1. I quote from Mary Morris (1993:xv), who attributes the view to the late John Gardner. See also Walter Benjamin's essay ''The Storyteller'' (Benjamin 1969:84-85).

2. Cf. remarks by Carol Fleisher Feldman (1987:213), and by Michelle Rosaldo (1980:32-34, 54-56).

3. The severed head that listens lurks uncannily in Walter Hasenclever's play Humanity (1963 [1918]). In this Expressionist drama, a murdered man shows up in a city carrying his head in a sack. He is discovered, brought to court, and branded as his own murderous assailant. Hasenclever's play, I should point out, antedates the listening heads described in this book.

4. The idea is that harmful spirit-beings will try every "open door" in the comb and become exhausted before reaching the door to the house.

5. If any reader is puzzled, I am alluding to The Raw and the Cooked , by Claude Lévi-Strauss (1969).

6. For overviews of the critique, and efforts to resume work beyond it, see Dirks, Eley, and Ortner (1994), and R. Fox (1991).

7. Rosaldo's treatment of Ilongot hunting stories (1986) shows a similar regard for the play of memory, history, experience, and circumstance.

8. Jane Atkinson takes the same approach in her work on the art and politics of Wana shamanship (1989:14).

9. Burur's violent feat does not appear to figure into R. Rosaldo's history of Ilongot headhunting (1980). By the reckoning of that work (1980:302), Burur's beheading of a victim would have taken place around the time a covenant of amity was being put into place. The way in which Burur remains problematically beyond the horizons of R. Rosaldo's study suggests that an unexplored set of difficulties surrounded the case and the celebratory event thereafter.

10. Sherry Ortner's use of "cultural schema" (1989)—when applied to rituals or narratives—strikes me as problematic, and draws our attention away from the very practices she might best illumine. Very simply, Ortner's cultural schema is a synoptic construct (reminiscent of Propp's syntagmatic structuralism) that puts aside a history of specific narrative events and eliminates the role of memory. Ortner then has to struggle to (re)ground the schema in types of practice, rather than in events and practices as such. For an important critique bearing on the use of cultural schemas and other basically basic stories, see Barbara Herrnstein Smith (1981).

11. The dominance of formalist, structuralist, and semiotic theories during the 1960s and 1970s drove the concept of theme into exile. Though there are problems with the term in light of contemporary interpretation theory (especially deconstruction), I think it is a concept worth recuperating for the purpose of my analysis. I note that Michelle Rosaldo (1980) used the notion of "theme" to explore Ilongot headhunting.

12. For a brief look at the cultural politics of ethnography in these communities, see George (1993a).

1. The same images have also played a significant and lingering role in nineteenth-and twentieth-century representations of an oppositional human figure (i.e. an "Other") suitable for ethnological study and contemplation. I have already mentioned this problem in Chapter 1, where I touch on headhunters and headhunting in the anthropological construction of the primitive and the exotic.

2. For a discussion of politics and the construction of minority religions in Sulawesi, see Atkinson 1983 and 1989.

3. See Acciaioli (1989:283-284) for a related application of the term among Bugis communities.

4. These recollections amount to a search into the past for a golden age, a process Michael Herzfeld calls "structural nostalgia" (Herzfeld 1990).

5. The membership of the league shifted through time, especially in response to the growth of Mandar principalities and to the advent of Islam. For example, Tabang (a community lying to the east of Mamasa) is often mentioned as a member of the upland league. Accounts that include Tabang usually omit Tu'bi, a community that appears to have been absorbed into the Mandar sphere following its conversion to Islam.

6. Toraja genealogies collected by C. H. M. Nooy-Palm and L. T. Tangdilintin (Nooy-Palm 1979) identify him as Patiang Boro from the Sa'dan settlement at Napo. Genealogies from Rantepalado also trace Pongka Padang's ancestry back to Sa'dan deities.

7. People of Bambang say that these raids brought about the collapse of Pitu Ulunna Salu (cf. Smit 1937). The violent event may have had roots in the political instability of the lower peninsula and the Dutch attack on Mandar in 1872 (Sutherland 1983a). Also, Bigalke's portrait of upland turmoil (1981) shows the impact of the regional coffee-, slave-, and weapons trade of the late nineteenth century.

8. Their request was granted since the time of my research. The ToSalu settlements were grouped to form a single desa, Bambang Hulu, while the ToIssilita' ones comprised Bambang Hilir. This administrative partitioning of Bambang, it should be noted, reflects and follows upon the long-standing organization of Protestant assemblies in the district. See the Epilogue for subsequent developments.

9. Enang, the last surviving parengnge (Dutch-appointed district head) in the area remarked to me that mappurondo rituals persisted in the village of Mambi until 1936.

10. Krüger (1966) identifies Kijftenbelt of the Protestantsche Kerk at Makassar as the figure who oversaw the introduction of Christianity to the region that included Bambang.

11. Thus, the Hindu Dharma bureaucratic apparatus has no means to shelter highly localized religions. There is no provision for setting up offices for the benefit of followers who live in but one desa or kecamatan.

12. The structure of the house has changed radically in the last 80 years. Formerly, the great majority of banua had as many as eight hearths. A nuclear family resided at each hearth, and maintained their own rice barn in the yard. The families were related to one another through sibling or parent-child ties. The first campaigns for single-hearth, single-family dwellings were under the direction of Dutch administrators and missionaries. Indonesian officials and religious institutions renewed these campaigns after independence. The success of their efforts has contributed to the opening of new residential sites within village territory and the appearance of very rudimentary dwellings. Nonetheless, most villages still have several double- or triple-hearth houses, often in which young married couples, or divorced or widowed females make a home with siblings or parents. Further change has taken place with the placement of the house. In keeping with local cosmology, houses once faced upriver into the rising sun, toward the dwelling place of the gods. (When a house had multiple hearths, it was lengthened along an east-west axis, rather than north-south.) This pattern underwent abrupt change in 1980 when civil administrators decided to "beautify" hamlets. Henceforth, houses were to face one another along a central lane or yard.

13. Lacking the human and material resources to act as a practicing ritual community, these households usually take part in ceremonies staged by kin in neighboring settlements.

14. For a discussion of the coffee trade in upland Sulawesi prior to the twentieth century, see Bigalke (1981:29-59).

15. Given the way a "sense of place" enters into local social organization, it would not be off the mark to identify Pitu Ulunna Salu as a "house society" of the kind described by Errington (1987) and Waterson (1986).

16. I was unable to obtain women's viewpoints on polygyny.

17. Incest tabus also apply to classificatory siblings at the parental generation, i.e. first cousins of ego's parents.

18. There is evidence in the ceremony for understanding somba as a swap for a wife: It is presented in thirds, of which the first two are proclaimed, "Not enough!" Only with the third presentation do the bride's kin announce that a proper amount has been given. In this sense, somba might be best described as bridebarter.

19. Some landless households have appeared in the last twenty years in the villages of Masoso and Rantelemo, two settlements that are almost exclusively Christian.

20. Jane Atkinson (1990) reports that the Wana of eastern Central Sulawesi also use the term bela to denote husband and wife. In the Wana case, too, the term is unmarked for gender. In each case, the term implies that the husband-wife relationship shares formal, functional, and moral features with the ties appearing within unisexual cohorts and task groups.

21. For theoretical debate over male/female, public/domestic, and encompassing/encompassed hierarchies, see Ortner (1990); M. Rosaldo (1980); and Yanagisako (1979).

22. If I am correct in my assessment, the mappurondo case weakens any claim that cultural assertions of male dominance are universal.

23. The most recent morara recalled by eider men are, by village: Minanga, 1977; Rantepalado, before 1906; Saludengen, not since c. 1870; Karakean, 1937; Limbs, twice before 1906; Rantelemo, before 1906; Ulumambi, 1933 (with Salutabang); Salutabang, 1933 (with Ulumambi); Masoso, before 1906. Knowledge of how to properly stage morara has seriously eroded, to the point where one village, Rantepalado, even refuses to hold the ritual for fear of making liturgical errors that will invite the wrath of the spirits.

24. Cf. with Bugis sumange' . See Errington 1989.

25. Informants describe morara as an enormous potlatch that demands a month of elaborate preparation. Guests and hosts consume huge surpluses office and meat, stage mock battles, and exchange gifts of cloth and other valuables. Excluding preparation time, morara takes about two weeks to perform. Elders explained that the ritual requires that a head be taken, and the few morara song texts that I have been able to get suggest thematic, metaphoric, and generic parallels with those of pangngae.

26. The succession of ceremonies is as follows:

The sequence is presented in simplified form. ToIssilita' practices include (1) sacrificial variations that call for yet more elaborate terminology in the case of mekolâ and (2) an alternate form of ma'bua' called ma'bua' tumbâ (also called ma'bua' mata ). I am uncertain of the status of medio, a female-domain ritual, common to ToIssilita' and ToMambi communities, that vanished before 1940. It appears not to have been part of the sequence, but may account for why ToIssilita' communities presently lack a counterpart to the ToSalu metinda'. An eyewitness account of medio suggests the elaborateness of a parri' ceremony.

27. The curing rarely involves trance or other shamanic techniques. Shamanism appears to have been in practice among males until about 1950. My language assistant, for example, reported a curing in his natal household during his childhood. In his account, a male shaman from another village treated an uncle. The shaman howled and beat his chest, and then pulled a crab from the uncle's body. I was not able to verify claims that there remain two or three persons versed in shamanic techniques at Ulumambi.

28. I treat this issue at length in "Music-making, Ritual, and Gender in a Southeast Asian Hill Society" (George 1993c).

1. My thanks to Johannes Fabian for pointing this out to me (pers. comm. 1991).

2. It may be helpful to think of the head and surrogate as different editions of the "same" text. (See McGann 1991).

3. For other confrontations with the textuality of history in the archipelago, see: Boon 1977, 1990; Bowen 1991; Siegel 1979; and Steedly 1993.

4. I here include ethnographic studies and surveys, travel accounts, and mission diaries.

5. An investigation of colonial juridical records would probably turn up something to the contrary.

6. In their critique of Freeman's analysis, Davison and Sutlive (1991) point out that Iban discourse often little support for symbolically associating head with phallus. They argue that the metaphors of Iban ritual comprise a "frugiferous" model of headhunting in which trophy heads are equated both with the fruits of the mythical ranyai palm and with the infant offspring of the demon Nising . Though Davison and Sutlive show a willingness to explore ritual discourse for an understanding of Iban headhunting practices, they remain certain that in local thinking, seed-laden enemy heads bring about prosperity. For Davison and Sutlive, this is only natural. In their analysis, Iban headhunting has its roots not in psyche and libido, but in the Bornean rain forest. Concerned to swap an ecological determinism for a psychological one, Davison and Sutlive never seriously challenge the logic in which severed heads make fields and communities thrive.

7. Note, too, that the throng turns its heterophonous shouts against the babalako, beheading the chant.

8. I believe this viewpoint is implicit in some of the comments of M. Rosaldo (1977, 1980) and R. Rosaldo (1980, 1984).

9. The headhunters of the mappurondo community are not alone in linking heads and prosperity through the vow. Writing about headhunting practices of the ToMangki (or Galumpang Toraja), a group living to the north and east of the Salu Mambi region, A. C. Kruyt (1942:543) explained that, "One pledged to seek a head if the rice the following year were successful. If one then had a good harvest, he would make preparations to go."

10. Hospitality, as Michael Herzfeld writes, "signifies the moral and conceptual subordination of the guest to the host" (1987:77). Seen from this perspective, hospitality in pangngae further underscores the subordination of the victim.

11. For comparative materials, see Hoskins (1989), who shows how Sumbanese headhunting practices factored into eras of exchange, alliance, and trade between rival groups.

12. For the southeastern peninsula, see Kennedy (1935), J. Kruyt (1924), and Schuurmans (1934); for the eastern peninsula and central mountain region, see Adriani and Kruyt (1950), Atkinson (1989), Dormeier (1947), Downs (1955), A. C. Kruyt (1930); and for the western and northern mountain regions, see Kennedy (1935) and A. C. Kruyt (1938). Key accounts and commentaries for the so-called "Southern Toraja" region are: Bigalke (1981), Goslings (1924 [rep 1933]), A. C. Kruyt (1923; 1942), Nooy-Palm (1986), Tangdilintin (1980) and Volkman (1985).

13. It should be kept in mind that grief, fertility, and the welfare of the community may have a scope that goes beyond the local, and may in fact articulate regional political strains.

14. Compare: Mambi and Bambang, pangngae ; Galumpang, pangaye ; Sa'dan mangaung ; Bare'e, menga'e ; Indonesian and Malay, kayau . In the Bare'e case, discussed by Downs (1955), some association is made between the term for headhunting ( menga'e ) and the term for harvesting rice ( menggae ). I am skeptical of Downs's claim that the two words are synonyms. Nonetheless, I would allow for Bare'e wordplay and a metaphoric thinking that connects harvesting rice to taking heads. For further notes on cognate terms for headhunting in the Malay world, see Maxwell (forth.).

15. One Mandar Bupati , or regent, in my acquaintance was celebrated for having killed his first man over loss of face at the age of ten.

16. Daetta, the fourth Mara'dia of Balanipa, was the first of the region's elite to enter Islam (DepDikBud, n.d.[d]). Islam appears to have prevailed in nominal fashion until the beginning of the twentieth century, at which time reformist Islam began to purify and deepen local belief. For a persuasive discussion of how Islam influenced political life in South Sulawesi, see Andaya (1984).

17. Given what I can reconstruct from oral histories and written sources, the historic exchange system between the uplands and the coast basically conforms to the hypothetical model advanced by Bronson (1977).

18. Hoorweg appears to have been unaware of the ritualized agricultural calendar at Pitu Ulunna Salu.

19. Famines and harvest shortfalls were not unusual in the highlands, and continue to this day. Oral histories suggest that food shortages were common by 1870. Dutch authorities, too, mention this problem in their reports (e.g. anonymous [Militaire Nota] 1924).

20. A comparable situation may have obtained m parts of Borneo. Dr. Peter Kedit of the Sarawak Museum (pers. comm.) claims that Iban headhunting often took place in association with bejalai , journeys undertaken for profit or social prestige (cf. Freeman 1970).

21. The ma'paisun chants would lend this claim credibility.

22. " Masae! Tellu bulam !" "A long time. Three months!"

23. I write about "the grotesque" in the same manner that some write about "the ludic" or ''the carnivalesque.'' What the grotesque is, or might be, varies from society to society. Its specific aesthetic, rhetorical, and political functions are by no means stable, but are manifest as particular cultural and historical projects. For example, Bakhtin (1984) argues that the Renaissance grotesque differed profoundly from Romantic grotesque. Further, he sees the grotesque as an oppositional mode of language and practice set against official discourse. Below I will argue that the discourse of the grotesque in pangugae is official discourse.

24. Cf. Bakhtin 1984.

25. Cited in Holman (1972:246).

26. My use of the terms "marked" and "unmarked" owes much to discussions in Caton (1987), Nagy (1990), and Waugh (1982).

27. Some commentators on Bornean headhunting practices (Davison and Sutlive 1991; Freeman 1979; and Metcalf 1987) mention the way women will take the trophy skull and dance with it between their knees. For these scholars, such dance-gestures suggest that the head is a symbol of fecundity or birth. I find that gestures like this are intended to mock, humiliate, and subordinate an oppositional other.

28. See Bakhtin (1984) and Canetti (1978) for reflections on laughter, terror, power, and survivorhood.

29. I borrow the term "undominated discourse" from James Scott (1990:175).

30. This resembles a Nietzschean turnabout of cause and effect (Culler 1982: 86-87). As the feeling of pain leads someone to look for a cause and thereby discover an offending pin, so too does the headhunter reflect on his tradition of violence and look for its cause in the past.

31. In the spirit of celebrating the indeterminacy of causes and origins, a claim might be made that the head is a surrogate coconut. Indeed, the rites of the past may have involved any number of symbolic objects that could (1) summon or render the body politic of the "other" and (2) withstand disfiguring abuse. Nonetheless, I think we have to be mindful of "the gradient and direction of flow of metaphor and symbolic substitution from one domain to another " (Stallybrass and White 1986: 196, emphasis theirs). Though it is true that that gradient and direction are subject to historical change, I think it unlikely that a severed head would displace the coconut except in a project of restoring a prior sociosymbolic formation.

1. The first term is applied to rites that share the emphases of maringngangi'. The latter term applies to the headhunting rites that follow upon the death of the tomatuatonda'.

2. Some reports from Borneo (see McKinley 1976) suggest that sacrifice of a slave or dependent provided an acceptable alternative to headhunting in some settlements. In contrast, the communities of Pitu Ulunna Salu seem to have drawn around themselves a social line within which dehumanizing violence was unacceptable. In this light, the human sacrifice at Rantebulahan may be read as headhunting out of place. Note, too, the double out-of-placeness of human sacrifice in the Rantebulahan episode: violence happened during pa'bisuam baine , the sphere of domestic rites run under the authority of women, not during the village rituals directed by and toward men.

3. It should be obvious that world religions deliver a similar message. As discussed in Chapter 2, the Indonesian state insists that participation in a monotheistic world religion is the keystone to progress-oriented citizenship.

4. I do not mean to suggest that the "boundaries of the human" were free of instability prior to the advent of Dutch and Indonesian administrations. Yet the transcendant citizenship of the contemporary era opposes and subverts pangngae in ways quite different from the oppositional threat posed by the encroaching lowland societies of the precolonial period.

5. Some societies commemorate sacrifice and the valorous loss of life in ways that still violence. Societies that remember and enshrine their anguish run the risk of becoming spawning grounds for violence. See Feldman (1991) and Girard (1977).

6. The transformation of emotional discourse under different political regimes and crises has been explored in Iran by Good and Good (1988), and in China by Kleinman and Kleinman (1991).

7. As discussed in Chapter 3, McKinley (1976) makes a strong case for the relevance of knowing enemy names

8. Ghost-souls ( anitu ) of the recently deceased may linger in the village, brooding or playing pranks on the living. Ancestral anitu are also summoned when household rituals get under way. In no instance do the anitu show such suffering and disquiet that they would bring about further deaths in the village.

9. For some key anthropological works and surveys on the politics and social-structural implications of emotion, see: Abu-Lughod 1990; Lutz and Abu-Lughod 1990; Lutz and White 1986; Myers 1986; and M. Rosaldo 1984a.

10. Michelle Rosaldo (1980:138) also writes that, when asked why they take heads, the Ilongot occasionally mention ' uget ,' "bad feelings."

11. Renato Ronaldo's use of the term "piacular" probably derives from Durkheim's discussion of "principal ritual attitudes" in The Elementary Forms of the Religious Life (1965 [1915]:434-435). Although the term usually suggests expiation or atonement for wrongdoing, Durkheim applies this term to rites ''celebrated by those in a state of uneasiness or sadness" and "whose object is either to meet a calamity, or else merely to commemorate it and deplore it." For him, rituals of mourning are important examples of piacular rites. If I am not incorrect, Rosaldo has called headhunting a "piacular sacrifice" in order to underscore its connection with mourning. In passing, let me note that a recent verbal or transcriptive slip threatens to erase the genealogy of Rosaldo's thinking about Ilongot headhunting: ''Their ritual is a kind of peculiar [sic] sacrifice. . . . in the ritual attitude, it's probably more like sacrifice than one would imagine at first" (R. Rosaldo 1987:253; italics mine).

12. I find a special set of problems and analytic dangers in efforts that confound violence and catharsis. Construed as catharsis, violence appears "natural" and appropriate to the ends and needs of personal emotional balance. Such a perspective cannot grasp violence as a product of history. The problem and some of its consequences are spelled out in Walter Benjamin's "Critique of Violence" (1978). Cathartic violence needs to be interpreted in light of its capacity to shape and be shaped by changing historical conditions.

13. It seems to me that Metcalf (1987; forth.) misreads the Ilongot case and Rosaldo's essay, which takes Metcalf and his co-author R. Huntington to task for conflating death, mourning, and mourning ritual in their monograph Celebrations of Death (1979). Metcalf (1987:4; forth.:28) insists that the Ilongot have no head-hunting ceremonies even though the Rosaldos' descriptions of the buayat and the song-and-boast fests that followed beheadings clearly indicate otherwise. As regards Renato Rosaldo's essay, Metcalf (forth.:26) argues that Rosaldo believes: (1) that the force of emotion is the "principal shaper of human behaviour", and (2) that the Ilongot headhunter's deadly violence is the "natural" outlet of a universal emotion—the rage that comes with grief. Metcalf distorts Rosaldo's more measured claim: that an apprehension of anguished rage can offer us a glimpse of how the Ilongot sometimes turn to violence in coping with painful experience. Rosaldo never reduces Ilongot headhunting practices to pure rage, nor does he argue that sentiment forms the basis for culture and conduct. And although relying on his own sense of loss and grief to fathom Ilongot experience, Rosaldo cautions against thinking about emotions and their cultural formulation in terms of human universals (1984:188-189).

14. I thank a reviewer for reminding me that "from a psychological point of view, a symbol could just as easily trigger a cathartic release as the thing the symbol represents. Whether the symbol actually triggers catharsis would have to be actively investigated."

15. There is a huge literature on ritual transformation. Among recent statements, a few of the more important include: Geertz 1973; Kapferer 1979a, 1979c; Obeyesekere 1990; Ortner 1978; Peacock 1968; Tambiah 1985; and Turner 1969, 1974.

16. See Just (1991) for a discussion of "other-control" in the display and management of emotions on the Indonesian island of Sumbawa.

17. The lyrics hint at the vulnerability of local women. They suggest that upland women may be potential, if very mistaken, targets for a headhunter's violence. See comments in Chapters 5, 6, and 7.

18. Among the village elders of Minanga there is one who holds the office of toma'gasinna ([?] "the one who is like a top" [i.e. going out and back]). He is obliged to accompany the cohort and procure the right kind of bamboo for their tambolâ flutes (see following text). For Minanga, it is not unusual for the babalako and toma'gasinna to occupy the roles of too'na and lolona , respectively.

19. From a structural viewpoint, the ToSalu village holds fast to the topuppu, keeping him among women and weakened, grief-stricken men.

20. For example, special caution is exercised to keep necessary discussion of the ritual out of the earshot of women and girls. On the evening before departure, headhunters customarily refrain from having sex; many husbands find a place of sleep away from their wives. Moreover, the headhunters may not eat from the sheaves of young rice taken by their wives, sisters, and daughters just before the most recent harvest. Violation of any of these tabus is said to make the headhunter vulnerable and perhaps bring illness upon the women concerned. In principle, a man's wife and close female relatives should not know that he intends to join the cohort of headhunters. But, of course, women are free to figure things out.

21. In earlier times an inauspicious reading might have led the cohort to postpone their journey. Presuming that pangngae has always been an obligatory rite, one would conclude that negative signs from the "wrong chicken" would never cause a raid to be canceled altogether.

22. In the past, the headhunters would look and listen carefully for bird omens. Calls heard off to the left, birds seen flying to the left, and hawks trailing the cohort from behind all portended failure or death. Herons flying upstream toward the pack signaled that the victim's territory could soon be reached. In these days of surrogate heads, however, the importance of bird omens has faded.

23. Coconut palms do not thrive in Bambang because of the altitude and climate, and the nuts must be purchased at a local market or requested from villagers in Mambi, a place where the palms do grow.

24. The flutes are roughly a meter long and 3 inches in diameter.

25. See George (1993a, 1993c) for discussions of secrecy, gender ideology, and local musical culture.

26. Women seldom journey beyond local markets or the villages of relatives. From their view, going off on pangngae is to step off into unknown territory. It needs to be kept in mind, too, that many older mappurondo women do recall the not-so-distant past when their menfolk traveled off to the coast. But the years since 1970 have probably been the most tranquil and safe in the last century. Any confidence the headhunters and their wives may have about the journey is likely a very recent phenomenon.

27. Local discourse associates the coals and fires of the hearth with wives and mothers. In local cosmology, the debata of fire is also the debata of gossip. If I am not wrong, the tabu against lending coals to another hearth in the husband's absence puts a metaphorical check to the spread of women's knowledge. If coals cannot pass to another house, neither can knowledge of a husband's absence. Paradoxically, the refusal to lend coals would be a sure tip that some male in the house was away on the headhunt.

28. "Lelelele" is the name for male sumanga' (spirit, elan vital) in the sacred language of women's ritual. It is also the sound of a man's sumanga'.

29. My use of the term illocutionary derives from the pioneering work of Austin (1962) and Searle (1969) in speech-act theory, and from the work of Tambiah (1985) on a "performative approach" to ritual. Following Tambiah's scheme (1985:135), the shout and the drone comprise a "performative act" subject to "constitutive rules," but with uncertain perlocutionary effects in terms of participants' emotional states.

30. During two years of research I saw no instances of latah , the startle-hysteria widely reported throughout the Malay world. At the same time, persons are at pains not to startle others with frightening noises, or with the sudden and painful news of a loved one's death.

31. In 1984 I had the opportunity to catch a glimpse of a crisis of sorts in a ToSalu village in the Salu Mokanan watershed of Rantebulahan. The fires of pangngae were already burning when a villager passed away. Instead of dividing the village in two, the elders put a halt to pangngae and began funerary rites. They even took a chicken dedicated to pangngae by sacrificial prayer, and offered it up in death ritual. What was for some a misappropriation of sacrificial meats became topic of bitter debate. My closest friend was especially disturbed: "They better acknowledge their mistake. If the crops wither away next year or any other misfortune takes place in this village, it is sure that that chicken will be the reason." The mistake appeared to be a compound one—halting pangngae after it was under way, and "confusing" life and death in taking meats dedicated to the enlivening rites of pangngae and consuming them in the mournful ceremony of funerary ritual.

32. So as to leave no confusion, Rosaldo is not saying that volatile emotions caused headhunting to emerge as a cultural practice.

33. Atkinson (1989:27, 1990:68) reports that the Wana of Central Sulawesi link vows to headhunting and duels.

34. See, too, Renato Rosaldo's account of how, in the weeks following the accidental death of his wife Michelle, he vows to return to writing anthropology (1984: 184). Regarding arguments I make elsewhere in this book (see Chapter 6), note that Rosaldo's vow, as a generative force in writing "Grief and a Headhunter's Rage," must surface in that account as subject matter.

35. Although I have yet to come across supporting metaphors or comparisons in mappurondo discourse, it seems to me that being released from the grip of a vow could be likened to being released from the grip of personal anguish. If this happens to be the case, the samaja vow potentially otters a model for grief-work. In any case, I do not find it a matter of accident that this particular illocutionary act should become so entangled in the cathartic scenes of Ilongot and mappurondo ritual life.

1. See Chapter 2 for a general overview of the mappurondo social order. Social hierarchy and egalitarianism are subject to historical forces, and no doubt have changed over time. I have in mind the influence of slavery and debt bondage, forms that were prevalent through the nineteenth century. The communities of the present may in fact be more egalitarian than they were in the past. I should note, too, that several ToSalu described a now-vanished system of social rankings that seem to have been quite similar to those that traditionally informed Sa'dan and Mamasa society. Frankly, I find those remarks suspect, for they were all made by Christians who had enjoyed schooling in the Mamasa district. That is, I think these historical views belong to a decades-old cultural politics aimed at drawing Bambang within the orbit of the Toraja.

2. For a deeper exploration of secrecy in the mappurondo enclave, see George (1993a).

3. The essay itself appeared as an adornment on a larger essay entitled "Secrecy" (Simmel 1950; see the translator's note on p. 338).

4. See Taussig (1987) for an account of envy and its embodiment in a culture of terror and fear.

5. One woman I knew in Salutabang recalled that the last malangngi' to be staged in her village took place around 1950. On that occasion, 40 women took part as dancers in the secret group.

6. A husband's impotence was a common source for complaint among village women of my acquaintance. See, too, the mamose speeches described below.

7. See R. Rosaldo (1986) for Ilongot ideas about the parallels between courtship and headhunting.

8. Judging by reports from villagers throughout Bambang, antiphonal riddling and song challenges were far more common during the pangngae ceremonies of the past. Whole hamlets would take part, with choruses in separate households challenging or answering each other in song. Elders reminisced about times when whole mountainsides were swollen with singing, as house after house, hamlet after hamlet, would take part in the antiphonal and agonistic din of voices. The antiphonal contests of today are far briefer and more muted.

9. The term means, roughly, "narrative song." Kelonoson (or kelonnosong, kelong ossong ) crops up as a song term throughout Sulawesi and Borneo. For example, Rubenstein has published a Sarawak kelon about headhunting (1985:237). Tammu and van der Veen (1972) list the Sa'dan kelong as meaning "song" or " pantun ," and osso' as meaning "to arrange'' (as story, history, etc.); cf. gelong , a song in the maro ritual. Zerner and Volkman (1988) further describe the Sa'dan gelong as an octosyllabic form common to maro ritual. Stokhof (1986:6-8) mentions that the Makassan kelong is a quatrain (8-8-5-8 or 8-8-8-8) that deals with themes from daily life, and which traditionally was recited on adat occasions. The Mandar osong he describes, following Azis Syah, as a "fighting song.''

10. Usually, households throughout the village donate the foodstuffs for the tarakan. In Rantepalado, however, recent practice has been for the immediate families of the tobarani to provide the tarakan foods. An elder from there reports that the village has grown unhappy with this procedure and is planning to return to their former practice of taking food shares from all mappurondo households in the village.

11. Nonetheless, a speaker might feel real satisfaction in having made a speech deemed more eloquent or moving than that of another.

12. The same clichéd slogan is heard in Sa'dan Toraja ritual (Coville 1989:121).

1. Yet insofar as representation is a form of violence, one might say that a head-hunt not only begins in violence but ends in violence as well. At the same time, this chapter should provide grounds for claiming that pangngae begins in representation and ends in representation.

2. For a detailed linguistic and musicological description of these liturgical registers, see George 1989:226-264.

3. A list of key works on oral poetry is impossible here. Some places for the reader to start: Bauman 1977, J. Fox 1988, Hymes 1981, Lord 1960, Ong 1977, and Tedlock 1983. See also the important recent statements by Bauman and Briggs (1990, 1992), which reflect the influence of Roman Jakobson and Michael Silverstein.

4. Bauman and Briggs (1990) call this process of making an autonomous text entextualization , a concept applied with great insight by Kuipers (1990) to a body of ritual discourse in eastern Indonesia. The term—which has become a bit of a cult object itself—has its roots in the writings of Roman Jakobson and Paul Ricoeur, and in the unpublished work of Michael Silverstein.

5. The social modalities Schutz has in mind sound much like the social bonds of communitas described by Victor Turner (1969, 1974).

6. For a more detailed discussion of this issue, see George 1993c.

7. Commemorative rites and traditions, of course, can be the context or subject of political contest, and may lead to radical rejections of the local and the communal (cf. Thomas 1992). Still, what seems to be at stake in these contests are claims to continuity and a place in the world.

8. Cf. Battaglia 1990:188.

1. Headhunting rituals subsequently kept, but played down, thematic linkages to mourning.

2. As noted above, the rest of Bambang planted in August 1983.

3. I left the Salu Mambi region in July 1985 and have no firsthand data about the pangngae ceremonies that followed this harvest.

4. Full details of the comparison can be found in George 1989.

5. Wherever one goes in the Salu Mambi and Salu Hau region, one discovers that the mappurondo communities have a long history of anxiety about polluting objects, technologies, and practices that come from afar. At the same time, exogenous objects—severed heads, cloth, porcelain bowls, exotic words—have an important place in ritual practice. This tension between sacred and polluting objects that make their way into the community has suggestive links to geographies of power and conceptions regarding its exogenous sources (cf. Atkinson 1989).

6. The language of entranced women is always an exogenous tongue and, quite often, Mandar. Readers must be curious, as am I, about the significance of Mandar as a woman's trance language in pangugae. Mandar is the language of the ambushed victim, after all. Let me risk a speculative remark. It seems fitting that women in trance would use the language of those who were traditionally the region's dominant political power. Women's trance-critique is thereby authorized by the signs of a dominant power.

7. I note one exception: the melodic structure of the sumengo tomatua at Salutabang is quite similar to that of the ToIssilita' melody heard at Saludengen. For a basic musicological analysis of sumengo melodic structure, see George 1989.

8. In this song, the women praise the mountain communities and their headhunters. They sing about the andulan (alternate spelling, andulen; see also sumengo 17 and 20 in this group, and note 9, below), which is a variety of areca prized for the yellow leaves and trunk that make the tree stand out in the upland forests, and which is considered emblematic of the mountain region's virtue and glory. The women liken the headhunter to a telo-telo, a swift-flying bird that nests in the mountains but searches for fish in the sea (i.e., it finds prey along the coast, as do the headhunters). Tammu and van der Veen's Toraja dictionary (1972) indicates that telo-telo can also refer to decorative buds, tassels, and leaf tips, though I am unaware of this usage at Bambang. If it does hold true for Bambang, the sumengo then appears to play with the term's semantic possibilities, alluding both to the tips of the andulen tree and to the bird that flies over the seas.

9. Rembu andulan (blades of andulan) are the leaves that the headhunters use to make the streamers that spill from the throat of a tambolâ. The swaying of the tamboll streamers indicates that the warriors are on the move, making their return trip.

10. This song is puzzling to me. I believe it may refer to the mourning shirt worn by women who have been grieving for the dead. Yet my language assistant argued that it was a song of a woman's longing for her husband. tie has been gone so long that his shirt is foul with mold; the invitation to take it off is an invitation to lovemaking.

11. The descending stars are the torches carried by villagers who pour down from the hamlets to meet the returning headhunters.

12. The song mentions the balida , a batten used in weaving, as the instrument that brought down the victim. Villagers customarily think of tradition and the ongoingness of adat as a blanket that is continually being woven. In keeping with this metaphoric logic, the "balida" by which the enemy was slain was a war knife used for the sake of adat.

13. Using, as it does, metaphors having to do with weaving, the first song comments on the headhunt and the providential gifts it will bring about. One weaves sau' lambe' , the cloth for a striped blanket, in lengths of approximately 3 meters. The threads run the length of the cloth, from one end of the loom to the other, and then back again, and are thus quite long. The back-and-forth run of the thread is here a metaphor for the headhunting journey. When making such a blanket, women first lay out a skein of thread between two stakes ( tandajan ) planted in the ground. The blanket in the song has its beginning at Takapak, a Mandar village, and is completed at Rantetarima—the place where uplanders receive ( tarima ) the returning headhunters. The reference to Da'ala seems to point to the name of a Mandar village. Still, it is worth noting that the term echoes the Muslim expression Allah ta'ala , roughly, "Allah the Perfect."

14. See George (1990), Fabian (1992), Street (1984), and Tedlock (1983) for discussions about the theoretical implications of the dialectic between the written and the oral.

15. I leave aside the problem of chronological or temporal privileging. That is, I tended to favor the interpretation I had heard first as being the authoritative and conditioning one.

16. The discourse of contemporary headhunting ritual thus fails to illuminate the past with the imagery of husband and wife. Interestingly, this is precisely the imagery used in local myth to explain the social and political origins of the highland region (see Chapter 2).

17. In a slightly different version of this chapter subsection (George 1993b), I have argued that differences in song interpretation are symptomatic of the same historical forces that have led one of these two communities to tolerate village exogamy and the other to avoid it.

1. For key discussions and case studies about adat, agama, and local ritual practice, see: Acciaioli 1985; Aragon 1992; Atkinson 1983; Bowen 1991; Geertz 1983; Hefner 1985; and Kipp and Rodgers 1987.

2. I thank Mary Steedly for reminding me about the importance of this term. Kepercayaan corresponds with ritual expression that does not have the status of religious practice (i.e. agama) and shows itself recalcitrant to reformulation as secular custom (i.e. adat). See Steedly (1993:69-70).

3. Beneath a somewhat different rubric, I have elsewhere examined the process of oppositional collaboration as it relates to secrecy and concealment in the mappurondo community (George 1993a). See, also, Chapter 2, and below.

4. In fact, two mappurondo communities still exist in Mambi, and still observe pangngae. One, at Saluassing, consists of eight households; because it is too small to hold pangugae on its own, it joins in festivities at Minanga or Saludengen. The other community is in Salukepopo', a village deeply divided over religious issues; there, mappurondo households make up a little over 40 percent of the total, and are successful at holding pangugae annually. I regret that I had no opportunity to join in the ritual there.

5. I made it a habit to remove myself from the regulated soundscape of mappurondo settlements when working on the musics of pangngae, unless, of course, pangngae was in progress.

6. Papa' Amang's parody of singing style is also a surreptitious dismissal of the song's meaning. The music and meaning of the sumengo are noise and nonsense to the car of this civil officer.

7. Typical scenarios for conversion to Christianity are: (1) marriage, in which a husband always follows the religion of his wife; (2) entry into middle school, which requires a student to profess a recognized religious faith; and (3) parents "following" their children into Christianity, if all of the latter have converted.

8. Comprising the following principles: (1) the belief in one supreme God; (2) a just and civilized humanitarianism; (3) the unity and integrity of Indonesia; (4) consensual democracy; and (5) social justice for all Indonesians.

9. Both Makatonan (1985) and Tamadjoe (1983) discuss the emergence of Kondo Sapata' at midcentury.

10. I have real suspicions about the references to the "golden stake" ( tana' bulawan ) in Tamausa's account. The term alludes to a former hierarchy of prestige and social rank that in my researches appears to have been absent or muted in most head-water settlements. I suspect that the ranking system was peculiar to Mamasa, or has been derived (through cultural literature) from the social order of the Sa'dan Toraja around Makale and Rantepao.

11. Subsequently, the same author would prepare a similar outline for the Kabupaten Office of Education and Culture (Samar & Mandadung 1979). In this later Indonesian version he restores references to headhunting with the explanation that "this pangngae dance merely portrays difficulties as if one were fighting an opponent" and constitutes a "former theory/practice" of war (1979:17).

12. For many readers, the English in this pamphlet will seem awkward and strange. I have no wish to make fun of its flaws or its viewpoints, but retaining the original English seemed the best way to convey a sense of textual encounter, and to acknowledge the tactical struggles and choices of authors trying to work in a language not their own.

13. The claims of these scholars rest largely on linguistic evidence. In their findings on the Bare'e Toraja, menga'e (headhunting) is synonymous with menggae (harvesting). I am suspicious about this, and wonder whether the investigators' interest in cognate word roots might not have sparked the "discovery" of the synonymous relationship between these two words.

14. The mappurondo failure to become "Hindu" contrasts with the successful campaigns of larger, more viable groups in protecting ancestral religion as local variations of Hinduism. See: Boon 1990; Hefner 1985, Kipp and Rodgers 1987, Steedly 1993, and Volkman 1985). Compare, meanwhile, the case of the Meratus Dayak (Tsing 1993:273).