significant. This dread relic, this grotesque, did more than fascinate and trouble. Rather, it served as a mnemonic object and privileged icon for the commemorative work of communal self-representation. The violence of headhunting could be storied . . . storied in such a way that its commemorative retelling or reenactment led to the reproduction of the communal body. Headhunting was and is a form of memorable violence that can be recollected and narrated with regenerative effect.

I do not mean to suggest that the violence of headhunting works as a mnemonic device or technique in the conventional sense of these terms. It is not an antidote to forgetfulness. It is not equivalent to the formulary devices used by rhapsodists and rhetoricians in oral cultures (Ong 1977, but cf. Finnegan 1988). What I do think the violence of headhunting has offered is a place to begin. Taking a head has been a good way to start up a story of difference and domination. Showing the grotesque trophies from these astonishing, violent events downriver, calling forth a gathering to see heads and hear tales about them, is a way to convene the community and affirm its ongoing presence. In this sense, violence serves, in the words of Eugene Vance (1979:383), as a "generative force in the production of its own discourse."

The ritualized ambush and beheading of a victim downstream, I want to suggest, is already an act of commemoration. It does not take place outside of the commemorative frame. Taking the head of someone downriver within the context of pangngae is an act of memory. There is no need to resort to

ideas about bi-presence or consubstantiation, as I once did, to grasp that reenactment confers a kind of legitimacy on both the past and the present. Mimetic stagings of the headhunt display the ongoingness of the community and its moral tradition. Reenactments of pangngae convey a sense of collective survivorship, and with it, a sense of power, especially in striking down an adversary. As Canetti (1978:227) remarks, "There is nothing that can be compared with it, and there is no moment which demands more repetition."

Scenes and images of physical violence are at the heart of this commemorative rite. They are there as a transcoding of representational violence, as precondition to festive and triumphant sociality, and as a display of the community's enduring presence. But this violence, as I noted earlier in this book, has a shifting and ambiguous relationship to regional history. Why head-hunt in the political terrain of the past? Perhaps it is to claim and control an autonomous history in an effort to evade or resist the contemporary order. Perhaps the discourse of Indonesian citizenship allows the mappurondo communities no way to overcome grief, no entry into manhood. Ideas about communal regeneration—objectified in the facts and metaphors of fertility, growth, and abundance, and of diurnal and nocturnal cycles—no doubt provide a partial rationale for ritual violence. If the commemorative act is the principal means of regenerating communal sociality and presence, then the community must look to the local past to find an event suitable for recollection and reenactment. It is one of the ironies of pangngae that the mimetic passion for sameness threatens to aggravate the incongruities between ideology and circumstance, between the past and the here-and-now. One wonders whether the rites of pangngae afford any means of working with this incongruity.

As the ritual action moves back to the village and toward a close, mimesis and memory displace violence. If pangngae begins in violence and ends in commemoration, choruses must sing of the ceremony's commemorative purpose and its ephemerality. The commemorative functions of pangngae find clearest expression in the sumengo tomatua, which frames every liturgical act and series of song exchange:

The idea of pleasing the spirits sacralizes commemoration and disguises the political and social interests that are advanced by the narrative of the

headhunt. By appealing to these transcendant forces, the song authorizes the ma'denna and kelonoson, the ma'paisun chant, and the sumengo exchanges, and in so doing legitimates their representational and illocutionary dimensions. In short, all choral and liturgical genres are reenactments, whereas mamose speeches are not. However formulaic they may be, the latter are indices and examples of spontaneity and contingency, and need not be repeated or memorized in subsequent rituals. (In fact, it would be wrong to do so.)

As we have seen already, violence yields to festivity. Celebratory excess emerges as a key theme in pangngae, and the choruses sing about singing as the headhunt ends in song:

Mimesis and the regeneration of community, in this rite, involves a restitution of the voice silenced by labor and grief. In celebrating an absent violence, the community hears its own voice, its presence filling the void of the present.

Yet such commemorative festivity, vibrant with song and laughter, can stir up eddies of nostalgia and loneliness, a sharpened sense of longing, loss, and absence. Such feelings, called rio-rio , are customarily felt by older singers and spectators. Commemorative song may remind them of the rituals of their youth, rituals peopled by friends now dead. For persons such as these, one senses that the community's enduring presence, fostered by commemorative ritual, creates rather aching reflections on their place in the world. Others may find that the ceremonies allow them to recapture their youth, and still others to live up to the desires and accomplishments of their elders. My point is that the participants in commemorative rites are positioned in different ways to the project of regenerating the community through violence and song.

To move on, then, let me summarily remark that commemoration obliges us to consider both its subjects and its objects, the commemorating and the

commemorated. Commemoration cannot be dissociated from the sociality of remembering, nor can it be dissociated from the events that it recollects and reenacts. The stability or instability of commemorative tradition, and its stabilizing or destabilizing effects are largely empirical issues. What I should underscore in the case of pangngae is that the discourse of this ritual leads to the narrative and liturgical genesis of the mappurondo polity in each village.

Commemoration and Other Ritual Themes

The community's practical commitment to ancestral tradition rests on the principle and problem of reenactment. Failure to hold pangngae would disrupt the moral coherence of the mappurondo polities and threaten to prevent the communities from convening and reproducing themselves as social and ideological bodies. But the obligation to hold the ritual brings the mappurondo communities up against desa officers, police, and school administrators, authorities who usually sanction such events with great reluctance. Meeting doubt and censure from such figures can only remind the mappurondo communities of the autonomy they no longer have, a lost autonomy that pangngae paradoxically celebrates. Still, the mappurondo settlements see their fate tied up in this ritual, just as it is tied up in the encompassing state order.

Because the ritual headhunt mediates "official" order and disorder in the mappurondo communities, we will do well to examine how the commemorative surfaces of pangngae refract a range of communal themes and interests. Like other rituals, pangngae is a theater in which the drama of prestige politics can be played out: villagers exchange gifts of meat and other foodstuffs, swap valuables for praise, and make sacrificial exchanges with the spirit-world—all opportunities for advancing the status of the individual or the household, all arenas for contest, discord, and conciliation. Yet head-hunting ritual has special significance as the discursive means for the communal management of mourning, masculinity, and communal prosperity. Altogether, then, it is a crossroads for several distinct social projects (cf. R. Rosaldo 1984).

The work of the Rosaldos has left me convinced that their Ilongot acquaintances found the celebratory discourse following a headhunter's assault more significant than the act of violence per se. They found strength, vitality, and a sense of well-being in commemorative speech and song. The

songs called buayat , in particular, were the most basic source of liget (anger and passion) and, in most instances, were instrumental in casting off grief and distress (M. Rosaldo 1980:54-57; R. Rosaldo 1980:163 and 1984:190). The celebratory song and oratory of the Ilongot are obvious rhetorical acts used to manage emotions, social relations, and violence, and the commemorative discourse of pangngae, in this sense, has much in common with Ilongot headhunting celebrations. Yet one key difference is clear: in the current social order, the discourse of pangngae does not incite physical violence. Unlike the buayat, the songs of pangngae do not stir up turbulent feelings. Rather, they amount to an aesthetic reformulation of dynamics now suppressed in current ritual practice. The commemorative lyrics of the sumengo, for example, constitute a running commentary on what should take place in pangngae, but cannot. In short, the mappurondo villagers show themselves to be contemplative lyricists and playwrights of ritual violence. Their ritual has the shape and feel of Brechtian epic theater, where the goal is to force disjuncture and open up reflexive space between representation and reality.

As previous discussion has shown, pangngae no longer sparks the cathartic violence that might allow someone to vent anguish over recently deceased friends and kin. Rage, of course, is not the only way people experience grief. A sense of heaviness or passivity, and feelings of sorrow, make up a burden that lingers on long after anguished rage has run its course. Laden with familiar memories and meanings, and calling forth an enlivening kind of sociality, the commemorative songs have a real potential to provoke feelings of relief. Singing, laughing, feasting, and speechmaking can go a long way in taking people's minds off their sorrows and in helping them get on with life. At the same time, the themes and topics of commemorative celebrations run counter to the moods, sentiments, and conceptions associated with death and grief. The celebration of pangngae also demonstrates the capacity of communal tradition to endure, to prevail when those who have wielded it pass away. The death of an elder, for example, not only stirs an emotional crisis among the bereaved, but marks a potential loss of knowledge and a loosening of authoritative ties to the past. Every death in the village's adult population threatens to orphan the community and cause its moral foundation to diminish. It is a wound to the social body and a wound to ada' mappurondo itself. The reenactment of pangngae requires village men to pledge themselves to communal tradition and to prove their readiness as ones who know and can guard ada' mappurondo. By demonstrating the surviving power and continued well-being of village tradition, mappurondo headhunting ritual fosters a sense of moral and cultural continuity, and socially vests it in the ranks of men and the ideology of manhood.

By the same token, the ritual is an outcome and index of village prosperity. Reciprocal exchange with the spirits, mediated through ceremonial sacrifices of rice and meat, and through the headhunters' oaths, helps propel the fertility and well-being of the village. But understood as a gesture of providential favor, prosperity is contingent, nonetheless, on the commemorative work of pangngae:

Further still, cosmographic idioms and the polemics of upstream-downstream encounters give the discourse of prosperity a political dimension that stretches beyond the village. The commemorative violence of pangngae, in this light, is simultaneously a display and a bid for divine favor in a threatening and unstable world. Reenacting pangngae, very broadly, links social and ideological reproduction together with the fertility and abundance of the settlement. The commemorative project, then, not only provides a means for surmounting grief and death, but legitimates and brings into being the continuity of the community (cf. Bloch and Parry 1982; Hertz 1960).

Although pangngae figures significantly in the experience of men and youths, it is not, strictly speaking, a rite of passage for males. The ritual does not bring about an irrevocable change in a person's social status. It is not obligatory for a male to take part in pangngae as a tobarani for him to become an adult. Nor is a man's participation as a headhunter a precondition to seeking a wife. Rather, the ritual produces and celebrates an exemplary image of masculinity that can help a man understand and recognize his place in the world. As he moves from childhood, through adolescence, and on to adulthood, his unfolding sense of self and his claim to knowledge and experience will find measure in the headhunt (cf. Crapanzano 1992; R. Rosaldo 1980). The commemorative discourse of the ritual, then, evokes and ratifies a notion of manhood, and connects it to the reproduction of the mappurondo polity in each village. The commemorative sociality of the ritual, with its implicit forms of hierarchy (manifest not only in songs and singing but also in the way the community gazes upon the headhunter), becomes a coveted experience, the object of men's envy. Thus the commemorative discourse of the ritual carefully elicits and constrains a form of envy that in other contexts might be quite threatening to social order. Women, as well, are subject to the commemorative scene, where they are figured as singers and adulators. The biographical path of women, then, necessarily takes them

through moments where they celebrate manhood for the sake of regenerating the polity.

What I hope is clear in the preceding pages is that the headhunt is plotted in such a way that commemoration displaces violence. The commemorative discourse of pangngae creates a sense of closure—however illusory and in-firm—confirming its plot, its structure, its purpose, and its themes, even as it exposes them to the contingencies and political interests of the moment. Closure, as Barbara Herrnstein Smith (1968:2) reminds us, "is not always a matter of endings." The headhunt does not merely stop or come to an end, but finds possibilities for completion and coherence in the ongoing work of commemoration. Following Smith's speculations on closure, we can perhaps say that the discourse of pangngae situates violence and memory within a structure of gratification. Posed against the indeterminant, the contingent, the incoherent, and the open-ended, the ritual forms an enclosure where procedure and will can produce designed effects, as for example in the tension and resolution of mourning and the vow. And although we may associate commemoration with recollection and looking backward, it is also prospective—it offers a structure of anticipation. Memory—as a form of sociality and as a form of something remembered—is kept in motion.

Violence animates commemoration; commemoration animates violence. The twin-faced structure is also essential to mappurondo social and political order, in the sense that it legitimates ritually remembered violence as a source for the community's anxious claims to autonomy. Yet mimetic efforts to reproduce this remembered violence and its social effects take for their model not a scene of primordial violence, but, rather, prior reenactments. The commemorative rite becomes subject to emulation. Within this context, villagers exploit discursive ambiguities and transcodings for individual and communal ends. Played off of ever-changing social and historical circumstances, stagings of pangngae edit as well as perpetuate the collective memory of the headhunt.^[8] Every headhunting rite is a select and open-ended version of itself, motivated by and placed in relation to the immediate goals of the community and its members. In short, every ritual performance is situated. For this reason, I now turn to the matter of the variability, instability, and open-endedness of pangngae, with the hope of discovering how the discourse of ritual violence is produced and reproduced in a climate of censure, doubt, and social difference.

7
The Songs of the ToSalu and the ToIssilita' The Horizons of Textuality and Interpretation, 1983-85

The commemorative discourse of pangngae, the way it narrativizes and immortalizes violence while reconstituting the community, is not invariant. For many anthropologists, such as Roy Rappaport (1979), the most obvious aspects of ritual have been its invariability, formality, and liturgical order. Yet in my two years of work on headhunting ritual, I invariably encountered difference, variation, and change in ceremonial performance and discourse. As I moved from settlement to settlement, and then returned to them in the following year, I began to grasp how ritual speech and song are deeply circumstanced. The instability and variability of ritual forms, it seems to me, are symptoms of the immediate social purposes and interests of different groups living in different times and places, a reminder that the projects and themes of ritual discourse are always situated. As a result, thinking about ritual language and tradition as idealized and invariant structures seems less useful than examining the ways communities make and authorize "ritual texts" for ongoing projects of interpretive and pragmatic work.

Not surprisingly, one of the most basic lines of social difference in Bambang found reprise in the headhunt: that between the ToSalu and ToIssilita'. Friends in the mappurondo communities were quick to point out several liturgical or musical differences that distinguished ToIssilita' performances from ToSalu ones. They acknowledged, for example, that the sumengo melodies sung by one of the two groups differed from those sung by the other.

And as regards liturgical action, it happens that the ToIssilita' rites always included the anointing of the headhunters by the tomatuatonda', whereas the ToSalu had dispensed with this gesture. Acknowledging these distinctive ways of conducting the ritual served both ToSalu and ToIssilita' in the work of mutual definition. Cultural differences served to mark long-standing social differences. Thus the headhunting rites of pangngae did nothing to bridge the mistrust between "the people of the ground" (ToIssilita') and "the people of the river" (ToSalu), a mistrust that stretched back over a century to the first ToSalu raids on Bambang.

Christians and the sole Muslim household in Bambang, I should remind readers, do not observe pangngae, and their local strength and influence has much to do with the waning of pangngae. For example, the ToSalu villages of Limba, Masoso, and Rantelemo are predominantly Christian communities. Pangngae last took place in Limba in 1953, and in Masoso in 1969. By 1984, only one hamlet in Rantelemo still ran pangngae, and then, only sporadically; the handful of mappurondo households there preferred to join in celebrations in the neighboring village of Karakean. Mappurondo ritual traditions in these settlements are moribund.

The scale and enthusiasm with which people conduct pangngae elsewhere in Bambang directly reflect the results of the rice harvest. For instance, the rice crop planted in August 1983 and harvested in January 1984 was a failure throughout the watershed. Devastating losses to rats and disease created an unusually severe subsistence shortfall. That year, the headhunting ceremonies in the ToIssilita' settlements of Minanga and Rantepalado were muted by disappointment and worry. The mappurondo community at Salutabang responded to the same crisis not by holding maringngangi, as did Minanga and Rantepalado, but by staging mahhaha banne , literally, "putting blood on the seed." Involving a greater sacrifice of meat than does maringngangi (minimally, a pig), this version of pangngae seeks to solicit blessings from the debata for the yet unplanted rice crop. By way of contrast to the crop failures of 1984, the 1985 harvest was an abundant one, for both ToSalu and ToIssilita' communities. The headhunting rites that followed the harvest that year were exuberant events, marked by large numbers of youths taking part as topangngae and by hosting guests from other villages. Indeed, the good fortune of the mappurondo communities in Minanga, Rantepalado, Saludengen, and Salutabang was especially striking that year: no one died.^[1]

That must have come as a relief, for there were other disturbing circumstances and events that were on peoples' minds around this time. The murder of a villager caught up in a family dispute in Minanga in 1981 had repercussions throughout Bambang. No one had come forward to admit to

the wrongdoing, and no one had made an effort to calm the debata for this startling crime. Minanga was a community on edge and under suspicion. Pua' Kallang, the Indona Bambang, died in Rantepalado during June 1983. Some whispered that his death and the crop failures of 1984 had been brought on by the crime that had shaken Bambang and that had so seriously breached the juridical tradition of adat tuo (discussed in Chapter 2). During the pangngae celebrations of 1984 and 1985 in Rantepalado, Pua' Kallang's son Dessibulo would speak, for the first time, as the probable successor to this traditional office.

Upriver, the rains of March and April 1984 brought about more than a dozen landslides. The worst of these occurred in May just below Salutabang, when a mountainside slipped into the river, burying scores of terraces and coffee gardens and flooding still others beneath the lake of water that then formed behind the fallen earth and rock. Searching for some cause behind their misfortune, villagers in Salutabang uncovered two cases of incest within the Christian community, and sent one of the wrongdoers away from the settlement with a formal curse:

Wrapped in thorns and burnt
Wrapped in thorns and washed away
Banished to places where there is no adat

Debate over the incest cases reopened wounds between kin-based social factions and between the mappurondo and Christian camps.

Government initiatives persuaded those in Minanga to plant two rice crops in 1983, in April and November.^[2] Both crops were failures, but the mappurondo community at Minanga felt obliged to run pangngae after each harvest, in October 1983 and June 1984. The agricultural initiative thus had the effect of accelerating the timing of social and ceremonial events pursued by the mappurondo households. Many were apprehensive about the decision to hold pangngae that often. Yet, the experiment was tried once again in 1985, with the mappurondo community in Rantepalado taking part, too. Crop losses that time ran between 40 and 80 percent of the normal yield.^[3] Along with this particular initiative, the government encouraged farmers to experiment with new rice strains, expensive fertilizers, and new cash crops, like chocolate. In short, the 1983-85 period found the mappurondo communities—especially those in ToIssilita' settlements—testing the limits and effects of economic rationalization.

The same period saw the civil administration launch programs to spur compliance with tax laws, to distribute ID cards, and to regulate the coffee trade. Over and against these desa-wide programs, there were events and

developments that sustained the long-standing tensions between the ToSalu and ToIssilita' districts. The person appointed as the new Kepala Desa (desa head) in 1984 was a ToIssilita' man from Rantepalado. Replacing an officer of ToSalu descent, the new Kepala Desa brought the political leadership of Bambang back home to Rantepalado, much to the delight of the ToIssilita'. Not long after, the settlement won project funding for the construction of a small bridge. A few kilometers away, the ToIssilita' settlement of Saludengen was selected as the site of an agricultural survey sponsored by the Provincial Governor's Office. Each participating household would receive a small stipend. The ToSalu, for their part, did not see much in the way of favors, nor did they realize significant assistance in the wake of the landslides that caused burdensome agricultural losses in Salutabang and Masoso. The ToSalu consoled themselves with the expansion and improvement of the marketplace at Rantelemo, right in the heart of ToSalu territory—a project initiated under the former desa administration. The market bred new interests and new strategies among the ToSalu, and began to draw them away, if only slightly, from the market activity in Mambi. The ToIssilita', meanwhile, continued their trade principally in Mambi.

The social and political differences between the ToIssilita' and the ToSalu led to no violent disputes but did sustain the mutual wariness, dissatisfactions, and grudges of past decades. Desa administration and the organization of the local Geraja Toraja Mamasa—divided into the ToIssilita' synod called Bambang Hilir (Bambang Downstream) and the ToSalu one, Bambang Hulu (Bambang Upstream)—provided new institutional contexts for the expression of these social and historical tensions. At the same time, the ToSalu and ToIssilita' each had their own internal strains and interests. The ToIssilita' settlements in Minanga and Rantepalado, for example, had locked horns in a land dispute, a dispute that seemed to aggravate competing claims to political ascendancy in Bambang's history. The fine harvest of 1985 allowed Minanga to host an entourage of mappurondo leaders from Rantepalado during pangngae. But this gesture toward seeking a resolution of the dispute failed. Saludengen remained isolated and closed during these years, hosting no one during pangngae. Nonetheless, Saludengen sent a chorus of singers to Minanga, at the request of the latter, to help enliven the pangngae ceremonies there in 1985 (see below). Upstream in the ToSalu area, villagers from Salutabang provoked comment as they continued to expand their land ownership outside of the settlement, obtaining terraces and gardens in Masoso, Rantelemo, Limba, and as far away as Rantebulahan. Like their ToIssilita' counterparts downstream, Salutabang hosted ToSalu guests

from Ulumambi and Salubulo during pangngae ceremonies of 1985, with an interest in promoting or activating alliances between distant friends and kin.

Although not all of these tensions and concerns found direct expression in local headhunting ritual, they certainly offered an ambient sense of whether Bambang was doing well or not, and in 1984 it was not. Mappurondo villagers in most instances read misfortune and conflict in their particular locales, and in Bambang more generally, as the outcome of a crisis in ancestral traditions. The state of ritual tradition was on people's minds, all the more so with an ethnographer trying to fathom their concerns.

The ToIssilita' of Minanga and Rantepalado

Villagers throughout Bambang acknowledge the ToIssilita' settlement of Minanga as the place where ritual speech and song have their origin. The name Minanga itself means "rivermouth." Used metaphorically it connotes the mouth of tradition. Whether by virtue of historical fact or a wish to vest songs with the authority of tradition, ToSalu and ToIssilita' all agreed that Minanga was the birthplace for the songs and liturgies of pangngae. In support of this claim, the ToSalu point to the language of their own ceremonies: all the songs and liturgies of pangngae are in ToIssilita' dialect, a dialect the ToSalu consider noble and refined. But when it comes to headhunting ritual, Minanga, ironically, has lost its voice. Having dwindled to a community of less than 35 households, the mappurondo collective has lost many of its knowledgeable singers and ritual leaders to the church. Singers are few, and the repertoire of sumengo is quite small (less than half of what may be encountered elsewhere), oriented now to themes of adornment, envy, and celebration and lacking signs of violence. Indeed, Minanga celebrated pangngae in 1983 and 1984 with just tambolâ and drums—there was no head.

The first time I ever took part in pangngae was at Minanga in 1983. In response to my interest in the songs of pangngae, an elderly woman explained to me that she and other villagers had forgotten the tune to the ToIssilita' sumengo, a tune remembered only in Saludengen. Claiming to recall lyrics from their past, the singers instead follow the ToSalu sumengo melody (see Figures A and B, Appendix III). In fact, the most prominent singers among the four or five tomantokko at Minanga are an aging mother and a 40-year-old son of ToSalu birth and upbringing. They are not excellent

singers as such, but they do have a knack for remembering lyrics and leading a song. In a sense, the sumengo tradition at Minanga has coalesced around this pair of singers.

The melody and lyric play of the sumengo together form a site where cultural difference, loss, and domination find articulation. Have the ToIssilita' at Minanga appropriated ToSalu songs for their own ends? Or have ToSalu songs appropriated the ToIssilita' memory? There is some truth in each perspective. What is clear to me is that people of ToIssilita' descent at Minanga find irony in the fact that their village, which holds the authority to forge sacred speech and song, has become forgetful and stammering, and obliged to borrow the traditions of the ToSalu. Relations between the ToIssilita' and ToSalu are not so antagonistic that the "cultural exchange" of the sort I am describing cannot go on. After all, the ToSalu themselves say that the songs that they sing in pangngae have roots in ToIssilita' tradition. In comparing lyric repertoires from the different villages, I found that Minanga showed vastly greater overlap with ToSalu communities than with other ToIssilita' settlements.^[4] In this light, it seems that singers at Minanga were not so much putting ToIssilita' lyrics to a ToSalu tune, as refiguring ToIssilita' lyric tradition and the ToIssilita' past through ToSalu texts-in-performance.

The sumengo performances at Minanga tended to be faltering and spiritless. Songs fell apart, voices cracked or went astray, and seldom did more than two or three people form a chorus. Through 1984 I had the impression that the polity celebrated in pangngae had withered to the point where there were few singers and little to sing about, and the situation was not much different at Rantepalado, where the mappurondo community had de-dined to just 30 households. Tobarani returned to the village with tambolâ only; the celebration took place without a trophy head, and villagers did not bother to perform the ma'denna or kelonoson, the songs that celebrated the violent feats of the topangngae. The community as a whole no longer provided foodstuffs for dipandebarani, leaving it to the families of the to-pangngae to obtain sufficient rice and meat to honor their husbands, sons, or brothers. The repertoire of choral songs in Rantepalado was exceedingly spare, and it usually fell to men and guests from other villages to perform them. The key songleader was none other than the ToSalu man who figured so prominently in Minanga's celebrations. Not surprisingly, the songs owed far more to ToSalu traditions than to ToIssilita' ones.

For mappurondo communities in ebb and disarray, holding pangngae runs the risk of further demoralizing the polity by exposing weakened communal bonds. In celebration of the reassuring harvest of 1985, people at Minanga made a gesture toward revitalizing and reclaiming village tradi-

tion. In a calculated effort to retrieve the forgotten melody of the ToIssilita' sumengo, Ambe Assi' and the other mappurondo leaders asked elders in Saludengen to send a chorus of singers to take part in festivities. Six young Sa-ludengen men took part in Minanga's ritual, lending their voices to an exuberant evening of song. Just as striking was the effort of Ambe Dido, the young babalako, in writing down lyrics, not to the communal sumengo, but to the kelonoson and ma'denna, two songs that he wished, or was expected, to lead. As we shall see, this was not the first time someone had decided to write down the ritual speech and song in the mappurondo communities. At Minanga, however, this effort to resist forgetfulness, to salvage an autochthonous tradition and authority, was—in light of the murder just four years earlier—a nervous acknowledgment of regenerative violence. Just as in Rantepalado, the kelonoson and ma'denna had not been performed in Minanga during 1983 and 1984, a further sign of the erasure of violence from the idealized representations of pangngae in these two communities. Consulting elder men in his village, Ambe Dido began a project of reclamation that conjured once more the violent acts of the topangngae.

The ToSalu at Salutabang

In contrast to Minanga and Rantepalado, Salutabang is something of a "frontier" settlement. Once the site of a ToIssilita' village, the area fell into the hands of the ToSalu around 1870. In the view of contemporary ToSalu, village history dates back to those first settlers—five or six generations back, roughly speaking. Thus, the community is historically and genealogically shallower than the ToIssilita' settlements downstream. In 1984, just under 60 percent of the villagers professed to be Christian, leaving 54 homesteads oriented to ada' mappurondo. As in most every village, tensions between Christians and the mappurondo community arise from time to time. When relations are calm, villagers say the settlement is kalebu tellu , "round like an egg" (that is, characterized by external and internal unity); when they are strained, the village is kalebu lemo , "round like an orange" (characterized by a surface unity masking internal divisiveness).

Political and ritual authority in Salutabang are seriously eroded. During the past 75 years, the position of tomatuatonda' has often been in dispute, and in 1985 at least four ritual offices went vacant. Many local specialists have converted to Christianity without passing on their knowledge of various household rites or their authority to stage them. In fact, for rituals other

Fig. 20.
Ambe Dido, a young babalako, looks on during pangngae celebrations in
1985, a notebook of lyrics by his feet.

than pangngae, the mappurondo community at Salutabang often has to "borrow" ritual specialists from other ToSalu villages. Even then, some mappurondo households are reluctant to host the more elaborate women's ceremonies for fear of making liturgical errors that would invite the wrath of the debata.

History has left the bonds of communal tradition and polity at Salutabang discernibly loose, a situation sometimes requiring the personal ends and narrow initiatives of individuals to serve broader collective interests. Pursuing his vow made earlier in the year, Ambe Tasi' sponsored ma'pa-tuhu'i botto , "having the hamlets follow," in 1985. The extensive feasting connected with this particular headhunt did much to advance Ambe Tasi's status in the community. At the same time, the ritual kept the community in motion with respect to its ceremonial obligations. Yet missing in this celebration was any effort to make a mamose speech with the purpose of publicly dedicating oneself to the village and mappurondo tradition. Interest in speechmaking was little stronger than it had been the year before. That year, there was but one, rather inchoate speech: assuming the enraged look befitting a tobarani, the man rushed his opening words, got tongue-tied, clenched the handle of his machete for a moment, and then abruptly sat down.

I have already hinted that writing and reading are becoming significant features in mappurondo ritual life, and, of course, the current demoralization of the mappurondo communities derives in part from ideas about written texts. Lacking a book of sacred scripture, mappurondo villagers cannot claim to be inheritors to a legitimate religious tradition as defined by Christians and Muslims. Although sacred texts and textual practices have provided a stick with which to beat local ancestral tradition, a few people in the mappurondo community have determined that writing may afford them rescue. This response stands in real contrast to attitudes and tabus that lingered into the 1970s, views that prohibited the use of writing, kerosene lamps, radios, or tape recorders at ceremonial occasions.^[5] Several elders, for example, were eager for me to write a comprehensive account of their ritual practices, turn it into a book, and thereby legitimate their religion. Others, most of them under 35 years of age, take a different interest in writing: they have begun to write down materials so as to be able to memorize them and perform them. That is to say, villagers write and read materials with oral performance in mind (cf. Fabian 1992).

No one in Salutabang, for example, uses writing with the idea of preserving or storing ritual texts. In fact, writing in no way guarantees the accumulation and safeguard of traditional materials. One of my greatest disappointments was discovering that a young Christian woman, Indo Kenias, had

compiled several notebooks of ritual song and speech during her school years, only to have used the paper later to make house decorations. Relaxing in her home one day I gazed up at the ribbons of paper-cuttings that festooned the rafters. Looking more closely at these doily-like decorations, I recognized the remnant inscriptions strewn through them. That day I muttered something about loss. Looking back, I note that it was but another case in which ritual speech was rendered as adornment.

Some persons have memorized written liturgical texts and then discarded them later. Take the case of Ambe Sope, a man about 30 years of age who had just inherited his father's role as topuppu, the specialist who oversees headhunting ritual in Salutabang. I happened to be there the first time Ambe Sope presided over the ceremonies, in 1984. Throughout the rite he kept looking down at a note pad to his left. Toward the climax of the ritual, he put aside his notes, took an offering up into the loft, and began chanting the ma'paisun to roughly 75 debata, all in the proper sequence. Later that day Ambe Sope said to me, "When I was up in the loft, I could see the names of the debata written in front of my eyes:' Although I have no way of being sure, I think his comment is telling evidence that a written text had played a crucial part in his rehearsing and memorizing the performance sometime beforehand. A year later, during his second headhunting ceremony, Ambe Sope worked without a written text of any kind. He was able to recompose the liturgy from memory. Whether he still "saw" the names of the debata before him on that later occasion, I do not know.

Not all ritual specialists learn their materials this way. On another occasion, in another village, I watched a young babalako in his debut, so to speak. This one worked without written notes. An elder sat at his side throughout the ceremonies and prompted him through key performances. When it came time to chant the ma'paisun, for example, the elder was right there, quietly running off the proper names of the debata from memory into the waiting ear of the young babalako. The babalako merely called out each name as it was given to him.

For Ambe Sope, writing played a key part in delivering a sound, smooth, liturgical performance. As the authority in whom liturgical conduct had recently been vested, it was important for him to appear knowing and skilled. It is especially significant, too, that he and other villagers used writing for mastering performances that are considered invariant from one context to the next. But writing did not make these kinds of chants, songs, and speeches invariant. To the contrary, it was their perceived invariance that made them especially suitable for being put into writing. Writing did not provide these genres with a fixity or authority that had been lacking. Rather,

writing inhabited seemingly stable and authoritative forms of the "already said."

In Salutabang, no one has bothered to write down the sumengo. The songs remain in oral tradition and are learned, rehearsed, recollected, and improvised in ritual performance. Villagers build up their repertoire of songs simply by listening to the tomantokko and the choruses and singing along. Sumengo performance at Salutabang usually takes the form of a ludic and oftentimes agonistic polylogue between small choruses made up of men or women. (On occasion, a choral group will include both men and women.) The choruses try to outperform one another in what might be termed song duels. The idea is to get the better of another chorus by coming up with a lyric rejoinder, sung loudly and in a manner that "chases" or "steps on" an ongoing performance. Thus, it is not unusual to hear two or three sumengo being performed simultaneously. When several choruses join in, an exuberant, swirling overlay of song fills the ceremonial house. The songs pulse like the buzzing of cicadas, and it is just such moments that bring villagers delight.

In the song play between men and women, no effort is made to sing the songs in any particular order. With the exception of the sumengo tomatua (which figures into the ritual's liturgical sequence), song choice is wholly up to the tomantokko, who use their skills and memories in the tactics of play. The songs do not follow a set narrative line, nor are they performed in set thematic or topical clusters. For instance, I once listened to a chorus of Salutabang men sing about the anguish caused by the headhunters' raid:

Meanwhile, a chorus of women sang in praise of the headhunters:

Before these singers had finished, yet another group of women started up with a song about the felling of a victim:

And while that song continued, the chorus of men performed a flattering rejoinder to the first group of women. Picking up the reference to parrots, the men invited the women to keep singing:

Perhaps these songs remain unwritten at Salutabang because they offer the antithesis of that which is desired in liturgical speech. Although the sumengo appear to form a body of stable and easily recalled lyric sets, they are performed in improvisatory play. They are caught up in the disjunctiveness of the game, and are aimed at producing difference and inequality between choruses (cf. de Certeau 1984; Lévi-Strauss 1966:32). They thrive in the din of voices and drums. The liturgical chants and songs conjoin the officiating entourage of men (led by the topuppu and the tomatuatonda') with the collectivity—as in the antiphonal repetitions of the ma'denna—or effect communion between the human and spirit worlds—as in the ma'paisun. Liturgy demands silence: drums may not sound, sumengo may not be sung, and the throng must remain quiet unless joining in antiphonal response to the entourage of elders.

When it comes to sumengo performance, it is play, not writing, that sustains the individual and communal memory. Remembering becomes a mode of play and tactical maneuver, as singers try to gain advantage over one another. The immediacies of song-play, meanwhile, gather up and deploy remembered scenes from the headhunt. Apprenticeship and leadership in choral struggle count as important poles of experience and sociality. Yet the fundamental tension in song exchange lies between men and women. In its most common (or stereotypic) form, the agonistic play of memory and voice pits groups of husbands and groups of wives against one another. Following de Certeau (1984:22-24), we can perhaps look to sumengo exchanges for schemes and repertoires of action that are useful precisely because they are detached from the complexities that entangle men and women in everyday life. They "teach the tactics possible within a given (social) system" (de Certeau 1984:23). There are strong signs here that the tactics of song play are the tactics of gender play. And this leads us to see that in Salutabang, it is gender play that sustains the communal memory. Commemorative representations of the headhunt issue from the fleeting struggles and inequalities of husbands and wives locked in contest.

The song exchange can sometimes have an edge to it. For example, it

is not uncommon to hear insults, teasing, bragging, and parody when the ToSalu sing the sumengo. These songs are invariably done in solo style, that is, without a chorus. Here a tomantokko calls another a coward:

Another song cautions the listener against conceit and arrogance:

Pushed too far, a singer might improvise an insult:

It was only in Salutabang that I heard such aggressive lyrics, and it was there, too, that I heard sumengo parodies. The common way to parodize the songs (and the ritual itself) is to make a grotesque of sumengo vocal style by exaggerating pitch and vibrato. Singers will even, occasionally, have some fun by asking for water or coffee to the tune of the sumengo's solo line.

In my experience, it was usually women who delivered aggressive, parodic, or humorous verses. Certainly some of their humor was made in innocent fun, yet at other times one had the sense that the singer was disparaging the entire event. Although I never got to the bottom of it, this seemed to he the case during the 1985 pangngae ceremonies. While taking part in a'dasan, Indo Gahu, the most prominent sando baine (or, specialist in women's ritual) in the village, sang in a manner that was as wild and mocking as it was deliberate. Later that week, her singing brought her censure. As the evening of umpaningoi ulunna ("playing with the head") drew to a close, a houseful of villagers and guests gathered at Indo Gahu's home to sing and eat. Indo Gahu resumed her parody of headhunting song. Not much time had gone by when a guest from Ulumambi, who herself was a sando baine, was "taken by the debata" (diala debata ) into trance. After dancing for a few moments the woman collapsed in a swoon and began to

say something (see Figure 21). People crowded close around her, fanning her face and listening intently to the strange words coming from her mouth. Whispers of "Mandar! She's speaking Mandar!" swept through the room.^[6] A man listening to the entranced woman conveyed the spirits' critique: the debata were angry with those belonging to the house; they had shown little respect for the ritual taking place. Shamed and warned by the words of the woman in trance, Indo Gahu ceased her efforts to amuse and mock throughout the remaining ceremonies of pangngae.

Of the four villages in which I had a chance to witness pangngae, Salutabang presented the largest and most diverse body of sumengo. The rites of 1984 ("putting blood on the seed") and 1985 ("having the hamlets follow") differed in purpose and scale, but showed thematic consistency when it came to the sumengo and the liturgical songs of the elders. Signs and themes of violence were easily discerned: the trophy head, the ma'denna, ma'paisun, and kelonoson, and nearly a score of sumengo conjured a scene of death and dismemberment. Yet the tropes of violence did not emanate from a powerful narrative order. Rather, they took form as the lyric blows of an agonistic poetry, a poetry that entailed a "virtuosity in handling the familiar" (Ong 1977:225). Their rhetoric was forged from an alloy of representational and performative violence.

The ToIssilita' at Saludengen

On the face of it, the mappurondo community in the village of Saludengen should be an ideal shelter for ritual tradition. For one, the community is determinedly endogamous, admitting no males from other villages as sons-in-law, and letting no daughter move off in wedlock. The community has also sequestered itself topographically: after several years of bitter dispute with Christian residents of the village regarding ritual and ritualized planting practices, the community exchanged terraces with the churchgoers and moved upstream along the banks of the Salu Dengen, leaving the Christians to occupy contiguous sites downstream—a tactic not unlike the one used to check the advance of Islam during the last century (see Chapter 2). Historically, the community has been a center of resistance to exogenous forces that threatened mappurondo tradition. For example, it refused to accommodate Dutch-sanctioned leadership in the district (Smit 1937), and it served as headquarters of local resistance during the rebellions

Fig. 21.
"Mandar! She's speaking Mandar!" 1985.

of the late-1950s. With respect to internal politics, succession to the position of tomatuatonda' has been undisputed since the beginning of the century. The village also has a full complement of ritual specialists. Saludengen, then, is the most striking case in which a mappurondo community has resisted social and ideological admixture from without.

The piety and resolve of this community are acknowledged by ToIssilita' and ToSalu throughout Bambang. For those in Minanga and Rantepalado, ToIssilita' tradition resides in Saludengen in authentic and practical form. Emblematic of its authenticity is the ToIssilita' sumengo melody, now lost in Minanga and Rantepalado. The ToIssilita' melody and its accompanying textual structure (see Figures C and D, Appendix III) are quite distinct when compared to the ToSalu sumengo, differing especially in terms of choral patterning.^[7] In contrast to the closure of the ToSalu sumengo, ToIssilita' song structure creates the illusion of an "unfinished" or "interrupted" text. Performing a song, a chorus will go through its lyrics from beginning to end. They then begin to repeat the third lyric phrase or line (see musical lines VII-IX, Music Figure C, Appendix III). But the repetition of that phrase goes "unfinished," for the chorus always stops singing upon reaching the sixth syllable. For example:

Usually, the close to the song does not fall so neatly on a word break:

As soon as the singers come to a halt, a different tomantokko and chorus take up another song. The performance consequently gives the illusion of one chorus interrupting the other (specifically, males interrupting females and females interrupting males, as I shall explain below). In my view, the alternating interruptions of male and female choruses are an artful representation of ludic dialogue between husbands and wives. The very structure of the performed text anticipates the conventionalized "interruption" from another chorus. Agonistic song exchanges—which make up the performa-

tive milieux for the ToSalu sumengo—are incorporated as a formal internal feature of the ToIssilita' genre.

By March 1985 I had yet to witness ritual life in Saludengen. In fact I knew very little about the village, having visited there but once, and then in the context of running a household survey, during the planting season of 1984. A chance to take part in pangngae there following the 1985 harvest was one I did not want to miss. Given the rhetoric of authenticity that I had encountered when discussing Saludengen with other ToIssilita', I looked forward to going to the village and working with a sacred oral tradition unsullied by writing. My romantic and utopian views set me up for a real surprise: since 1980, the villagers of Saludengen have put down sumengo lyrics in writing and have forged them into a song cycle. The cycle posed a fresh set of problems and mysteries with respect to representations of violence and to the character of mappurondo ritual tradition. My time in Saludengen was limited to just a few days, but nonetheless I hope I can convey some of the complexity that the cycle poses for the understanding of local headhunting ritual.

The Sumengo Cycle At Saludengen

Villagers at Saludengen reserve the evening before the concluding ceremonies of dipandebarani to perform the sumengo cycle. Replacing what is elsewhere called umpaningoi ulunna, the song cycle consists of 67 sumengo that depict a headhunt from start to finish. The community repeats the whole cycle three times through the course of the night—an effort that takes about eight hours. When performing the cycle, villagers divide into two large semi-choruses, one throng exclusively male, the other exclusively female. The two groups do not intermingle but sit on opposite sides of the house centerpost. In performance the semi-choruses take turns, and the sumengo cycle thus unfolds as an alternating dialogue between men and women.

The cycle has no plot in the conventional sense of the word. A well-knit plot is replaced by a series of tableaux, images, scenes, or "stations" that capture particular moments or episodes in pangngae. There is no development of character, no foreshadowing of events, no moral conflict. The cycle instead presents a sequence of themes and events, and in this sense has the cast and tenor of epic theater. At the same time, the community of singers also imposes a dialogic counterstructure on the cycle's narrative. The narrative moves forward through the alternating voices of husbands and wives. For that reason, specific songs assume a gendered voice and perspective. As

will become clear below, a song dialogue between men and women not only provides a performative context for the cycle, but enters into its very constitution as a narrated scene of encounter. Representation and performance fuse as one: the dialogue that narrates and quotes is also a dialogue that is narrated and quoted.

The narrative cycle begins with a liturgical invocation to the debata, and in the course of the next two songs makes a problematic and ambiguous statement about the fate of the community and the topangngae. The 25 songs that follow recount the headhunters' departure, the felling of a victim, the headhunters' trek homeward, and their sudden arrival at the village. Thereafter, the sumengo portray a long conversation between the headhunters and their wives, consisting of questions and answers (including riddles). The remaining songs depict the ritual celebration that follows a head-hunt, and close with a melancholy coda.

It falls to men to begin the narrative by singing the first three sumengo. When choruses repeat the cycle later in the night, women will sing the second song:

The lyrics to the second song, readers will recall, hold out hope for the measure of prosperity that will let the uplanders assert their power over the coast. The third sumengo shatters the illusion of hoped-for prosperity and power. The realities of power bring the uplanders humiliation and failure.

The cycle next moves on to songs about the departure of the headhunters: