1. My interviews in Cotabato were conducted primarily in Magindanaon, but also in Tagalog and, occasionally, in English. It was not uncommon for more than one language to be used in the course of a conversation.

2. A very important exception to that general trend is provided by a historian of the Philippines, Reynaldo Ileto, in his work Pasyon and Revolution . Ileto inquires of the Philippine Revolution begun in 1896, "how did the [lower classes] actually perceive, in terms of their own experience, the ideas of

nationalism" enunciated by their elite leaders (1979, 4)? It is the very same question I ask of the contemporary Muslim separatist movement in the Philippines, Ileto and I agree that the meaning of these movements was not the same for rank-and-file adherents and movement elites. There are parallels in this book as well with Ileto's focus on folk narratives, religion, and popular beliefs in preternatural powers. Our answers to the question about the perceptions and motivations of ordinary adherents do differ, Ileto launches his analysis from the position of wanting to know "how the traditional mind operates, particularly in relation to questions of change"(1979, 2). He tends to view the behavior of lower-class participants in the revolution as "attempts to restructure the world in terms of ideal social forms and modes of behavior" (1979, 8). Together with most contemporary anthropologists, I am quite skeptical of the existence of a "traditional mind" (of course, they and I have had the benefit of more than twenty years of hindsight since Ileto's initial writing). I also take a different view of the relationship between culture and collective action. I will argue that the independent perceptions and motivations of rank-and-file adherents of the Muslim separatist movement represent not only imaginative but strategic endeavors. That is to say, Muslim subordinates have exercised their political imaginations to envision a more perfect (though not always more "traditional") world, but also to voice their resistance to the official aims of movement leaders, especially when those goals seemed to contradict other, more local considerations.

3. The term "ordinary Muslims" is a problematic one (see Peletz 1997), and I need to define my particular usage before proceeding further. Throughout this work I use the phrase "ordinary Muslims" in three intermingled senses to reference three (near-perfectly) overlapping categories of persons. "Ordinary Muslims" refers first to those Cotabato Muslims (the great majority) who comprise the subordinate (or "lower") classes—those who occupy similarly disadvantaged positions in the regional system of resource distribution. In respect to relations of production almost all may be classified as peasants, low-skilled wage workers, or petty producers or service providers in the urban informal economy. Second, "ordinary Muslims'' refers to those Cotabato Muslims who are not political elites. By my definition, political elites either occupy commanding positions at the head of a social grouping or are able to exert significant influence on those occupants as the result of their control of political and economic resources. In its third sense, "ordinary Muslims" refers to those Cotabato Muslims who possess neither a strongly self-conscious ethnic identity as Moros (Philippine Muslims), nor a highly objectified Islamic consciousness. The ethnoreligious consciousness of ordinary Muslims in Cotabato is discussed in detail farther on (see especially chapters 5 and 6). It should be assumed that, despite these similarities, significant differences in religious orientation and practice may be found among ordinary Muslims in Cotabato.

4. This notion of multidimensional "social fields of force" is taken from William Roseberry (1994, 357). Roseberry's own use of the metaphor is a reworking and extension of an image originally provided by E. P. Thompson (1978).

1. The thrust of these efforts has shifted somewhat over time. Earlier works pointed out the contradictions and contestations contained within systems of domination (see, e.g., Abu-Lughod 1986; Comaroff 1985; Ong 1987; Stoler 1985). More recent treatments, responding to perceived simplifications in the literature on everyday resistance generated mostly from outside anthropology (see, e.g., Scott 1985, 1990; Willis 1981), have focused on "rethinking resistance" and its relation to domination (see, e.g., Abu-Lughod 1990; Kaplan and Kelly 1994; Kondo 1990; Lagos 1993; Linger 1993; Reed-Danahay 1993; Woost 1993).

2. This occurred earliest and most prominently in the work of Eugene Genovese (1974) on plantation slavery.

3. Perry Anderson notes, for example, that in some of his writings Gramsci contrasts the terms "hegemony" and "domination"—equating hegemony with "consent" and domination with "coercion.'' Elsewhere, however, "Gramsci speaks of hegemony, not as a pole of "consent" in contrast to another of "coercion," but as itself a synthesis of consent and coercion . . . This version cannot be reconciled with the preceding account, which remains the predominant one in the Notebooks" (1976, 24-25).

4. Raymond Williams's interpretation is also shared, in many of its general features, by such other analysts of the cultural politics of Western capitalism as Ernesto Laclau (1977) and Stuart Hall (1985).

5. In his 1977 work, Williams defines hegemony similarly as "a whole body of practices and expectations . . . a lived system of meanings and values—constitutive and constituting—which as they are experienced as practices appear as reciprocally confirming" (1977, 110). He notes also that hegemony "is continually resisted, limited, altered, challenged by pressures not at all its own" (112).

6. Williams seems to recognize this problem in his remark that "[t]he sources of any alternative hegemony are indeed difficult to define" (1977, 111).

7. Some of these anthropological usages display a similar internal inconsistency in speaking, on the one hand, of hegemony's power to shape the experience of subordinates and, on the other, of its susceptibility to challenge by those same subordinates. Comaroff and Comaroff argue that hegemony—"that order of signs and practices . . . that come to be taken-for-granted as the natural and received shape of the world"—is "habit forming" and so "is rarely contested directly" (1992, 23). At the same time they note that hegemony "is always intrinsically unstable, always vulnerable" (1991, 27). Michael Woost declares that hegemony comprises "both domination and resistance" and is best "understood as the tendential outcome of a struggle for order" (1993, 503). He also observes that "hegemony works behind the backs of those it dominates . . . in the sense that those who become the objects of the dominant ideological discourse are not freely formed subjects/individuals who deliberately decide whether or not to assign legitimacy to a social order" (1993, 516). Maria Lagos criticizes the tendency "to separate resistance from hegemony . . .

as if they were two different processes with different dynamics" (1993, 53). Daniel Linger defines hegemony as "the maintenance of a political structure through the cultural shaping of experience," while also declaring that "hegemony thrives on discontent" (1993, 4).

8. Numerous other anthropological treatments of nationalism have been produced. Some of these (see, e.g., Dominguez 1989; Handler 1988; Spencer 1990; Verdery 1991) have focused predominantly on the discursive productions of nationalist elites rather than on the popular reception of nationalist ideologies at the community level. Others have examined popular responses to national ideas without recourse to the notion of hegemony (see, e.g., Bendix 1992; Bowman 1993; Sluka 1989, 1995) or have applied it in a somewhat different manner than Raymond Williams (see, e.g., B. Williams 1991). Bruce Kapferer has developed the separate notion of "ontologies" (1988) to explain the motivational force of nationalist ideologies—a concept which, though clearly distinguishable from it, resembles hegemony in many respects.

9. Michael Woost faults Scott for failing to recognize the "shifting, uneven character of hegemony" (1993, 503). Deborah Reed-Danahay is troubled by a "disturbing simplification" in Scott's distinction between an official and a hidden transcript of power relations (1993, 223.) Aihwa Ong, more specifically, accuses Scott of "misrepresent[ing] social realities in rural Malaysia as a simple dichotomy between a national hegemony and a resistant village subculture" (1995, 188).

10. A separate, ethnography-based critique has been recently leveled at Scott by Sherry Ortner (1995) and (in more detail) by Michael Peletz (1997). It is that by not considering ambivalence (mixed emotions) as a significant phenomenon among subordinates, Scott produces an "ethnographically thin" account of dominant-subordinate relations (Ortner 1995, 190). More generally, studies of domination and resistance that do not devote analytical attention to the problem of ambivalence run the risk of being "highly anemic with respect to their treatment of the cultural psychology of the social actors who are at the center of their inquiries"(Peletz 1997).

This critique may appear similar to one above but is quite separable. Ambivalence on the part of subordinates may arise from divided interests (stressed by Ortner) or as the result of moral constraints (emphasized by Peletz) or from various other sources, and may under certain circumstances produce political paralysis. But such paralysis is not brought about by the workings of hegemony. Ortner and Peletz seek primarily to "thicken" the ethnography of domination and resistance by adding the element of psychological ambivalence on the part of subordinates. See chapter 9 for a discussion of ambivalence as a factor influencing the behavior of Campo Muslim residents.

11. It has become increasingly commonplace to find in postcolonial situations mobilized peasants or urban workers fighting for or actively supporting armed ethnonationalist movements aimed at exchanging one set of state-level elites for another. Most of the armed secessionist struggles of recent occurrence have been carried out by populations exhibiting significant internal disparities in social power. Those struggles include, as a modest sample, armed movements for the creation of independent states or autonomous regions for Kurds,

Eritreans, Palestinians, Sikhs, Sri Lankan Tamils, Kashimiris, Basques, Shans, Ibos, Croatians, Abkhazians, Chechens, Sudanese Christians, and Philippine Muslims.

12. The distinction between nationalist killing and nationalist dying is not an unimportant one. There is, for one, a very great symbolic separation between the "giant exorcism" of the murderous Sinhalese rioting described by Kapferer (1988, 101) and the ritual self-sacrifice of Irish Republican Army's hunger strikers. In addition, the communal rioting provoked by nationalist passions that has erupted in India, Sri Lanka, and elsewhere is a far less dangerous undertaking for individual members of nationalist mobs than is rankand-file participation in an armed separatist movement, even given that a fair proportion of the violent activity of separatist insurgents may be directed against unarmed civilians.

13. An ethical issue often accompanies the methodological problem. There is always the risk, even after an armed struggle has subsided, that eliciting and recording detailed narratives from non-elite adherents concerning their attitudes toward, or actions on behalf of, a separatist movement may endanger them in various ways.

14. For examples of such studies see Bowman (1993), Feldman (1991), Lan (1985), Mahmood (1996), Pettigrew (1995), Sluka (1989, 1990, 1995), Swedenburg (1990, 1991, 1995), and Zulaika (1988). Of these, the works by Bowman, Lan, and Sluka (1989) are principally focused on ordinary adherents and local communities. Those by Feldman and Zulaika take as their subject not nationalist struggles per se but political violence, specifically the ways in which violence is culturally constructed, ritually enacted, and reciprocally exchanged between political antagonists. Mahmood and Pettigrew have produced very different works, each of which focuses primary (though by no means exclusive) attention on the political and military leadership of the Sikh separatist movement. Swedenburg's work (1990, 1991) considers how popular memories of an earlier armed revolt have "undergone secondary revision for the sake of the [contemporary struggle for the Palestinian] nation" (1991, 177). David Lan's work, on nationalist guerrillas and peasants in Zimbabwe, is neither concerned with armed separatism nor, strictly speaking, with ethnonationalism. It is a study of a "popular" revolutionary struggle—one of the last of the twentieth century's "peasant wars" (Wolf 1969) of national liberation. It also remains one of the very best ground-level ethnographic accounts of an armed insurgency.

15. Similar observations have been made in theoretical form by Brackette Williams (1989) and ethnographically by Bowman (1993).

16. In a more recent work, Swedenburg provides more detailed depictions of the "popular" memories of Palestinian revolt he was able to elicit from informants—memories that "led a submerged existence in the everyday realm of private conversation rather than being expounded in the public arena" (1995, 27). He is more concerned, however, with how "popular memory and official Palestinian histories fused into a fairly unified picture" (1995, xxvi). Swedenburg defends his lack of interest in uncovering the "objective truth" of local resentments and rank-and-file dissension by noting that ''solidarity requires us to

learn from and (to a certain extent) be tactically complicit with the silences... of the people with whom we live and study" (1995, xxviii).

While Swedenburg's stance may be an appropriate tactic in his particular circumstances, it represents a counterproductive strategy for the anthropological analysis of nationalism in general. By accepting public silence as genuine consent to the official versions of events and thereby ignoring the profound political tension between local concerns and nationalist goals, we achieve neither solidarity with, nor understanding of, those ordinary citizens caught up in nationalist movements, citizens who, like Slavenka Drakulic in Croatia, feel diminished by nationhood.

17. That refashioning may already be seen under way in various forms in the work of Hermann Rebel (1989) and William Roseberry (1989, 1991, 1994), as well as Derek Sayer (1994).

1. Previous scholars have referred to the language and its speakers as Maguindanao (Mastura 1984; Stewart 1978), Magindanao (Beckett 1982; Ileto 1971; Mednick 1965), and Magindanaw (Llamzon 1978). I adopt the usage of Fleischman (1981b), who found in his linguistic research that Magindanaons "usually refer both to themselves and their language as /magindanawn/and their land as/magindanaw/" (Fleischman 1981b, 57). The results of his survey of native speakers also showed an "overwhelming preference" for the spelling of their language as "Magindanaon." Informal data from my own fieldwork support Fleischman's findings.

2. The estimation of the Muslim population of the Philippines has involved a good deal of numerical uncertainty and political controversy. The figures used here, taken from the 1980 Philippine census figures as reported in Ibon Facts and Figures (see IBON Databank 1981), should be viewed as suggestive rather than definitive. For a discussion of the logistical and political problems associated with counting Philippine Muslims, see O'Shaughnessy (1975), "How Many Muslims Has the Philippines?" See also Majul (1979) for an alternative perspective. For examples of widely varying population estimates for Muslim ethnic groups compare the figures reported in Gowing (1979), Llamzon (1978), and Majul (1979).

3. Three other Magindanaon dialects are spoken outside the Cotabato River Basin by the descendants of settlers who emigrated from there. TagaBiwangan ("from the left") is a place-name referring to those Magindanaon living to the south (or "left") of the Cotabato Basin along the long narrow seacoast that abuts the Tiruray Highlands. Taga-Kawanan ("from the right") is the term used for Magindanaons who have settled to the north on the northwest shore of Ilana Bay. Speakers of the Sibugay dialect are descendants of Magindanaon immigrants to Sibugay Bay on the Zamboanga Peninsula (Fleischman 1981b).

4. An intermediate dialect also exists, that of the Nagtaganen (or people in the middle) of the smaller Kabuntalan Sultanate, located where the Pulangi

splits into two branches. For our purposes, however, it may be subsumed under the Tau sa Ilud dialect.

5. There exists no complete ethnography and very little ethnographic description of the Magindanaon. Published ethnographic accounts of the Magindanaon are limited to an ethnomusicology study by Ernesto Maceda (1961) and an ethnographic appendix to a Ph.D. dissertation by James Stewart (1977). There is also a published proto-ethnographic account (in Spanish) by Blumentritt (1893). Jeremy Beckett (1993) has written on Magindanaon political culture. In folklore studies, Clement Wein (1984, 1985) has translated Magindanaon oral literature. Historical works on the Magindanaon are more abundant. Reynaldo Ileto (1971), Cesar Majul (1973), Michael Mastura (1984), and Ruurdje Laarhoven (1989) have written on the history of the Magindanaon sultanates. Ghislaine Loyre (1991) has contributed an ethnohistorical account of the institutions of the downriver (Magindanao) sultanate and Jeremy Beckett (1982, 1993) has focused on the Magindanaons under Spanish and American colonial rule. There is also a wealth of primary historical sources in the form of European reports from the seventeenth to the twentieth century in English, Spanish, French, and Dutch (see, e.g., Bernaldez 1857; Combes in Blair and Robertson 1903-19; Dampier 1906; Forrest 1969; Montano 1886; Nieto 1894).

6. Population estimates for the Iranun differ wildly. They range from the improbably high figure of 429,000 offered by Gowing (1979) to the impossibly low figure of 6,517 reported in the 1986 regional socioeconomic profile for Central Mindanao (National Economic Development Authority 1986) Both estimates apparently reflect substantial confusion in separating Iranun speakers from neighboring speakers of various dialects of Maranao. They also indicate that one of the capabilities that the Philippine state lacks is the capacity to enumerate, with any degree of accuracy, its peripheral populations. The range I offer may be taken as simply a rough guess.

7. Recent linguistic evidence has challenged prevailing assumptions that the Iranun are a dialectical subgroup of the Maranao and are relatively recent (circa 1765) migrants from Lake Lanao to the Cotabato coast (see, e.g., Ileto 1971; Kuder 1945; Mednick 1965; Warren 1981). After clearly establishing that Iranun is a language separate from Maranao, Fleischman (1981b) reports, based on measurements of cognate percentages, that the Iranun language is centrally located between Magindanaon and Maranao. This discovery lends strong support to the belief held by the Iranun themselves that their language is the original language from which the other two languages diverged (Fleischman 1981b, 70). Thus, when the divergence occurred, surely far earlier than 1765, it was the Maranao and Magindanaon who separated from the Iranun at the coast to move to the inlands and uplands, and not the opposite.

8. The propensity of the precolonial Iranun to travel long distances as maritime marauders, along with the fearsome reputations they gained as such, has led to the problem of distinguishing in historical records and oral traditions actual Iranun speakers from a variety of other sea raiders erroneously identified as "Iranun." The Tausug and other peoples of Sulu still use the term "Iranun"

to refer to all Muslims from Mindanao. And in the eighteenth and nineteenth centuries, the English used "Iranun" (or Illanun) to designate any "Sulu pirate" (Warren 1981).

9. For an ethnohistorical examination of Iranun defiance of external rule, see McKenna (1994). Few other studies of the Iranun exist. Eric Fleischman (1981a) has written a short account of traditional Iranun leadership, and Warren (1981) devotes a long and richly detailed chapter to Iranun raiding in the eighteenth and nineteenth centuries in his history of slave raiding and external trade in "the Sulu zone."

10. Tiruray women who married Muslim men usually became Muslims. In addition, Some Tiruray men recently have been converted to Islam.

11. Those four provinces are Maguindanao, Sultan Kudarat, North Cotabato, and South Cotabato. Of those four, only the first three lie within the Cotabato River Basin. General Santos City, a tertiary city on Sarangani Bay in South Cotabato, has grown rapidly as the extractive and plantation industries of South Cotabato—tuna, timber, and pineapples—have developed.

12. This approximate percentage is based on Philippine census figures for 1980 as reported in the 1986 regional socioeconomic profile for Central Mindanao (National Economic Development Authority 1986). As of 1980, Maguindanao Province had a population reported at 536,546 persons. Of this number, 396,400 (or 73.8 percent) were reported to be ethnic Muslims. In the region as a whole, there are reported to be 92,000 ethnic Tiruray. Almost all of these are located in Maguindanao Province.

13. For the purposes of statistical coverage, the Cotabato Basin may be defined as the area lying within the political boundaries of the three provinces of North Cotabato, Maguindanao, and Sultan Kudarat. Those provinces include large areas outside the Cotabato Basin, but most of their developed agricultural land lies within the basin itself.

14. As an example, in 1986 the basin produced more than 620,000 metric tons of rice on 182,480 hectares, a 15 percent increase over 1982 production, on a somewhat smaller harvest area (National Economic Development Authority 1986). These figures represent an average 1986 yield of 3.39 metric tons of unhusked rice per hectare—one of the highest in the country—and an increase from approximately 2.9 metric tons in 1982. It bears noting that the average yield for the predominantly Muslim Maguindanao Province, at 2.38 metric tons per hectare, was significantly lower than that for the two majority Christian provinces of North Cotabato and Sultan Kudarat, at 3.69 and 3.64 metric tons, respectively. While about 38 percent of the potential rice area in the latter two provinces (41,363 and 24,468 hectares, respectively) had been irrigated by 1984, less than 12 percent of the potential rice land in Maguindanao Province (14,358 hectares) was under irrigation by that year (National Economic Development Authority 1986).

15. More than one geographer has noted that Cotabato City's "eccentric" location on a low swampy interfluve between the two main distributaries of the Pulangi River has deterred its development into a major commercial center on a par with Davao or Zamboanga (Wernstedt and Spencer 1967, 554; Burley 1973, 216).

16. Four of the firms are engaged in timber extraction and processing. The remaining two produce cornstarch and corn oil. There are also a number of medium-size agribusiness operations—primarily fishponds and coconut plantations—in the vicinity of the city.

17. The considerable ethnic diversity found in Cotabato City was specifically addressed by Chester Hunt in a 1957 survey article entitled "Ethnic Stratification and Integration in Cotabato."

18. Hardly any Tiruray reside in Cotabato City.

19. In Davao, General Santos City, and Iligan, as well as in the cities of northern Mindanao, the language of public discourse is generally Visayan. In Zamboanga City, Tagalog is beginning to replace Chavacano as the language of commerce.

20. See, for example, the proceedings of the seminar on "The Mindanao Problem" published in the January 1987 issue of the Philippine journal of current affairs, Solidarity (Jose et al. 1987). See also Bauzon (1991), whose juxtaposition of "Islamic" and "Liberal" paradigms is simply a recasting of the oftrepeated argument that Muslim-Christian conflict in the Philippines is the consequence of clashing worldviews.

21. On more than one occasion, I witnessed Muslims with strongly separatist political sentiments standing to join in the singing of the national anthem of the Philippines at public meetings or seminars—a practice ingrained through education at state schools.

22. Muslim Chinese-mestizos have played prominent roles in Cotabato City in the present century as business entrepreneurs and politicians. See Hunt (1957) for a discussion of the changing character of the Chinese-mestizo community in Cotabato in the 1950s.

23. This information, and most of the quantitative data on the community that follows, was obtained by means of a household survey of Campo Muslim carried out to enumerate the population and obtain basic census information. Every household in the community was surveyed by myself and two research assistants who were also Campo Muslim residents. Information requested from household heads (self-identified) included age, birthplace, occupation, ethnolinguistic identity, length of stay in Campo Muslim, reason for migration to the city, number of occupants of the dwelling, and their relationship to the household head. Data were also collected by observation on house size and type and the presence or absence of water taps, electricity, toilets, radios, televisions, and refrigerators.

Delimiting Campo Muslim for census purposes was a fairly straightforward procedure. The community is bounded on two sides (front and back) by water, and its borders with the communities that adjoin it are clearly marked by roads or open spaces.

24. The terms "Babu" and "Bapa" (literally, "Aunt" and "Uncle'') are used as terms of address and reference for community elders. The names Babu Imun and Bapa Akub, and most names used in this book, are pseudonyms.

25. Some years earlier, an enterprising Muslim owner of a dwelling overlooking the nearby Matampay River had constructed crude pay toilets, which were used primarily by Campo Muslim residents.

26. The Mindanao Cross , Cotabato City's weekly newspaper, would commonly conclude reports of armed robberies in the city by noting that "the suspects escaped into Campo Muslim." Community residents reject the characterization of Campo Muslim as a den of thieves, arguing that the robbers are not from Campo Muslim but that they actually flee through the community to reach the edge of the city and beyond.

1. For examples of political ideologies in the Islamic world similarly based in sanctified inequality see Bujra (1971) and Combs-Schilling (1989). For a discussion of the doctrinal support for sanctified inequality in Islamic (and specifically Philippine Muslim) tradition, see Majul ( 1973, 3-6).

2. Two statements by Nur Misuari, founder and Central Committee chairman of the Moro National Liberation Front, may serve as illustration. In a 1975 policy paper entitled "The Rise and Fall of Moro Statehood," Misuari wrote, "[O]ne unalterable fact of history remains a cornerstone of the present revolutionary movement. This historical fact is inextricably linked to the Bangsa Moro people's inherent desire to be left free and sovereign having their own honoured place in the community of nations. Their national aspiration is nothing more than to enjoy again the prerogative of chartering their own national destiny with justice for all and to see the democratization of the wealth of their homeland" (emphasis mine; quoted in Mastura 1984, 111). In a 1977 speech titled "Cultural Genocide in the Philippines,'' he stated: "The Muslim people and homeland have 500 years of Islamic culture and civilization. They were once free, sovereign and an independent nation. As a matter of fact, they were once one of the strongest powers in Southeast Asia" (quoted in Majul 1985, 136).

3. I use the term "precolonial Cotabato" to refer to the entire period prior to the American colonial occupation of the region in 1899. Although the Spaniards had consolidated colonial control of the northern Philippines by about 1600, they never accomplished the complete political subjugation of the southern sultanates. Downstream Cotabato was occupied in 1861 and formally included as a Spanish possession, but Spanish colonial control never extended much beyond a rudimentary military occupation of strategic points on the Pulangi River (see chapter 4). Political and economic relations within Cotabato seem to have been only slightly affected by Spain's formal possession of Cotabato in the late nineteenth century. It is therefore not inaccurate to extend the "precolonial" period in Cotabato to the turn of the twentieth century.

4. This is a simplified compendium of the opinions of a number of public figures in Cotabato. For an example of this perception in written form, see Damaso (1983). For a rare dissenting opinion, see Lingga quoted in Gowing (1979, 241).

5. I employ the term "traditional" here not in any essentialist sense but as a shorthand reference to a host of local sociopolitical practices, beliefs, forms, and expectations, some of which are described in detail farther on. Some "tra-

ditional" arrangements in Muslim Cotabato are more ancient than others and not all are self-consciously or universally regarded as "traditional culture" (in Magindanaon, adat betad ). They do, however, form a sociocultural domain largely uninfluenced by Western colonialism or the political domination of Christian Filipinos.

6. Datu Mohammad Adil had a long career in the Philippine Constabulary, and when I met him in 1986 he held the rank of lieutenant colonel (ret.). His followers referred to him as "Datu," and I use that honorific in these pages. Most Christians and members of the Muslim elite in Cotabato city referred to him as "Major Adil," referring to the rank he had held for a number of years. In 1992, he was invested with the title "Sultan sa Kutawatu" (Sultan of Cotabato) by the members of the high nobility of Cotabato.

7. Although the ethnographic material presented here reflects a broad range of positions and perspectives, it does present certain limitations. For one, it is somewhat removed, unavoidably, from the precolonial period; most narratives of past relations concern the very late precolonial or early colonial period. Another is that the voices of women, as well as detailed information on the social positions and political actions of women in the Cotabato sultanates, are mostly lacking here. There is somewhat more information on women available for the American colonial period, and I do attempt to make the voices of the women of Campo Muslim heard in the chapters concerning present-day Cotabato.

8. The term "bangsa," or "bansa," is found throughout the Malay world and has been used to refer to descent lines, descent groups, nations, castes, races, or estates. There seems to be some disagreement about its origin. Dewey (1962, 231) suggests that the term is Chinese in derivation. Milner (1982, xv) identifies "bangsa" as having a Sanskrit origin. This latter foreign derivation seems the more plausible one.

9. The term "sultanate" refers here to a political institution based on an Islamic legitimating ideology and headed by a sultan—a formally hereditary leader who possesses the authority to bestow titles and appoint individuals to specialized subordinate offices.

10. The diminution of the significance of local ancestry is by no means an automatic outcome of the introduction of a political system based on an Islamic model. Gullick (1958) and Mednick (1965) each describe systems where local ancestry was not devalued with the adoption of an Islamic political idiom. According to Mednick (1965), the Maranao, after the introduction of Islam, developed a single, complex status system based equally on local ancestry and descent from the prime Islamic ancestor (also Sarip Kabungsuwan). The Maranao sociopolitical system also remained relatively uncentralized. Gullick (1958) describes the ruling class of the Malay states, which were relatively more centralized than the Cotabato sultanates, as having two principal groups: members of the royal line, from which the sultan was chosen, and members of a number of nonroyal but noble local lineages that produced local officials (datus). The noble status of these local lineages was based not on their kinship links with the royal lineage but on their right to fill various chieftainships. This conferral of high status on particular lineages on the basis of ancient

agreements is paralleled in the special situation of one local descent line in Cotabato, the Tabunaway bangsa (see below).

11. Of the twenty tupus listed by Mastura (1984, 34) as having been under the administration of the Buayan Sultanate, only three are named after apical ancestors and two of these ancestors are immigrants who arrived after Sarip Kabungsuwan. Twelve of the tupus are named after their localities, one is not a descent line at all but a category for foreigners ( rafu ), and four represent caste-like groups of craft specialists who held a special dependent status under the sultan.

12. In a Maranao version of this myth, Sarip Kabungsuwan accidentally marries his long-lost sister, a more exact means for establishing an exogenous and unadulterated aristocratic lineage from a single apical ancestor (see Mednick 1965, 97). Similarly structured origin myths for sultanates are found throughout the Malay world. For illustrations from Perak and Maluku see, respectively, Sullivan (1982, 1) and L. Andaya (1993, 53).

13. The intermediate Kabuntalan Sultanate was established in the mid-eighteenth century by a branch of the Magindanao royal house (see Mastura 1984, 36).

14. I am grateful to John Bowen for pointing out the strong parallels between the political and status-ranking systems of the Magindanaon and those of the Bugis of Sulawesi. The Bugis system of rank gradation is quite similar to that found in Cotabato, especially in its assignment of intermediate descentrank to children of unequally ranked parents (see below). Unlike in Cotabato, the central Bugis myth of noble descent is pre-Islamic, describing the Bugis nobility as descendants of "mysterious beings called " tomanurung " [in Bugis, literally, "One who descends"]" (Millar 1989, 43). For recent works in English on the Bugis see Millar (1989) and Pelras (1996).

15. This is the datu version of the original agreement between Sarip Kabungsuwan and Tabunaway. In the version told by present-day Tabunaway descendants (dumatus), the agreement included a clause declaring that the descendants of Sarip Kabungsuwan could not be proclaimed as sultans without the consent of the descendants of Tabunaway, and if this were not done the descendants of Tabunaway would have the authority to cancel the proclamation and take the leadership themselves. The datu version mentions only participation (implying ceremonial participation) in the investiture of the sultan. Mastura (1984, 5) notes that "dumatu" is the future tense of the verb "datu" (to lead) but does not offer an explanation for that designation. As the descendants of Tabunaway, the former ruler of Cotabato, the past tense of "datu" seems a more appropriate appellation. However, if Tabunaway is believed to have reserved the right to reassume the leadership of Cotabato under certain conditions, the term ''dumatu" ("will lead") is descriptive.

16. Traditional Magindanaon status groups have been categorized by scholars in various ways. In general, analogies to classical or feudal European stratification systems have engendered more confusion than clarification. To cite an example, the endatuan have been described as "serfs" (Beckett 1982, 411), "freemen" (Stewart 1978, 244), and "commoners" (Mastura 1984, 33); the dumatu (see below) have been classified as "nobles" (Damaso 1983, 76),

"lesser nobles" (Stewart 1978, 244), and "commoners" (Beckett 1982, 411); and the ulipun have been mentioned as "vassals" (Mastura 1979), ''servants" (Mastura 1984, 33), and "slaves" (Beckett 1982, 411). To avoid similar confusion, I endeavor to track as closely as possible the literal sense of the Magindanaon terms for these social categories.

17. Jeremy Beckett's (1982, 411) treatment of slave status in precolonial Cotabato differs from my own. He lists banyaga as occupying the fourth tier of the stratification system. He then suggests that since the status of an ulipun, at least in theory, could be changed with the clearance of his or her debt, that the ulipun do not constitute an estate. In support of my categorization l add to the reasons outlined above that the term "banyaga" has the literal meaning of "foreigner" or "alien." Gullick (1958, 104) employs a classification similar to mine for slave status in the Malay states. At the same time, it should be noted that actual social relations between ulipun and banyaga—and between those two unfree statuses and the other social strata—were more complex than the necessarily simplified description I present here. For comparison, see W. H. Scott's (1982) detailed account of the subtleties and complexities of the Tagalog slave system.

18. The conceptualization of maratabat among the neighboring Maranao is dramatically different from its use among the Magindanaon. Among the Maranao, maratabat primarily denotes rank honor and sensitivity about rank. It is a central and compelling social value that reflects pre-Islamic cultural traditions. Offended maratabat demands retribution that often takes the form of violent retaliation. The most distinctive aspect of Maranao maratabat is its relation to lineal descent and corporate kin responsibility. To defend one's maratabat is to uphold one's descent line. Both the responsibility for defending maratabat and the culpability for insulting it extend beyond the individuals involved in any particular incident of soiled maratabat. The pursuit of retribution for an offense to maratabat can last for generations. Individuals are socialized to seek revenge for long-past injuries to descent line honor by retaliating against a direct descendent generations removed from the original perpetrator. In a similar manner, badly soiled maratabat may be avenged immediately against a close kin of the offending party. In 1986, Manila newspapers luridly reported the murder of the seven-year-old daughter of a Magindanaon security guard at a Manila mosque by three young Maranao men whom he had publicly insulted the previous day. The reports did not explain the cultural logic behind the tragedy. The association between functioning local descent groups and the cultural intensification of maratabat (both lacking among the Magindanaon) is illustrated by a Maranao saying: "A man who has lost his bangsa has no maratabat" (Saber et al. 1974).

19. "Pamalung ka sa kaing sa saken na kundang aku na manik sa tulugan na entayn muna salegan u bulawan datumanung?"

20. Members of the ruling families of the three principal sultanates of the Cotabato Basin were linked by common descent and occasionally intermarried. Datu Kasim relates that those ties were recognized in the special three thousand-peso (or -dollar) bridewealth payment for marriages between two individuals of pulna rank. This represented one thousand pesos for each of the

three major royal houses of Cotabato. Compare Gullick's description of the "inner circle" of the Malay royal lineage (1958, 59-60).

21. The Cotabato sultanates had neither the number of offices of the Sulu Sultanate (see Majul 1973) nor the elaboration in ranked titles found among the Maranao (see Mednick 1965). Mastura (1984, 35) lists eleven offices (including that of sultan) arranged in three orders of rank for the Magindanao Sultanate.

22. Kinship links through males were, however, favored over those through females (Stewart 1977; Beckett 1982).

23. Those who received such "gift wives" ( tawakim ) were not required to provide bridewealth in return.

24. See Mednick (1965, 140-41) for a description of the use of a tarsila in the tracing of individual descent among the Maranao. For a fascinating, detailed illustration of written genealogies employed in a similar fashion elsewhere in the Islamic world, see Eickelman's (1976, 183-210) discussion of the uses made of silsila (name-chains) by the elite patrilineal descendants of a Moroccan marabout.

25. Similar mechanisms are found, for example, in the traditional political system of the Bugis. For men regarded as extraordinary achievers, the rule of female hypergamy may be suspended and noble ancestors rediscovered. Millar's observation about the Bugis holds equally well for Cotabato: "Given the Bugis assumption that descent-rank is correlated with character, it is natural that low-ranking people with exemplary achievements frequently are thought to have a high-ranking ancestor about whom people do not know. It is also natural that they use the strategy of 'marriage-up' to adjust for these achievement/descent-rank anomalies" (1989, 5).

26. Anthony Reid notes that for Southeast Asia in general in the sixteenth and seventeenth centuries, "state pageantry was the most effective way in which the citizenry was incorporated into the hierarchic state" (1988, 181). "For the majority of the population [royal and religious] festivals served three important purposes: participation in the majesty and hierarchy of the state; economic activity, such as marketing and rendering tribute; and entertainment" (ibid.).

25. Similar mechanisms are found, for example, in the traditional political system of the Bugis. For men regarded as extraordinary achievers, the rule of female hypergamy may be suspended and noble ancestors rediscovered. Millar's observation about the Bugis holds equally well for Cotabato: "Given the Bugis assumption that descent-rank is correlated with character, it is natural that low-ranking people with exemplary achievements frequently are thought to have a high-ranking ancestor about whom people do not know. It is also natural that they use the strategy of 'marriage-up' to adjust for these achievement/descent-rank anomalies" (1989, 5).

26. Anthony Reid notes that for Southeast Asia in general in the sixteenth and seventeenth centuries, "state pageantry was the most effective way in which the citizenry was incorporated into the hierarchic state" (1988, 181). "For the majority of the population [royal and religious] festivals served three important purposes: participation in the majesty and hierarchy of the state; economic activity, such as marketing and rendering tribute; and entertainment" (ibid.).

27. The Bugis formally acknowledge the distinction between authority based on "proven superiority" (Millar 1989, 34) and that based in inherited nobility by recognizing a distinct category of individuals known as tau matoa , or "outstanding leaders" (1989, 6). That category crosscuts the status ranking system inasmuch as not all tau matoa are members of the high nobility. Many tau matoa do eventually obtain noble status for themselves or their descendants by means of the social mechanisms noted above.

28. Shelly Errington reports similar conceptions of highly potent individuals as inherently, and often unintentionally, dangerous in Luwu, South Sulawesi: "[T]he potent stinging energy of rulers and high nobles exists quite apart from their intention" (1989, 61).

Beliefs in the possession of special divine powers by rulers (in Malay, daulat ; in Javanese, wahyu ) that could be harnessed for the welfare of the community are found throughout Islamic Southeast Asia and have been shown to

have pre-Islamic origins (see, e.g., B. Andaya 1975, 1979; L. Andaya 1975a, 1975b, 1993; Gullick 1958; Reid 1993).

Throughout the Malay world we find related belief in the punitive effect of such powers, with supernatural punishment (in Malay, timpa daulat ) automatically befalling those who disobey or disrespect rulers (see, e.g., B. Andaya 1975, 1979; L. Andaya 1975; Gullick 1958). Such beliefs are also found beyond the Malay world. Errington reports them for Luwu, noting that "a person who failed to get off a horse or close an umbrella when passing in front of a high noble's house" would suffer supernaturally inflicted malady or misfortune (1989, 62). In other precolonial Islamic polities such beliefs seem to have been less developed (L. Andaya 1975) or nonexistent (L. Andaya 1993). I have found no explicit evidence for belief in supernatural sanctions for violations of rank honor in precolonial Cotabato.

29. Datu Adil, for example, tells of how as a boy he was sent by his father to a guru to learn the kamal arts.

30. In the preface to his work on Malay political culture on the eve of colonial rule, Milner (1982) describes his analytical goals—goals that seem representative of this interpretive approach as a whole: "I wanted to understand Malay political activity in Malay terms. In order to investigate the process of change during the colonial period, it was necessary to examine first not political institutions or the flow of "real power," but what Clifford Geertz has described as the "meaningful structures" by means of which Malays gave shape to their political experience. I needed to explore Malay political culture" (1982, viii).

31. The three most important informants cited by Geertz in his book Negara belonged to the traditional Balinese ruling class. Two of the three were members of the core nobility (1980, 142). For her work Meaning and Power in a Southeast Asian Realm , Shelly Errington relied on informants most of whom were the descendants of the former ruling family of the kingdom of Luwu in South Sulawesi (1989, 22).

32. Similar criticism has been made of those using an exclusively dyadic alliance perspective in examining modern political relations in the Philippines (see, e.g., W. Wolters 1984; Davis 1986; Kerkvliet 1990).

33. Beckett reports an oral tradition that the followers of Datu Utu, the last independent Sultan of Buayan, deserted him for his rival and former protégé, Datu Piang, because he refused to open his granaries to them during a time of famine (1982, 399). I suspect, however, that this is less an illustration of a norm of redistribution and the consequences of its violation, than of a typical response by followers to a critically weakened ruler who had lost most of his coercive power (see Ileto 1971, 95; and below).

34. Because an adjudicator's payment was most often taken as a percentage of the fine assessed, impartiality in the dispensation of justice was probably a practical impossibility (see Mednick 1974, 19-20).

35. Based on the ethnohistoric data she collected, Ghislaine Loyre notes that "lilt would . . . seem that the [Luwaran] was hardly known and hardly used except by the legal specialists who advised the Sultans" (1991, 69). She also remarks that physical punishments, including executions, were very often

commuted to fines by adjudicators, citing for example the report of an eighteenth-century chronicler that "robbers could have a choice between having their hands cut off or paying three times the values of the stolen property" (1991, 47). While she explains such commutations as "intended to put a limit to violence" (1991, 46), it seems more probable that they reflected the political ecology of precolonial Southeast Asia—a region with plentiful land and relatively low populations in which local rulers competed for relatively scarce followers (see Reid 1988; and below).

36. Anthony Reid (1993, 268-69) provides several additional examples of arbitrary power from various precolonial Southeast Asian polities but notes that this form of political culture was not immutable: "[T]here were other times and places—Melaka around 1500, Banten and Patani around 1600, Aceh apparently in the 1580s and certainly in the 1650s, sixteenth-century Banda—when the great merchants were secure against arbitrary power and tended to build fortified compounds and brick warehouses" (1993, 269).

37. Forrest was a British East India captain who spent eight months at the capital of the Magindanao Sultanate in 1775-76, ninety years after Dampier's visit. His mission (ultimately unsuccessful) was to attempt to arrange the establishment of an English factory at the Magindanao Sultanate. His account of his visit is considered the best and most complete description of Cotabato during the eighteenth century.

38. Datu Adil recalled tales of how Utu's fierce and unpredictable nature made even his datu vassals fear for their lives: "Datu Utu would develop a craving for venison and send his datus out to hunt for him. The datus would encircle a valley, send in slaves as beaters, and shoot any deer they flushed. It was said that any datu who let a deer escape by missing his shot would immediately flee the territory rather than face the wrath of Datu Utu."

39. These accounts present a one-sided picture of class relations. Covert acts of resistance, such as those detailed in James Scott's Weapons of the Weak (1985), are absent in historical accounts and are not recounted in oral traditions. Subordinate behavior such as false compliance, foot dragging, and pilfering (see Scott 1985, 29) is conducted individually and ""offstage""(1985, 25) and is often not susceptible to public retelling. Of course "offstage" telling, such as relating the "secret" sins of particular datus, is itself a form of resistance. One of the few examples of overt resistance related to me was a story, told by Datu Adil, concerning the melitan , a castelike group of potters who were bound to the Sultan of Buayan. The sultan, at one time, had a member of the melitan put to death arbitrarily. In response, the melitan as a group refused to make any more pots and were able eventually to extract a promise from the sultan not to execute any more melitan without legal cause.

40. Like the janissaries of the Ottoman Empire, these bodyguards were Christians who were enslaved, trained, and formed into a military corps by a Muslim ruler. While the janissaries were acquired from the subjected Christian communities of the Balkans by means of the devshirme , a tribute of boys paid to the Ottoman sultan, the Cotabato bodyguards were taken in slave raids by Muslims on the Spanish-controlled Christian communities of the central Philippines.

1. Those sultanates most immediately affected by the fall of Melaka to the Portuguese include Patani, Aceh, Banten, Perak, Pahang, and Johor (Reid 1993).

2. The paired positions of tribute-taker and tribute-provider correspond generally with two classes that may be identified as nonproducers and direct producers. However, not all tribute-providers were direct producers. In addition, while the two class positions were mutually exclusive, the social positions were not, some individuals being both tribute-takers and tribute-providers.

3. As in the rest of the prehispanic Philippines and much of insular Southeast Asia, the division drawn between freemen and slaves often carried more social than economic significance (see, e.g., Warren 1981; W. H. Scott 1982, 1994). The economic positions of slaves and freemen were often indistinguishable, and, in the case of debt slavery, an individual might move from freeman to slave and back again more than once in his or her lifetime.

4. Wang Ta-yuan's Tao I Chih Lueh ( Summary Notices of the Barbarians of the Isles ), written in 1349, is the earliest recorded account that specifically mentions the Cotabato Basin. According to William Henry Scott, Wang describes "Mintolang" (Mindanao) "at the mouth of the Pulangi River in Cotabato [as] a strategic location with good communications to the sea and an abundance of rice and grain" (Scott 1984, 73). Cotabato at this time was exporting forest products such as sandalwood, ebony, and animal hides; and absorbing, in addition to a wide range of cloth and metal goods, such luxury items as gold, silk, and porcelain. In the Ming annals, begun in 1368, the rulers met in Mindanao were designated by the Chinese term for monarch (Wang) while overlords in Luzon in the northern Philippines were referred to as chieftains (W. H. Scott 1984, 78).

5. As noted in the previous chapter, the term "datu" refers both to hereditary members of a traditional nobility who claim the right to rule, and to actual leaders who command followers, and who are usually, but not always, members of the high nobility. Because of a system of cognatic kin reckoning, and various other sociopolitical factors, there are always more individuals able to claim membership in the traditional nobility than there are active leadership positions. Stewart (1977) distinguishes between these two meanings of "datu" by capitalizing the term when referring to positions of political leadership and using the lower case when speaking simply of the designation of rank. I have not systematized my references in that manner, and instead either modify "datu" with the term "ruling" when referring particularly to political leadership or assume the distinction is made clear from the context of the passage.

6. Forrest notes in reference to slaves being used as units of valuation in Cotabato: "Talking of the value of things here, and at Sooloo, they say such a house or prow [ prahu , or ship], etc. is worth so many slaves; the old valuation being one slave for thirty kangans [or bolts of imported cloth]" (1969, 280).

7. For detailed descriptions of the politics and economics of debt-bondage in the Malay states see Gullick (1958) and Sullivan (1982). Although Sullivan is pointedly critical of Gullick's functionalist approach to the indigenous

Malay political system, his depiction of debt-bondage in Perak is in essential agreement with Gullick's overview.

8. In her Triumph of Moro Diplomacy: The Maguindanao Sultanate in the 17th Century , Ruurdje Laarhoven provides a vivid illustration of the practical limits on a sultan's political authority. She relates a Dutch account of the downfall of Sultan Kuday, who had succeeded to the Magindanao throne under a cloud of controversy in 1699: "[I]n January 1701, the sultan issued orders to completely close off the river at both ends of the [sultanate], and had it heavily guarded. The purpose was to have complete control over his subjects' movements, because he 'had introduced a new invention on how to raise money'"(1989, 103). It was reported to the Dutch ambassadors that the Sultan's scheme to require his subjects to purchase passes to travel into and out of the sultanate, and his threat to reduce any violators to debt-slaves, caused a massive defection of his datus and their followings. "As a result, Sultan Kuday was left only with 30 men, most of them slaves, in Simuay. The state council thus elected Anwar as the new king" (Laarhoven 1989, 103).

9. For a detailed account of the preferential access of seventeenth-century Magindanao sultans to external trade opportunities, see Laarhoven (1989).

10. It is tempting to view the two rival power centers on the Pulangi River as typifying van Leur's (1955) dichotomy between inland states and harbor principalities in the Malay world. The upriver sultanate was relatively inward-looking and drew resources from intensive rice cultivation, while the downriver sultanate was focused externally and founded on maritime trade. However, as Ileto (1971) notes, the downriver (Magindanao) sultanate also had an agricultural base in the delta and the upriver (Buayan) sultanate had other exits to the sea aside from the Pulangi River which allowed maritime trading and raiding.

11. Forrest uses the term "Haraforas" to refer to the Tirurays he describes. Dampier, writing a century earlier, employs a similar term—"Alfoores"—in reference to a distinct population "under the subjection of the Sultan of Mindanao" (1906, 333). These designations are undoubtedly related to the Molukan term "alifuru." As described by Leonard Andaya (1993), the alifuru were the interior inhabitants of Halmahera, distinguished from the ngofagamu —the common people (literally, ''people of the land"). "Alifuru" was used generally to designate a number of distinct non-Muslim peoples of the interior highlands who were attached as client groups to various Muslim rulers. The reports of Dampier and Forrest suggest that the term was used in the same way in Cotabato. While the term "alifuru" apparently remains in use in Halmahera, I found no evidence for its continued use in Cotabato. For additional evidence for social and cultural linkages between Malukan and Mindanao in the sixteenth and seventeenth centuries, see Andaya (1993), Laarhoven (1989), and Majul (1973).

12. Beeswax was an especially desired item for external trade, particularly with Europeans. Most of the beeswax from Cotabato was shipped by Chinese junks to Manila for the candle-making industry (Wickberg 1965). It was also the primary export item in the Cotabato-Dutch trade (Laarhoven 1989). Also important at various times were gutta-percha and almaciga, two tree saps used

in insulation and varnishes, respectively, and traded via Sulu and Singapore to European markets.

13. Among the Chinese articles carried to Cotabato in the late eighteenth century, Forrest specifies "especially kangans [bolts of cloth], beads, gongs, china basons [ sic ] with red edges; deep brass plates, five in a set; deep saucers, three and four inches in diameter; brass wire, and iron" (1969, 281). Forrest also remarks on rates of exchange between the Tiruray and Magindanaons, most specifically in the following passage: "One day, near Tubuan, a Harafora [referring in this context to a Tiruray] brought down some paddy from the country: I wanted to purchase it; but the head man of the village, a Magindanoer, would not permit him to sell it to me. I did not dispute the point; but found afterwards, the poor Harafora had sold about three hundred pounds of paly [ palay , or unhusked rice] for a prong, or chopping knife" (1969, 282).

14. Dampier also found goldsmiths, blacksmiths, and carpenters in the capital and a thriving shipbuilding industry (1906). When Forrest visited in 1775, he counted at least 370 buildings in the three settlements that made up the capital, as well as "many Magindanao mechanics, vessel builders, and merchants" (1969, 178-84).

15. Little information exists on the activities of the Iranun prior to the mid-eighteenth century. Laarhoven relates reports from Dutch visitors to the Magindanao Sultanate in 1700 that the Iranun were portrayed to them as dangerous rogues and "one of the least trusted and wildest groups in the territory of the sultanate" (1989, 111).

16. Some Iranun cruisers ( mangayaw prahus ) were as much as one hundred feet long, held three banks of oars, and carried close to two hundred men (Warren 1981, 179).

17. Most Filipino captives were sold initially at slave markets in Cotabato or Sulu. Before 1800, they were usually purchased by Bugis or Brunei merchants for transshipment to the Dutch port cities of Batavia, Malacca, Makassar, Palembang, or Banjarmasin. There they were resold to Dutch or Chinese households as servants, boatmen, laborers, and concubines. By 1800, however, Jolo Island, the capital of the Sulu Sultanate, had become the most important slave center in the entire region, absorbing the majority of Filipino captives brought there (Warren 1981).

18. Long-distance sea raiding carried out from Cotabato predated European penetration of the region and was probably very similar to that found elsewhere in the precontact Philippines. Raiding, or mangayaw , was a socially approved activity throughout the Philippines (and all of insular Southeast Asia) before the Spanish occupation. The primary objective of these (often reciprocal) raids was the acquisition of slaves for ransom, sale, or sacrifice (W. H. Scott 1982, 91). Seaborne raiding cannot be neatly separated from maritime trading. William Henry Scott, in fact, suggests the term "trade-raiding" (1982, 85) be applied to these activities because raiding victims were often former trading partners and captives were usually treated as commodities and sold as chattel slaves.

19. Laarhoven cites Dutch reports (again from 1700) to the effect that the Magindanao sultan did not have jurisdiction over the Iranun, "for they were in

the hands of their own chiefs" (1989, 112). The Dutch visitors did witness certain Iranun datus apparently rendering tribute to the Magindanao sultan. Laarhoven notes, however, that "no mention was ever made of the Iranun datus submitting to the obeisance ceremony" (1989, 111).

20. For a detailed account of the Chinese in seventeenth century Cotabato, see Laarhoven (1987).

21. In his description of Magindanaon laws, Forrest notes that the "Chinese seem to be excluded from the benefit of law: those in power often forcing kangans [bundles of trade cloth used as currency] upon them, and making them yearly pay heavy interest" (1969, 277).

22. Prior to the late nineteenth century, the Spaniards attempted only one additional full-scale military assault against Cotabato. Their 1639 campaign was thwarted by the downstream Sultan Kudarat (by every account an extraordinary military and political leader), with the assistance of the Iranun datus of the coast (Majul 1973; Laarhoven 1989). A principal tactic in Sultan Kudarat's anti-Spanish strategy was simply to withdraw his warriors and populace inland, thus forcing the Spaniards away from their coastal supply bases in order to pursue their attack (Laarhoven 1989).

23. See, e.g., the instructions given to Captain Esteban Rodriguez de Figueroa, who, in 1578, was commissioned to subdue the Moro sultanates of Sulu and Mindanao. They direct Rodriguez to promote trade with the Moros, explore their natural resources, Christianize them, and compel them to acknowledge Spanish sovereignty, in that order (Blair and Robertson 4: 174-81).

24. The only evidence I have seen from Cotabato is found in occasional references in the letters sent by Jesuit missionaries from the mission at Tamontaka. A typical passage, from January 1894 reads: "Three Sharifs [Islamic teachers] arrived with the mail boat [presumably from Jolo]. The same as always! When will they [referring to the Spanish colonial government] be convinced that they are those who oppose not only our ministry but also our dominion?" (letter of Mariano Suarez to the Mission Superior quoted in Arcilla 1990, 378).

25. In his impressive work, The Sulu Zone 1768-1898 , James Warren (1981) examines and rejects the argument that Muslim raiding and slaving in the sixteenth to eighteenth centuries should be viewed within the framework of the "Moro Wars" as "retaliation against Spanish colonialism and religious incursion" (1985, xvi).

26. Laarhoven cites a late-seventeenth-century Dutch source who reports that "the trade with Manila never stops" (1989, 147). Dampier, who visited Cotabato in the same period, remarks of the Magindanaons that "their trading vessels they send chiefly to Manila" (1906, 340).

27. Cesar Majul's 1973 book, Muslims in the Philippines , is a Muslim nationalist history and is best evaluated in the context of the political environment in which it was written (see chapter 8). It compares favorably with Philippine nationalist histories, which virtually ignore the existence of Philippine Muslims (see Agoncillo 1969; Constantino 1975). Majul's corpus of work in general (see especially his 1985 work, The Contemporary Muslim Movement in the Philippines ) is an impressive collection of carefully researched historical and political writings on Philippine Muslims.

1. In addition to the materials cited below, see, e.g., Beckett (1982), who quotes a 1927 report from a colonial administrator entitled "Who's Who among the Datus (1982, 405). See also the chapter on "Moros" by former Governor General Forbes in his 1928 work entitled The Philippine Islands .

2. In his Knowledge and Power in Morocco: The Education of a Twentieth-Century Notable , Dale Eickelman offers a more general methodological justification for the use of "capsule social biographies," noting that they are a particularly useful means of "understanding wider social and political realities" in complex historical settings (1985, 14, 15).

3. Evaluating the events at Bud Dajo twenty-eight years later, Vic Hurley, a writer generally supportive of American military efforts in the Philippines, had the following to say: "By no stretch of the imagination could Bud Dajo be termed a 'battle' . . . There appears to be no justification for the intensity of the bombardment at Bud Dajo, and many Americans who witnessed the battle concur in this belief" (1936, 186).

4. In 1913, the last large-scale military action by American troops against Philippine Muslims took place, also in Sulu. At the battle of Bud Bagsak, approximately five hundred Muslim rebels (who were resisting an American disarmament policy) died after their fort was bombarded and stormed. American casualties were limited to fourteen killed and thirteen wounded (Gowing 1983, 240).

5. The short-lived tribal ward system seems to have been a compromise between the views of colonial administrators such as General Wood, who generally disdained Philippine Muslims and had little use for the traditional nobility, and those, like General George W. Davis, who in a 1901 report argued: "It seems to me that the worst misfortune that could befall a Moro community and the nation responsible for good order among the Moros would be to upset and destroy the patriarchal despotism of their chiefs, for it is all they have and all they are capable of understanding" (Report of Brigadier General George W. Davis to Luke E. Wright, Vice-Civil governor of the Philippine Islands, 4 December 1901, Bureau of Insular Affairs Records, file No. 5075-2, National Archives).

6. The "ama ni" form of reference is itself an honorific sometimes used as an alternative to the term "datu."

7. The very first armed challenge to American rule in the entire Muslim Philippines occurred in 1902 at Parang on the Cotabato coast, where Iranun fighters fired upon an American military patrol, killing one of them (Gowing 1983, 84). That initial attack was followed by a series of guerrilla-style raids over the next two years against U.S. forces (1983, 154). Sporadic Iranun armed resistance continued under various local leaders into the 1920s, meaning that Iranun insurgency began earlier and lasted longer than any other armed anti-American opposition in the Muslim Philippines. None of the Iranun efforts, however, attained the scale of Datu Ali's uprising or of other movements elsewhere in the South. American reports typically characterize Iranun armed defiance as "banditry" (see McKenna 1994).

8. Volkman (1985) reports very similar occurrences of elites substituting slave children for their own in colonial schools during the early years of the Dutch occupation of the Toraja highlands in South Sulawesi. Datu Adil informed me that certain of the slave children sent to American schools in Cotabato in the early colonial period went on to become some of the very first Magindanaon teachers and bureaucrats, and that their slave origins are a very closely kept secret.

Collaborating datus such as Balabaran also sent their slaves when asked to provide recruits for the Philippine Constabulary, a colonial police force with American officers, organized in Cotabato in 1904. The principal mission of the constabulary throughout the Philippines was to apprehend insurgents, identified officially only as "brigands" or "outlaws" after the inauguration of civil government in 1901 (White 1928).

9. A similar story is told that Sarip Kabungsuwan, the legendary founder of the Cotabato sultanates, could kill a man simply by pointing his finger at him. While Kabungsuwan ruled Cotabato on his own, Piang required external assistance. In the story about Piang, finger-pointing alone is not enough to kill a man. Also needed is an incantation (in Spanish) marking the intended victim as an enemy of the colonizers.

10. Letter of Joseph Ralston Hayden to Dr. Barr, September 21, 1926. Joseph Ralston Hayden Papers, Box 28, Folder 26, Michigan Historical Collections, Bentley Historical Library, University of Michigan, Ann Arbor. In the same letter, Hayden reports making the acquaintance of "Sunset" Cox, a colorful American colonial character—former American soldier, Philippine Constabulary officer, mercenary, and journalist—who had recently sold his services to Datu Piang as a publicist.

11. The Hayden Papers contain an unpublished manuscript on the life of Bai Bagungan by C. Montera (Box 27, Folder 30, Michigan Historical Collections, Bentley Historical Library, University of Michigan, Ann Arbor). While the prominent participation of Magindanaon women in economic and political life in the modern period is not an uncommon occurrence (see below), there is disappointingly little information available about the economic and political roles of Magindanaon women in the precolonial period. A Jesuit missionary writing in 1888 did offer the following intriguing comment about the wife of the Sultan of Barongis (a small upriver sultanate uncolonized at the time), giving indication that prominent public roles for women (aristocratic women at least) are not just a "modern" phenomenon: "The sultana is still young, very alert, and speaks with great self-assurance. She attends all the bicharas [political consultations] of some importance, and as a matter of fact, she is the one who wields the baton in the sultanate" (Ramon Bea to the Mission Superior, October 4, 1888, quoted in Arcilia 1990, 282).

12. While I have been unable to verify Datu Adil's account, it seems fairly certain that Datu Ortuoste was raised at the Jesuit mission in Tamontaka on the southern branch of the Pulangi River. That mission operated from 1862 until the end of the Spanish occupation of Cotabato. In 1872 the mission opened an orphanage for "ransomed slave children" (Arcilla 1990, xx). In that year a severe famine forced a number of Cotabato datus to sell their juvenile slaves. The mission continued the practice of purchasing children from Chinese

middlemen or directly from Muslims, usually in periods of epidemic or famine, for more than twenty years.

Datu Ortuoste's surname is of interest because it is identical to that of one of the most important officials of the Spanish colonial period in Cotabato, Don Pedro Ortuoste. Don Pedro was the official interpreter for the colonial government in Cotabato. While not an especially high-ranking office, the position carried a significant amount of actual power (Arcilla 1990). Don Pedro spoke Magindanaon and is reported to have been on very good terms with Cotabato Muslims (1990, 31), but it seems unlikely that he was the natural father of Datu Ignacio Ortuoste and there is no indication of any such connection. He was more likely a foster father or godfather. Missionary letters from Tamontaka note that the prominent Spanish families of Cotabato town sponsored "ransomed" children, acted as godparents at their baptisms, and provided them with both Christian names and Spanish surnames (Arcilla 1990).

13. An article in the Philippine Herald from September 11, 1933, describes in detail the funeral of Datu Piang and notes that Ignacio Ortuoste acted as "toastmaster" at Piang's "necrological service" at the burial ground (Bureau of Insular Affairs Records, File No. 5075 (post-1914), National Archives).

14. Datu Adil is referring here to spirit crocodiles ( mga pagali ); see chapter 8.

15. An early marriage was to the daughter of the Sultan of Kabuntalan (Beckett 1977).

16. An illustration of Datu Sinsuat's regional influence is found in a story told by Datu Adil of an event witnessed by his father circa 1930.

After the death of Sultan Mastura, all the leading datus of Cotabato gathered in the gambling house of Datu Sinsuat in Cotabato City to decide who the next Sultan of Magindanao should be. As established by tradition, a special panel of the nobility had been chosen to find the best hereditary candidate. The panel reported that they were undecided and asked Datu Sinsuat his opinion. Sinsuat pointed to a young man in the back of the room, Datu Esmael. He said, "Esmael is the most handsome and fair-skinned of all the candidates and he has sufficient blood ties. He should be sultan." The others agreed and Esmael was named Sulutan sa Magindanao

17. Quoted in Provincial Circular No. 98, January 15, 1935, by D. Guitterez, provincial governor of Cotabato. Joseph Ralston Hayden Papers, Box 27, Folder 30, File 1, Michigan Historical Collections, Bentley Library, University of Michigan, Ann Arbor. A letter from Cotabato Governor Guitterez to Director Guinguna (contained in the same file) notes that Datu Sinsuat was "apparently worried" about the directive. I have found no evidence, however, indicating that the directive substantially curtailed the incidence of traditional adjudications by Sinsuat or other Cotabato datus.

18. Datus Piang and Sinsuat each chose early on to collaborate with American authorities in order to gain an advantage over powerful and aristocratic competitors by allying with dominant outsiders.

19. Beckett states that it is "difficult to locate the source of this expansion," noting that there were very few Christian settlers in Cotabato until the

1930s, that tenancy rates were quite low, and that very few datus emerged as major landowners (1982, 403). He suggests, and I agree (see below), that the expansion of production must have taken place within the framework of traditional production relations.

20. Ileto (1971) suggests that the relative peace that existed between Datu Utu, the last independent Sultan of Buayan, and the Spaniards for roughly twenty years (ca. 1864-84) "was, to a great extent, due to the commercial rapport between them." After the Spaniards took control of the delta, commercial trade between sa laya and sa ilud went on much the same as before, with the Spaniards merely taking the place of the Magindanao Sultanate. Agricultural production was intensified upriver not only to supply Spanish garrisons but to compensate for the production shortfall in the delta caused by the migration of a great part of the delta's population upriver in advance of Spanish forces (1971, 30-31).

21. American colonial discourse on Philippine Muslims is peppered with such expressions as "Moroland," "Moro Policy," "the Moro Problem" (see below), "Moro country," and "Moro bandits.'' Specific Muslim ethnolinguistic groups were virtually always distinguished by their geographic location—as "Joloano Moros," "Lanao Moros," or "Cotabato Moros"—rather than by the names they called themselves (see Gowing 1983; Thomas 1971).

22. The use of this phraseology by representatives of the state to refer to various difficulties encountered in attempting to rule Philippine Muslims has exhibited remarkable longevity. Its first appearance in writing seems to have been as the title of Saleeby's 1913 essay. It was then used throughout the colonial and commonwealth period (see the various citations to the term in the index to Thomas 1971). In 1954, the Philippine Senate appointed a committee to study the "Moro Problem." The recommendations of that committee led to the creation of the Commission on National Integration in 1957 (Majul 1985; Tamano 1974).

23. Saleeby was certainly aware that Datu Piang, Datu Sinsuat, and others of the leading datus of Cotabato during his stay there had little in the way of blood ties to the high nobility of the Cotabato sultanates. Nevertheless, he seems to have been much taken with the idea of a traditional Muslim aristocracy in Cotabato. He spent a good deal of time with Datu (later Sultan) Mastura, who, while a prominent member of the high nobility of the Magindanao Sultanate and a direct descendent of Sultan Kudarat, was little more than a local dignitary. Saleeby describes Datu Mastura as "the best-informed datu of Magindanao" and declares that he possesses "the most reliable of the royal documents that have been preserved" (1905, 36). Mastura allowed Saleeby to copy those documents (most of them tarsilas ), and they formed the basis of his 1905 Studies in Moro History, Law, and Religion .

24. Thomas remarks that "Saleeby's approach to governing and integrating the Muslims was referred to as the 'ideal' by others involved in Muslim policy-making, but few put it into practice" (1971, 15).

25. Carpenter to Secretary of the Interior Rafael Palma, January 27, 1919, Bureau of Insular Affairs Personal File—Tarhata Kiram, National Archives, Washington, D.C.

26. Memorandum submitted to Mr. Jorge Bocobo, August 19, 1935, P. 7, Joseph Ralston Hayden Papers, Box 29, Folder 24, Michigan Historical Collections, Bentley Historical Library, University of Michigan, Ann Arbor.

27. Ibid., p. 9. It should be noted that in contrast to Saleeby, who made his proposals to American colonial authorities, Kuder is here making an argument to Christian Filipino officials that it is in their self-interest to maintain and expand education programs among Philippine Muslims.

26. Memorandum submitted to Mr. Jorge Bocobo, August 19, 1935, P. 7, Joseph Ralston Hayden Papers, Box 29, Folder 24, Michigan Historical Collections, Bentley Historical Library, University of Michigan, Ann Arbor.

27. Ibid., p. 9. It should be noted that in contrast to Saleeby, who made his proposals to American colonial authorities, Kuder is here making an argument to Christian Filipino officials that it is in their self-interest to maintain and expand education programs among Philippine Muslims.

28. Salipada K. Pendatun to Vice-Governor General Joseph Ralston Hayden, August 15, 1935. Joseph Ralston Hayden Papers, Box 27, Folder 32, Michigan Historical Collections, Bentley Historical Library, University of Michigan, Ann Arbor.

29. Datu Adil told me that the initial "K" in Pendatun's name stood for Kuder, a middle name Pendatun had given himself to honor his former teacher.

1. Wolters reports a near-doubling of registered voters in the first fifteen years of independence, from 4.3 million in 1947 to 8.4 million in 1961 (1984, 143). A graph presented in Carl Lande's classic study of the structure of Philippine politics indicates a threefold increase in votes actually received by congressional candidates (from roughly 2.2 million to 6.5 million) within the same period (1965, 29). This expansion of the electorate in the postwar period (from 22 percent of the population in 1947 to 29 percent in 1959) was not due to a broadening of the criteria for inclusion—the poll tax and wealth requirement for voting were eliminated prior to the war—but apparently to an increase in voter registration in rural areas (Lande 1965; Wolters 1984).

2. For a comprehensive survey of immigration to Mindanao from other parts of the Philippines from the Spanish period onward see Wernstedt and Simkins (1965).

3. The seven agricultural colonies established by the Americans in Mindanao (six of them in Cotabato) included Muslims as well as Christian settlers. It was thought, in accord with the colonial "Moro Policy" outlined in the previous chapter, that Muslim farmers would learn more advanced methods (of both farming and family life) through imitation of their Christian neighbors (Pelzer 1945).

The agricultural colonies were relatively expensive to administer and had limited success. In 1918, colonial authorities instituted a new program that provided free transportation to selected prospective immigrants. Despite extensive advertising, this program also had disappointing results, primarily due to a shortage of immigrants willing to take up the offer of free land and free transportation (Wernstedt and Simkins 1965; Gowing 1979). One reason for their reluctance was undoubtedly the popular Filipino image of the untamed, bloodthirsty Moro. Advertising directed at prospective Christian homesteaders anticipated this problem by painting a very different picture of the indigenous inhabitants of Mindanao. The following passage (entitled "The Moro") from a brochure circa 1920 advertising immigration to Cotabato illustrates the attempt:

The Moro is first of all a farmer . . . Although entirely ignorant of the great world outside his rancheria he is a reasonable human being . . . Sometimes he evades the orders of the Government but when caught he meekly submits and considers it rather a joke on himself when punished . . . He welcomes the Christian Filipino colonist to his country, extends the hospitality of his home and table, and asks nothing but that his religion and tribal customs be not interfered with. ( Cotabato, Largest and Most Fertile Province in the Philippine Islands: Paradise of the Homeseeker from Over-crowded Luzon and Visayas , Bureau of Insular Affairs Records, file no. 26741 (Post-1914), National Archives)

4. Hukbalahap is an acronym for the Tagalog phrase Hukbo ng Bayan laban sa Hapon, meaning the People's Army to Fight the Japanese.

5. The problem of titling already occupied land was aggravated by the acceleration of timber and pasture concessions granted by the Philippine government to corporations and individuals in the late 1960s to increase government revenues. Such leases were granted without apparent regard for the rights of prior inhabitants—usually Muslims or other ethnic minorities (George 1980).

6. Both Hunt (1957) and Mastura (1979) note that the descendants of the earliest Christian residents of the city, the families of soldiers and "presiderios" from Zamboanga, still live in Manday, the majority Christian neighborhood that borders Campo Muslim.

7. This Muslim-Christian geographic alignment is another instance of the river-versus-road residency pattern found throughout Cotabato. It exists, I imagine, for the same reason found elsewhere: because the indigenous Muslims were the first to occupy the riversides and the immigrant Christians were the first to occupy (or file for legal ownership of) the roadsides. This seems a more plausible explanation than the one often repeated by Christians in the city: that Muslims simply "like to live by water."

8. It was reported in 1957 that a large part of the copra production of southern Mindanao and Sulu was being shipped to Borneo and exchanged for contraband goods, with American cigarettes the prime item of exchange (Hartendorp 1961). Trade figures from 1958 released by the Borneo government (Noble 1977, 67) show the Philippines ranking second only to Japan in its absorption of Bornean imports. According to Noble, that percentage was due primarily to the cigarettes and other items that left Borneo legally but entered the Philippines as contraband.

9. For an account of the establishment of quality cigarette production in the Philippines and its effect on the smuggling trade see Lewis Gleeck's (1989) The Rise and Fall of Harry Stonehill in the Philippines .

10. It was suggested by some that the mayor, who owned a fleet of cargo ships, may also have had a personal interest in the removal of Bird Island.

11. In 1967, Mando Sinsuat lost the mayorship to a Christian candidate in a race where he was also opposed by his half-brother, Datu Mama Sinsuat. The combined number of votes of the two Sinsuat brothers was greater than that of the Christian winner of the election.

12. The 1952 Cotabato Guidebook remarks of Datu Udtug's early career in his home municipality of Pagalungan that he "was very instrumental in helping the Christian settlers in getting homestead lots even if it was sometimes inimical to the interests of his brother Islams [ sic ]" (Millan 1952, 257).

13. Until the mid-1950s, Cotabato and the other majority Muslim provinces of the South were governed as "special provinces" and had their highest officials appointed by the central government in Manila (Gowing 1979, 186).

14. A front-page obituary for Datu Udtug in the Mindanao Cross (January 1, 1983) observes that under his governance "law and order was at its best in the province accented by close Christian-Muslim relation [ sic ]. Even in his official set up, his formula was: Muslim governor—Christian vice-governor—2 Christians to I Muslim in the 3-man provincial board."

15. The Mindanao Cross obituary of Datu Udtug just cited (January 1, 1983) continues by noting that "Kudin Dataya, his longtime private secretary and later executive secretary, remembers him as top in man-to-man diplomacy. He seldom delivered speeches. He approached people personally."

16. Numerous reports of the killing or wounding of members of prominent Muslim families in Cotabato may be found in issues of the Mindanao Cross from the 1950s and 1960s. Datu Adil related his personal involvement in a long, violent feud between his family and the Sinsuats beginning in 1949, in which his father was killed. For a vivid account of a similarly structured armed feud between the Masturas and Sinsuats in 1940, see Horn (1941, 166-68).

17. In his analysis of political relations in Central Luzon, Wolters notes that the central state resources flowing to (or through) local officeholders in exchange for their delivery of votes were actually of two types: "The transactions that occurred under these vertical alliances showed a mix of personal and more public aspects: personal in the exchange of votes for private gain such as credit, protection, prevention of audits, renewal of licenses, etc.; more public in the channeling of credit to the network of party followers; and generally public in the bestowal of pork-barrel funds for the whole community, e.g., the building of roads and bridges, the delivery of concrete, the erection of schools."

18. Autocratic rule at the local and provincial levels also removed much of the incentive to spend any significant proportion of the pork-barrel funds received for the benefit of the electorate, whose votes were, in most cases, simply appropriated from them. A story often told by ordinary Muslims in 1985 concerned a highway between Cotabato City and Marawi City that, according to official records, had been built three times but still did not exist in fact.

19. The political positions of Pendatun and Datu Udtug did occasionally diverge. In the 1963 Cotabato mayor's race, Pendatun supported Datu Mando Sinsuat, the son of Datu Sinsuat, who was the official candidate of Pendatun's political party. Although nominally a member of the same party, Datu Udtug refused to back a Sinsuat—the Sinsuats having always been his political foes—and endorsed Datu Mando's Christian opponent instead.

20. Philippine political parties date from the early American period. Only one of those early parties—the Nacionalistas—became a genuinely nationwide

entity. During the Commonwealth period it dominated electoral politics, but after independence in 1946 it was challenged by the Liberalistas, and a two-party political system developed, remaining in place until the declaration of martial law in 1972. In his well-known work on the Philippine two-party system, Lande (1965) finds it to be characterized by the extreme fluidity of party membership and weakness of intraparty solidarity; pronounced ideological similarities (including virtually identical official policies and internal structures) between the two major parties; and the virtual absence of any permanent rank-and-file party membership among the electorate.

21. Although Datu Pendatun was also the son of a sultan, he was not identified nearly as strongly with that cultural legacy as was Datu Udtug. One ready example is the very different terms of reference used for the two men. Salipada Pendatun is virtually always referred to as General or Congressman Pendatun, while the term of reference for Udtug Matalam is invariably the traditional one—Datu Udtug. Another indicator is that while I was told a number of stories about the magical powers possessed by Datu Udtug—including the report that he had a sixth finger on the palm of his right hand, which he used to perform marvelous feats—virtually no specific accounts were given of the supernatural abilities of Congressman Pendatun.

22. Similar combinations of relatively assimilated and relatively tradition-oriented datus may be found among all of the prominent Muslim political families of the province, particularly among the Sinsuats, Ampatuans, and Masturas.

23. The use of the term "Muslim Filipinos" to denote Philippine Muslim populations was also adopted by a number of scholars of the Muslim Philippines—see, e.g., the titles of the works by Gowing (1979), Gowing and McAmis (1974), and Mastura (1984). By contrast, Cesar Majul's 1973 Muslim nationalist history is conspicuously entitled Muslims in the Philippines .

24. Hadji (female: hadja ) is the honorific title given to a Muslim who has made the pilgrimage to Mecca ("Hajj"). I use the English spelling usually seen in contemporary Cotabato. As the Magindanaon language lacks either an initial "h" or a "j" sound, the title was formerly rendered as kagi and that version is still often heard today.

25. The observations of Hunt about mosque attendance in Cotabato City in 1953 are relevant: "[O]nly a small percentage of the faithful attend group services with any degree of regularity . . . The worshippers give every evidence of piety, but they cannot comprise more than fifteen percent of the Moros in town at the time. Leading datus do not seem to feel that their position demands regular attendance at public services" ([1958] 1974, 202-3).

26. Some of these, such as the "Knights of Mohammad," were clearly self-consciously Muslim versions of Philippine Christian voluntary organizations.

27. Datu Adil, who was present at the 1955 MAP conference, recalls that Edward Kuder was one of its principal organizers.

28. Datu Blah Sinsuat's introduction to the 1952 Cotabato Guidebook offers a narrative that illustrates some of the points discussed above. In that year, Datu Blah was serving as Cotabato's representative in the lower house of the Philippine Congress. A passage from his introduction demonstrates the manner

in which datu politicians presented themselves to Christians as both advocates for and supervisors of Cotabato Muslims.

I consider now the most opportune time to state that the unselfish Native who welcomed his Christian brother in years past should also receive the gratitude of the people of this province, all of whom are immensely enjoying the great opportunities here, the peace and quite [ sic ] that have been theirs since their coming. The harmonious relationship pervading among the populace of Cotabato is attributable only to the willingness of the Native to offer a little of his share of the natural wealth of his land of birth to his Christian brother so that both may not live in want.

In this single rather extraordinary passage, Congressman Sinsuat manages to reinforce Christian perceptions of Muslim cultural backwardness, to solicit resources to "improve" Muslim communities, and to insinuate that peace may not continue should his suggestions go unheeded.

1. Outlaw activity commonly referred to as "banditry" was widespread in Cotabato and throughout the Muslim South in the 1950s. Much of it involved rather straightforward criminal activity, especially highway robbery and cattle rustling. Other incidents, however, have the appearance of social banditry and a few, including the famous "Kamlun Uprising" (Tan 1982, 68) of 1952 in Sulu, approached the level of genuine armed insurgency against the state. It was this latter episode of "banditry," along with a similar armed uprising in Kapatagan in Lanao, that prompted the formation of the Congressional Committee in 1954 (Tamano 1974).

2. Congressman Alonto was the keynote speaker at the First National Muslim Convention sponsored by the Muslim Association of the Philippines in 1955. In that speech he echoed the language of his committee's report, declaring to the delegates: "Let us purge ourselves of our defects," and proclaiming: "We need a thorough spiritual rejuvenation . . . If we are good Muslims we are automatically good citizens . . ." (Muslim Association of the Philippines (1956, 31).

3. The invasion, or infiltration, of Sabah apparently was to be made as part of the Philippine government's prosecution of its claim to Sabah. That claim was first announced by President Diosdado Macapagal in 1962 and was based on an 1878 transaction between the Sultan of Sulu and an Austrian businessman (Noble 1977).

4. George (1980, 197) reports that Misuari was "one of the founding fathers" of the Kabataang Makabayan (Nationalist Youth), the largest and most active leftist student organization of the pre-martial-law period. The Kabataang Makabayan was organized in 1964 by Jose Maria Sison, who in 1968 founded the Maoist Communist Party of the Philippines (CCP). Shortly thereafter, Sison organized the New People's Army and initiated a new armed communist insurgency against the Philippine state, one that eventually grew to be far more extensive and successful than the Huk Rebellion of the 1950s. A detailed account of the beginnings of this insurgency, including a description of radical student politics at the University of the Philippines in the 1960s, may be found in Chapman (1987).

5. Another featured speaker at the 1955 Muslim Filipino Conference was an Egyptian emissary, Sheikh Hassanal Baguri. In his speech, Sheikh Baguri announced the commitment of the Egyptian government to underwrite the advanced Islamic training of numbers of Philippine Muslims: "Our dear Muslim Filipino brothers, we are here declaring in the name of our government that we are ready to send you teachers for your schools and that we are even ready to establish an Islamic institution in the City of Manila for Islamic studies. We are ready to accept your sons and daughters to study in our universities and to give them all the facilities in our hands" (Muslim Association of the Philippines 1956, 40).

6. For an absorbing account of Islamic education and politics in Cairo in the mid-1960s see Gilsenan (1982).

7. The Muslim separatist movement was, to a significant degree, ignited by the aspirations engendered by both secular and Islamic higher education. The coalescence (at least for a time—see below) of Middle East- and Manilaeducated activists in the MNLF leadership represents a distinctive variant of a pan-Islamic development recently analyzed by Dale Eickelman (1992): the relationship between mass higher education, political and religious activism, and transformations of religious authority.

8. The Ampatuans were descendants of Datu Ampatuan, a datu of the early colonial era who, like datus Ayunan and Balabaran had few blood ties to the precolonial high nobility of Cotabato. Datu Ampatuan, however, claimed Arab descent through his great-grandfather. He was a lieutenant of the anti-American Datu Ali but later allied with the pro-American Datu Piang. He succeeded to Datu Piang's seat on the Cotabato Provincial Board in 1917 (Beckett 1977, 1982).

The English term "clan" is widely used in the Philippines to refer to kinbased political factions, especially those with pronounced dynastic tendencies (see, e.g., Francia 1988). Its use in Cotabato has exactly the same meaning and is not intended to describe any actually existing corporate descent groups.

9. The Mindanao Cross reported on April 29, 1967, an announcement by Congressman Pendatun that he was running as the Liberalista candidate for governor because Datu Udtug was in "failing health."

10. By mid-1970, Datu Udtug had changed the name of the Muslim Independence Movement to the "Mindanao Independence Movement" in order to include Christians. Prior to this he had already modified his stand to one in favor of regional autonomy in a federal-state framework (see George 1980).

11. In an August 17, 1968, interview with the Mindanao Cross , Datu Udtug expressed puzzlement at the "war talk" creating anxiety in Christian settlers in North Cotabato and "causing many families to sell their property."

12. In 1966, after years of maneuvering, Datu Udtug succeeded in having the capital of the province moved to Pagalungan, his home territory, only to have it returned again to Cotabato City by Governor Datumanong immediately after his assumption of office.

13. Some writers (see, e.g., Gowing 1979; Majul 1985; Mercado 1984), in seeking a direct causal connection in the flow of events between 1968 and 1970, have suggested that the manifestos and activities of the Mindanao Independence Movement were principal precipitating factors for the wave of intense communal violence in Cotabato that began in 1970. There is little evidence for that proposal, however, and although the formation of the MIM may have marginally intensified Muslim-Christian tensions in the province, it was probably only a minor contributing factor. It is more productive, I believe, to view both the sectarian conflict and the formation of the MIM as effects of more fundamental political and economic pressures in the province.

14. "Ilonggo" is a term commonly used to refer to speakers of Hiligaynon (also called Ilonggo). Hiligaynon speakers originate from the provinces of Iloilo and Negros Occidental (on the islands of Panay and Negros) in the central Visaya region of the Philippines. In 1970, Ilonggos made up about 10 percent of the Philippine population. Ilonggos also composed a significant percentage of postwar migrants to Cotabato.

15. For a rare firsthand account of the sectarian conflict in the Cotabato Valley, especially as it occurred in and around Midsayap, see Stewart (1977, 254-61).

16. The Ilonggo politicians who, in these speculations, were supposed to have invented and supported the Ilaga were, in fact, themselves divided by separate political parties, aspirations, and interests. Not all were Nacionalistas, and those who were supported two different Ilonggo Nacionalista candidates for governor (see below). Also difficult to explain when considering these hypotheses is what sitting Ilonggo mayors, most of them ruling municipalities with large Christian majorities, would have to gain by initiating Ilaga terror within their areas of influence.

17. The experience of Doroy Palencia, a Christian Liberalista politician and longtime confederate of Datu Udtug, in the 1971 election is instructive. Palencia had been elected to serve on the three-person provincial board for every one of Datu Udtug's terms as governor and also under Simeon Datumanong. He stated in 1986 that he had been "strong" among Muslims during elections and that Christians also supported him. He failed in his bid for the first time in 1971, because, as he put it, the sectarian violence had caused both Muslims and Christians to mistrust him.

18. Although sectarian conflict did occur during this period in neighboring Lanao, and to the west across Ilana Bay in Zamboanga del Sur, by far the greatest number of violent incidents occurred in Cotabato.

19. Premier Kadaffi announced in 1972 that he would send "money, arms, and volunteers" to aid Muslims in the Philippines (quoted in Schlegel 1978, 48). There is also evidence from informants and elsewhere to suggest that

Kadaffi was already providing arms to Lucman, and through him to MIM, well before this announcement (Noble 1976).

20. Datu Udtug had, in fact, run for governor in the 1971 race as an independent Nacionalista candidate.

21. The Marawi uprising was, by most accounts, spontaneous and idiosyncratic. It reportedly took by surprise the MNLF leadership in Lanao Province, some of whom held high positions in the Marawi City government (see Mercado 1984; George 1980).

22. The Bangsa Moro Army has probably never had as many arms as men. A rebel commander informant noted in an interview that in his first armed encounter with the military, one hundred rebels shared seventy guns, some of them homemade.

23. An August 11, 1973, story in the Mindanao Cross reported 924 Muslim "surrenderees" from Tran, noting in passing that all of these were women, children, and elderly men.

24. In late 1973, what remained of the original province of Cotabato after the splitting off of South Cotabato in 1966 was subdivided by presidential decree into the three provinces of North Cotabato, Maguindanao, and Sultan Kudarat. The division coincided with the division of votes in the 1971 local elections. North Cotabato and Sultan Kudarat Provinces were 80 to 90 percent Christian and Maguindanao Province was 85 percent Muslim. All of the mayors in Maguindanao were Muslim, and virtually all in Sultan Kudarat and North Cotabato were Christian. The unusual shape of Maguindanao Province was the result of including pockets of Muslim population in areas such as Buldun, Buluan, and Pagalungan; and excluding the Christian communities of Pigcawayan, Pikit, and Esperanza (see map 3).

25. Iranun narrators are fond of pointing out that Iranun fighters are the only Muslim rebels who never surrendered to the Philippine government. While this exaggerates the facts, it is the case that Iranun commanders and their men were very disproportionately represented among those rebels who remained under arms as late as 1985. It may also be noted, in regard to Iranun armed involvement in the insurgency, that the mostly Iranun municipality of Subpangan on the coast north of Cotabato City was the site of some of the very first as well as the last armed engagements of the rebellion. Iranun commitment to the rebellion is, I believe, related both to their recent past as cigarette smugglers and to their long history of resistance to external domination (McKenna 1994).

26. The composite term "Bangsa Moro" has sometimes appeared in MNLF literature as one word and at other times as two. Current spokespersons for the Moro Liberation Front in Cotabato have stated a preference for "Bangsamoro" because of its emphasis on "bangsa," which they translate as ''nation."

27. A policy statement from the first issue of Maharlika , an MNLF newsletter, clearly illustrates the national, rather than specifically ethnic or religious, character of the MNLF appeal:

From this very moment there shall be no stressing the fact that one is a Tausug, a Samal, a Yakan, a Subanon, a Kalagan, a Maguindanao, a

Maranao, or a Badjao. He is only a Moro. Indeed, even those of other faith [ sic ] who have long established residence in the Bangsa Moro homeland and whose good-will and sympathy are with the Bangsa Moro Revolution shall, for purposes of national identification, be considered Moros. In other words, the term Moro is a national concept that must be understood as all embracing for all Bangsa Moro people within the length and breadth of our national boundaries. (Quoted in Noble 1976, 418)

28. In a 1975 lecture delivered at the National Defense College, Brigadier General Fortunato U. Abat, then commander of the Central Mindanao Command of the Philippine Armed Forces, expressed the widely held Christian viewpoint on the "datu system":

[T]here are irreconcilable features in the cultures of the Muslims and Christians. On the one hand, we have the Muslims and the Islamic religion, the datu system serving as their government, the lack of education and different customs and practices. While they arouse pity, they usually do not command the respect of the socially superior Christians . . .

1. The barangay system was created by presidential decree on December 31, 1972, shortly after the imposition of martial law. Its purpose was to create basic political units—barangays—which could be used as instruments to further the objectives of martial law and at the same time provide controlled outlets for political participation.

The barangay is actually a very old precolonial political institution in the Philippines. It was a unit of thirty to one hundred houses under the authority of an autocratic headman (datu). The institution was adapted by Spanish colonial administrators to suit their needs (with the name "barangay" eventually changed to barrio ). The term and institution were resurrected by the martial law regime as part of its attempt to equate nationalism with autocratic leadership. All barangay "captains" were, initially at least, appointed.

Modern barangays are intended as local-level political units and meant to be composed of a maximum of five hundred families. Cotabato City in 1985, however, contained only five barangays for its eighty thousand or so inhabitants. All were much larger than the ideal. Barangay Bagua, for example (the barangay in which Campo Muslim is located), contained four communities—Manday, Campo Muslim, Lugay-lugay, and Bagua—each with more than five hundred families. The reason for so few original barangays in the city is not clear. That more had not been created, despite widespread recognition of their

need, was largely due to the strong opposition of the five current barangay captains. In around 1990, Cotabato City added a large number of new barangays but retained part of the former structure, designating the original five barangays as "Mother Barangays." Thus Campo Muslim is now its own barangay though still considered for some purposes a constituent unit of "Mother Barangay Bagua." For more information of the formation and functioning of the barangay system under the martial law regime, see Lapitan (1978).

2. The Civilian Home Defense Force was also created by presidential decree shortly after the declaration of martial law. CHDF units are paramilitary entities associated with particular barangays and under the formal direction of the barangay captain. CHDF members receive arms and a minimal stipend. During the rebellion, CHDF units were effectively controlled by the Philippine Army.

3. Kalanganan is the name for all of the area that lies between the Pulangi and Tamontaka Rivers and west of the city proper. Today it is included in the Cotabato City limits as the only entirely rural barangay.

4. Akmad's father, as a man of the coast, was almost certainly an Iranun speaker. That he was also a member of the Magindanaon aristocracy may be due partly to intermarriage. However, it also may be due to the fact, related by both Mastura (1984) and Ileto (1971), that in 1879 the Magindanaon title of Amirul Umra was conferred by the Sultan of Magindanao upon an Iranun datu from Malabang in an apparent effort to shore up the rapidly fading power and prestige of the downriver sultanate (see Ileto 1971, 42).

5. Questions on the interview schedule did not attempt to ascertain whether respondents held legal title to land. My principal concern was to find what percentage of residents held rights in agricultural land, regardless of whether those rights were legally recognized by the state. Qualitative research data suggest, however, that most respondents do hold legal title to the land in which they have indigenously recognized rights.

6. The members of Candao's CHDF unit apparently also saw their own role, to some degree, as protecting the community from army intrusions. Residents tell of one occasion in which the army brought an armored personnel carrier to the Manday bridge, intending to enter the community with it. Candao's men refused to let it pass, and when challenged entered the street with guns drawn to stop it. The standoff ended only when the military police arrived and negotiated a settlement.

7. As Commander Jack explains it today, "I was captured because the Tripoli Agreement had been signed, the cease-fire had started, and Hadji Murad [the Cotabato rebel commander] and his men had come down from the hills for peace talks. The military were afraid that we would join forces, so they detained me. Their official reason was that I was being detained for safekeeping."

8. The MNLF has never claimed responsibility for any of these terror bombings and none has ever been solved. There is a strongly held belief among some Christian as well as Muslim city residents that at least a portion of the bombings were the work of government agents attempting to discredit the MNLF.

9. Jeffrey Sluka reports responses from rank-and-file IRA members in Northern Ireland quite similar to those I found in Cotabato: "[T]he major reason [Republican guerrillas] give for why they turned to armed struggle is because they say that repression and state terror drove them to it. That is, when asked how they came to join the IRA, they do not usually refer to Republican ideology and goals, but rather they tell personal histories of their experience with repression and state terror" (Sluka 1995, 85).

10. In a recent illuminating discussion of musical code-switching, Mark Slobin employs the term "domestication" to refer to the sort of borrowing I have described—a process whereby "music is brought into the subculture from the superculture" (1993, 90). Slobin's musical ''superculture" is conceived as a hegemonic system encompassing a music industry, governmental regulation, and "a set of standardized styles, repertoires, and performance practices" (1993, 33). Musical subcultures, or "micromusics," are "small musics in big systems" (1993, 11). The mass-marketed music of the Philippines, today overwhelmingly sung in English and largely produced in the United States and England, corresponds to Slobin's musical superculture, while Magindanaon music fits his definition of a "micromusic."

11. The English term "cowboy" is used by young men in Campo Muslim and throughout the Philippines to describe an individual (and occasionally an action) thought to be unusually rugged or reckless. That usage is derived from the Philippine-made Westerns that were extremely popular movie fare in the 1960s and 1970s.

12. This song, "Mana Silan Cowboy," is one of the few I recorded that dates itself fairly precisely. The time range indicated in the opening phrase(1971-79) indicates that this song was written after the 1977 cease-fire and intended primarily for public performance outside the rebel camps.

13. The early rhetoric of Muslim separatism did not emphasize jihad as a component of Bangsamoro ideology. Stress was placed instead on national identity. Only years later, after the cease-fire, did separatist leaders appeal to the concept of jihad, usually in the context of a broader Islamic renewal, as evidenced in the following passage from a 1985 declaration: "All Mujahideen . . . adopt Islam as their way of life. Their ultimate objective in their Jihad is to make supreme the WORD of ALLAH and establish Islam in the Bangsamoro homeland [emphases in the original]" (Salamat quoted in Mastura 1985, 17).

14. The "wide green land" of the first stanza of song 2 is a description of the Cotabato River Basin and a metaphorical reference to Cotabato as a whole. In his collection of Magindanaon folk songs, Clement Wein (1985) includes a song said to have been composed circa 1950 and sung to the melody of "Green Valley" which begins with the same couplet found in song 2 (1985, 117).

15. Compare Jeffrey Sluka's observation for Northern Ireland that "the British government and their Security Forces have applied military and judicial repression against the Catholic communities they believe support the Republican insurgency and . . . this has served to alienate the population and created and continuously reinforced popular support for the Republican movement" (1995, 76).

16. As late as the last month of 1977, one year after the cease-fire, a military air strafing killed fifteen Muslim civilians (six of them young children) and wounded many more in Kalanganan, within the Cotabato City limits ( Mindanao Cross , December 17, 1977). As had happened on previous occasions, the military apparently mistook a wedding party for a gathering of rebels and opened fire on it.

17. Amulets (agimat) and especially muntia —rare stones with magical protective powers—were particularly popular with rebel fighters. Commander Jack possessed a muntia consisting of a petrified egg sac from a spider wrapped in a cloth and hung on a thong around his neck. Among its other powers it could detect and neutralize poisons placed in liquids.

18. Commander Jack recounts a very similar story from personal experience concerning assistance received from a tunggu a inged in another manifestation: "Once during the siege of Tran, I was eating ripe mango with my companion. I heard a bird call 'Awa, Awa' ['Awa' means 'leave' or 'get away' in Magindanaon]. I told my companion, 'Quick, we have to move.' He did not believe me. I jumped into our foxhole and just then a jet appeared overhead and dropped a bomb right where we were. My friend was blown to pieces."

19. Although these and other independent representations by Muslim followers accord with Islamic doctrine in that they are accounts of divine compassion shown for those fighting for Islam, their strongly folk-Islamic elements were disapproved of by Islamic clerics and discounted by some rebel leaders.

20. Although not given precedence in the authorized lexicon of the Muslim separatist movement, "inged" and "jihad" do have certain hegemonic connotations. As a term that describes a community but also a traditional political entity, "inged'' suggests hierarchy and domination: a community ruled by an autocratic chieftain. Similarly, while "jihad" refers to the defense of the community against alien invaders, it also connotes armed mobilization at the behest of a local ruling elite. These terms, then, are part a "common meaningful framework . . . for talking about . . . domination" (Roseberry 1994, 361), yet may be given quite different emphases by subordinates and superordinates. More significantly for this case, ordinary adherents of the separatist movement have used these terms (or their particular colorings of them) to understand the rebellion, even though movement leaders have not employed them in official separatist rhetoric.

1. Madrasahs in the Muslim Philippines typically offer programs at the "elementary, preparatory and secondary levels of Arabic education." A madrasah that includes all three levels (a twelve-year program), or only the secondary program, is termed a mahad (in Arabic, ma'ahad ) (Hassoubah 1983).

2. Alim is the Arabic term for a scholar who has mastered a specific branch of knowledge—a learned person, or savant. In the southern Philippines the term is used to refer to someone who has educational qualifications for, and is knowledgeable in, the teaching of Islam. The plural of alim is ulama . Ustadz

(from the Arabic ustadh , meaning "teacher") is the most usual term of address in the Muslim Philippines for an alim. (Glassé 1989; Hassoubah 1983).

3. This statement seems to suggest that in 1950 there was no Magindanaon speaker available in Cotabato who spoke Arabic well enough to translate directly for the Maulana.

4. Ustadz Abdul Gani Sindang arrived in the Philippines in 1950, along with another missionary from al-Azhar. Their first school in Malabang closed within a year and he proceeded to Cotabato City (Muslim Association of the Philippines 1956, 106). He is almost certainly the same missionary referred to by Hunt in his brief account of Muslim religious education in Cotabato in the early 1950s: "In 1950, a formal Islamic school was set up in Cotabato, housed in a residence donated by a local datu and headed by a Muslim missionary sent by the Egyptian government. In addition to learning to read the Qur'an in Arabic, the students are taught to understand the language" (1974, 205).

5. This traditional maktab system of Islamic education in Cotabato was quite similar to the pondok schools of Malaysia (see Nagata 1984) and the pesantren schools of Java (see Geertz 1968).

6. For an account of the politics of Islamic preaching elsewhere in the contemporary Islamic world, see Patrick Gaffney's (1994) ethnographically rich investigation of the complex connections between religious rhetoric (as expressed in Friday sermons) and political dissent in Upper Egypt.

7. Partial funding for the establishment of the English program was obtained from the Philippine government's Ministry of Muslim Affairs.

8. Hassoubah notes elsewhere that, despite the intensification of Islamic instruction in the Muslim Philippines, "the quality of education in the madaris , with very few exceptions, leaves much to be desired by way of being at par with standard schools in the Middle East or even compared to the quality of the Philippine public school system"(1983, 74).

9. In the context in which it was usually used, the term "Shia" was used pejoratively to mean a heretical Muslim radical influenced by the Shia Islamic government of Iran.

10. Tantawan is the Magindanaon name for the main hill of Cotabato City, commonly known as P.C. Hill because it had served as local headquarters of the Philippine Constabulary during the American colonial period.

11. For a detailed account of the operation of the Christian Children's Fund in Campo Muslim, including an analysis of the injurious effects of child sponsorship on the community, see McKenna (1988).

12. The mean estimated income of community households in 1985 was 1,333 pesos ($71.82) per month. That of CCF recipient families was about 900 pesos (less than $50.00) per month. Household income was calculated as the sum of the monthly earnings of the household head at his or her primary occupation, additional earnings of the household head (reported by just under 20 percent of respondents) from productive land, "sideline" jobs, or other sources, and monthly contributions from other household members (reported by slightly more than 60 percent of respondents). It is plainly a very inexact figure. Respondents often had no precise knowledge of contributions of other household members and, because of the high incidence of uncertainty in their

own economic endeavors, were sometimes able only very roughly to estimate their average daily income as well.

13. As this passage and the previous one suggest, it was most often women—the mothers of CCF recipients—who were forced to confront the pressures and contradictions involved in attempting to obtain CCF resources for their children. Their predicament illustrates what Sherry Ortner has recently referred to as the "multiplex identit[ies]" and "compounded powerlessness" of subaltern women—in this case as women, as poor, and as Philippine Muslims (1995, 184). By attempting to assist their children they placed themselves in an emotionally wrenching double bind, anguishing on the one hand about the imagined loss of their children to unseen and unreachable "sponsors," and on the other about the public disapproval of the ulama and MILF. At the same time, the responses they have made to their dilemma—seeking to contact sponsors directly to tell them the ''true" stories of their families, insisting that those seeking to remove the CCF program provide another in its place (see below)—represent (as Ortner also notes) creative efforts on their part to "formulate projects and . . . enact them"( 1995, 185).

14. As reported in the Mindanao Cross , Sandiale Sambolawan, the Muslim governor of the province (and a member of the Ampatuan clan), was disturbed by the student demonstration and enraged by the banner portraying the "Ministry of Munafiq Affairs" ( Mindanao Cross , February 28, 1985).

15. Included among these six were the son of Datu Udtug Matalam, as well as two younger members of the Sinsuat and Ampatuan families.

16. A number of Magindanaon datus attended the ceremony, some of them traveling in the traditional manner on decorated boats to Lanao.

17. Similar ritual feasts ( kanduli, kenduri, kenduren, slametan ) held for various occasions by other Southeast Asian societies are described by Geertz (1960), Reid (1984), and Bowen (1992).

18. Datu Adil, fostered as a child by Edward Kuder, always preferred to speak to me in English. While strongly opposed to many of the efforts of the independent ulama, Datu Adil is also a vigorous proponent of Muslim autonomy and a very vocal critic of the martial law regime and its local supporters.

19. Wahhabism, an eighteenth-century Islamic reform movement in the Arabian peninsula begun by Muhammad ibn "Abd al-Wahhab, strenuously opposed Sufism and advocated puritanism in religious practice. Wahhabism played an important role in the creation of an Arabian state and remains the dominant variant of Islam practiced in the Kingdom of Saudi Arabia (Cole 1975; Eickelman and Piscatori 1996).

20. The lyrical form of the dayunday resembles that of the Magindanaon bayuk, a style of romantic poetry that almost certainly predates the arrival of Islam in Cotabato. The practice of romantic song duels between men and women was apparently found throughout the Philippines at one time. Anthony Reid, citing a seventeenth-century Spanish account, notes that spontaneous contests of romantic poetry and music (called balak ) were "enormously popular in the central Philippines up to early Spanish times" (1988, 148).

21. Dayunday lyrics are customarily sung in an archaic form of Tau sa Laya (upriver) Magindanaon, the same form used for traditional ballads and

poetry. Downriver audiences typically find dayunday lyrics difficult to understand. At the performances I attended I found that I was able to make out only occasional words or phrases. My downriver companions reported that they understood, on average, about 25 percent of the lyrics. Difficulties with aural comprehension did not in the least hinder the enjoyment of downriver audiences. Campo Muslim residents advised me that the dayunday must be appreciated in its totality: the music, the showmanship of the performers, the costumes worn by the female performers, the nonverbal interactions among the singers, the repetitions of standard phrases—all combine to provide the entertainment experience so appreciated by ordinary Muslims.

22. Another of the Friday sermons of Ustadz Ali had as its theme the statement that "Allah will help the weak person to claim justice."

23. Clearly, a number of the behavioral reforms suggested by the independent ulama—their efforts to prohibit "emergency" marriages, the dayunday, and other entertainments—primarily reflected their concern with regulating the behavior of women. Those efforts did not, in 1986, extend to any specific attempts to prescribe either the proper dress or work activities for women.

24. Imam Akmad was by no means the only community elder to question the reform efforts of the independent ulama. A passage from my field notes relates another specific instance: "Bapa Hasan is a traditional healer. His cousin, Ustadz Murid (an important ustadz inside [in the MILF]), criticized him for these traditional beliefs, but Bapa Hasan showed the ustadz the passages in the Qur'an where it referred to such healing practices."

25. As suggested by the passage quoted above, the MILF acquiesced to community members on this issue, at least temporarily, for "practical" reasons.

26. Eickelman and Piscatori observe about varying Islamic interpretations of zakat:

There is little agreement [about zakat] other than using it for humanitarian or charitable purposes: The Qur'an encourages Muslims to spend their mal (wealth) "out of love for Him, for your kin, for the needy, for the wayfarer, for those who ask, and for the ransom of slaves" (2: 177). Some Muslims argue that it is a voluntary act of faith; others argue that it is obligatory . . . An indication of the degree to which doctrine is malleable is the specific political use of zakat in a resolution of the Organization of the Islamic Conference (1981, 699). It endorses collection of zakat to support the work of the Palestine Liberation Organization, rather than, for example, to support Palestinian widows, orphans, and refugees as might be expected. (1996, 16-17)

27. It may be noted in this context that the common term used to refer to the rebels—both current ones and those who fought during the active rebellion—was not "mujahideen" but the Spanish-derived "rebelde."

28. One respondent, in fact, cited the new president of the Philippines, Corazon Aquino, as the most powerful datu he knew.

29. It should be noted that the public position of the MILF in regard to voting in the 1986 presidential election was primarily one of indifference

rather than pointed opposition. As a (nominally) separatist organization, the MILF simply commented that it had no interest in elections of officials to a government that it considered illegal, with the implication that its supporters need have no interest as well. The following excerpt from an editorial carried in the January 1986 issue of Tantawan, the newsletter of the Kutawato Regional Committee of the MILF, illustrates official MILF attitudes toward electoral participation: "Many are asking this column whether the Moro Islamic Liberation Front (MILD takes sides in the forthcoming election [for president] or any election in the future. Our answer to this is solid NO! The MILF does not believe in elections to correct injustices, oppression, exploitation, persecution and aggression in society" [emphasis in original].

30. The ulama and professionals who formed the aboveground component of the counterelite were able to acknowledge publicly their support for the MILF without fear of sharp reprisal only because of the formal cease-fire that still obtained between the MNLF and the Marcos government and the expressed interest of various Arab oil-supplying states in a continued dialogue between the government and the rebels on the issue of the full implementation of the Tripoli Agreement.

31. More than one Campo Muslim resident remarked on the main irony surrounding the ustadzes' adamant disapproval of dayunday performances. The dayunday first gained popularity during the period of the armed rebellion when military repression was most severe. Popular legend states that it was invented by an upriver rebel commander and his sweetheart. Dayunday performances were among the very few popular diversions available during those dark years when virtually all ordinary Muslims suffered as a result of the Bangsamoro Rebellion. Yet when the ustadzes, who were so closely associated with the leadership of the rebellion (and in many ways personified its aims), were able to speak openly after the cease-fire, one of their very first pronouncements was to denounce the dayunday.

32. In a recent work, Michael Peletz (1997), provides a fascinating discussion of the ambivalence of ordinary Muslims toward Islamic resurgence in contemporary Malaysia. While Campo Muslim residents expressed ambivalence toward Moro nationalism (both desiring and distrusting it), I did not find the same sort of responses in respect to the Islamic renewal efforts of the independent ulama. Instead, as I have noted, some aspects of the renewal program were accepted and others resisted, in some cases openly. The primary reason why the Cotabato case differs from that described by Peletz for Kelantan is that the independent ulama were openly opposed in Cotabato by established Muslim politicians who advanced an alternative ideology of moral authority. The existence of this powerful opposition allowed ordinary Muslims the opportunity for "fence sitting" and permitted potential mediators such as the community imam to explore the middle ground between the polar positions.

1. The Communist Party of the Philippines (CPP) was established in 1968 by a group of young intellectuals as a breakaway party from the older,

Moscow-oriented Partido Komunista ng Pilipinas. In 1969, the CPP formed the New People's Army as its military wing. In 1973, a revolutionary front, the National Democratic Front (NDF), was established by the CPP and other underground organizations, including the Christians for National Liberation. Observers agree that the NDF was effectively controlled by the Communist Party of the Philippines. The armed insurgency conducted by the New People's Army did not become active in more than a very few parts of the country until the end of the 1970s and was not considered a major threat by the government until the early 1980s (Schirmer and Shalom 1987).

2. For a detailed account of the extraordinary staying power of certain datu politicians in Cotabato, see McKenna (1992).

3. A presidential election was held in June of 1981, not long after the formal lifting of martial law. President Marcos ran for reelection unopposed by any meaningful opposition candidates. He won the election with about 86 percent of the vote (Bonner 1987).

4. Despite the KBL money distributed at the barangay level during the 1986 presidential campaign, I heard of not a single case of vote-buying in Maguindanao Province and of very few attempts among Muslims in the city. This was due to the absence of a tradition of vote-buying in Muslim Cotabato prior to martial law but also because of the coercive apparatuses available under martial law in Cotabato for repressing the popular vote.

5. This use of Magindanaon in a public speech by the soon-to-be appointed governor of the province was in itself unprecedented. English is the language of public political discourse in the Philippines. Every one of the speeches made at the KBL rally just two weeks earlier had been delivered in English (occasionally interspersed with Filipino, the other official national language). Even Datu Udtug Matalam, who spoke almost no English, endeavored to deliver his short and infrequent political speeches primarily in that language. Zacaria Candao's use of Magindanaon in his first public speech of the post-Marcos era—one delivered to Christians as well as Muslims—sent the clear message that, for the very first time, a Muslim nationalist had attained an official position of power in postcolonial Cotabato.

6. In Filipino, "Kalimutan ang Nakaraang Politika at Harapin ang Krisis."

7. Two additional factors may explain the special attention given to Nur Misuari and the MNLF by the media and central government. Misuari hailed from Sulu and his power base was there, particularly in Jolo. Philippine Christians (as reflected in the mainstream national media) know very little about Philippine Muslims. They do tend, however, to associate Muslims with seagoing people, colorful boats, and faraway islands. Tiny Sulu thus received a degree of media attention far out of proportion to the percentage of Philippine Muslims living there (less than one-third of the total).

The national government was also particularly concerned with Sulu because of its proximity to the very permeable national border with Sabah, Malaysia. Insofar as the islands of Sulu have overwhelmingly Muslim populations, are fairly easily defensible, and form the border between the Philippines and a powerful neighboring Muslim state, Muslim armed separatism in Sulu always posed the greatest immediate security threat to the territorial integrity of the

Philippines. For a comprehensive discussion of the unique position of Sulu Muslims see Kiefer (1987).

8. I was never able to determine the origin of the curious, and untranslated, English term "prayer rally" for the MILF's mass demonstration. The rally was not called in order to pray for divine assistance in achieving the full implementation of the Tripoli Agreement, or for any similar divine favor. Insofar as prayer ( salat ) in orthodox Islam consists solely of devotional worship, not supplication, it would have been quite surprising had that in fact been the goal of the rally. The term was likely borrowed from Roman Catholic practice and used with a Roman Catholic audience—the mainstream media and national administration—in mind. The "prayer rally" label accentuated the formal leadership of the rally by Muslim clerics and emphasized the religious character of the Muslim Islamic Liberation Front.

9. The previous record had been set just a few months earlier when more than thirty thousand people attended the Cotabato City campaign rally for Corazon Aquino on January 26. No one I spoke with at that time was able to remember a crowd as large as that ever having gathered in the city plaza.

10. I do not mean to suggest that the provincial voting resembled the North American ideal for democratic elections. Presumably there was a fair amount of vote-buying and other formally illegal activities (see below). However, it was the first-ever opportunity for individuals in a number of municipalities to cast their own ballots for freely chosen candidates without impediment from local authorities.

11. Such an attitude on the part of established Muslim political figures is certainly not unique to Cotabato. Eickelman and Piscatori provide examples from throughout the Islamic world of political leaders using various arguments against the mixing of religion and politics, including the rather extraordinary 1993 statement of King Hassan II that Moroccans should "render unto God that which is God's and unto Caesar ( Hiraql ) that which is Caesar's"(2996, 52).

12. Despite the implication in Congressman Mastura's speech of widespread ulama participation in political office-seeking, only one Cairo-educated cleric ran as an IPP candidate in either the provincial or city races. He won his race as the highest vote-getter of all candidates running for the eight-member provincial board. While many ustadzes were active IPP officers or members, no other ustadz ran for political office in Cotabato in 1988. I am not certain why this was so. A number of possible reasons come to mind, including the intention of the IPP to avoid just the sorts of charges leveled by Congressman Mastura.

13. There were some fifty-six thousand registered voters in Cotabato City in 1988—approximately twenty-nine thousand of them Christians and twenty-seven thousand Muslims. By contrast, Muslims comprise roughly three-quarters of the population of Maguindanao Province. Because of confusion over new rules promulgated by the National Commission on Elections after 1986, it was unclear until less than two weeks before the gubernatorial election whether or not Cotabato City residents would be able to vote in the provincial elections. They had never been allowed to before, and many Christian city residents, especially those associated with city government and the

Chamber of Commerce, preferred it that way. They felt threatened by provincial involvement in city affairs and did not want the economically independent city to become entangled in Muslim-dominated provincial politics. A Supreme Court decision shortly before the election exempted Cotabato City from the new rules, and city residents did not vote in the provincial elections.

14. Campo Muslim residents reported, and reviews of back issues of the local newspaper confirmed, that there was no attempt by any candidate in the 1980 city elections, or in three previous city elections, to appeal, in any specific fashion, to the urban poor.

15. The use of English here suggests that the term "urban poor" had become part of Philippine political vocabulary along with "poll-watcher," "people power," and "snap election."

16. After months of lawsuits and judicial hearings no substantial evidence of ballot box tampering or related voting fraud was established to support the charges made against any candidate, and the winners were officially proclaimed.

17. Voter turnout figures by precinct for the 1988 election were not available before I left Cotabato City. However, figures for the 1986 presidential election—which showed a similarly high level of preelection interest and a similarly low overall turnout (57.5 percent)—may be useful for illustration. Voter turnout in the seven precincts at which most Campo Muslim residents were registered was 54 percent. In seven similarly sized precincts in a middleclass Christian area of the city, 67 percent of registered voters cast their ballots.

18. Mary Hollnsteiner, in her 1963 book, The Dynamics of Power in a Philippine Municipality , provides a definition of the term "lider":

The word "lider" though originating from the English word, "leader," has been incorporated into Tagalog speech . . . with a very precise connotation . . . [It] refers to the person with a large following in a barrio [barangay] who utilizes this support during political campaigns, where he pledges himself to campaign for a certain candidate or group of candidates. These candidates call him their "lider" referring to his dominance over his particular followers rather than to any superordinate position he holds in relation to the candidates. On the contrary, the "lider" . . . is a staunch follower of the candidate he is supporting. The "lider'' has no official position as such but is repaid by candidates with favors which can in turn be distributed to his followers, reinforcing his position. (1963, 41, fn. 16)

As shall become apparent, Muslim liders in Cotabato City did not always exhibit all the attributes or behaviors included in Hollnsteiner's definition.

19. Fifty pesos (approximately $2.50) was reported to be the minimum cash price being paid for votes during the 1988 election. It was the equivalent of two day's wages for many of the residents of Campo Muslim.

20. Neither of these two individuals accepted the offers made to them. In my conversations with them, I was able to determine that their refusals were not based on any moral discomfort associated with brokering votes per se, or with brokering votes for Christian candidates. Instead they had to do with

concerns about their personal reputations. The purok leader planned to run for public office himself eventually and did not want to become indebted to, or associated with, that particular candidate. Nur Miskin, the Islamic activist, was attempting to build an organization and did not want it associated with "politics." He was also concerned with the potential problems associated with the sublider role. As he told both the lider and me: "If I took the money and distributed it all to my men, many would still say that I received much more. I would be considered just like Bapa Pantal [a well-known lider of Angka Biruar, the leading Muslim mayoral candidate]. Some of his followers are accusing him of taking money from Angka Biruar and not distributing it."

21. Although a Christian, this candidate was a member of a Chinesemestizo Muslim family, had long-standing connections with certain prominent Muslim figures in the city, and was the "unofficial" IPP candidate for mayor. The IPP nomination panel had refused to endorse him, but some prominent IPP members apparently favored his candidacy.

22. It should be remembered in respect to this first sort of breakdown that while vote-buying was a long-established practice in multi-ethnic Cotabato City (as in the rest of the Christian Philippines), it was relatively rare in rural areas of Muslim Cotabato prior to 1986 insomuch as autocratic local rule usually made it unnecessary. In the 1988 elections, the incidents reported to me of subliders failing to distribute a candidate's funds invariably occurred in Muslim rural communities (such as Bokhana) within the city limits.

23. Votes gained by the eight leading candidates represented about 75 percent of the total votes cast in the six precincts.

24. It is doubtful that any of the Muslim candidates for mayor expected to attract significant numbers of Christian voters that would be alienated by the candidate's embracing the IPP. Their political calculus was more likely based on the knowledge that some Christian city residents had recently expressed anxiety at pronouncements of the IPP and MILF concerning their intention to have Cotabato City included in a Muslim autonomous region without holding a popular referendum on the issue. Given the level of Christian apprehension at the time, a candidate's affiliation with the IPP would be self-defeating in that it would certainly prompt a higher voter turnout among Christian city residents and might even provoke the leading Christian candidates to cooperate to prevent an IPP victory. As it happens, a government-sponsored referendum on the issue of autonomy did occur in November of 1989. Maguindanao Province was one of four Muslim-majority provinces in the Philippines that voted to form the Autonomous Region of Muslim Mindanao, while Cotabato City voted against inclusion in the autonomous region.

1. See Chatterjee (1993), Comaroff and Comaroff (1992), Fox (1985, 1989), and Scott (1985, 1990) for examples of subordinates reinterpreting dominant representations.

2. Fox reports that the term "Sikh" refers to several cultural identities prevailing in the Punjab in the nineteenth century, identities that "subsumed a

range of quite different religious beliefs and social practices" (1985, 7). The British, however, regarded only one Sikh identity—the Singh variant—as the significant form and, in fact, believed the Singhs to be a separate race. In the early twentieth century, urban-based reformers "appropriated the Singh identity fostered by the British to launch an anticolonial protest" (1985, 12). In doing so they themselves merged the Sikh and Singh identities, promoted a single image of Sikh orthodoxy, and directly challenged Sikh collaborators with British rule.

3. Kahn is not alone in objecting to Hobsbawm's notion of "the invention of tradition." See Friedman (1992) and Kapferer (1988) for additional critiques.

4. For details on that armed uprising, which resulted in the deaths of thirty-five Muslim insurgents (Princess Tarhata and her husband escaped), see Thomas (1971, 73-76).

5. William Roff has described the source and consequence of that dynamism succinctly: "[T]he recognition of [the] non-congruence [between ideal and social reality] by both prescribers ( ulama ) and backsliders acts as a dynamic force within Islamic cultures, resulting in what can be seen as dialectic constantly engaged in translating synchronic tension (the aspect taken by the lack of fit at any given moment) into diachronic 'oscillation' (social, cultural, political, or ideational change in one direction or another)" (1985, 9).

6. For an engaging account of recent generational conflict among the Sama (a Philippine Muslim group of the Sulu archipelago) over "ways of knowing Islam," see Horvatich (1994).

7. William Roseberry presses a similar point in his writings on hegemony. Like Rebel, he proceeds from an explicit recognition of "differential experience in terms of . . . structures of inequality and domination" (1989, 48). The "common understandings and modes of interaction" that emerge across this differential experience ''can never encompass" all of it. "Cultural production is not limited to those who control the means of cultural production. Experience constantly intrudes" ( 1989, 49).

8. Caution is required when assuming the widespread living of lies by political subordinates. The absence of ideological incorporation is just as much an empirical question as is its presence.