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.