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.