Chapter 3
Islamic Rule in Cotabato

Representations of the Precolonial Order

Two origin myths animate Muslim nationalist politics in Cotabato. The first is an ancient myth that explains social disparity (past and present) among Cotabato Muslims by sanctifying it. This myth of sanctified inequality delineates an aristocracy entitled to rule Cotabato Muslims on the basis of their ancestral ties to the legendary figure who first brought Islam to Cotabato and, through him, to the Prophet Muhammad.^[1] The second is a modern myth about ancient origins. The myth of Morohood is shared widely by Muslim nationalists throughout the Philippines and states that a transcendent Philippine Muslim (or "Moro") identity was uniquely forged among the various Muslim ethnolinguistic groups of the southern Philippines in the course of their struggle (begun more than four hundred years ago) against Spanish attempts to subjugate them. That shared oppositional Islamic identity, together with the tradition of aristocratic rule, is considered to constitute the primordial foundation for the contemporary Muslim separatist movement.

In this chapter and the two that follow I will consider these two myths in their historical contexts. While they tend to be complementary and mutually reinforcing (the anti-Spanish resistance posited in the second myth was inspired and organized by the Muslim nobility constituted by the first), they have not had equal motivational force in the struggle for Muslim separatism. The myth of Morohood is universally embraced by Muslim nationalist ideologues. The myth of sanctified

inequality, on the other hand, has been at least indirectly challenged by some separatists. It has also been appropriated by prominent Muslim opponents of the separatist cause. Nevertheless, the two myths remain securely coupled in nationalist ideology and so will be analyzed in tandem.

The revanchist or revitalistic spirit that publicly motivates most ethnonational movements almost always includes assertions of the preferability of traditional governance. The Muslim separatist movement in the Philippines is no exception to this tendency. While depictions by Muslim nationalist ideologues of the precolonial past have been mostly vague, they invariably extol the moral and material advantages of the premodern order.^[2] Assessing such claims, and the responses of ordinary movement adherents to them, requires the investigation of past political and economic relations. This chapter examines the ideological expression and practical operation of the system of Islamic governance in precolonial Cotabato.^[3]

Precolonial Governance and its Idealization in Cotabato

In contemporary Muslim politics in Cotabato, the past inhabits the present in a manner insufficiently expressed by the term "traditionalism." Various idealized versions of the precolonial past are used as models for present-day political action by a wide range of Muslim political actors, from separatist insurgents to Philippine government officials. Recourse to an imagined past for present political guidance is, of course, a common element in contemporary nationalist movements. The extraordinary aspect of its occurrence in Cotabato is the extent to which not only various Muslim separatist factions but also Christian politicians and state functionaries share the perception that the political fate of Cotabato Muslims is tied to their traditional nobility as a result of ancient and immutable bonds. Both proponents and opponents of Muslim separatism tend to assume that ordinary Muslims reverence their highborn leaders (or datus ) for their sacred ancestry and could not imagine politics without them. In this view, any political arrangement to enhance Muslim self-determination in Cotabato must include prominent consideration of the role of the traditional nobility. It is considered preferable to attempt to reform the unsuitable practices of contemporary datus rather than to try to undermine their influence, because the bond between traditional leaders and followers is the epitome of Muslim political culture in Cotabato.^[4]

Little historical scholarship exists by which to evaluate the validity of these perceptions. Although a number of the available scholarly works are quite astute in their political observations (see Ileto 1971; Stewart 1977; Beckett 1982; Adil 1955), none has provided a comprehensive analysis of traditional sociopolitical relations. I offer such an analysis here in order to provide a historical standard by which to measure contemporary political relations between Muslim leaders and followers.

My depiction of political relations in the Cotabato sultanates relies on ethnographic information garnered from interviews I conducted in and around Cotabato City in the mid-1980s. My primary consultants in this matter were four individuals, three of them elders, who were unusually knowledgeable about traditional arrangements.^[5] Two—Datu Mohammad Adil and Datu Kasim—are descendants of the Magindanaon high nobility.^[6] Imam Akmad, an Iranun speaker, is the imam (prayer leader) of the Campo Muslim mosque, and Hadji Abbas is a descendent of the dumatus , a special status group in the Magindanao Sultanate (see below). Each of these informants shared with me oral traditions learned from their elders, as well as their own inferences about the nature of traditional governance in Cotabato.^[7]

I supplement ethnographic material from Cotabato with primary indigenous historical sources from the Muslim Philippines and works based upon them (Saleeby 1905; Adil 1955; Glang 1969; Majul 1973; Hooker 1983; Mastura 1984; Loyre 1991). I rely as well on historical evidence from Western sources (Blumentritt 1893; Combes 1903–19; Saleeby 1905; Dampier 1906; Forrest 1969; Arcilla 1990) and on scholarly works based on those sources (Ileto 1971; Majul 1973; W. H. Scott 1982, 1984, 1994). In addition to information pertaining peculiarly to Cotabato, I also use European source materials and scholarly works that describe precolonial polities in other parts of the Muslim Philippines (Mednick 1965, 1974; Warren 1981; W.H. Scott 1994), the non-Muslim Philippines (W. H. Scott 1982, 1984, 1994), and elsewhere in Islamic Southeast Asia (Gullick 1958, 1987; Milner 1982; Sullivan 1982; Andaya 1993).

From Heterarchy to Hierarchy:The Decline of Local Descent Groups

Pre-Islamic Magindanaon social organization consisted of a number of localized cognatic descent groups, or bangsa ,^[8] that were associated with particular autonomous or semiautonomous ingeds, or localities.

The Magindanaon bangsa were most likely similar to those of the Maranao, the neighboring Muslim ethnolinguistic group of the Lanao Plateau. As described by Mednick (1965), membership in a Maranao bangsa provided access to rights in a specified area—most importantly, usufruct rights in land, or the right to exploit a certain natural resource—as a result of an ancestral association.

In pre-Islamic Cotabato, these large, localized descent groups produced chiefs who, under certain conditions, were able to extend their power beyond their own bangsa and become chieftains or, in exceptional cases, rulers of harbor principalities. With the coming of Islam, however, the bangsa were crosscut, and eventually attenuated, by a separate, societywide aristocracy whose members claimed descent from a common ancestor, Sarip Kabungsuwan, an émigré nobleman. With the establishment of the Cotabato sultanates,^[9] local chieftains became principally interested in establishing the maximum number of descent links to Sarip Kabungsuwan, the prime ancestor. Their new status aspiration was to establish their rank positions in respect to the paramount rulers of the sultanates, who ruled by right of their membership in the aristocratic lineage (barabangsa ) founded by Kabungsuwan. With increased status delineation and relative political centralization, the significance of local ancestors and the power of local corporate descent groups were greatly diminished. While local identity remained strong, local ancestry retained only secondary importance as a criterion for local rule.^[10]

Magindanaon bangsa survived as tupus , or local descent lines. Informants report the existence of forty-two Magindanaon tupus, although none could name more than twenty of them.^[11] It appears likely that while localities (ingeds) remained largely independent entities under the sultanates, local descent lines gradually lost both depth and meaning. The descent line of Sarip Kabungsuwan, the point of reference for ruling datus, became not only the paramount descent line but the only significant one. Tupus seem to have been insubstantial categories that did little more than focus loyalty to place. Bilateral kindreds replaced bangsa as the largest effective kinship units for ordinary Muslims.

Islamization and the Myth of Sanctified Inequality

The coming of Sarip Kabungsuwan to Cotabato is the charter event for the claims of the rulers of Cotabato in the historical period to nobility

and moral authority. The story of the arrival of Sarip Kabungsuwan is told in the tarsilas , the written genealogies that link the royal houses of Cotabato with their progenitor. The term "tarsila" is derived from the Arabic "silsila," meaning name-chain, and the genealogical accounts in the Magindanaon tarsilas not only linked living datus with Sarip Kabungsuwan, their apical ancestor, but also contained supplementary sections tracing the ancestry of Sarip Kabungsuwan to the Prophet Muhammad. The tarsilas were written on paper in Magindanaon or Malay in Arabic script and were possessed by the sultans and all leading datus. Individual tarsilas were frequently recopied for updating or to replace worn copies.

According to the tarsilas, Sarip Kabungsuwan was the son of Jusul Asikin, the daughter of the Sultan of Johore, and Sarip Ali Zain-ul Abiden from Mecca. Thus, Kabungsuwan was the offspring of a princess of the Melaka royal family and, more significantly, the son of a sharif (the original Arabic form of "Sarip"), and hence a direct descendent of the Prophet Muhammad. Majul (1973) suggests that the historic Kabungsuwan must have sailed to Mindanao some time after 1511, when the Melaka royal family, to which his mother belonged, was driven from Melaka by the Portuguese and established a sultanate in Johore.

The tarsilas relate how Kabungsuwan arrived by chance at the mouth of the Pulangi River and began to convert local chieftains and their followers to Islam. He married the daughters of some of these chieftains, thus establishing in Mindanao a barabangsa, or royal lineage, whose members claimed a sacred genealogy, tracing their origins to the Prophet Muhammad. The tarsilas report that the first wife married by Sarip Kabungsuwan in Cotabato (Putri Tunina) was found in a bamboo stalk. Kabungsuwan's son, Sarip Makaalang, married a woman who emerged from a crow's egg. The existence of these supernatural children raised the status of the female descent line, thus further distinguishing the barabangsa line from autochthonous lineages.^[12]

Sarip Kabungsuwan founded the Magindanao Sultanate at the coast and his rule passed to his son, Sarip Makaalang. The upstream Buayan Sultanate originated, according to the tarsilas, from a union between the daughter of Sarip Kabungsuwan and a Buayan chieftain. Because the male descent line carries more genealogical weight than the female, the Magindanao Sultanate was able to designate itself the premier royal house of Cotabato, while the rulers of Buayan for most

of its history used the title raja (prince) rather than sultan.^[13] The traditional system of hereditary ranked statuses is also traced to the arrival of Kabungsuwan. This system, in its most general configuration, comprised four tiers. At the apex were the datus—rulers or descendants of rulers. Their ascendant status stemmed from Sarip Kabungsuwan's assumption of sovereignty over Muslim Mindanao (and its dependencies) and from their ability to trace descent from Kabungsuwan and, through him, to the first ruler of Muslims, the prophet Muhammad.^[14]

Passing over, for a moment, the second level, we find on the third tier the endatuan , meaning literally "those who are ruled." The endatuan were the subjects of datus rather than of the realm; they were considered residents of particular ingeds (settlements) and subordinate to specific datus. The status of the endatuan seems to have been negatively defined. They were the followers of (and likely bore some kinship relation to) local chieftains. But unlike their rulers, they were unable to demonstrate sufficient kinship links with Sarip Kabungsuwan, the prime datu ancestor, to qualify as datus.

The second tier was occupied by the dumatus. This intermediate status has not been reported for other Philippine Muslim populations. The dumatus, as they describe themselves today, were neither datu nor endatuan—neither rulers nor ruled. The dumatus are the descendants of Tabunaway, a legendary Magindanaon chieftain who welcomed Sarip Kabungsuwan to Cotabato. The tarsilas record that Tabunaway acknowledged the sovereignty of Sarip Kabungsuwan and his descendants in exchange for certain privileges. The first entitlement was that neither he nor his descendants would pay tribute to any datu. Hadji Abbas expressed the distinction between endatuan and dumatu in his inged this way: "At harvest time the datu sent sacks to the endatuan who were obliged to fill all the sacks the datu gave them. The dumatus were not sent sacks and did not have to provide rice to the datu." The second dumatu entitlement was that no datu could be proclaimed as sultan without the participation of a Tabunaway descendent.^[15] The dumatus have kept their own genealogical records of the Tabunaway descent line, principally to preserve their privileges vis-à-vis the Magindanaon aristocracy. Theirs is the only tarsila in Cotabato that does not trace descent from Sarip Kabungsuwan. The special status of Tabunaway descendants has allowed them to maintain, more so than any other group, their separate bangsa by remaining ancestor focused, self-ruled, and relatively corporate.

The fourth and lowest tier comprised the ulipun , or "disfranchised" (W. H. Scott 1982, 142) persons.^[16] Ulipun were debt-bondsmen whose unfree status resulted from punishment for a legal offense, from failure to pay tribute or repay a debt, or from being sold by relatives (or occasionally by self-sale) into servitude in times of economic crisis. Ulipun status, unlike the other three estates, was, in general, neither ascribed nor permanent. It could, however, be inherited and, in practice, the system inhibited self-redemption by debt-bondsmen, probably because datus favored maintaining a high proportion of bound followers. As Warren (1981) describes the system of debt slavery in another Philippine Muslim polity—the Sulu Sultanate—the debt-bondsman's service to his creditor "did not generally count towards repayment of his debt" (1981, 216). Ulipun status was likely similar in most respects to the position of debt-bondsman described by Warren. Despite their inferior status, the ulipun, as Magindanaons and Muslims, were societal insiders and possessed certain rights, including the right to own chattel slaves (banyaga ). Those ulipun who were members of a datu's retinue were treated as household dependents and were occasionally able to rise to positions of significant responsibility. Banyaga, on the other hand, were despised outlanders who were not included in the system of social rank. They were captives acquired for the most part outside the territory of the sultanates, were usually not Muslims, and had no recognized rights and no social status other than as acquired persons—the property of others.^[17]

The central organizing principle of this system of ranked statuses was maratabat (from the Arabic "martabat," or "rank"). Among the Magindanaon, "maratabat" primarily connotes rank arid secondarily the honor due to rank. Maratabat is the quantifiable essence of status rank and is measured most commonly as a monetary valuation when determining the proper amount of bridewealth (bantingan ) to be exchanged at marriage or the amount of wergild (bangun ) required to avert a feud. It is expressed generally as a gradation of four standard sums required for bridewealth within each of the four ranked estates; datu maratabat is valued at one thousand units, that of dumatus at seven hundred, the maratabat of the endatuan at five hundred, and that of ulipun at three hundred units. These quantities were expressed to me in Philippine pesos by informants but almost certainly referred originally to Mexican silver dollars, used as media of interregional exchange in Southeast Asia as recently as the early American period (Forrest 1969, 279; Gullick 1958, 20 n. 1; Miller 1913, 341).^[18]

Ideological Incorporation and Rank Competition within the Aristocracy

In precolonial Magindanaon society, a close correspondence obtained between datu status and political and economic predominance. As detailed in the chapter that follows, the Magindanaon nobility composed a ruling class that extracted tribute in various forms from ordinary Muslims and other subject peoples. Many aristocrats also invested the wealth so obtained in trading or raiding expeditions and further enriched themselves. Ideological instruments existed that promoted ruling class closure and authorized existing social arrangements. The foremost of these were the tarsilas themselves. With the single exception of the tarsila of the descendants of Tabunaway (which does not trace descent from Sarip Kabungsuwan), the Magindanaon tarsilas provide a charter for an aristocratic ruling class. They provide "proofs of legitimacy" (Majul 1973, 3) for a tribute-taking elite, most importantly by asserting descent from the Prophet Muhammad through a sharif line and thus justifying the role of datus as political and religious leaders.

In addition to the written genealogical accounts, there were formal and informal oral traditions that advanced an ideology of aristocracy. Taritib (from the Arabic "tartib," meaning "order" or "sequence") refers to a body of oral guidelines or regulations concerning the proper criteria for the choice of a sultan and the procedure for his installation and, generally, to the protocol governing relations between the sultan, datus, and subordinate classes. Taritib may still occasionally be heard in Cotabato orated by elder members of the royal houses on Magindanaon radio programs.

Aside from taritib, there were a number of anecdotes, adages, and precepts concerning the Magindanaon nobility that circulated within the ruling class. The highest-ranking members of the aristocracy were perceived to have "Arab" features, that is, relatively light skin and aquiline noses. Datu Kasim told me that the phrase "Watu na Ulu" (head of stone) referred to the conviction that every Magindanaon community required a strong leader identified with the nobility. The precedence of highborn status over acquired wealth was affirmed in various narratives (compare Gullick 1958, 65). Hadji Abbas related one such narrative told to him by his parents. It concerns a welldressed young man of modest rank who attended a celebration at the house of the datu and sat in a prominent place. An elder relative of the datu could only afford poor clothes and so sat far from the datu. When

the datu noticed this, he had his armed retainers remove the well-dressed man and seat the datu's poor relation in his place. A proverb accompanied the story: "You wear a rich garment and I only a poor one; but when we arrive at the court of the datu, there we shall see which of us shines with the golden radiance of nobility."^[19]

Other cautionary proverbs were sung by bayuk performers—praise singers who performed at special ceremonial occasions of the nobility. The familiar metaphor used in these proverbs was that of the small bird who endeavors to fly very high but in the end falls back to his old nest. The admonition was to young men of questionable status who wished to marry high-ranking women but lacked proper genealogical qualifications.

The principal impulse for the status-related behavior of individual members of the traditional Magindanaon ruling class was rank competition. Gullick's observation for the Malay states, that "the concept of differential status was one of the main interests and values of the ruling class" (1958, 66), is equally accurate for the Cotabato sultanates. The sultan was the apex of the ruling class in each state and the point of reference by which members of the aristocracy determined their relative status. Sultans, ideally, were distinguished by their pulna status; those individuals designated as pulna were able to trace direct ancestry from Sarip Kabungsuwan through both parents.^[20]

Within the datu estate as a whole, claims to status rank were predicated upon the quality and quantity of descent lines linking an individual to Sarip Kabungsuwan. The quality of a descent line was determined by its association with an inherited, ranked title. These titles originally corresponded with specific offices conferred by a sultan upon particular individuals. Such titles were inherited and often lost any direct connection with a functional administrative position while remaining significant status markers.^[21]

Judicious marriages were vital for maintaining or increasing family status among the Magindanaon nobility. The rule of hypergamy—mandating that a woman could only be given in marriage to a man of equal or higher rank—ensured, for instance, that the offspring of a woman of pulna status would always remain pulna. However, it also meant that women flowed up through the ranking system as parents endeavored to contract marriages for their daughters that would increase the status of their grandchildren's line. Because tribute-takers were also wife-takers, the rule of hypergamy also encouraged particular forms of class exogamy as datus obtained junior wives from among

their non-elite followers. With descent figured cognatically, so that kinship links through females mattered, a formally straightforward status system was, in practice, quite complex.^[22]

Because of the rule of hypergamy and the practice of polygyny, a sultan or other titleholder would likely have numerous recognized offspring whose status ranks ranged from pulna to the lowest status levels of the datu estate (through unions with endatuan or ulipun wives). This was advantageous for a ruler, who was thus possessed of a number of daughters of differing ranks whom he could bestow on various datu allies without violating the rule of hypergamy (Beckett 1982).^[23] Because the Cotabato sultanates lacked a rule of primogeniture, succession to a title was open to competition. Potential conflict was limited, however, by taking into account the status of the mother as well as the father of the claimant. In this way, the inheritance rights of rival claimants—half brothers, for instance—might be decided on the basis of fine distinctions in their relative maratabat, or rank (Beckett 1982).

The effect of the combination of rank gradations within the datu estate, cognatic descent reckoning, and status group exogamy was to render the ideal of discrete valuations of maratabat based on social categories impractical and largely irrelevant. With the children of differently ranked parents allocated to a rank intermediate between those of their parents, the system of prestige stratification generated an "intergraded spectrum of rank" (Stewart 1977, 290) rather than sharply defined status groups. Just as marriage was the major vehicle for rank enhancement, bridewealth determination was the primary occasion for the public evaluation of a family's maratabat. Bridewealth negotiations were complex and delicate affairs necessitated by the status refinements produced by almost every marriage union (compare W. H. Scott 1982).

As instruments for the determination of relative status, tarsilas were considered private warrants rather than public documents. Although they served to define and delimit the aristocracy, their principal purpose was to assign rank within it. A tarsila, in other words, was possessed primarily to substantiate the rank of its holder, but it might also be employed to ascertain the rank of other individuals in specific situations for the purpose of assessing—in an ad hoc manner—one's maratabat relative to a particular other.^[24] Titleholders employed tarsilas to seat guests at ceremonial occasions according to rank. Hadji Abbas related the following story about the use of tarsilas by individual

datus to assess relative rank: "When boats passed by the house of a datu, their passengers were required to fold or lower their umbrellas unless they were of higher rank. If a passenger did not do so, the datu immediately had his expert check the tarsila. If the man was found to be of lower rank the datu would order his men to sink the boat." Tarsilas were also used by datus individually to enforce the status privilege of the aristocracy as a whole. As described by Datu Adil, certain garments and dress styles were reserved for the nobility: "Women of high status and those of lower status wore their malongs (sarongs) differently, and carried themselves in different ways. High status women would carry their malongs over one arm. Lower status women would wear theirs more in the manner of men. If a woman of questionable status was found wearing her malong in the style of a high status woman the tarsila was examined. If she was found to be a commoner she would be fined forty pesos by the local datu. If she was unable to pay she would be made to work for the datu for one or two years."

The myth of sanctified inequality was the basis for an ideology of nobility embraced by a ruling class that nonetheless regularly contradicted some of its basic doctrines: aristocratic endogamy, the association of paramount rule with the purest bloodlines, and the indelibility of the lines separating social strata. As in other political belief systems, rather than provoking confutations these irregularities generated mechanisms (such as genealogical manipulation or special appointment) for the "post hoc ennoblement of the powerful" (Beckett 1982, 398).^[25]

The "struggle for status" (Geertz 1980, 116) within the Magindanaon aristocracy fostered coherence among the ruling class by mitigating the many centrifugal tendencies inherent in a political field where local leaders competed for followers and slaves in an underpopulated region. The notion of a common noble ancestor provided a mechanism that bridged the gap between various ingeds (localities) so that individuals were motivated to marry and otherwise relate to one another in order to maintain their positions within the status group. The spatially dispersed members of the datu stratum were integrated by ties of supposed shared ancestry that crosscut local descent lines. As the source of all aristocratic titles and the focal point for the system of prestige stratification, the office of sultan (if not always the current officeholder) was treated with deep respect by members of the datu stratum. Ruling class coherence even extended between sultanates (whose elites shared the ruling ideology) and probably helped to

smooth intersultanate trade relationships and temper the level of endemic intraregional hostilities.

Ruling Ideas and the Incorporation of Subordinates

The post-Althusserian social theory of the past twenty years offers a broad range of analyses of the reproduction of social disparity within complex political structures. While analysts, as we have seen, disagree sharply on many of the basic issues, general accord is evident on at least one point. It is agreed that it is inaccurate to measure the incorporation of political subordinates merely by assessing their belief or disbelief in the legitimizing myths of a ruling class. Political structures are significantly reproduced by other means (and here there is disagreement): by nondiscursive practices, by coercions, by compensatory beliefs or the "common sense" consciousness generated as part of the everyday operation of systems of domination (see, e.g., Bourdieu 1977; Foucault 1978; Abercrombie et al. 1980; Williams 1980; Corrigan and Sayer 1985; Elster 1985; J. Scott 1985, 1990).

The contemporary descendants of the Magindanaon aristocracy favor the view (shared by a range of political actors) that Magindanaon subordinates have perfectly shared (and continue to share) the dominant myth of sanctified inequality. Datu Alunan Glang notes that "down to this day, many of them still hold the datus in characteristic religious awe and adulation" (1969, 33). Datu Michael Mastura seconds this sentiment when he writes that "much of Magindanao's sociopolitical history can be evaluated in terms of the tradition for loyalty to the datus or at least in the veneration of descent groups belonging to the Magindanao core lineage" (1984, 40). Bai Rebecca D. Buan ("Bai" is the form of reference for a female member of the nobility), in a 1978 letter to the editor of a Cotabato newspaper, takes a peculiar pre-Islamic slant when she comments that "the datus of olden days . . . were the living gods of their people" (Mindanao Cross , April 7, 1978).

An examination of the historical evidence and oral traditions from Cotabato suggests that ideological adherence was more fragmentary and symbolic power relations more complex than imagined by contemporary datus. Looking first at the more formal constructs of the myth of sanctified inequality, we find that the tarsilas had little relevance for subordinates except on the occasions when they were used to restrain the behavior of a social-climbing commoner. As essentially private doc-

uments, the tarsilas were viewed rarely, if ever by subordinates (Majul 1973, 3). Commoners were more likely to hear tales of nobles past and present performed by professional reciters and praise-singers at weddings and other celebrations. The primary audience for these performances was aristocrats, and their first purpose was rank competition rather than the edification of commoners.

A second official forum for the transmission of the dominant ideology would have been the mosque, particularly at Friday community prayers. As putative sharifs, sultans and ruling datus were religious as well as political leaders. Imams (prayer leaders), themselves datus, were appointed by rulers, and they paid customary obeisance to those rulers before beginning public prayers in the mosque. In their Friday orations, imams asked special blessings for the sultan, his family, and his royal predecessors (Saleeby 1905; Loyre 1991). Accounts of public religious practice suggest, however, that the mosque was not a primary arena for interclass instruction. William Dampier visited Cotabato for seven months as an officer aboard a British privateering vessel, the Cygnet . His 1697 account, New Voyage Round the World , provides the best ethnographic description of Cotabato in the seventeenth century. It includes the following description of mosque attendance: "Friday is their Sabbath; but I never did see any difference that they make between this Day and any other Day, only the Sultan himself goes to the Mosque twice . . . The meaner sort of people have little Devotion: I did never see any of them at their Prayers, or go into a Mosque" (1906, 344–45).

The entire community did gather for such ceremonial events as royal weddings, installations, or lustrations. These were certainly (as elsewhere in Southeast Asia) occasions for displaying the power and splendor of the ruler, but they offered limited opportunity for the direct inculcation of dominant beliefs.^[26]

The ideology of sanctified inequality was tracked by more elemental (and more indigenous) representations of rule. These were symbolizations of individually exercised power rather than sanctions for collectively claimed authority. Jeremy Beckett notes that the term "datu" had two meanings: "ruler" and "one entitled to rule on account of his descent from datus" (1982, 396). The ideology of nobility emphasized the latter meaning. In practice though, the regular occurrence of feuds, wars, and succession struggles ensured that the personal attributes of a datu, especially his ability to command fear and deference, remained a key factor in his political successes.^[27] Datu Adil informed me that the

ideal datu acquired recognition without demanding it and obtained tribute without exacting it—not because of his elevated rank but as a result of his personal force. That regard for personal power and its projection was often represented in the names of prominent datus, three of which—Mangelen, Makapuges, and Dilangalen—may be glossed as "the Controller," "the Enforcer," and "He Who Cannot Be Moved." It was also evident in some inherited titles. The full title of one of the highest offices in the Magindanao Sultanate, "Amirul Umra a Tibpud a Bangias a Dimakudak u Pagilidan," translates as "Commander of Decisive Power Who Treads upon the Seashore."

The substantial power of a ruling datu was projected symbolically through his supernatural abilities. As with the political skills that secured his position, those abilities were individually acquired, not innately possessed. The single exception to the pattern of intentional acquisition of supernatural abilities was the supernatural potency that inhered in the person of the sultan. That endowment of potency was equally capable of benefaction or injury. As related by Imam Akmad, "The sultan would bathe ceremonially in the Pulangi River to drive away the three evil spirits of Satan [Saytan]. Ordinary people were careful never to bathe downstream from the sultan. It was believed that the sultan was toxic, that the water that flowed past his body would make a person ill."^[28] All other aristocrats, I was told, acquired their supernatural abilities through the study of esoteric arts known collectively in Magindanaon as kamal or ilmu ("special power" or "special knowledge"). Kamal was taught by special masters and was effectively limited to members of the aristocracy. Kamal arts included the abilities to repel blades and bullets, leap great distances, and disappear at will.

Ethnographic evidence suggests that these symbolizations of power were appreciated, and often embellished, by Magindanaon subordinates. When asked about the supernatural powers exercised by past rulers, contemporary datus had few stories to offer, and those they did tell usually emphasized how the powers were acquired from specific teachers rather than relating how such powers were used to secure or maintain political power.^[29] By contrast, Magindanaon commoners related numerous narratives of supernatural power—stories that diverged from those of datus in interesting ways.

I often heard stories of datus of old (and a few in living memory) possessed of extraordinary physiognomies: terrible visages, monstrous bodies, extra eyes and fingers. Rulers also exhibited various marvelous

powers—those directly related to the kamal arts but others as well: prognostication, remote vision, and the ability to strike a man dead without touching him. In these stories, neither extraordinary appearance nor exceptional powers was explicitly linked to noble descent. They were attributed only to prominent and powerful members of the nobility and served as both metaphors for and explanations of the personal power of individual rulers. The narratives related by non-elite Muslims tend to be unclear about, or indifferent to, the ways in which kamal powers were acquired by rulers. They focus instead on linking the power of command with the command of supernatural powers. Taken as a whole they form a subordinate mythology of power analogous to but distinct from the dominant mythology of nobility.

While the myth of sanctified inequality is advanced by many in Cotabato today as the fundamental axiom of traditional governance, it is more likely that symbolic representations of power, some of them generated endogenously out of the experiences of subordinates, were more significant for social reproduction.

Interpreting Political Relations in Precolonial Cotabato

How are we to characterize the political relations represented by the myth of sanctified inequality? In recent years, the most favored interpretive stance for reconstructing the precolonial polities of Southeast Asia has been to privilege local epistemologies and search self-consciously for authenticity (see, e.g., Geertz 1980; O. W. Wolters 1982; Milner 1982; Errington 1989; Bentley 1986).^[30] Proponents of this approach tend to focus interpretive attention on sacred texts, courtly ceremony, and the formal schemata of social hierarchy as crucial bases of traditional political order. These phenomena are analyzed not as instances of symbolic power relations but rather as expressions of broadly shared "structures of experience" (Milner 1982, 113) signifying that "political systems are expressions of culture" (O. W. Wokers 1982, 9).

The works cited above—and particularly those by Clifford Geertz and Shelly Errington—present evocative and finely detailed portraits of traditional Southeast Asian polities and demonstrate the value of an anthropological perspective for illuminating Southeast Asia's precolonial past. Nevertheless, their aversion to exploring linkages between political culture, material relations of domination, and the reproduction of disparities in social power raises the suspicion that they don't

provide the full picture. That impression is reinforced by the fact that these works nearly always present an exclusively elite view of traditional arrangements. What anthropologists should also be able to offer as a complement to Southeast Asian historical sources is a village-level view of the traditional social world; yet ethnography in these works is not employed to examine possibly divergent viewpoints from the peasant periphery, but relied upon only for oral description of the "exemplary center" (Geertz 1980, 11; see also Errington 1989).^[31]

The works by Milner and O. W. Wolters mentioned above share this top-down perspective and its limitations. Concern with shared "structures of experience" tends to overlook the fact that traditional nobilities and their mostly rural subjects had very different life experiences, and that, as Pierre Bourdieu reminds us, those occupying disparate economic and political positions in a social configuration tend to produce distinct "practical interpretations" of objective conditions (1977, 116). It is precisely those sorts of divergent practical interpretations that I have attempted to draw out and analyze in my own reconstruction of the precolonial social order in Cotabato.

In contrast to the emphasis on the unifying effect of shared culture found in the works just cited, various depictions of political relations in the precolonial sultanates of the Philippines have emphasized material remuneration and dyadic negotiation between leaders and followers as the principal bases for political adherence (see, e.g., Mednick 1974; Gowing 1979; W. H. Scott 1982). The writers of these works have been tempted to view precolonial political relations through the same "dyadic alliance" (Lande 1965) lens so prominently utilized to analyze contemporary political activity in the Philippines (see, e.g., Lande 1965; Lynch 1984; Hollnsteiner 1963). Exclusive attention to "rational actors" and personalized relationships obscures the profoundly constraining effect of political-economic structures on the negotiation of vertical dyadic ties.^[32] As we shall see, the larger political context in precolonial Cotabato—one marked by armed coercion and legal insecurity—diminished much of the negotiating advantage of subordinates, despite the fact that land was plentiful and followers were a relatively scarce, and thus highly valued, resource.

William Henry Scott (1982) characterizes the tacit agreements binding datus and subordinates in the Philippine sultanates as follows:

The datu's power stems from the willingness of his followers to render him respect and material support, to accept and implement his decisions, and to obey and enforce his orders, and is limited by the consensus of his

peers. Followers give their support in response to his ability and willingness to use his power on their behalf, to make material gifts or loans in time of crisis, and to provide legal or police protection and support against opponents. . . . Failure to discharge such duties may result in the quiet withdrawal of cooperation and support, so that autocratic behavior on the part of any datu is the result rather than the cause of subservience on the part of others. (W. H. Scott 1982, 139–40)

Scott's depiction of the sultanates as virtual free markets for the negotiation of political linkages between individual datus and followers is not well-supported by the evidence from Cotabato and elsewhere.

Relations of subjection in the Philippine sultanates exemplify what Orlando Patterson (1982, 18) has termed the "personalistic idiom" of power. In such social settings power relations are humanized, often employing the principle of kinship, but not mystified. Domination is direct and discernible. "No dependent in such societies ever loses sight of the stark and obvious fact that he or she is directly dependent on a more powerful party" (Patterson 1982, 19).

In the Philippine sultanates, as elsewhere in such systems, dependency was often euphemized as the receipt of protection. An alternate term for Magindanaon "endatuan" (one who is ruled) was "sakup," meaning one who is protected. While commoners had a genuine need for protection, the search for a powerful datu protector was driven less by the benefits to be gained from dependency than by the hazards associated with freedom from it. The social context for the protection offered by datus is outlined in Charles Wilkes's 1842 depiction of the predicament of subordinates in the Sulu Sultanate:

the untitled freemen . . . are at all times the prey of the hereditary datus, even those who hold no official status. By all accounts these constitute a large proportion of the population, and it being treason for any low-born freeman to injure or maltreat a datu, the latter, who are of a haughty, overbearing, and tyrannical disposition, seldom keep themselves within bounds in their treatment of their inferiors. The consequence is that some lower class of freemen are obliged to put themselves under the protection of some particular datu, who guards them from the encroachment of others. The chief to whom they thus attach themselves is induced to treat them well in order to retain their services, and attach them to his person, that he may, in case of need, be enabled to defend himself from depredations, and the violence of his neighbors. (Wilkes 1842 quoted in Mednick 1974, 21)

Wilkes's observation indicates that some of the commoners compelled to bind themselves to a datu for their own defense might be utilized to

protect that datu against his enemies. Similar arrangements pertained in the Cotabato sultanates, according to my Magindanaon consultants. Although the officers of the armed forces of a datu or sultan were usually members of the nobility, the soldiers were principally drawn from the endatuan. In an unpublished paper, Muhammad Adil, who holds the title of "Sultan of Kutawatu," describes the traditional protective relationship this way: "Usually the nucleus of an 'inged' is the house of the ruler who in the early days was both ruler and protector but later on, when the caste system was fully developed, was only a ruler, protected and defended by the people" (Adil 1955).

Another benefit cited by W. H. Scott is the datu's redistribution of food or goods in the form of gifts or loans in times of crisis. None of my elder Magindanaon consultants could recall accounts of datus redistributing food in periods of shortage, although all could remember the details of tribute collection and some remembered stories of droughts, epidemics, or other natural calamities so severe that families were forced to sell their children.^[33]

I heard only one story, told by Hadji Abbas, of gifts provided to followers by datus, and those coincided with debts incurred to other followers: "In previous times in our inged the datu would sometimes donate the remainder of the bantingan [bridewealth] of young men who were unable to raise the entire amount. However, he never actually paid any money to the family of the bride but only promised it as a debt owed by him."

The most frequent service performed by datus, according to W. H. Scott, was adjudication and the provision of legal protection. Adjudication was the principal official function of Magindanaon datus, but, according to published sources and Magindanaon memories, it was almost always performed for immediate compensation and very rarely conducted impartially.^[34]

The Cotabato sultanates employed a set of written legal codes known as the Luwaran ("selection"). The Luwaran consisted of selections from the Shafi'i school of Islamic law combined with customary (adat ) law. Saleeby (1905), who first translated these codes into English, estimated that the Luwaran was compiled in the mid-eighteenth century. He also had the following to say about Magindanaon jurisprudence: "The Moros are not strict nor just in the execution of the law. The laws relating to murder, adultery, and inheritance are seldom strictly complied with . . . [and] Moro law is not applied equally to all classes. Great preference is shown to the datu class" (1905, 66).

Stories told me by various Magindanaon elders tend to concur with this assessment. After detailing the various customary fines assessed by a datu for murder or adultery, Hadji Abbas related with some hesitation the "real situation" in his inged in former times: "The elders [mga lukes ] told secret stories of the datu in our place. He was one of the highest officials of the Sultan of Magindanao. That datu was very harsh with the common people but he showed great leniency when passing judgment on his relatives, some of whom were notorious rustlers of water buffalo. Those water buffalo often found their way to the land of the datu. I was also told that troublesome wives brought to the datu by their husbands for judgment were sometimes raped by the datu or his men."

Datu Adil, in detailing the misdeeds of "bad" datus provided similar stories. According to him, the justice administered by these datus was inequitable and their fines oppressive. He also recalled stories of datus raping young unmarried women accused of unlawful (usually premarital) sexual activity or elopement. A young man caught thus would be fined for seduction and elopement. The punishment prescribed in the Luwaran is one hundred lashes each for the man and woman and mandatory marriage (Saleeby 1905, 71, article XL). These accounts suggest that partiality and inequity may have been the rule rather than the exception in datu adjudication.^[35]

Finally, Scott suggests that followers were able to withdraw support from datus who failed to discharge their obligations. Mednick concurs when remarking that "about the only choice a freeman seems to have had was the right to attach himself to a leader and to abandon him if he chose" (1974, 21). Other evidence indicates that even this choice was very narrowly constrained. It seems likely that the open competition among datus for followers both moderated the behavior of datus toward their followers and encouraged followers to look for better arrangements if unsatisfied with their present situation. All the same, the actual ability of followers to change leaders was probably severely limited. Both leaving one's current datu and joining another were risky undertakings because of the advantages and interests held by datus collectively. A datu could convert his follower into a debt-bondsman (ulipun) practically at will and without warning; and, because all notable datus had debt-slaves, members of the ruling class respected each other's property rights in acquired persons and tended to honor requests for the return of escaped debt-bondsmen (Ileto 1971; Gullick 1958). Followers were undoubtedly aware of these possibilities and

presumably avoided behavior (including obvious dissatisfaction with their situation) that would raise the suspicions of the datu or his lieutenants as to their constancy. That followers occasionally deserted their datus for new ones is certain. Nevertheless, switching datus was not nearly as commonplace and unproblematic as envisioned by W. H. Scott. The exchange of one leader for another probably occurred most commonly en masse, under conditions where a datu was so severely weakened that followers were certain he could neither protect them against other datus nor retaliate against them if they abandoned him.

Alternative Sources of Political Control

It is worth noting that in both historical accounts and oral traditions of precolonial politics, the most consistent theme concerns the notably autocratic and arbitrary nature of traditional rule. Dampier (1906) explains the "laziness" of seventeenth-century Magindanaons by suggesting that it ". . . seems to proceed not, so much from their natural Inclinations, as from the severity of their Prince of whom they stand in awe: For he dealing with them very arbitrarily, and taking from them what they get, this damps their Industry, so they never strive to have anything but from Hand to Mouth" (1906, 334).

Dampier provides several examples, including one of the Sultan's ploys to extort money from subjects: "Sometimes he will send to sell one thing or another that he hath to dispose of, to such whom he knows to have Money, and they must buy it, and give him his price; and if afterward he hath occasion for the same thing, he must have it if he sends for it" (1906, 342).^[36] Thomas Forrest, in his 1775 journal of his visit to Cotabato, also notes the capriciousness of the Magindanao sultan (Forrest 1969, 278).^[37] Both Dampier (1906, 370) and Forrest (1969, 289, 291) describe swift and harsh punishments for disobedient followers.

Oral traditions primarily concern Datu Utu, the nineteenth-century Sultan of Buayan. Ordinary Muslims continue to tell stories of Datu Utu as the man who reduced recalcitrant followers to "human ducks" (itik a tau ) by crushing their knees and depositing them to live in the mud beneath his house.^[38] Beckett also reports folk memories depicting Utu as a "monster" (1982, 399) and remarks that he held his dominion together by terror; yet the many examples of the severity of traditional rule suggest that Datu Utu's tyranny was exceptional only in its heinousness.

Those examples also suggest that, while datus might rely upon endatuan to defend against external threats, a local ruler's control of the means of destruction was a fundamentally important mechanism for dominating followers and suppressing dissent. Datu Adil (1955) describes political relations within a typical inged in precolonial Cotabato: "To express one's opinion unasked on any question invites not only dire consequences but almost certain condemnation or even death. One who expresses his opinion, especially if in protest against any despotic act of the ruling tyrant usually brings death upon the hapless one and slavery to his family. In extreme cases the offending subject may be publicly executed to give an example to the whole people. If however, he is a member of the ruling family or the 'barabangsa' class, he may only be banished from the kingdom."^[39]

All ruling datus maintained core groups of armed retainers. At least two paramount officeholders in precolonial Cotabato (Forrest 1969; Ileto 1971) possessed personal retinues of thirty or more armed warriors composed entirely of banyaga slaves. These troops—the Southeast Asian equivalent of Ottoman janissaries—were fed, clothed, and provided with wives by the officeholders.^[40] They formed a force of trusted and privileged soldiers whose reliability derived from their status as outsiders with no rights or relatives within Magindanaon society.

There was an additional social factor, noted by Beckett (1982), that may have significantly promoted compliance to datu rule. That was the fact that, in general, the most exploited groups in precolonial Cotabato lived, or originated from, outside Magindanaon society. These were the upland client groups such as the Tiruray and the banyaga slaves acquired from Christian settlements and the more distant highlands. Beckett (1982, 398) implies that the existence of exploitable external groups functioned as an indirect form of remuneration, whereby the commoner followers of a powerful datu were able to share in the tribute, plunder, or captives taken from outsiders. Followers undoubtedly gained some direct or indirect material benefits from the exploitation of outsiders, but the way in which that exploitation functioned to foster the compliance of commoners may have been as much psychological as material.

The evidence for the material benefits to commoners from tribute and captives seized abroad is lacking. Although commoners were legally able to own slaves there are no historical records to indicate what percentage of them actually did. Neither is it known to what extent commoners shared in plunder or tribute (although in the case of

tribute it is highly unlikely that they partook at all). Furthermore, it is important to remember that the boundary that separated commoners from chattel slaves was a status divide and not one of class. Multiple commentators on the Philippine sultanates have pointed out that the material conditions of life for banyaga slaves were not appreciably worse, and were in some cases better (the banyaga bodyguards of the sultan are a case in point), than for the average commoner (Ileto 1971; Mednick 1974; Warren 1981). The presence of disdained aliens may have worked to sustain the stratification system largely through its psychological effect on subordinates, who were inclined to draw the most meaningful social dividing line below rather than above themselves and identify with insider Muslims as opposed to outsider pagans and Christians.

Summary: The Bases of Traditional Rule

In synthesis, there existed in the Cotabato sultanates, from at least the mid-sixteenth century, a distinct ruling class that, although often absorbed in internal status competition, was nevertheless integrated by a well-developed ideology of nobility—one that entitled its members to exercise authority over, and extract tribute from, the Cotabato populace.

The fundamental basis of power of datus was their control over subordinates, both those legally free as well as unfree ones. Individual datus strove to acquire the greatest possible number of endatuan, ulipun, banyaga, and client Tiruray from whom to extract surplus directly. On this foundation they were able to build impressive personal retinues of armed and unarmed dependents. An ambitious datu with a large force of followers could further expand the scope of his power and wealth by subduing or sufficiently impressing lesser datus so as to convert them into vassals; by leading or financing large-scale external raiding and trading expeditions; or by entering into strategic alliances with other powerful datus in order to influence state-level or even regional politics.

What was the social cement that kept the Magindanaon political order intact? It was undoubtedly an amalgam of armed force, material remuneration, and cultural commitment. Datus effectively controlled armed violence, and the use or threat of physical coercion was a consequential source of social control in the Cotabato sultanates, even though it is the least acknowledged in written indigenous records. The

precolonial lords of Cotabato were, on the whole, capricious and, in some cases, ruthless autocrats. Domination was exercised directly and overtly; and if it was not openly acknowledged by subordinates, or only expressed metaphorically (with atrocious actions denoted by ogreish appearance), it was because obvious dissatisfaction would likely have invited additional repression.

Remuneration was also a significant source of political compliance in the Cotabato sultanates. However it did not play the key integrative role envisioned by W. H. Scott (1982) and others. Goods controlled by a datu were very unevenly distributed, with those closest members of his retinue receiving by far the greatest share. Neither were relations of redistribution between traditional leaders and followers simply a premodern version of the political clientelism found in the modern Philippine state. While followers were as essential to precolonial datus as they are to present-day politicians, datus had more means at their disposal to acquire clients and forcibly retain them once they had them. Subordinates were unable to attach themselves to more than one datu at a time and their ability to change datus was severely limited. In addition, despite the claim for such in the ruling ideology, the provision of legal and armed protection by datus was mostly illusory.

Finally, cultural commitment was an important factor for sociopolitical cohesion in precolonial Cotabato. Commoners (endatuan) and even debt-slaves (ulipun) presumably identified themselves, if only nominally, as Muslims. They felt themselves to be insiders—members of the dominant sociocultural category in the Cotabato Basin—and thus superior to Christian or pagan banyaga slaves or Tiruray clients. They were attuned to the symbolic role of the sultan and acquainted with the rudiments of the ideology of nobility. Evidence suggests, however, that ruling ideas were not shared nearly as well, nor as completely, by Cotabato subordinates as is imagined by those anthropologists who have been principally concerned with political culture in precolonial polities in Southeast Asia.

There is evidence also for the existence of independent perceptions and representations of the social order by Cotabato subordinates. In this unofficial transcript, power relations were not naturalized as the product of a self-evident hierarchical order. Instead, social power was denaturalized by according it a supernatural essence. Such images portrayed a profoundly unequal distribution of social force, vividly illustrating the belief, based on direct experience, in the power of certain rulers "to prevail in any encounter" (J. Scott 1990, 73). While such

beliefs may have had considerable ideological effect, they bear scant resemblance to the products of the hegemonic process posited by most analysts of cultural domination. In particular, subordinate images of datu rule were applied not only independently, but selectively, as only some ruling datus were endowed by Muslim subordinates with supernatural attributes. Moreover, as we shall see for contemporary Cotabato, such endogenous symbolizations of power also contain a large measure of counterhegemonic potential in that what has been endowed independently may be unilaterally withdrawn.