PART THREE—
BOGANG AND REMBAU IN THE POSTINDEPENDENCE ERA

5—
Agricultural Production and Income

In Rembau and the rest of Negeri Sembilan, the first half of the twentieth century witnessed an increasing household dependence on cash-cropping activities and the labor inputs of males. This chapter delineates the acceleration of such trends following Malaya's independence in 1957. It commences with an overview of economic and ecological change over the past few decades, and then proceeds to an assessment of transition and decline in rice cultivation. Analyses of cash cropping and of the progressive centrality of money in intravillage economic relationships follow. The final section addresses themes of economic stratification, including the factors working against the cultural realization of stratification and other features of twentieth-century social process.

Economic and Ecological Change:
An Overview

Since independence Rembau and Negeri Sembilan have experienced an intensified reliance on cash income and production for the world market, and a corresponding decline in the importance of subsistence activities

such as freshwater fishing and wet-rice cultivation. The general direction of these shifts was fairly well established as early as the 1920s, and discernible even before the turn of the century. Nonetheless, postcolonial history has fostered a nexus of economic dependence—most notably on monocropping, Chinese middlemen, and the vagaries of natural and world-market forces—that appears qualitatively different from that which prevailed under British administration. Additionally, the postcolonial era has seen the demise of traditional institutions based on reciprocal exchange and mutual self-help and their replacement by inflexible sharecropping and rental arrangements of a predominantly instrumental and self-interested sort.

Two mutually reinforcing developments are particularly relevant to these phenomena at the outset. The first involved the large-scale clearing of catchment areas and other forest lands deemed suitable for rubber cultivation by rural producers, and Malay smallholders in particular. Beginning immediately after Malaya achieved independence in 1957, this practice effected a radical break with colonial-era restriction schemes, which had all but precluded the purchase of commercially valuable state lands by Malay villagers. The subsequent alienation of lands in accordance with the new policy meant that many Malays who had previously been sharecroppers or landless in terms of rubber acreage were on their way to becoming smallholders in their own right. Less desirable, however, were the unforeseen ecological consequences of these massive government-sponsored deforestation projects. Rivers silted and apparent climatic shifts occurred, interfering with the irrigation requirements of wet rice and eventually resulting in greatly decreased yields. As a consequence of this unintended trade-off, rice cultivation over the past decade especially has been only nominally worthwhile in terms of yields relative to investments of time, labor, and capital.

The second development worked in much the same direction and centered on the introduction, beginning in the mid 1960s, of quick-maturing and high-yielding varieties of rice. The planting of these strains was at first optional, but it often came to be mandatory in the sense that the cultivation of traditional rice varieties was soon discouraged and ultimately prohibited by extension officers and other state agents involved in local-level agriculture. The shift to new strains heightened villagers' dependence on cash incomes and local as well as international market prices, for the successful cultivation of these hybrids has always necessitated relatively expen-

sive modern chemical inputs (commercial fertilizers, herbicides, and pesticides). The widespread use of these chemicals also killed off the readily available supplies of edible fish once found in flooded rice fields and the lower stretches of streams. Fish, which along with rice has always been the mainstay of the Malay diet, thus had to be purchased daily. This chain of developments therefore rendered income from the tapping of rubber trees virtually essential for household maintenance.^[1]

Analyzing just how each set of changes independently affected the local ecosystem and economy is problematic. Because the introduction and rapid spread of high-yield rice followed quickly on the heels of large-scale deforestation, for example, the individual significance of either event is difficult to assess. Villagers, moreover, tend not to include either of these variables in their interpretations of declining rice yields, progressively insurmountable irrigation problems, and apparent climatic changes; rather, they explain these shifts by reference to the abandonment, since about 1965, of an elaborate ritual complex known as berpuar , which was traditionally overseen by shamanic specialists (pawang ). The accounts of local historians are thus not especially helpful in this instance. The same holds for villagers' characteristically vague assessments of when the decline in local rice productivity began. Even so, there is justification, on both analytic and sequential grounds, for examining these two developments in turn, and for treating them as separate and distinct (though interdependent) phenomena.

Large-scale deforestation commenced in the independence year of 1957 and continues still today. In the village of Bogang in 1962 alone, for example, roughly seventy-two acres of land lying upstream from homestead areas and rice fields was cleared of all growth, parceled into twenty-two plots, distributed to about as many owners, and planted in rubber. Since that time numerous tracts of land in the same upriver area have been earmarked for residential purposes or banana cultivation and stripped of all vegetation. In consequence, and owing in particular to the destruction of subterranean root structures, a good deal of catchment area topsoil has been lost in the course of the past two decades' rains. Most of this soil eventually washed into the single riverine network watering the village's sawah, thus contributing both to the silting up of streams and to their periodically overflowing their banks. As a result, decreased amounts of water were channeled through irrigation canals and into the fields. In further re-

stricting rice cultivators' tenuous but critically important control of irrigated water, all such developments diminished the effectiveness of the hydraulic system and the returns on labor and capital inputs. The immediate outcome was greatly reduced yields. The larger trend, as we shall see, has been for households to work fewer holdings and smaller plots and, frequently, to abandon rice production altogether. Such trends are by no means unique to Bogang but exist in other parts of Rembau and elsewhere in Negeri Sembilan as well (Kahn 1981, 545, 547, 554; Stivens 1985, 19 et passim).

It is also possible that climatic change followed deforestation, as a result of the heightened exposure of soil and ground moisture and an attendant acceleration of evaporation rates. Experienced cultivators in Bogang hold that annual rainfall has fallen dramatically in the past decade or so relative to the years before independence. Composite figures of monthly precipitation averages over multiyear periods, however, do not bear out this contention. But neither do they disprove it, insofar as the available data do not allow for fine distinctions within these periods and could easily conceal a trend toward a precipitation decrease in the last ten or twenty years. Also, as the figures are for rainfall measured some ten miles from Bogang, they might fail to reveal contemporary variation and slight historical shifts in both the timing and the overall amount of precipitation for different micro-environmental zones.

In any event, topsoil erosion and leaching, along with the silting of streams and possible shifts in the volume and periodicity of local precipitation, have interfered with the irrigation requirements of wet rice. This fact is quite apparent in Bogang, where the network of streams serving the community carries far less water than it once did. During most of the sixteen months of my fieldwork, for instance, the Bogang River was a mere trickle, rarely exceeding three to four feet from bank to bank or two feet in depth. Only after torrential rains lasting at least a few hours did its volume increase. Village elders repeatedly lamented the stark contrast with the situation "of long ago," or even a generation or two back, when much of the river was commonly over seven feet deep.

Bogang elders are not the only people aware of recent water flow problems. For good reason, issues of hydraulic control and overall water supply have been of major interest to government planners as well ever since independence. Official concern with the effects of large-scale deforestation on

irrigation was no doubt instrumental in prompting the state to upgrade and expand government-sponsored drainage and irrigation services throughout Negeri Sembilan during the 1960s. In any case, by 1966 the government had embarked on an ambitious scheme of replacing the traditional brushwood and earthen waterworks with concrete dams and canals so as to improve yields, reclaim acreage in swamp lands, and otherwise expand the area suitable for rice cultivation. By the following year, and in pursuit of these same objectives, several thousand acres of Negeri Sembilan sawah had been given over to the double-cropping of rice on a trial basis. These experiments produced impressive yields and foreshadowed the rapid spread of double-cropping throughout the state.

For the residents of Bogang, the 1970–1971 period was historic on two accounts: quick-maturing, high-yield varieties of rice were introduced into the village during those years; and in 1970 a government-financed project providing irrigation facilities for approximately eighty acres of sawah was completed. At an officially estimated cost of over M$45,000, the project centered on the construction of twin-barrel culvert headworks along with about 4,700 feet of concrete-lined irrigation canals to replace ditches of traditional design (ARDIDNS 1970, 6). The dam itself was built upriver from all village residences and rice fields, but downstream from large areas of rubber acreage and other cleared land that had been alienated since the late 1950s. Whatever the initial effects of the new irrigation works on the flow of water into the fields, many of Bogang's elderly felt that the dam's location was ill chosen and that the diversion of stretches of stream leading from the dam site was a grave mistake. Indeed, more than a few elders feel that the project fell somewhere between a highly qualified success and a wholesale squandering of precious capital and labor, thanks to the dam's location.

Apart from local appraisals of the consequences and ultimate value of the project, no measures, such as reliable figures for consecutive annual yields, are available by which to judge its impact. Nonetheless, villagers assert that less water reached their fields subsequent to, and as a direct result of, the project's completion, and that per-acre yields over the past decade have declined 30 to 40 percent for some village holdings, and as much as 80 to 90 percent for others. The dam's placement, along with village attitudes that concrete canals do not require much maintenance and repair, could be crucial factors here. Western-style dams, too, unlike their

traditional Malay counterparts, are not of the permeable filter-dam variety that allows trickles of water to pass through and over the structure; rather, they stop water completely, thereby allowing silt to build up appreciably. The sudden release of heavily silted water in the course of sluice openings and storm-fed overflows can damage streambeds, cause a deepening of local riverine valleys, and lead to a progressive lowering of water tables in irrigated fields (Dobby 1950, 108). This situation could easily lead to crops depending more heavily on moisture provided by irrigation and would also support the local view that less water reaches the fields nowadays as compared with the 1960s and before.

Just as the new hydraulic facilities were completed, quick-maturing rice was introduced into the village, with double-cropping being initiated in an area of roughly sixty acres. The problems of meeting the different and more exacting irrigation needs of short-season, high-yield rice varieties may well figure in villagers' conviction that less water has reached their fields since the dam was constructed. Be that as it may, double-cropping emerges as the more critical of the recent shifts in rice agriculture and village economics, and the one that merits greater scrutiny.

Transition and Decline in Rice Agriculture

The Scheduling and Organization of Traditional Rice Production

An assessment of agricultural change and the requirements of double-cropping requires first of all a brief discussion of the characteristics of rice cultivation on an annual basis. Traditional forms of this agriculture involved varieties of rice with a maturation period of approximately six months. This single factor controlled much of the timing and organization of agricultural work and a good deal of social and religious activity as well. Of comparable significance were the broad givens of equatorial climatic cycles, especially the bimodal clustering of annual rainfall during two distinct monsoon seasons separated by hot, relatively dry spells of a few weeks' duration. Moreover, the village majority had only limited technological means and stores of capital at their disposal. Ox-drawn plows, though never all that common in Negeri Sembilan (R. Hill 1977, 121, 130, 133; cf. Dobby 1957, 96–97), were probably on the decline throughout most of the twentieth century, and certainly long before mechanized tillers

and four-wheel tractors were widely available to replace them. Even assuming some access to oxen, however, household cultivators relied largely on their own energy and that provided by fellow villagers on a mutual self-help basis, and on the simple hoes and harvesting implements they fashioned themselves or purchased locally. This approach to agriculture was highly labor-intensive and quite demanding as regards coordinating water and labor resources, but it had the advantages of presupposing minimal capital inputs and distributing household risks in production.

With some variation even among neighboring villages, preparations for padi cultivation in Rembau commonly began in mid June (see figure 4). The initial stages of work centered on the burning of rice thresh and weeds in the fields and the reconstruction of local dams and sections of irrigation canal that had been damaged by rain, the off-season pasturing of animals, and water-control activities associated with the previous harvest (R. Hill 1977, 41, 133). In Rembau and elsewhere in Negeri Sembilan, repairs both to dams and to the upper stretches of the village's irrigation system were held to be a communal responsibility, and each cultivating household was obliged to furnish one worker for these tasks (Dobby 1957, 93; Swift 1965, 40).^[2] The renovation of waterways adjacent to and running through sawah plots, in contrast, was shared among those with holdings most directly affected by the flow of water in that immediate area; much of this work proceeded on an individual or household basis.

In July, following the month or so of intermittent work required to repair the local irrigation system, villagers flooded the sawah by opening all dams, and constructed nurseries either in well-watered sections of their fields or on swampy ground adjacent to the planting area. Each nursery bed was prepared and tended either individually or cooperatively by several cultivators working proximate plots. Owing to the sexual division of labor, partible inheritance, and the spatial aggregation of lineage holdings, these task groups tended to be made up of women related as sisters or simply female enates of the same lineage branch or lineage. Both in terms of siblingship and other kinship norms and by virtue of shared responsibility toward the upkeep of common nurseries, cooperation among these women was of cardinal importance, especially during the most labor-intensive stages of the cultivation process (early in the season and at harvest time). Prior to transplanting, for example, cooperation consisted of reciprocal labor exchanges (tolong-menolong ) to ensure that each cultivator

^[

Figure 4.
The Agricultural Cycle in Rembau Under Single- and Double-Cropping Regimes
Note: Data on rainfall courtesy of Jabatan Pertanian, Negeri Sembilan.

successfully completed all work associated with plowing and harrowing, leveling, and diking plots. The extent to which this work sharing was essentially ad hoc or was instead planned and publicized well in advance is unclear; the only major difference either way would have involved, in the case of the larger and more formalized arrangements (or seraya ), a greater obligation to feed helpers. In any event, the transplanting itself, which was carried out in August some forty to fifty days after seed was sowed in the nurseries, drew on the labor of these same women (although nonenatic kin and neighbors might provide assistance here and in all other stages of cultivation).

The completion of transplanting marked the beginning of a five-month period when cultivators concentrated on weeding and, during the final four to six weeks before the harvest especially, pest control (birds, field mice, and the like). These tasks were probably conducted primarily on an individual or household basis; even so, at least some coordinated effort and collective responsibility among individuals operating contiguous plots was essential. If plots with ripening grain were not tended for a few days, for example, birds that flocked to them in the cool early morning and late afternoon hours in search of grain could cause severe damage to even the most carefully tended holdings nearby. More generally, there seems to be an inverse correlation between the number and maximally efficient dispersal of people in a given area of the fields during the month or so leading up to the harvest, and the amount of damage birds cause to the grain in those fields. The larger issue is the consistently positive and close association between the volume of labor inputs per land unit of sawah and the productivity level of that unit (Geertz 1963).

Toward February, villagers drained their fields by breaking the dams and strategic sections of their canals, and began the harvest. The principal harvesting implement was the tuai , a small knife that could be concealed in the palm of the hand so as to avoid frightening, offending, and thus giving flight to the life force of the grain (semangat padi ). Harvesting with a tuai was highly labor-intensive, since it involved cutting stalks one at a time. This procedure permitted villagers to choose between fully ripened stalks ready to be cut down and those requiring additional sun and maturation. It also contributed to the drawing out of the reaping process, allowing for greater flexibility in the timing and application of labor and, in turn, facili-

tating labor exchanges. Cultivators were thus intermittently freed from the harvesting requirements of their own fields and became available to assist kinsmen and neighbors on an immediate and short-term basis.

Following the harvest of all village grain by late February or early March, for roughly three and a half months villagers ceased all work directly related to padi cultivation and oriented their activities to somewhat separate concerns and objectives. Especially during the early part of that period attention shifted to the threshing, winnowing, and husking of rice and other preparations for long-term storage in household granaries. Other economic activities were also pursued, but these were not so much reinitiated then as simply given higher priority. This situation contrasts with the timing of rituals and feasts associated with marriage, which tended to be held during the off-season months following the harvest, when households were most likely to have large stores of rice and the means to feed guests and help underwrite the other expenses of wedding and affinal exchange. The clustering of weddings in the agricultural off-season also precluded conflicts in the scheduling and organization of the labor required for cultivation on the one hand and for a successful wedding on the other.

The Switch to High-Yield Variety Rice

Many of the recent changes and problems in local rice production stem from the shorter maturation period of high-yield variety rice—roughly half that of traditional strains. Thus, whereas before 1970 the period from transplanting to harvest covered approximately 180 days, the 1970s saw the same agricultural phase reduced to roughly 90 days (see figure 4). More broadly, the past ten years witnessed the reduction of the production cycle from nine to roughly five and a half months so as to accommodate a repeated semiannual cycle. This change amounted to the elimination of the three-and-a-half-month off-season from the village agricultural calendar and its replacement by two postharvest slack periods, each lasting only three or so weeks at the most. Viewed from February, a harvest month under both single- and double-cropping regimes, the bulk of March was now taken up with waterwork repairs and other preparations for off-season cultivation, just as the coming of April signaled the completion of all field diking along with the construction of nurseries and the sowing of

seed therein. Similarly, the six-week period from early May to mid June was given over to transplanting and related tasks, even though it too had once been part of the off-season. Other changes of this sort can be found in figure 4; I might emphasize, however, that the new agricultural cycle made scant provision for the timing of village weddings. One result is that conflicts between the demands of cultivation and labor obligations to kinsmen and fellow villagers sponsoring weddings have often been resolved in favor of the latter, potentially causing appreciable delay in one or another phase of cultivation and thus working against the attainment of maximum yields.

A striking feature of the shift from traditional to short-season rice is in fact the reduction of cultivators' flexibility in deciding when to initiate or terminate the various phases of the production process. Perhaps most significant is the timing of the postharvest renovation of local waterworks. Under double-cropping regimes, canal and dam repairs must be undertaken almost immediately after the previous season's harvest. Further, owing to the extreme sensitivity of high-yield rice to water level fluctuations in the flooded fields, these repairs and all subsequent canal maintenance must be performed with greater precision and coordination of labor than is necessary with traditional single-cropping. The timing and extent of these labor inputs become all the more critical given the extensive deforestation, topsoil erosion, and leaching of the past two decades, and the related fact that in recent monsoon seasons water levels in the government-built canals have been insufficient for the irrigation requirements of the new rice varieties. Even if we assume enough rainfall and water in the canals, however, critical problems remain: to wit, labor shortages generated by accelerated out-migration and near-universal school attendance appear to have undermined cultivators' ability to maintain the stretches of embankments and earthen canals on which they also depend both individually and collectively. Difficulties in mobilizing and effectively coordinating communal labor are also important here, although problems of this sort, and of village leadership on the whole, predate double-cropping (Swift 1965). Since the organizational and technological requirements of rice agriculture have clearly shifted over the past decade, the limitations of local leadership are more critical now than in the years preceding the switch to "miracle rice." Again we come up against the question of timing, for the inability of the political elite to enlist support for the cleaning and repair of

dams and canals can delay the production cycle by days or weeks, thus causing it to be out of sync with the seasonal pattern of rainfall.

The adoption of high-yield strains demanded additional capital investments during the initial stages of the agricultural cycle, but it also reduced some of the most physically demanding tasks associated with cultivation. The onerous work of uprooting weeds throughout the maturation period of the transplanted padi, for instance, was decreased significantly thanks to commercial herbicides recommended by agricultural extension officers. Like the pesticides and petrochemical-based fertilizers introduced at the same time, these herbicides were, and continue to be, essential components of the double-cropping packet.^[3] Indeed, the successful cultivation of high-yield strains fully requires chemical inputs of this sort, regardless of whether planting takes place every six months, once a year, or less frequently. These inputs presuppose cash expenditures of approximately thirty ringgit per acre or government subsidization or other direct assistance. Although such aid has been frequently provided during the past decade, it has not always been guaranteed in advance or forthcoming when most needed—and without government assistance, many cultivators cannot afford to purchase the necessary inputs.

The use of commercial pesticides necessitates in addition cash outlays for fish to satisfy household food requirements, since the active ingredients of these chemical compounds kill off most riverine life in the flooded fields. The recent destruction of sawah fish, moreover, is not confined to the holdings actually laced with pesticides, for the chemicals are water borne and so end up in virtually all parcels of land watered by the local canal system. Households without the means to buy modern pesticides thus lose doubly: not only do they end up with less than optimal yields, but they also cannot catch fish.

The shift to high-yield rice also corresponded with—although it did not directly cause—a significant decline in the labor investments required for the early stages of field preparation. I refer here to the increasingly widespread use of hand-held mechanical tillers and four-wheel tractors for plowing, in lieu of the traction provided by human labor and, in some cases, oxen. Since these machines can plow in a single day or less the same acre that it takes a person twenty-one days to turn over,^[4] they offer a distinct advantage in labor-short villages like Bogang. Here too, however, cash expenditures are unavoidable unless the government provides a sub-

sidy, for a mechanized plow or tractor costs some M$41 to $49 per acre to rent. Some villagers can easily manage expenses of this scale; others cannot, especially since the total cash required for preparing and maintaining an acre of sawah can amount to M$80 or $90. Many Bogang residents whose monthly household incomes hover around or fall below the village median of M$240 (see table 9, p. 187) consider such expenses prohibitive.

Of far broader concern is overdependence on government assistance. In Bogang, for instance, the government absorbed much of the cost of tractor rental for the 1978–1979 planting season, and most households engaged in cultivation benefited directly from the subsidy. Twice-yearly planting (but not high-yield rice) having been abandoned in Bogang by the mid 1970s, the next planting season should have commenced in September 1979, yet it was delayed a full six months. According to many community residents, the season was effectively canceled because the government reneged on its promise to deliver free plowing services to all interested cultivators. More likely than not, the following season, which began in March 1980, would have seen all village fields still lying fallow were it not for the government's decision once again to underwrite the costs of mechanized plowing. As it turned out, only 110 of the 215 acres of sawah in and around Bogang—a mere 51 percent—were turned over by the tractors sent there in mid March. Decisions concerning the extent of the acreage and the particular fields to be plowed derived from the village headman's appraisal of different households' current commitments to cultivation. He later acknowledged that the area actually planted would be closer to 50 acres than the 110 acres he once thought. But even this figure was optimistic, for by the time I left the village in mid May, it appeared that only 20 to 30 percent of the plowed land (that is, some 10 to 15 percent of the village total) would be worked that season.

Other recent transitions in the production process can be seen in the harvest and in the subsequent husking of rice. Such shifts were not a necessary concomitant of the adoption of new rice strains suitable for double-cropping. Nevertheless, they were part of the more encompassing trends toward greater efficiency in the application of human labor on the one hand and mechanization cum monetization on the other.

The first of these shifts involved the replacement of finger-knives (tuai ) with sickles. A few elderly cultivators still prefer tuai, of course, and they contend that yields are invariably superior—and the cooked grain much

tastier—if the traditional tool is utilized. Underlying this view is the belief that the life force of the padi is less frightened and offended by a delicate tuai cupped in the palm of the hand than by the awesome and painful sight of a large, gleaming sickle slicing the air and grain. Such beliefs have much greater currency than cultivators' choices concerning harvesting implements would lead one to assume. Many who opt for the labor-saving sickle, moreover, concede its inappropriateness on aesthetic and ritual grounds, allowing as well that the essential life force of the grain has long since fled downriver to the Straits of Malacca (cf. Lewis 1976, 83). Yet cultural considerations bearing on many domains of nature and spirits have been given the backseat of late, and they merely shrug, by way of explanation, that "Malays have become modern."

The labor requirements of working with tuai as opposed to sickles clearly differ, and without question the use of sickles in lieu of tuai has significantly decreased the extent to which cultivators seek extrahousehold assistance during the harvest. This fact is critical, for the harvest season constitutes one of the peak labor periods of the entire agricultural cycle, and the one in which households typically rely most heavily on the labor of other villagers. A massive reduction in the assistance needed during this stage of the cycle, in other words, is tantamount to an equally pronounced decline in the extrahousehold labor needed in rice agriculture in general. Thus, while the adoption of sickles no doubt resulted in a compressed harvest period and possibly larger yields, it also cut into labor exchanges significantly. Stoler's (1977, 88) data from Java indicate that the replacement of tuai with sickles can reduce the number of harvesters needed by as much as 60 percent, and I would expect a similar impact in Negeri Sembilan. Here, then, the large-scale decline in labor exchanges involving women is not simply a function of decreased padi production, but also a direct consequence of technological change.

The other recent shift in the production process appears in the displacement of traditional husking methods by machine milling, thus obviating the need for sisters and their close female kin and neighbors to work together in this endeavor. The customary postharvest preparation of grain involved the threshing of all padi obtained from the fields, followed by winnowing with shallow woven baskets (nairu ). These tasks completed, women stored their unhusked rice in household granaries a few yards from the house. Because long-season grain could be stored several years (unlike

"miracle rice," which "goes bad" after five or six months owing to discoloration and its greater susceptibility to weevils), household heads typically drew from the granary every few weeks however much padi the household would consume in that period of time. The grain was then spread out on woven mats in a sunny part of the compound for a few hours and dried. In the Inas area of the Kuala Pilah district adjoining Rembau, the husking itself was a two-stage affair entailing, first, the use of home-made milling machines (kisaran ) run by hand and, second, a foot-operated version of a mortar and pestle (lesong ) to complete the process (Lewis 1962, 264). This could have been the case in Rembau as well, although I never encountered kisaran and know only of the use of lesong.

Traditional milling occurred throughout much of the year, involving considerable informal labor exchange among the women of adjacent households. As a general rule, women seem always to have preferred working together rather than individually, particularly if the tasks at hand were strenuous, time-consuming, or easily distributed among two or more women laboring simultaneously or in turn.

The women assisting in the milling might receive tea and sweets or cakes from the household head, and they could of course count on reciprocal help when it came time to prepare their own grain later. As with many other instances of work sharing and labor exchange among women, these were as much social occasions as anything else, providing women with an additional opportunity to catch up on current gossip and news of village residents and local events. These work exchanges also met some of the women's needs for companionship and interaction, since women (and men) have apparently always felt that they should refrain from traveling about the village unless required to do so for reasons of domestic maintenance, production, exchange, or the realization of religious value.

The rhythmic and hypnotic thumpings of lesong are rarely heard in contemporary communities like Bogang. Today women prefer to patronize the truck-mounted mechanical mill that appears in the village on a regular (usually weekly) basis. Indeed, on most Tuesday mornings in Bogang anywhere from a handful to a dozen or more women—and at least as many bulging sacks of padi—gather next to the railroad tracks waiting for the miller (whose services are available at a government-controlled price of two cents per kati ). Mechanical milling would be in much greater demand if the satisfaction of household food requirements

relied more heavily on rice grown within the village. As it is, 57 percent (57 of 100) of the Bogang households for which I have data on the subject^[5] buy at least three-quarters of their staple rice, and a full 36 percent (36 of 100) rely entirely on store-bought rice.^[6] The bulk of this grain is purchased in nearby towns, and it is invariably husked and milled by the time it enters the community.

A few issues of broader significance require comment here. First, far less rice acreage is planted now than at any period in Bogang's history, as indicated by the effective cancellation of Bogang's fall 1979 planting and the mere 10 to 15 percent of village sawah that was planted during the following season. Second, the technological and organizational entailments of modern-day rice production are such that, even when land is worked, there is relatively little need for extrahousehold assistance and reciprocal labor exchange. These developments bear most heavily on women and have resulted in circumstances in which subsistence cooperation and exchange among sisters, female enates, and women in general are not only close to extinction but have also been superseded by rental (sharecropping) arrangements, which nowadays constitute the dominant links among households involved in padi production. We shall consider these points in turn.

The most dramatic trend in rice cultivation in Bogang over the past few decades is not the shift to new strains of quick-maturing grain but rather the sharp reduction in the total acreage of sawah worked on a regular basis. In the colonial period, villagers who did not plant all of their sawah were liable to receive summonses and fines from the District Office; this sanction, in conjunction with the high moral valuation of padi growing, was apparently quite effective in ensuring the annual cultivation of virtually all rice fields registered in the names of community residents. With the abandonment of this disincentive following independence, however, villagers had considerable choice regarding the extent of their holdings to be planted in a given season. This policy change might have made little difference in the amount of acreage a given household or community decided to work if yields averaged or exceeded some three hundred gantang per acre. Yet in the context of consistently low or radically declining yields, one would expect cultivators to respond to this shift in policy either by exploiting only the more productive of their holdings or, alternatively, by aban-

doning or renting out their own sawah and seeking work as sharecroppers on better land. In Bogang, where yields have fallen off drastically over the past two decades, such trends have been quite pronounced. Of all the households claiming rights over sawah, only 33.3 percent (30 of 90) planted their entire holdings in the 1978–1979 season.

The tendency of households with rights over sawah to work fewer holdings and less acreage appears to have coincided with a proliferation of tenancy arrangements in the form of sharecropping. This development stems in part from the presence within the village of fourteen households whose female members lack rights of proprietorship over sawah and have little if any chance of inheriting such rights in the future. As might be expected, most of these households are headed by women born outside of Bogang who moved to the village during the past twenty or so years. These women were not adopted into a local clan; some even hail from states other than Negeri Sembilan and so claim no formal descent affiliation whatsoever. These women are not only highly unlikely to inherit any sawah but also are effectively barred from purchasing any alienated sawah land, since virtually all of it is considered ancestral clan property, both in the eyes of the community and according to the titles held at the District Office. In decades past landlessness of the sort that has forced these women to labor as tenants was relatively rare, and in any case temporary, lasting only a generation or so. Until fairly recently women of extralocal birth could count on formal acceptance into one or another clan, or adoption by an adult woman in the village, either of which usually endowed them with usufruct and future inheritance rights over sawah and residential acreage. If subsistence guarantees of this nature did not prelude tenancy altogether, they certainly kept it in check. So, too, would a situation in which most households maintained all of their own sawah, albeit with periodic assistance from neighbors and kin, for any landless households that did exist would have relatively few opportunities to negotiate full-time sharecropping arrangements.

At present, in contrast, full-time tenancy is extremely widespread. Of all villagers engaged in season-long work during 1978–1979, a full 33 percent (30 of 91) were sharecroppers. The majority, or 53 percent (16 of 30), of these sharecroppers chose to work as tenants on relatively productive land rather than depend on their own marginal holdings.

Sharecropping arrangements involving sawah tend to be quite flexible and generally favorable to the tenants, particularly in comparison with those involving rubber. The latter are rigid and uniform, and regardless of the genealogical or social relationship between tenant and proprietor, tenants are invariably expected to provide owners with 50 percent of the income from their labor. With sawah, however, full-time tenants often retain well over half the produce from their efforts. Some proprietors encourage their tenants to keep as much of the grain as they see fit, or even all of it, and the workers do just that. Moreover, although genealogical proximity is likely to render the terms of sharecropping arrangements more favorable to the tenant, the converse is not always true. I know of a few instances where unrelated and socially distant tenants—some resident in Bogang, others in adjacent communities—were instructed to, and did, keep the entire yield from the acreage they tended.

At the same time, partners to sharecropping agreements do seem to prefer their relationships to be mediated by kinship. Of the forty-two cases of sharecropping in which Bogang residents cultivated local sawah, twelve persons (28.6 percent) labored on land held by a member of their own lineage or clan. Seven instances (16.7 percent) involved relationships in which tenants and owners were linked through affinal ties (that is, by virtue of one having a current or former spouse in the other's lineage or clan). In an additional five cases (11.9 percent) kin were connected in various other ways. While the eighteen remaining relationships (42.9 percent of the total) involved nonkin, many of these tenants were women without clan affiliation, locally resident collateral kin, or both. Owing to the circumstances of their birth or postmarital residence, individuals in this latter category are largely unable to avoid sharecropping arrangements with "strangers."

Even where demonstrable or putative kinship does not underlie the bonds linking tenants and owners, their relationships tend to be couched in kinship idioms and informed by the spirit of mutual aid and cooperation. One reason for this closeness is that virtually all local sawah is classified as ancestral property. Lands in this category are seen to exist because of the collective efforts of long-forgotten ancestors, who endured inordinate sacrifice for both their contemporaries and their descendants; the spirits of these exalted figures would certainly take offense at unkinsman-

like behavior concerning the use of ancestral property and the enjoyment of its benefits.

Yet in the final analysis, sharecropping relationships are forms of land rental, and villagers do occasionally use the term for "rent" (sewa ) to describe these arrangements. Idioms drawn from kinship, village citizenship, and modern commerce thus seem to be juxtaposed in a fairly comfortable alliance so far as most people are concerned. Nonetheless, a marked change in the relative significance of these idioms and their behavioral correlates is also evident. To appreciate this fact we need only consider that until fairly recently the institution of seraya provided villagers, especially political elite and ritual specialists, with the labor required to prepare fields and plant padi, just as assistance in the form of tolong-menolong enabled most cultivators to satisfy the bulk of their other extrahousehold manpower needs. As noted by Swift (1965, 48), the polysemic term seraya conveys meanings of "ordering" and "helping," in the sense of calling in one's debts and honoring them, respectively. It refers in particular to the custom whereby villagers formally announced both their readiness to begin production and their desire to organize a labor force for the agricultural tasks at hand. Participation in these endeavors was more or less obligatory for all who had benefited from the cultivators' ritual or political services during the year. Equally binding was the feeding of all workers by the proprietors of the land under preparation. All this was to change, however, when cash entered the local economy and debts began to be repaid with money rather than labor. Specifically, the combination of seraya and tolong-menolong gave way to a dependence on tolong-menolong and rental (sharecropping) arrangements, supplemented when necessary by short-term labor hire (Swift 1965, 48–49). More recently, this latter combination has been superseded by sharecropping arrangements alone, which are now the most pronounced of all interhousehold labor and economic relationships in padi growing.

Rubber, Cash, and Tenancy in the Village Economy

The majority of Bogang's households buy most of their rice, as well as virtually all of the fish, vegetables, tobacco, and other items essential to domestic maintenance such as condensed milk, tea, kerosene, cooking

utensils, clothing, and children's school supplies. All these purchases presuppose cash outlays. Furthermore, such items as mass-produced chairs, cupboards, kerosene stoves, and battery-operated radios have come to be regarded as basic furnishings and are also conspicuously present throughout the village. The same is true of household conveniences such as piped water and electricity. Twelve motorcycles and eight automobiles in the village—all privately owned—round out this list. The importance of cash in the local economy as a whole, and in terms of household living standards and prestige in particular, should be clear.

In these conditions, owning a moderately sized stand of rubber trees is of great concern to virtually all households in Bogang. A well-tended and high-yielding plot of three to four acres can net a conscientious tapper up to M$450 a month,^[7] assuming, of course, that the worker holds rights of proprietorship over the land. But while nearly all of Bogang's households rely rather heavily on earnings from rubber trees tapped either by their own members or by sharecroppers, only 51.9 percent (55 of 106) actually own rubber acreage. Tenancy involving rubber is thus extremely prevalent and of far greater economic significance than sawah sharecropping.

The very nature of tenancy in rubber tapping on the one side and rice cultivation on the other contrasts markedly. The idiom of rental is dominant in the former instance, with rigid and uniform terms of tenancy arrangements regardless of genealogical or other proximity between the principal parties. Moreover, households without rubber acreage are not necessarily headed by women who lack membership and kin in one of the local descent groups. As a general rule—and again in sharp contrast to the situation with sawah—the more rubber land owned by a household, the higher its income and economic status. Thus identification of "well-off" and "poor" households tends to rest on household ownership of commercially valued land planted in rubber. This fact might seem to suggest continuity in the traditional criteria for economic standing and prestige. But radical disjunctions are evident here: because most rubber plots are not formally designated as ancestral (or customary) and are not held by virtue of pedigree or descent group membership, they can in fact be (though usually are not) freely sold. This situation has helped to undermine the economic and cultural significance of social distinctions based on descent and origin point, and simultaneously has reinforced a competing and cross-cutting social taxonomy based on control of cash-yielding productive resources.

Significantly, however, many individuals who own rubber land and either cultivate it themselves or lease it out also labor as sharecroppers on other villagers' plots of sawah. These circumstances work against the cultural elaboration of social distinctions built up around notions of landedness and landlordism versus landlessness and tenancy. Similarly, even though it is fairly uncommon to find any one person or household whose rice and rubber lands are cared for by the same nonresident individual(s), it is nonetheless true that beneficent sharecropping terms in sawah production do counter community perceptions of such proprietors as uncaring or avaricious denizens of the community. The nature of productive relations centering around sawah thus tends to ameliorate the cultural, and to a lesser extent the social, impact of rigid tenancy arrangements in cashcropping (and the overall transformation of the local economy).

Bogang's community of rubber tappers consists of eighty-seven individuals distributed among sixty-nine households. This segment of the village population is younger and more predominantly male than the sawah labor force. If, moreover, we compare the extent of work undertaken by men and women in each of the two sectors (rubber and rice), the gender-based contrast would be more striking than the relevant figures indicate (see table 7). For one thing, females perform more sawah-related labor than males even when both participate in the same stages of rice cultivation. Furthermore, tapping, although obviously not confined to men, is generally considered a male activity since it falls within the rubric of cari duit (literally, "finding" or "getting money"). Hence, adult women tend to tap rubber only if they are widowed or divorced or if their husbands reside elsewhere. The few exceptions to this generalization lie with relatively low-income households in which both husband and wife must tap their own or someone else's trees to make ends meet.

Unlike rice cultivation, the tending and tapping of rubber trees is a highly individualistic and solitary endeavor. There are no peak labor seasons in which tappers seek out additional assistance to complete their work. Similarly, since rubber trees are perennials and yield for up to fifteen or twenty years after reaching maturity, seasonal preparations and maintenance of the sort required for the growing of rice are wholly unnecessary.

Villagers involved in tapping usually leave their homes shortly before daylight and walk or bicycle to the rubber fields, where they labor in solitude for three to six hours. These periods are not punctuated by much

conversation, since talk can be distracting and can lead to an imprecise paring away of the narrow strip of bark that allows the latex to flow into the collection cup. If cuts are too shallow, smaller volumes of latex will flow; if, however, cuts are too deep, vital tissues may be damaged, causing the tree to dry up prematurely and die.

The pace of tapping also prohibits casual socializing in the course of the morning's labor. Tappers must move quickly from tree to tree before the heat of the day thickens the latex. Then, after all trees are tapped the workers repeat their rounds, collecting the few ounces of liquid latex that have dripped into the cups in the hour or two since the strip of bark was first shaved off. Conversation only draws out this process. More generally, since tappers have work to do even after they have collected all of the day's latex, they are usually eager to return to the village as soon as possible.

The next phase of the tappers' daily activity is the only time they do seek the help of others—usually spouses or children. This is when they go to one of the village's two dozen or so rubber presses to complete the day's work. (These machines are privately held and made available to tappers either for a nominal monthly rent or for free, depending on the relationship between the press owners and the users.) Working at benches, the tappers pour their latex into metal containers resembling bread pans and mix in some formic acid, which transforms the liquid into a material the consistency of wet putty. They then dump the contents of each pan onto a smooth, clean surface and flatten it out by hand to prepare it for the manually operated press. The press squeezes out the excess water and further flattens each piece into a half-inch-thick sheet of uniform width and length. This phase of the processing requires two people, one to crank the wheel that turns the rollers and one to make sure the rough blocks of latex are properly fed into the press.

These tasks complete, the tappers return home with their sheets of latex and hang them out to dry. The sheets are left in the sun for a day or so before being placed for safekeeping beneath the house or on rafters in an interior room. They may be stored there for a few days or for months, depending on household cash requirements.

Latex can be converted into instant cash whenever the need arises. For this, villagers depend on Chinese brokers, who sell the rubber either to another broker or to a firm for further processing.

That no Malays are involved in the purchase of rubber from local producers is not particularly disconcerting to most villagers; residents of Bogang are acutely conscious of the broader pattern of Chinese dominance in virtually all entrepreneurial niches outside the village confines. Nonetheless, some men with whom I discussed the issue unabashedly preferred to do business with Chinese or Indians as opposed to Malays (cf. A. Wahab Alwee 1967, 31; Swift 1965, 32). One reason for this preference is that Chinese are explicitly businesslike, rarely disguising their priorities by gestures of hospitality or seemingly kinsmanlike behavior. Also, many villagers seem to welcome the possibility of pursuing self-interested interactions with non-Malays. This pattern testifies to a diffuse yet pervasive sentiment that considerations of material gain and self-interest should not figure into relationships among Malays. Although such ideals are increasingly difficult to uphold given the erosion of traditional norms and the concomitant monetization of much of the village economy, they are still central in rural Malay culture. This is partly due to the reluctant recognition of what might give way in circumstances in which the morality informing social obligation appears to impede individual or household gain.

Tenants, for instance, commonly emphasized that they sold their tapped rubber with the proprietors present owing to the proprietors' nervous concern that their workers might otherwise cheat them. In separate discussions, however, some of these same proprietors effectively disclaimed such anxieties, saying that their tenants handled all cash transactions on their own; a few also volunteered that they of course trusted their tenants, thus underscoring their reasons for not going along to the rubber agents' shops. Such delicate issues as trust and integrity are not openly discussed, but I suspect that flagrant abuses in this area are rare. Still, it is always possible that a tenant might deliberately underestimate the amount of rubber obtained from a proprietor's plot. I know of one Bogang resident who mistrusts tenants, and fellow villagers more generally, so much that he refuses to lease his rubber acreage at all, even though this results in the land lying unworked and hence in a substantially reduced household income. His position is that "(tenant) tappers end up with all of the meat, while the owners get only the sauce" (orang potong dapat daging, orang punya dapat kuah saja ).

Latent suspicions and tensions are additionally meaningful when we consider that 71 percent (39 of 55) of Bogang's tenant tappers work land

held by kin. Although the percentage of tenancy relationships between kin is larger in rubber tapping than in sawah sharecropping, this does not necessarily signify an across-the-board preference for entering into cash-based rental agreements with relatives (as opposed to "strangers"). Indeed, this comparatively high figure would seem to reflect limited, ambivalently realized tenancy options more than anything else. Of course, certain emotional satisfactions and other benefits, such as greater continuity in the relationship, may be derived from such arrangements between kin. The potential, however, for broad social disruption and spiritual discord is of much graver consequence when unresolved dissatisfaction exists with the relationship. Moreover, it seems true that villagers usually seek to separate ties of kinship from those based on the vagaries and self-serving nature of cash-based rental dependencies (cf. Swift 1965, 26, 62, 171). Thus, if a larger corpus of data and a broader range of tenancy opportunities were available, I would expect sharecropping arrangements focusing on rubber to involve more nonkin than they do at present.

In any case, most of Bogang's rubber tappers (63 percent, or 55 of 87) are tenants, and most of them labor for individuals categorized as kin. Here we see a radical departure from kinship and productive relations of the nineteenth century, which as a rule lacked the monetary coefficients of today, as well as the asymmetry emanating from differential access to scarce resources. Put differently, processes of twentieth-century economic development have entailed a profound restructuring of the bonds linking economic actors and all others once defined as social equals.

The infusion of cash into local economic relationships is manifest, then, both in the nature and incidence of rubber tenancy and in the way workers dispose of the product of their labor—that is, by selling it to agents of modern price-setting markets rather than using it for household consumption or intravillage distribution or exchange. Such dimensions of contemporary economic activity are profitably viewed alongside the essentially parallel trends in rice production. These trends in turn are best assessed in light of more encompassing processes that have witnessed the displacement of traditional subsistence pursuits and institutions by those centering on capital in the form of hard currency. The other side of this coin is the pronounced degree to which present-day households must rely on rubber tapping or other sources of cash income for the satisfaction of domestic needs.

Other Sources of Cash Income

Although the tapping and sale of rubber is the principal cash-producing activity throughout Bogang, many households draw on other sources of monthly income as well. These include cash remittances from nonresident kin (usually urban-dwelling children), money in the form of government salaries and pensions, and proceeds from local shopkeeping and other types of village trade. The relative importance of such earnings can be seen from table 8.^[8] These data show additionally what an enormous significance monies from extralocal sources assume in the local economy, and in household subsistence levels especially. Indeed, the very centrality of these extralocal funds in village consumption and domestic maintenance is compelling evidence of the trajectory of twentieth-century economic change, and of the progressively narrowing viability of rural economic institutions in particular.

Cash remittances from children and other kin living in the city represent the most common source of income aside from the working or leasing out of rubber acreage, with 56.7 percent (59 of 104) of Bogang's households receiving money in this form. Some households, in fact, depend for their very survival on money provided by their nonresident members: fully 16.3 percent (17 of 104) of Bogang's households obtain at least half, and 7.7 percent (8 of 104) more than three-quarters, of their income this way.

Monthly remunerations for current or previous government service are far less common than cash gifts from children. Two additional facts about the households receiving money in this form stand out: first, these households rank among the wealthiest in all of Bogang; and second, their adult male inhabitants predominate in the secular village councils that nowadays constitute the principal organs of community government. Hence, these groups are not only ideally situated within the local corridors of power, but also enjoy privileged access to all state- and federally financed projects of potential economic benefit to Bogang residents.

Most other sources of monthly income in Bogang derive either from trade, conducted generally within the village proper, or from specialized services. Most of these activities presuppose cash transactions between local residents—which, as noted above, tend to be marked by distinct ambivalence.

The ritual assistance provided by curers (dukun )^[9] merits brief mention

here, for these largely instrumental transactions tend to be heavily cloaked in metaphors drawn from kinship and religious fraternity. Throughout healing rituals, for example, patients and dukun alike steer clear of all conversation bearing on matters of remuneration for services rendered. At the close of the session, then, when the patient shakes the dukun's hand, a few ringgit are subtly transferred to the curer. Neither party acknowledges this somewhat inconspicuous but awkwardly managed payment, for it is cast in kinship and Islamic idioms and categorized as sedekah , denoting alms and other voluntary offerings such as those made to the poor and to deceased kin. Also relevant is that by the end of the initial treatment the patient is considered the dukun's anak ubat (literally, "medicine child" or "child through curing").

The content of these ritually engendered bonds does not amount to much nowadays, but dukun can still prevail on their anak ubat for small favors. All such ties are reminiscent of those linking villagers of earlier decades with shamanic pawang who oversaw the performance of community-wide rituals associated with the agricultural cycle. Pawang not only received generally unwavering public loyalty and labor services from all beneficiaries of their ritual expertise, but they were also presented with annual offerings of uncooked rice and other produce from each cultivating household in their jurisdiction. Of interest, too, is that many Bogang residents explained the mid-twentieth-century demise of pawang and their craft by reference to their profiteering at the expense of fellow villagers who were slow to wake up to the pawang's "unorthodox hoaxes" (see below, chapter 9). Some people believe that a narrow concern with capital accumulation motivates certain of Bogang's contemporary dukun as well, although others consider these same curers as trustworthy and altruistic practitioners of sacred arts.

One other source of cash income warrants remark, but it is only available for a few months each year. I refer to the local collection and roadside sale of forest perennials, such as durian and petai. Because these species grow in hilly, forested regions of the sort females prefer to avoid, they are commonly harvested by adolescent and adult males.

Even though most villagers engaged in the harvesting of durian seem interested mainly in supplying their domestic coffers with as much of the prized fruit as possible, 19 percent (20 of 106) of Bogang's households also recorded supplemental incomes of, on average, M$50 to M$200 from

durian sales. Durian tends to be sold to Chinese on a frequent but irregular basis throughout its season (late June through early August).^[10] But unlike petai, which are collected and sold during the same months, durian does not require that harvesters climb into the upper reaches of rain-forest growth; rather, the fruit simply falls to the ground when ripe.

Durian trees are often located deep in state land—and so can be harvested legitimately only by duly licensed persons—but they are also found on hilly orchard (dusun ) acreage alienated to individual proprietors. In the latter circumstances, sharecropping arrangements are analogous to those prevailing in the case of sawah—that is, they are inclined to be flexible and generally advantageous to the tenant. Similarly, the idiom of rental assumes relatively minor elaboration here, and is commonly supplanted by that of service, especially when young men gather durian on land vested in women of senior age or generational standing. Part of the reason for this is that many of the plots on which durian grow are designated ancestral and are held to have been planted by long-deceased forebears.^[11]

Most of the other produce obtained from the forest or from village gardens is earmarked for household consumption rather than exchange or cash sale. This includes riverine resources, small animals hunted with shotguns, and numerous varieties of banana, some of which are planted in small garden plots, others, along with chilis and ginger, on residential acreage. The same holds for the stands of fruit-bearing shrubs and trees dotting domestic compounds, such as pineapple, papaya, mangosteen, rambutan, guava, and coconut. Only coconuts serve an additional function, for they also figure into the gifts of uncooked food that women offer to kin and neighbors to help defray wedding and funerary feast expenses. Similarly, some villagers occasionally sell coconuts, for three to six cents apiece, but this is a marginally productive activity that may bespeak financial desperation.

The cultural muting of cash transactions, rental arrangements, and differential access to strategic resources warrants remark again at this juncture. I suggested earlier that the prevalence and typically favorable conditions of sawah sharecropping work against a cultural elaboration of rubber tenancy as involving rigid, instrumental, and asymmetrical relationships centered on the sale and purchase of labor. In this and other areas—durian sharecropping, for example—idioms of kinship, disinterested service, and village citizenship all serve to channel and constrain local perceptions of

the scope and force of twentieth-century economic change. Perhaps most striking is that these dynamics not only tend strongly to deemphasize socio-economic stratification within the community, but they also promote images of local society as composed of individuals and groups related as social equals.

Economic Stratification

In Bogang and the rest of Negeri Sembilan, present-day disparities in household subsistence levels reflect differential access to commercially valued rubber acreage and extralocal incomes, such as those obtained from salaried government employment. Access to one of these sources of wealth by no means precludes enjoyment of the other, however. Instead, the two tend to go hand in hand in a fairly common scenario wherein upwardly mobile males leave the state for a period of government service and return some years later with considerable savings, a handsome pension, or both, and hence the wherewithal to purchase rubber land for sharecroppers to work and the means to settle back to a comfortable and prestigious existence in the village.

As I mentioned before, many of the adult men of Bogang's wealthiest households receive impressive monthly pensions on account of extended service with the police, military, or railways of the Federated Malay States or Singapore or having been a teacher or clerk in a government office. These men are well versed in reading, writing, and formal speech, and generally quite adept at deciphering land application forms and other products of administrative bureaucracy. Such skills alone give them a marked advantage over the less urbane members of the community. Above all, though, these individuals comprise the rural segment of the new Malay elite. They tend to predominate on the secular village councils that nowadays constitute the principal organs of local government, and are therefore ideally situated to further enhance their own prestige and economic well-being. This is not to suggest an atomized Hobbesian society whose members forever strive to gain a competitive edge over their rivals, no matter what the social cost. It is simply to point out that, because state land and government aid are of limited availability and insufficient in amount to benefit the village majority, those holding positions on village councils, along with their allies, are the most likely to profit from federal and state-

level assistance. From this perspective, the rich do get richer while the poor, who continue to rely on income from tenancy or gifts from children, grow poorer, if only in a relative sense.

One of the most obvious indices of contemporary economic stratification in the rural sector appears in the broad spread and clustering of monthly household earnings (see table 9). In Bogang these sums range from M$13 to M$1,300, with a mean of M$286, which, although quite impressive by rural Southeast Asian standards, is rather biased in favor of the wealthy. Table 9 reveals that 16.3 percent (17 of 104) of Bogang's households bring in less than M$100 per month, and 30.8 percent (32 of 104) report monthly earnings below M$150. Similarly, over half, or 51.9 percent (54 of 104), of all households claim less than M$250 a month. In contrast, a striking 21.1 percent (22 of 104) receive upwards of M$450, with the vast majority, or 15.4 percent (16 of 104) of the village total, living on extremely comfortable monthly incomes exceeding M$550.

All such income differentials are expressed in household provisioning, if

only because the local cultural system accords prestige to those who invest wealth in consumer goods and other easily recognizable emblems of status. Table 10 presents data on the distribution of selected goods and services among households. Although no information on such matters as the frequency of feasts, the quality or style of residential quarters, or daily diet is provided, these also vary from one grouping to the next, being most pronounced at the two ends of the income continuum.

The existence of such variation raises two important questions. First, how has the erosion of traditional tenure institutions contributed to the twentieth-century concentration of resources within specific households and clusters thereof? Second, and relating to the cultural realization of socioeconomic stratification, why has local culture allowed for a selective recognition of such cleavage but simultaneously gone far toward denying it?

To take each of these issues in turn, I would first emphasize that present-day disparities in income levels and standards of living were unheard of in precolonial times. Although household subsistence levels have always varied according to differential access to strategic resources (cultivable land, for instance), rights over such resources were far more evenly distributed before colonial intervention and the commercial developments that followed. Numerous institutional mechanisms contributed to this more balanced situation, including sanctified conventions of tenure. Thus, a household's claims over agricultural land were only legitimate so long as its members (or nonresident dependents) contributed regularly to its exploitation. More precisely, if a plot of land lay fallow for three consecutive years, rights over it automatically passed to the kin group at large for reallocation to members in greater need. These nineteenth-century strictures embraced all acreage and were partially upheld by British authorities throughout the colonial interlude. However, these communalistic conventions hardly ever pertained to land planted in rubber or other export crops. Further, because most of the rubber land held by Rembau Malays since 1900 was not unambiguously classified as ancestral or customary property, it did not fall squarely within the divided title system underlying the effective operation of these communalistic traditions.

With the demise of divided title, no institutional means exist through which to counter, or even mitigate, either the disproportionate concentration of land among certain households or kin groups or the long-term

land shortages or landlessness of others. Thus we find a highly inequitable distribution of rubber acreage in Bogang (see table 11),^[12] with attendant income differentials. In a word, since contemporary households need not work their own rubber acreage in order to maintain rights over it, they have free reign both to lease it out to other villagers and to accumulate as much land as their capital and political connections permit.

Significantly, however, proprietorship over sawah allows for an appreciably different and far less radical perspective on all such developments regarding tenure. Rembau and other Negeri Sembilan women who relocate outside their natal communities effectively relinquish all rights to the proceeds of labor performed in their sawah. Additionally, these women essentially forfeit all claims to bequeath rights over such properties to potential heirs (their daughters), who may reside with them and have little intention of resettling in their former village. This situation not only contrasts sharply with otherwise comparable circumstances involving rights over rubber, but it also testifies to a certain conceptual continuity in the context of a profoundly transformed agrarian environment.

More generally, however we approach or attempt to assess economic stratification within communities like Bogang, we must appreciate that villagers themselves attach far less significance to specifically economic distinctions and cleavages among households than might be assumed. I have analyzed several factors that allow the conceptual deemphasis of social differentiation based on material wealth, especially noteworthy being the cul-

tural representation of sawah sharecropping in terms of reciprocal aid between social equals. The following chapter assesses other analogous phenomena, including cross-cutting kinship sentiments and loyalties, association with national political parties, and a long-standing feud concerning two dominant village lineages. Personal and social wealth, as we shall see, continue to be measured in terms of children; and prestige is still partly allocated according to descent affiliation and one's point of origin, in both a genealogical and a culturo-geographic sense. Ethnic variables are relevant here as well, especially the common attitude that the local poor—and the Malay population of the country as a whole—would enjoy an appreciably higher standard of living were it not for Chinese predominance in most entrepreneurial niches outside the village, and their disproportionate control of capital resources more generally.

The central point here, in short, is that the scope of economic differentiation in the village has yet to be realized in any correspondingly explicit ideological sense. Thus, even though processes of economic development have eroded innumerable social institutions and cultural forms, the latter have also served to channel and limit the overall impact of twentieth-century change.

6—
Kinship and Settlement Patterns

This chapter analyzes the ways historical transformations in the village and extralocal economy are realized in the composition, social ties, and spatial arrangements of individual households and lineages as well as Bogang's two "hamlets." Major shifts in economic and demographic factors have given rise to radical disjunctions in both inter-household relations and residential areas on the whole. While these discontinuities are especially apparent in the "new hamlet," their existence throughout Bogang points to villagewide tendencies toward atomized and largely autonomous domestic units consisting mostly of married couples and their offspring or previously married women and their (usually female) children. We will also see that such tendencies have been muted by the scope and force of constructs of origin-point, siblingship, and descent, and by mechanisms of feasting and adoption, all of which have played an extremely important role in promoting social and cultural continuity.

Identification with the Village

The residents of Bogang regard themselves as members of a single community of uniform culture that differs in various ways from other settlements.

This sense of belonging to a singular social body is enhanced by numerous secular institutions of relatively recent emergence: the office of village headman, the existence of village councils (and a hall for their meetings), a local chapter of the ruling political party, UMNO, and a community school. Of more ancient origin are the village mosque, with its trio of functionaries, the graveyard and the deceased ancestors buried there, and the shrines (keramat ) believed to house the physical remains and still-active spirits of Bogang's earliest culture heroes and sacred guardians. The corpus of myths pertaining to Bogang's founding and historical development obviously merits mention as well. Additionally, women boast a culinary expertise, expressed in a style of flavoring hot sauces, that identifies them as both distinct from and superior to the women elsewhere. Institutional and cultural factors of this sort reflect and bolster villagers' strong sense of identification with Bogang. So, too, do marriages involving village-level endogamy and the fact that former civil servants and other in-marrying males who have spent time outside the village typically return to Bogang and take up permanent residence there. The larger issue in all of this is the pervasive premise that all villagers are essentially kin (saudara ), if only in relation to the residents of other settlements (cf. Swift 1965, 143).

When villagers, including in-marrying males, travel in Rembau, they refer to themselves and are labeled by others as "Bogang people" (orang Bogang ). When villagers venture beyond Rembau, the social relevance of membership in the community of Bogang tends to give way to identification with Rembau, or simply Negeri Sembilan. This is not ethnographically unusual, although I would stress that much of an individual's social identity (and a good deal of negative stereotyping) derives from his or her association with a specific territorial domain. Village residents of a single lineage compound, for example, often characterize the behavior or social standing of large segments of Bogang's population by reference to where they live. Thus, many people, especially adolescent boys, who reside near a particular coffee shop are viewed by the members of neighboring compounds as "less than refined" (kurang halus ) or as troublemakers (orang jehat ). Similarly, the residents of another neighborhood are regarded as having marginal, even pariahlike, social status, even though their day-to-day behavior conforms to basic community norms. Still other villagers have the reputation of being sombong , that is, arrogant or unresponsive to social expectation, simply because of where they live.

Traditional patterns and ideologies pertaining to property and to post-

marital residence reinforce the association between particular enatic groups and specific territorial domains within the village. The nature and history of people's ties with residential land can also determine, both symbolically and concretely, how individuals and groups are classified. This situation should come as no surprise, since historical precedence in local settlement is still very important to the allocation of status and prestige. Persons claiming rights over ancestral land whose origins extend beyond village memory, for example, tend to enjoy greater prestige than those whose jural rights involve local property of more recent creation. Thus, although a woman's identity is still very much a function of descent-group affiliation and ranking, her residence within a particular part of the village, and the nature of her rights over homestead acreage, often tell more about her community standing and social universe. Similar circumstances prevail in the case of adult men as well, even though the situation of in-marrying males contrasts somewhat with that of married men of local origin.

In this chapter we will first consider Bogang's households—their significance for their residents, their kinship composition, and how their social space is defined. We then examine a single lineage and its compound, focusing on the relevance of such units for their members and in village society at large. Finally, we conclude with a comparison of Bogang's "old hamlet" and the recently settled "new hamlet."

Households and the Domestic Domain

The household is the origin point and locus of an individual's most valued and enduring personal relationships, public identity, and emotional mooring. The interior of a house—and to a lesser extent the surrounding yard—is also viewed as the private and somewhat exclusive domain of the individuals living there. In theory, then, whatever transpires within the confines of a house, or among its members, should not concern persons of other households: it is simply "none of their business." In practice, however, the concept of a domestic or private realm refers more to the relative social autonomy of individuals in their own homes, and to the fact that the adult members of a household are not usually beholden to anyone else so long as they appear to uphold basic community norms. But there is a catch here, and a problem for villagers: much of what goes on in the private domain is eminently public in that it can be heard and seen by immediate

neighbors, passersby, and others. Most houses, for that matter, rest on ancestral land over which sisters and other enatically related kin hold residual rights; this fact bears significantly on the married males of these households, and on the status of in-marrying males overall. In short, the notions of privacy and private domain are highly relative.

Household Size and Composition

Bogang's current population of 476 inhabitants is distributed among 106 separate households; the mean household thus consists of 4.5 individuals (see table 12). Households range in size from one to ten persons, although over three-quarters (78.3 percent) of these domestic units have six or fewer residents. Indeed, over half (57.5 percent) of the village's occupied houses claim from two to five members. The most common household size is five, accounting for 17.9 percent of all households. It should be noted, however, that these calculations and percentages reflect a focus on the cline of households in terms of size rather than the distribution or clustering of individuals in different-sized households. Specifically, twelve separate households in the village consist of a single individual, and sixteen others consist of only two members; the data are thus skewed in the direction of the smallest households. The final column of the table shows that

over half (51.9 percent) of Bogang's people live in households with six or more members, and that an impressive 29 percent of the entire village population can be found in houses having 8–10 residents.

While virtually all single-person dwellings are associated with the lowest income quadrant, so too are some of the largest households. Interestingly, the elders of these latter households are regarded as fortunate in light of the emotional riches and social capital they gain through coresidence with children, grandchildren, and other kin—however economically productive or draining they may be. The underlying sentiment is suggested by a local expression that, in this context, can be translated as "the more, the merrier" (makin, ramai, makin sedap ); that is, almost all adult villagers prefer to eat, sleep, and live in the company of their children rather than under a separate roof.^[1]

Household composition can be broken down into four basic patterns, three of which contain further internal subdivisions (see table 13). Each type, moreover, can be seen as a developmental variation on a single set of themes.

My typology of households based on interpersonal ties—that is, nuclear, conjugal, consanguineal, and individual—has no native counterpart in the sense of like indigenous categories. Nevertheless, the major analytic distinctions are useful for ordering the data. They also serve to emphasize features of domestic variation that are of considerable local relevance. Above all, however, they underscore the critical fact that a domestic unit consisting of a married (or formerly wed) woman and her dependent children is of far greater functional significance than is the same group plus the woman's husband. In short, while the term nuclear household (or nuclear family ) is typically associated with the latter configuration, I follow Goodenough (1970, 18–19) in restricting its use to domestic units consisting only of women and their resident dependents (that is, their sociological children, whether natural or adopted). The addition of the husband/father to the group does make a difference, of course, especially today. But this man's presence or absence may be—and in Negeri Sembilan definitely is—relatively unimportant to the jural standing, socialization, and overall development of children. Therefore, the "nucleus" of kinship consists of mother-child bonds and ties among siblings, and not the conjugal pair and progeny that Murdock (1949), Lévi-Strauss ([1949] 1969), and others regard as both the nuclear family and a cultural universal.

In any event, 24.5 percent (26 of 106) of Bogang's households are what I would term nuclear. Most are headed by divorcées or widows, as opposed to women whose current husbands live elsewhere owing to military service, distant employment, or cohabitation with a second wife. Aside from having no paternally elaborated counterpart (that is, consisting solely of a man and his children), this arrangement occurs in extended form only in terms of links through women. Hence, historical continuity in domestic arrangements is considerable.

Conjugal households, in contrast, account for 60.4 percent (64 of 106) of the village's domestic units. Most of these—or 47.2 percent (50 of 106) of the entire village—involve a married couple and their children, which also represents a common pattern of earlier times. It is historically very interesting that the twelve conjugal households consisting solely of a

married couple appear unlikely to include any children in the future. These are mostly elderly individuals past the age of reproduction; their children have either settled elsewhere in Bogang or, more often, migrated to rural development schemes or urban areas. Such two-person households were of course typical in previous decades and centuries, but they were mainly of temporary duration, consisting of newly married couples who had yet to bear or adopt children. More striking still, recently married couples are largely absent from the local community, as are natal residents of the village and in-marrying males in their twenties and thirties in general (see table 14). Consequently, we encounter both a greater percentage of single-person households than has probably ever before existed, and a bimodal age distribution that is undoubtedly far more pronounced than at any stage in history.

Even allowing far more single-person households at present, there is no

major break with the past as regards the gender, age grade, and marital status of those who reside by themselves, for most are formerly married women aged fifty-five and over.

Single-person households and conjugal households inhabited solely by a woman and her husband share a basic feature in common: they span but a single generation. The relatively large number of such households, twelve and sixteen respectively—or 26.4 percent of Bogang's domestic units—testifies to a villagewide trend toward the generational retraction of domestic groupings (despite lower infant mortality rates and greater longevity), and represents an appreciable shift from previous decades. The larger issue, referred to above, is the increasingly polarized age structure of the village population, due to accelerated rates of out-migration involving virtually all youth over the age of eighteen, regardless of socioeconomic standing.^[2]

In light of the great value placed on children and the profound dread of solitary existence, it is ironic that so many parents encourage their offspring to make lives for themselves outside the village. I recall one woman in her early fifties explaining forthrightly that she did not want any of her children (all of whom were males) to settle down to a rural existence. As she put it, "only the stupid ones stay in the village" (orang bodoh saja duduk kampung ). Unfortunately for her and others of her generation, many formerly resident children will probably never return to the village to live. This situation arouses considerable ambivalence and misgiving among parents, many of whom feel "left behind," if not abandoned, and also express anxious concern that their homes and compounds might stand empty, and their rice fields fallow, after their deaths.

Perhaps the most striking pattern in domestic group composition and variation in Bogang is the marked skewing in favor of females and links through women, which prevails at all levels. This skewing is most evident in the large percentage of households without any ever-married males—a full 29.2 percent (31 of 106), as compared with only 2.8 percent (3 of 106) for households without ever-married women. Expressed differently, more than ten times the number of households have no adult males than have no adult females. This situation highlights a basic social fact: women can and do live well enough without husbands and other adult males; men, in contrast, cannot really survive without wives or apart from other female kin.

As a result, whereas formerly wed women are in many instances committed to remaining single, their male counterparts invariably remarry as quickly as possible.

Structural Complementarity in the Domestic Domain

There are well-developed social and cultural links between women and femaleness on the one hand, and the domestic realm on the other. Females enjoy unquestioned, essentially unconditional rights in their households and are generally expected to remain in their own compounds unless subsistence, exchange, or religious activities call for their presence elsewhere. Males are far less strongly associated with the domestic domain and are never tied to just one household in the course of their lives; they have residence and household rights largely because of privileges acquired through ongoing relationships with women (their mothers or wives).

Female involvement in the subsistence sector and in myriad domains of exchange is obviously both permitted and encouraged, but it should neither interfere with domestic work—food preparation, cleaning, childrearing, and so on—nor result in "mixing too much" with women, let alone men, of other households. The possibility that women may be unable to fulfill their domestic duties is one of the several reasons why men prefer that their wives neither tap rubber nor seek other employment outside the home. The overriding issue, however, is less conjugal or domestic delinquency than the moral and aesthetic tenet that women should not be seen walking around the village, and should consciously minimize the time they spend away from their homes.

Especially striking to the outside observer is that women rarely go to certain coffee shops, and they go to others primarily during their "off hours," when only a few, mostly neighboring men will likely be there purchasing supplies or enjoying a cup of coffee or tea. Additionally, although women may participate in Friday mosque services, because weekly mosque attendance is only obligatory for males women tend to avoid the mosque on Friday afternoons. Indeed, with a few notable exceptions, female presence at the mosque is minimal as compared with males.

To point to coffee shops and the mosque as male domains is not merely to say that women maintain a low profile in these areas since they are expected to (and invariably do) busy themselves with household chores to

the exclusion of most other activities. It is to draw attention to more pervasive patterns, one being that men are allowed and encouraged to socialize outside the home to a far greater degree than are women. To appreciate this fact one need only recall the eminently valued tradition of temporary out-migration on the part of males (merantau ), and the absence of any corresponding tradition for females.

It remains to examine what goes on within the household. Here we will focus on the two activities villagers frequently cite as paramount: eating and praying. Most other individual and household concerns are in some way tied to these, and much of what villagers do and say relates to either the consumption and sharing of food or the immediate social relevance and ultimate (or afterworld) implications of conscientious prayer, or to both.

On the surface, it might appear that eating and praying in one's own home—as against, say, at a small-scale feast or a mosque gathering—would be relatively private activities. In one sense, of course, they are. Yet residential construction and the close proximity of most homes enable neighbors and passersby at any given time to see or hear which members of the household are praying and who—including nonresidents—might be eating there. Actually, the most relevant distinction here is that within the household husband and wife eat and pray together, whereas outside of their own home they virtually never do. More generally, husband and wife behave rather freely with one another in their own home, but they avoid association, let alone casual or vaguely intimate interaction, in all other contexts. This situation can be seen as paradigmatic: interactions between males and females, including husband and wife as well as brother and sister, should be limited as a rule and restricted to the household in particular. Males and females do in fact move in separate and typically non-overlapping spheres outside their own homes (aside from the partial exceptions encountered in local commerce). Only within the household do adults, or any males and females past the age of seven or eight, interact freely with members of the opposite sex. Men, moreover, do not usually visit one another or stop in on their nearest neighbors (that is, at the homes of their female affines and "brothers" through in-marriage) or other friends. And women usually socialize with their female friends either when the men are out or, when the men are at home, in a separate room. Thus, the only truly sanctioned cross-sex interaction occurs within the household, and then only between household members. This tolerance of male-

female interaction is in many respects the most distinctive feature of the domestic domain and what sets it apart from the rest of village society.

In their own homes, husband and wife joke with and browbeat each other, use terms of address that in the presence of others would be considered overly intimate (hence unrefined and vulgar), display verbal affection, and engage in sexual relations—none of which would take place in the homes of others, or in their own homes if others were present. There are definite limits to this relative indulgence, however; indeed, the social distancing based on gender that occurs within the household seems to provide the model for sexual segregation in the society at large. Above all, though, the intimacy and indulgence permitted between husband and wife privately reflects their status as social equals, a fact that is partially denied in many public and especially ritual contexts where females are culturally defined as less than equal to males and the "closer to nature" of the two sexes (cf. Ortner 1974; and below, chapter 8).

The complementary relationship of husband and wife is perhaps most characteristic of the conjugal bond. Because males are portrayed as fundamentally different from, and less responsible than, females, wives tend to be accorded numerous duties that might otherwise devolve upon their husbands. Safeguarding household jewelry and cash, for example, is typically a female task, even though these valued items may derive from husbands' earnings on land obtained outside of marriage. More generally, wives usually manage the cash resources of their households, including the periodic doling out of sums to their husbands. Men occasionally appear less than enthusiastic about such an arrangement, but they also seem resigned to its inevitability. As one man wryly explained, "It's like death; no one wants to die, but once you're dead what can you do?"

The complementary responsibilities and activities in relation to household income—such that men go out and earn cash while their wives manage it at home (and also perhaps bring in capital in their own right)—are informed by gender constructs that also contribute to a pronounced favoring of female heirs. These same gender constructs, I might add, are realized in labor exchanges associated with feasting and in the rearing and informal adoption of children, both of which invariably involve females much more than males. Such activities clearly transcend the domestic domain (as do the affinal exchanges effected largely by women) and are also accorded almost unparalleled social and cultural value. Such factors indi-

cate that the position and prestige of local women has not been seriously undermined by forces of colonialism and modern market institutions (see chapter 8; cf. Peletz 1987b). Moreover, they suggest that the relative exclusion of women from coffee shops and various mosque gatherings should be read not as an across-the-board devaluation of women but rather as an index of the complementary tasks, spheres, and more encompassing social roles of women and men both.

A more general point to bear in mind is that most women and men in Bogang belong not only to one or another household but also to more encompassing compounds defined in relation to a local lineage. We next consider the nature of such compounds, including the ways in which certain of the social dynamics we have explored are manifested in the ties among residentially clustered households, especially their female members.

The Hill Lineage and Its Compound

The internal structure and dynamics of the Hill lineage of the Lelahmaharaja clan are representative of the village as a whole, despite the group's marked prestige and power within Bogang. The following discussion will thus serve to elucidate the social entailments and other concomitants of compound residence and lineage membership. In one sense we are moving here from the least public facts and expressions of kinship—as revealed, for example, in eating and sleeping arrangements—to those of a more broadly social nature. For although the affairs of a lineage compound are, in theory, the somewhat exclusive concern of lineage personnel and in-marrying males of senior standing, they often end up being a matter of public record and villagewide interest. The lineage compound, moreover, serves both as the setting for collective exchanges and activities and as an important buffer between the wholly public realms of community life and the relatively private spheres of the domestic domain; its real capacity to mediate between these different interests, however, is ultimately limited by internal cleavages and contested claims to property, productive resources, and personal affection.

In comparison to other lineages and more encompassing descent units, the Hill lineage is extremely aggregated. Excluding those who have left the village altogether, all women and children of the lineage reside in a single compound. The Hill compound is in a centrally located area of Bogang's

"old hamlet," one which is conceptualized in relation to the lineage, and to the adult women of its nine households in particular. Just as this compound constitutes the more or less continuously occupied ancestral homeland of the lineage, so too is it fairly unlikely that any of its inhabitants will relocate to other areas of the village in the immediate future. Indeed, if they were to leave their current homes, these wealthy and upwardly mobile individuals would more likely settle in the state capital, Seremban, or in Kuala Lumpur than elsewhere in the village or in other rural environs.

One striking feature of the Hill lineage is the highly developed sense of group membership and exclusivity, not only among the women and their spouses but also among nonresident children (a good many of whom work and live in cities and return to Bogang only on free weekends and national holidays). The clustering of all lineage households in a single compound obviously contributes to this phenomenon, as does the compound's proximity to the homesteads of certain other groups. On one side is the sister lineage, Valley, with whom Hill has long been embroiled in a bitter feud (see chapter 9); on another, the "upriver" segment of the Biduanda Dagang clan, which tends to be regarded by Hill and gentry clans in general as a sociologically distant and low-status social entity with whom marriage should definitely be avoided. Although at least some Hill women acknowledge a degree of siblingship with one neighboring line of Biduanda Dagang (emanating from a Meccan-born apical ancestress discussed below), the Hill group is essentially surrounded by outsiders, many of whom are either longstanding enemies or the staunch allies of adversaries.

The Hill lineage's sense of group membership and exclusivity is largely a product of political and economic factors. Some of these factors hark back to ancient times and point to various mechanisms whereby prestige and wealth are reproduced over the generations. Elsewhere in this chapter, for example, we will encounter instances of adoption by a nineteenth-century Undang. This man of renown was of the Hill lineage, and in generating clientlike relationships through adoption he increased the lineage's prestige and political clout as well as his own standing as a patron. More generally, the mere fact that he rose to the supreme office of Undang not only renders him the most famous and successful of all human village ancestors, but also translated into considerable prestige and political opportunity for his children. It is far from coincidental that at the turn of the century a

good deal of village land was registered in his son's name, whereas virtually all other local acreage on the books was listed under women's names; that a generation or two later one of his descendants in the Hill lineage attained executive office at the state level; or that today the latter man's sister's son holds a powerful post in the federal government. This individual's brother, I might add, stands as the highest-ranked adat leader in Bogang.

The Hill lineage clearly has an impressive pedigree, and it has been very successful in realizing the capital benefits accruing from its genealogy, vertical ties, and enviable history of political connections. This is not to say, however, that the lineage necessarily acts in unison, with or without the aid of in-marrying males, to achieve economic or other objectives, or that all lineage households derive equal advantage by virtue of their association with the lineage. It is simply to bring into focus a set of political and economic factors that have contributed both to the lineage's prestigious and powerful position in contemporary Bogang and to its understandably well-developed sense of group membership and exclusivity. Regarding the latter point specifically, the notion of "us" versus "them" is more pronounced among members and spouses of the Hill lineage than among any other descent units or household clusters. This is not so much a behavioral or sociological distinction as a conceptual and emotional one; it is nonetheless an important factor in defining, generating, and solidifying political and economic relationships and interests.

The Hill lineage stands as the sole heir to the unique fortunes and legacies of Bogang's most illustrious (though notoriously unpopular) ancestor; yet we need not turn to history, or even to issues of pedigree or economic concentration, to find evidence of its unique placement in the village. Rather, concrete reinforcement of the lineage's separateness and genealogical homogeneity exists, in the form of low, rough-hewn, and partly overgrown fences—of barbed wire, bamboo, large branches, and lengths of old timber—that encircle the compound, setting it off from the fields, abandoned lots, and houseplots of neighbors. One also encounters fences of the same simple construction around certain houseplots belonging to members of the lineage. But these exist elsewhere in the village as well, and should not deflect our attention from the highly unusual occurrence of an entire lineage or lineage branch set off by a makeshift network of fences and occasional gates.

Malays seem to dislike the notion that "good fences make good neigh-

bors." What, then, do the fences mean, and why were they built in the first place? The second of these questions is—at least superficially—the more easily answered. The stock response is that fences keep village domesticates, particularly goats, out of residential areas, which are frequently dotted with stands of flowers and fruit trees, all easily damaged by grazing animals. While this explanation is true in part, it is equally apparent to villagers that the marauding goat population can sooner or later make its way into (or out of) any area of the village. Also, since many village gates are left propped open, goats are not really discouraged by the mere existence of fences. Nor for that matter are intruders in the form of adolescent boys who, especially at night, may surreptitiously enter and lurk about a compound either with the intention of making off with personal items left near the houses (sandals or sheets of rubber, for instance) or in hopes of catching a glimpse of an unmarried girl.

Fences, then, do not really serve the purpose for which villagers claim they are built. They do, however, demarcate domains of relative autonomy and social control—and this, more than anything else, explains why they are constructed and maintained, as well as what they mean to those who live within or outside of them. In short, members of the Hill compound are ultimately concerned with setting themselves apart from and above others. The century-long concentration of wealth, political clout, and prestige within the lineage has given its members and spouses more to lose, not only through theft, malicious mischief, and gossip, but also as a result of shifts in personal allegiances and the distribution of power. Similarly, since the lineage has more to gain by maintaining and, if possible, upgrading its unique position in the village, it must take greater care to demarcate its social and spatial boundaries so as to avoid identification with unrelated neighbors, sister lineages, or the village majority. Although these tensions between kin group and community are evident elsewhere in Bogang, they are especially strong in the case of the Hill group—as the fences indicate.^[3]

Thus far the discussion of the Hill lineage and its compound has proceeded in terms of social topography and lineage identity, and on a relatively abstract level. I have focused on its unique history and contemporary standing rather than its basic similarities with units of similar shape and structure. In continuing the discussion of the Hill lineage, I turn to data and issues having broader applicability insofar as they find generally comparable expression elsewhere in the old hamlet. These pertain to how the

lineage is conceptualized (hence, to the vertical scope and overall distribution of a particular segment of genealogical knowledge), where authority resides within the lineage, and what types of relationships obtain among the group's socially active core, adult women of the grandparental generation.

The Hill lineage embraces individuals of four generations and collaterals as distantly related as fourth cousins (see figure 5). All told, the lineage claims 107 living members, most of whom work and live outside Bogang. Of those members currently inhabiting the village, only one, a married man in his fifties, does not reside within the lineage compound. (This individual is the clan subchief, who lives in a separate area of the village among his wife's kin.) The eldest person of the lineage resides in the middle of the compound and is a wizened but still energetic and forceful woman of approximately seventy-five; the youngest is a four-year-old boy. The age structure of the compound's enatic and overall population is markedly skewed (as is that of the village as a whole) owing to out-migration on the part of adolescents and young adults of both sexes. Thus, well over half of the Hill compound's adult enates are fifty-five and over. This skewing is even more pronounced for the compound as a whole, since most of the adult women live with husbands at least a few years their senior.

The most remotely recalled figures and connective links in the genealogy are the sisters Selbiah and Maimunah (see figure 5). Interestingly, when the grandchildren of the one sister describe how they are related to the grandchildren of the other, they point to this tie of sisterhood (although they frequently also specify that their own mothers were first cousins). They do not speak of common descent from the mother of Selbiah and Maimunah, whose name no one seems to know; nor do they refer to her antecedents, who are equally anonymous but further submerged in the amorphous category of "ancestors" (nenek-moyang ).

Precise knowledge of the relationship between Selbiah and Maimunah is not evenly distributed throughout the lineage. Moreover, there is little likelihood (or concern) that such knowledge will be passed on, particularly since "technical" information of the sort necessary for the construction of a lineage genealogy is typically of much greater interest to the anthropologist than to anyone in the village. Consequently, the younger descendants of Maimunah cannot specify their exact connection with their generational counterparts in Selbiah's line, or vice versa. Similarly, since no one in this

^[

Figure 5.
Genealogy of the Hill Lineage of the Lelahmaharaja Clan, Showing Effects of Out-Migration

generation knows the names of Selbiah and Maimunah, they too will eventually fade into anonymity. When this situation prevails throughout the lineage—and not merely at its lowest generational levels—the members of the Hill lineage will no longer constitute a single lineage branch (pangkal ). Instead, in another twenty or so years we will find two such units. Of greater relevance than this structural distinction—which is recognized but not usually emphasized by villagers—is the underlying fact that the precise relationship of the two lineage lines will not be known. It is not clear how this will affect attitudes and interactions among enates within the compound. Still, it seems reasonable to assume that eventually internal differentiation will be more sharply perceived and that a corresponding decline in the group's sense of unity and common purpose will occur. This is not to suggest that at present no intralineage cleavages or antagonisms exist. As we will see, lineage unity is, to paraphrase the Geertzes' (1975) observations for Bali, something to be achieved and maintained (if sought in the first place), and not a historical or timeless given.

To assess Hill lineage solidarity and the bonds among its consanguines on the whole, we will focus on the relationships of the eleven married (or formerly wed) women who regard the compound as their home, and on the distribution of prestige and authority among them. Eight of these women, ranging in age from forty-five to sixty-three, are of the same (grandparental) generation; further, only one woman is senior to them. Thus, the lineage core, and its most socially active component, is essentially a group of classificatory sisters from three related sibling sets. Wan, the seventy-five-year-old lineage elder, is the widowed mother of three of these women. In relation to others of the core generation, she has the standing of a mother's sister (MMZD) and the additional prestige of being the oldest living member of the lineage. Her status as a haji further reinforces her esteemed position within the lineage and throughout the village community. So, too, do the acquired wealth and largely achieved status of her politically prominent brother, now deceased, and that of her sons, daughters, and sons-in-law. Three of her children, it may be noted, have traveled to Mecca with their spouses, and all three of her sons-in-law enjoy substantial incomes (from pensions or other government monies in conjunction with rubber lands leased out to tenants).

Wan's position within the lineage is, at first glance at least, rather elusive. Like most women of her generation, she maintains a low profile in

the compound and in the more public arenas of village life. She no longer works in the rice fields, for instance, but seems content to spend her time preparing spices, condiments, and sauces for the members of her house-hold and praying either at home or at the nearby village mosque. Aside from going to the mosque and occasional weddings, she rarely ventures outside the homestead area she inherited from her mother. Nevertheless, and although she does not involve herself in the everyday affairs of lineage households, she speaks with authority and exercises considerable influence over the lives of current and former compound residents, and over their interactions with outsiders. Thus, her own children and maternal nieces tend to defer to her wishes and directives, even though they are grandparents in their own right and of enviable social standing and influence in the larger community. The de facto locus of lineage authority, then, evidently need not be coterminous with the lineage core or its most socially active citizens or traditional political representatives. Instead, as in this case, authority may emanate from the eldest woman of the lineage, provided of course that she is not yet senile or otherwise incapacitated by old age. This situation occasionally poses serious problems for those of Wan's children who have extensive and active social networks or are politically well connected. All this will be clear when we consider Wan's role in generating and perpetuating the feud between the Hill group and its sister lineage, Valley, which has involved much of the entire village population.

Wan is the focal point and principal source of authority in the lineage. It is therefore fitting (though by no means inevitable) that her house is in the very center of the compound, allowing for a strategic view of most other residences and thus facilitating knowledge of the daily movements and activities of many of her kin. Moreover, since two of Wan's daughters live on either side of her, while the third and youngest lives with Wan (and will inherit the house), Wan is in constant touch with the most intimate members of her social universe. For this reason, and because she enjoys such easy access to the village mosque, she has refused numerous invitations to resettle in Kuala Lumpur to live with her wealthy son. There, she claims, servants do much of the cooking and other chores and would leave her with nothing to do all day long. The distance from her son's house to the nearest mosque also discourages her from moving, particularly since she sees her days as numbered and feels that old people like herself are best advised to pray conscientiously and to attend the mosque on a regular,

preferably daily, basis. Even if daily mosque attendance in Kuala Lumpur were feasible, Wan would have to depend on the transportation provided by her son or a salaried servant in his household. In Wan's view, such dependence is definitely to be avoided.

Wan's household is distinctive in that neither of its female members actively participates in bilateral food or labor exchanges with other women in the compound. This is wholly acceptable, however, in light of Wan's advanced age and the fact that the adult daughter with whom she lives has long been afflicted with severe emotional disorders (widely attributed to the sorcery of a rejected suitor). Yet Wan's household continues to sponsor ritual feasts (kenduri ) when the occasion demands and has also been actively involved in the informal adoption of children born to other women of the lineage. A brief discussion of these institutions will provide additional insights into the nature of intralineage relations.

Feasting

Whenever Wan sponsors a ritual feast, all women of the lineage assist in the preparations, including the fixing of whatever uncooked food offerings are deemed appropriate by convention or contingency. Her two healthy daughters, as well as any female grandchildren who happen to be home at the time, shoulder the heaviest responsibilities before and during such feasts; all women of the compound, however, share in devoting the better part of a day (or even a few days) to helping out. This participation is less a function of Wan's esteemed position in the lineage than a basic imperative of lineage membership. Hence, whoever else might provide labor before, during, or after a ritual feast, women of the lineage usually show up earlier, work harder, and stay later than anyone else. They never complain, for they know that their efforts will be reciprocated by lineage sisters when the time comes; further, they share in the prestige that the lineage and the compound's population as a whole accrue from such feasts.

The only time that all locally resident members of the lineage and its compound work together and converge in a single locale is during ritual feasts. Although women perform the bulk of the labor involved, unmarried males of the lineage contribute as well. In-marrying men, in contrast, often do little more than pray together beforehand and then share in the feast itself. If the scale of the feast is such that additional dining or cooking

structures are required, however, in-marrying men will gather the bamboo and other construction materials and erect temporary platforms and rain shelters adjacent to the sponsor's house.

Expressed differently, the institution of feasting provides villagers, especially women, a clear sense of who their most immediate and generally responsive kin are. Given the ritual and religious significance of the kenduri complex, and the fact that feasts are the most lively and enjoyable occasions for persons of all ages, it can also be said that the bonds among compound residents, and adult females in particular, are both sanctified and celebrated on these occasions. More generally, kenduri serve the crucial purpose of mobilizing lineage sentiment and galvanizing the group's sense of common purpose, shared destiny, and overall social identity.

Whether these latter concomitants of kenduri were more or less pronounced in decades past is, unfortunately, difficult to gauge. Some villagers claim that invitations to kenduri are more selectively extended now than a few decades ago, and that kenduri have become highly politicized. Although such claims are true, they pertain more to solidifications of cleavages and alliances among lineages (and factions formed around their principal luminaries) than to dissensions within lineages. This fact, too, will become quite apparent when we examine the feud between the Hill and Valley lineages.

By any measure, however, the shape of twentieth-century history has overly burdened the feasting complex; for even though the complex persists, most other traditions conducing toward the same ends have suffered irrevocable devitalization. The institution of feasting is therefore expected to accomplish far more than is in fact possible—especially given the proliferation of criteria for allocating prestige, the larger issue being increased cultural heterogeneity and socioeconomic stratification.

Integrally related to this latter issue, and of more immediate concern, is the relatively pronounced economic independence that women belonging to the same lineage enjoy. No Hill lineage women, for example, cooperated with one another during the 1978–1979 cultivation season. Those who did plant rice or lease out their fields chose to work for or engage as sharecroppers women from other lineages or clans. Similarly, all households in the compound derive their livelihood not from lands classed as ancestral but rather from recently acquired acreage (mostly in rubber) held on full title either individually or by a married couple. Hence, while women

of the lineage do share mutual rights over homestead plots and sawah, these rights no longer entail intralineage bonds of agricultural cooperation or exchange. These dimensions of traditional relatedness used to involve not only an objective economic interdependence among the women of a lineage but also a conscious and ritually celebrated sense thereof. This has virtually vanished, however, and given way instead to atomistic and largely autonomous household units whose outside material support comes, if at all, not from immediate collaterals who are also close neighbors but from nonresident children. Little remains of the prior nexus except comparatively unproductive land rights, which, although a trove of great pride and symbolic significance, can be a source of bitter contention. This is particularly true when economic needs and incentives no longer counterbalance divisive intralineage antagonisms and status rivalries involving sisters, cousins, in-marrying males, and compound households on the whole. In sum, the balance of counterposed forces lying at the heart of lineage dynamics during the 1800s and before has become increasingly tenuous over the past century and at present is highly susceptible to invidious intralineage distinctions and other centripetal forces. We will return to this theme in due course.

We have seen that the Hill group is bound together by a single genealogical and territorial point of origin, allegiance and deference to Wan, relatively dense shared property rights within the lineage, and labor exchanges related to feasting. It remains to examine the other factors that unite these women and set them off from members of other lineages, such as the prevalence of informal adoptions within the lineage.

Adoption

The institution of (informal) adoption^[4] serves to redistribute rights over children and enables married couples and formerly wed women to mitigate the social and emotional consequences of sterility, infertility, and infant mortality. It also allows the women of economically hard-pressed households to find "better homes" for some of their children. Of primary concern here, though, is not why adoption exists in the first place, but rather the types of genealogical bonds between natural and adoptive mothers.^[5]

Figure 6 depicts the distribution and direction of adoptions in which

Figure 6.
Genealogy of the Hill Lineage of the Lelahmaharaja Clan, Showing Adoptions Involving Locally Resident Women as Natural or
Adoptive Mothers
Note: Each arrow indicates the transfer of a child from its natural mother to the adoptive mother. Arrows point to adoptive
mothers.

adult women of the Hill lineage have stood as natural or adoptive mothers. Virtually all the transactions occurred between women of the lineage, the majority between women currently associated with one of its two major lines (or future branches). To the extent that sisters and first cousins transfer rights over children with one another more often than do women related more distantly, the institution of adoption also generates a conceptual cleavage within the lineage that is otherwise largely unmarked and generally devoid of explicit sociological correlates. Of perhaps greater structural significance, this line of cleavage will be the most directly affected when all genealogical information pertaining to the uppermost level of the enatic pedigree—that is, the sibling link between Selbiah and Maimunah—is lost. The distinction currently marked primarily by adoption will then assume a qualitatively different hue, for it will simultaneously coincide with and define the existence and boundaries of two separate lineage branches. The clustering of adoptions within analytically isolable segments of a largely undifferentiated lineage thus foreshadows the emergence therefrom of two (or more) less encompassing and structurally discrete units of descent.

Adoptions, then, tend to involve natural and adoptive mothers affiliated with the same lineage, and are typically aggregated within separate descent lines that have the current or impending status of lineage branches. These patterns testify to the pronounced preference—in the transfer of rights over children, but also in most other realms—for dealing with one's closest kin. Thus, and not surprisingly, natural and adoptive mothers are most often related as (natural) sisters. This social fact concretizes another level of distinction within the lineage, just as it engenders differentiation within the lineage branch itself by bounding, isolating, and solidifying sibling sets composed of sisters. Viewed in terms of either genealogical unit, and with respect to clans as well, the tendency for less inclusive groupings to emerge from larger ones that simultaneously constrain various dimensions of their separate identities and activities receives its penultimate expression at this level. The ultimate manifestation of such tendencies appears, of course, in the sibling set itself, most notably in the tensions and dynamics among adult siblings of the same sex.

The extreme modalities of affect and social interaction are most evident precisely in the area of same-sex sibling relations. Of greatest significance here is not that adult women linked as, for example, sisters depend on and

support one another for the satisfaction of innumerable needs—which, in many contexts, could easily be pursued through other social avenues. Rather, this situation of interdependence often exists alongside antipathy, bordering at times on thinly veiled loathing, which is nonetheless almost always submerged—and in any case is effectively denied through feasting exchanges and other public behavior—whenever the moral obligations of female siblingship require it. The sentimental configuration among sisters is thus perhaps best summed up by the notion of hostile dependence. Of course, sisterhood should not be characterized in principally negative terms; even so, the generalized ambivalence that attaches to most interpersonal bonds happens to be exceptionally high in the case of women's attitudes toward their sisters. Hence, whereas the distribution of adoptions within a lineage points to the relative intensity and durability of the different social prerogatives and duties of female kin, it concurrently highlights what could be judged the most problematic and potentially divisive of all kin relations. The more general, theoretical point is that even while sibling ties among same-sex individuals often serve as conduits for the expression of lineage unity and solidarity, they may simultaneously undercut all such unity and positive affect (cf. Kelly 1977).

The ambivalences and tensions among women linked as sisters, and among same-sex adult siblings in general, exist in large part because of the structural equivalence infusing all such parallel-sex relationships. This equivalence has been discussed elsewhere and will be taken up below from a different perspective. Here I need only emphasize that relatedness of this nature gives rise to heavily weighted moral expectations among all parallel-sex siblings, which can be onerous at times as well as exceedingly difficult to fulfill.

I might also point out that sibling tensions of the sort discussed here are far less pronounced among individuals who have never wed, simply because their relationships tend to be mediated indirectly by their parents, especially their mother. These single people, moreover, usually either remain within their natal homes or leave the village altogether. Hence, because they do not establish separate living quarters within the lineage compound or elsewhere in the community, their interests tend to be focused on and defined in terms of a single household. Subsequent to marriage, in contrast, a partial but nonetheless marked divergence of interests occurs among siblings owing to their new conjugal and domestic responsibilities.

The extent to which these may be wholly incompatible with enatic ties is considered in chapter 7. Suffice it to say that marriage and the creation of new household units often aggravate preexisting strains among siblings of the same sex, or in any event create a context conducive to their expression. This despite the fact that lineage continuity and social reproduction on the whole obviously hinge on the bearing of children, which, in this society at least, presupposes legitimate marriage. More generally, the institution of adoption must be viewed against all such dynamics, for, even if only at the conceptual level, adoption helps bridge the structural distance among sisters that is an inevitable consequence of their social reproduction and genealogical aging (cf. Kelly 1977).

Male Lineage Members

The scant attention devoted thus far to the men of the Hill lineage reflects the minimal role they play in most lineage affairs, owing partly to out-migration and the preeminence of female enates in all but a few domains of compound and household activity. As may be recalled from the Hill genealogy, none of the five men of the grandparental generation lives in the lineage compound. In fact, only one of them, who holds the title of clan subchief (buapak ), actually resides in the village. A comparable situation prevails among the married men of junior generational standing, none of whom have settled in Bogang either. Aside from the clan subchief, then, the eldest male of the lineage who lives in Bogang is roughly eighteen years old.

This situation affords little opportunity to discuss sibling antagonisms among adult men of the lineage. It warrants remark, however, that demands for political loyalty underwrite virtually all categories of brotherhood but are frequently honored in the breach (see chapter 7). In this latter connection we should bear in mind that the logic of succession to traditional as well as modern political office makes only partial provision for the structural equivalence of brothers. In the realm of clan titles, for example, rules of succession specify the particular localized clan or lineage that may furnish a candidate for office but usually do not indicate which fully enfranchised adult members merit preferential treatment as potential officeholders. Much of this holds as well for modern political titles, one critical difference being the far greater prestige that attaches to persons

holding positions on certain community councils and, of course, to the village headman. All of this makes for profound resentment on the part of men whose brothers attain roles of leadership while they themselves muddle along as "mere villagers." Here again, then, we see the extreme modalities of same-sex siblingship, which on the one hand engender lineage unity and solidarity but on the other lead to a divisiveness that impedes the realization of common lineage interests.

One further point concerning political offices tied to lineage membership is that the exercise of legitimate sanctions and duties is often necessarily constrained by factors of residence and domestic sentiment. This situation can pose serious dilemmas for politically empowered males who find themselves obliged to conform to the wishes of their mothers or wives, even when this amounts to dereliction of official duties vis-à-vis other lineage members or the clan or village as a whole. Chapter 9 provides ample illustration of these dynamics; I need therefore only cite Wan's position of centrality within the lineage and her profound influence over her adult sons, one of whom is clan subchief. In the course of Bogang's highly charged "adat crisis," this son chose not to censure certain transgressions of his brother, Wan's other son, and in consequence has been boycotted by the other two lineages of the clan. It is difficult to ascertain precisely to what degree the clan subchief's public posture on this matter stemmed from filial loyalties and obligations, as opposed to those associated with immediate brotherhood. Even so, it is clear that narrow kinship loyalties of this sort not only compromise the nature of political representation and leadership more generally, but simultaneously undermine the broad spirit of kinship, and of siblingship in particular, that underlies clanship and village citizenship.

These dynamics all attest to considerable historical continuity. So, too, does the fact that adult males of the lineage still act as guardians over their unmarried sisters and close female cousins, despite that fact that many of these women do not reside in the lineage compound, or even in Bogang. Such guardianship is especially evident in the realm of sexuality, for brothers and enatic cousins continue to be charged with, and to take seriously, the responsibility of safeguarding the virginity and moral standing of unmarried women—upon which rests the honor and prestige of the whole lineage. This situation goes a long way toward explaining the sense of out-

rage and betrayal that young men display when close female enates engage in sexual impropriety. As far as male guardians are concerned, such action amounts to a form of treason, for it constitutes a grave violation of siblingship trust and other sacrosanct norms. Cross-sex siblingship will be further discussed in the following chapter. I need only add here that complementarity rather than equivalence continues to characterize this variant of siblingship, and that this complementarity usually precludes the buildup of antagonisms encountered among adult siblings of the same sex.

The preceding discussion should convey a sense of the structure and dynamics not only of the Hill lineage but also of the other lineages and compounds in Bogang. It is important to recall that Hill constitutes one of three lineages of the Lelahmaharaja clan, which in turn is but one of three major clans represented in Bogang's old hamlet. We will now look at other settlement patterns and household clusters in the old hamlet, both in general and in relation to their new hamlet counterparts.

Variation in Settlement Patterns and Household Clusters

Map 3 provides an overview of Bogang village settlement in relation to such physical landmarks as the nearby mountain range, the local river and its tributary, and the railroad tracks connecting Singapore with Kuala Lumpur and points farther north, as well as such public structures as the mosque and the prayer house, the village meeting house, and community stores or coffee shops. Most important, the map reveals a pronounced clustering of domestic units in accordance with the descent affiliation of household heads. The dotted lines around occupied houses signify common membership in a local clan (or lineage); hence, the households in which most females and unmarried males reside are surrounded by those of enatically related kin. The principal exception lies with the twenty-two households located across from, and on the upriver side of, the road running roughly parallel to the mountains. The settlement pattern of this area of the village, known as the new hamlet (kampung baru ), differs markedly from that of the old hamlet, and it stands at the other end of the continuum from the residentially clustered Hill lineage of the Lelahmaharaja clan.

^[

Map 3.
Contemporary Bogang, Showing Settlement of Major Descent Units

The Old Hamlet

In sharp contrast to the new hamlet, then, most household units in the old hamlet lie very close to households comprising enates in the same lineage, and the same lineage branch in particular. Given the much larger spread and population of the old hamlet, this phenomenon is clearly villagewide. Viewed from a different angle, lineages and lineage branches (and to a lesser extent clans) tend to be aggregated into one or two territorially and conceptually distinct regions within the village.^[6]

As we have seen, the settlement pattern and enatic composition of the Hill compound testify to that lineage's enviable pedigree and long history of impressive political connections. Additionally, patronage activities of former Hill luminaries account for the genealogically heterogeneous composition of one of the two principal settlements of the Biduanda Dagang clan.^[7] Let us now briefly consider the Biduanda Dagang clan's two "neighborhoods."

All but a few of the women associated with the Biduanda Dagang clan reside in one of two village neighborhoods, the upriver or the downriver. The two segments of the clan are of such disparate origin and pedigree that intermarriage is permissible as far as many (but not all) clan members are concerned.

The upriver segment of the clan is not only highly variegated in terms of the parentage and ancestral homelands of adult females (and in-marrying males), but it also embraces three separate lines of descent. Hence, we are not dealing here with a single lineage compound, or even with a compound that invariably and "throughout history" has been associated with a particular clan.

Fatimah, the head of the house set off from all the others (see map 3), is the fifty-five-year-old granddaughter of a Chinese woman who was born in Singapore and adopted while quite young by the nineteenth-century Hill lineage luminary who became Undang. Fatimah's parents and collateral ascendants are all dead, and she has neither siblings nor any natural children. For these and other reasons, Fatimah views herself as having no kin either within the village or outside. In this regard she is extremely atypical; all the more so since she adopted a former neighbor's daughter and reared her as her own for a number of years.

The second line of descent within the upriver segment of the Biduanda Dagang clan is represented by the consanguineally related individuals of three separate domestic groups. The enates of these clustered households also reckon their descent from a woman who was adopted, albeit in quite different circumstances, by the same nineteenth-century Undang who adopted Fatimah's grandmother. In this instance, though, the adoptee was originally a slave (hamba ) who had been purchased in Mecca and brought back to the village to serve her master and his household. Eventually the lowly woman was emancipated, cleansed in a public purification ritual, and given land by her master. Owing to genealogical revision by the former slave's living descendants, this arguably benevolent man is now regarded—at least by members of the three households at issue—as having been merely the adoptive father, rather than the father and former owner, of the Mecca-born apical ancestress. Other villagers, however, if questioned on the matter, tend to emphasize (although in whispers, and with occasional tones of pity) the group's ignoble origins and permanently tainted pedigree. At the same time, some female members of the adjacent and extremely well connected Hill lineage do acknowledge a degree of siblingship with the enates of these Biduanda Dagang households. This recognition of siblingship stems not from common descent, or even from residential proximity, but rather from rights enjoyed by women of both groups over lands once residually vested in, and perhaps partially cleared by, the same nineteenth-century Undang.

The five remaining households of the upriver segment of the Biduanda Dagang clan represent the third separate line of descent in that settlement area. As with the previous example, the women and children of these households constitute both a local lineage and a lineage branch. Similarly, even though they trace their descent from a Biduanda Dagang community who fled wartorn Sungei Ujong in the late nineteenth century, their presence in Bogang, and in this region of the village in particular, also presupposed political largesse on the part of long-deceased gentry leaders, including perhaps the same Undang mentioned above. Their clan affiliation and their relatively recent immigration to the area cause them to be regarded as of decidedly lower birth and social standing than most other villagers. Their peripheral social status, together with that of the other upriver Biduanda Dagang women, is likewise reinforced and reproduced over the

generations by their practice of marrying either unrelated males of the same clan or itinerant Javanese (see chapter 7). The Javanese men in particular invariably lack descent affiliation, collateral kin, and competence in standard Malay (let alone the local dialect); they also occupy the most marginal positions in the community's social hierarchy.

The second and only other Biduanda Dagang community, the downriver segment of the clan, is residentially more isolated (see map 3), being located in a physically separate and conceptually distinct enclave known as the "banana grove" (tanjung pisang ). For reasons not entirely clear to me, they bury their kin in a graveyard located some distance away from the inhabited and agricultural areas of the village, even though the Bogang community cemetery lies only a few hundred yards from one corner of their compound. This practice bespeaks and simultaneously reinforces both their social and residential segregation from the village mainstream and their historically conditional ties—as Biduanda Dagang—with village land.

It merits remark, however, that the members born in this neighborhood seem less directly targeted, as compared with the remainder of the clan, by gentry concerns to avoid marriage with anyone of the Biduanda Dagang group. One possible explanation for this disparity might lie with the relatively homogeneous pedigree of the "banana grove's" enatic core; specifically, that its members are all related through ties of matriliny and are not held to have descended from either slaves or comparatively recent refugee immigrants. It is also relevant that none of the males who have married into this community in local memory have been of Javanese birth or ancestry.

The Lelahmaharaja clan, for its part, is the best-represented descent unit in the entire village, in terms of both numbers of households and the percentage of Bogang's population that it includes in its membership. The clan comprises three lineages that are viewed as related like sisters. In sharp contrast to the situation in other clans, resident members of the Lelahmaharaja clan are acutely aware of both the lineage to which they belong and the modern constraints and benefits entailed in lineage membership. These facts can be explained in part by the settlement within the village of all local clan officers—that is, the clan subchief and the three lineage leaders. Had these men married women from other communities, as occurred

with the heads of the Sediaraja clan, they probably would have gone to live in their wives' natal compounds. If so, they would be less able to mobilize supporters and galvanize lineage sentiment. The more decisive issue, though, is the intraclan feud alluded to earlier, which involves all three lineages and amounts to an alliance of two of them, the River and Valley groups, against the wealthy and extremely well connected Hill group.

The Lelahmaharaja clan's River lineage is dispersed in three noncontiguous settlements, all close to the railroad tracks. Each residential area seems associated with a distinct branch of the lineage. Although a generation or two back female elders probably could specify the genealogical ties among all lineage members, those links are no longer known, owing to terminological practices (teknonymy, birth-order names, and an explicit prohibition on using the names of elders in address) that limit the scope of vertical genealogical knowledge. Nevertheless, adult women do acknowledge their common lineage affiliation and the bonds of classificatory sisterhood that link them and mark their distinctiveness vis-à-vis persons and segments of other Lelahmaharaja lineages.

The clan's Valley lineage claims as its ancestral homeland the area currently settled by the members of eight Valley houses. The adult women of an additional three Valley residences located on the other side of the railroad tracks were also born in that area of the village, as was the head of another household in the new hamlet. The dispersal of the Valley lineage in three distinct and distant areas of the village testifies to the purchase by in-marrying males, a generation or two ago, of noncontiguous plots of homestead land; yet it seems not to have caused an appreciable erosion of lineage solidarity or a decline in the conceptual significance of lineage affiliation. Various factors account for this continuity, the most pronounced being the long-standing feud between the luminaries and spouses of the Valley lineage on the one side and those of the Hill lineage on the other. These two sister lineages occupy adjacent ancestral compounds in a spatial arrangement analogous to that usually obtaining between actual sisters and their households. But here, as in many other areas, siblingship is fraught with extreme ambivalence. Especially in this case, status rivalry, big-man politicking, and competition for socially valued resources (such as personal supporters and state lands suitable for rubber trees) all contribute to the dilution—some would say the very denial—of the prescriptive amity of

sibling relations. What emerges, however, is not a clan with three undifferentiated major lines of descent, but rather one in which these lines (or, more accurately, the members, spouses, and political leaders of two, and to a lesser degree all three) stand in antagonistic opposition to one another. Status competitions and antagonisms of a comparable sort do of course prevail among individuals, particularly women, of single lineages, and among actual brothers as well as sisters (although rarely between cross-sex siblings). But these dissensions do not reverberate so profoundly throughout village society. For that matter, neither cleavages between the "haves" (orang senang ) and "have-nots" (orang susah ) nor between supporters of the two opposed political parties are as pronounced or as public as the well-institutionalized feud between these two sister lineages of Lelahmaharaja.

The absence of any such feud or rivalry in the Sediaraja clan could help to explain why a good number of Sediaraja adults were uncertain about the name of their lineage or contradicted themselves on the matter in the course of formal interviews.^[8] In short, even though all but a few of the households associated with this demographically well-represented gentry clan are aggregated in two residential areas, the jural status of the clan's larger settlement, boasting some fifteen households, remains less than clear. It might be that the Sediaraja descent grouping glossed Upland encompasses two lineages rather than a single unit of that order. This would be consistent with otherwise problematic native statements that there are three Sediaraja lineages within the community.

In any event, the overwhelming majority of Sediaraja women were born in Bogang and take great pride in the undisputed conviction that their maternal predecessors "always" lived there and were the "rulers" (raja ) of the village, along with Lelahmaharaja ancestors, for hundreds of years prior to British administration. At this level, issues of origin point, historical precedence, and descent affiliation are unambiguous and are likewise a matter of prestige and public record.

One of the most general themes to be gleaned from the material presented here is that concepts of origin point, siblingship, and descent continue to assume critical importance not only in how individuals and collectivities orient themselves toward their social environment, but also in how they categorize both themselves and various areas in the village. Yet although these generalizations apply throughout the community, they refer

most specifically to the old hamlet, the settlement pattern of which is highly differentiated in terms of kinship (especially descent) affiliation. It remains to consider how these phenomena are manifested in the new hamlet, where no such differentiation occurs.

The New Hamlet

All but two or three of the twenty-two occupied houses across the dirt road from the old hamlet were built since the 1950s, and the vast majority of houseplots there are also recent acquisitions from the government. Another particularly striking fact about the new hamlet is that most of the adult women who reside there live at considerable distance from their sisters and other close female enates. This often reflects a conscious decision by married women, their spouses, or both, to live apart from the wife's natal compound and thus beyond the usual reach and frequent meddlings of the wife's mother and sisters, collateral kin, and male in-laws. A sizable portion of new-hamlet women hail from outside Bogang, and some were born in Malacca and other neighboring states that have no matriliny or adat perpatih traditions. Insofar as formal adoption into a clan has occurred very rarely for the past few decades, a relatively high percentage of the adult females in the new hamlet lack both clan affiliation and locally resident collateral consanguines.

These comments should not suggest that the inhabitants of the new hamlet are emotionally, socially, or economically disadvantaged because they live among neighbors with whom they have no genealogical ties. True, other "more established" villagers of the old hamlet might regard them as such, or at least as generally distinctive in this regard. But it is important to recall that proximate residence, casual neighborly exchanges of food and labor, and a system of reference and address bolstered by a kinship and religious ideology that stresses ties of siblingship over those of descent, all serve to convert nonkin into kin and to minimize social distance on the whole. In brief, there are no grounds for assuming that new-hamlet women encounter any difficulty in mobilizing the requisite labor resources or contributions of food for funerary or engagement and marriage feasts. More generally, and as elsewhere in Bogang, where neighbors tend to be close enates, many new-hamlet women behave toward one another—at least on the surface—as members of a single lineage.

Lacking, however, is the sense of closure, corporateness, and shared pedigree implied by common lineage membership. Probably, too, women of the new hamlet do not think and feel about one another as do lineage sisters, let alone women who have the same mother or maternal grandmother. Although villagers would state the matter in terms of flesh (or milk) rather than blood, they do subscribe to a version of the Western view that blood is thicker than water. Terminological and behavioral conventions aside, in other words, villagers clearly recognize the distinction between kin and nonkin and behave accordingly. To do otherwise—that is, to collapse the distinction altogether—would be a grave offense to relatives, particularly close kin; it should be remembered that many of these women do have sisters and other enatic kin elsewhere in Bogang or in neighboring communities.

My contention that kinship cuts both ways is also directly relevant to the new hamlet, for the possibly less-encompassing, affectively weaker, and more conditional sort of amity characteristic of relations among new-hamlet women may well coincide with a less pronounced degree of ambivalence than that found among sisters and other closely related female enates. Whatever else these women have in common, such as disparate origins, or commonly lack, such as sawah, they generally do not share rights to property either with their immediate neighbors or with other members of the new hamlet. Nor are many of them destined to be bound together in future co-heirship, for most homestead plots and agricultural holdings claimed by new-hamlet inhabitants are held by unrelated individuals or married couples on the basis of full title over which neighbors and other villagers have no jural interest or control. Consequently, a major source and symbol of petty disputes and ongoing feuds elsewhere in Bogang—namely, joint claims to provisional or residual rights over land—is largely absent from the new hamlet.

Also noticeably inconspicuous in the new hamlet, or at least comparatively insignificant, are social commonality and differentiation based on descent or other kinship constructs. To the extent that this area of the village stands as a separate social community, one can say that kinship is relatively unimportant to its internal organization.

New-hamlet adoptions, for instance, do not replicate conceptual cleavages of the type associated with adoptions in the old hamlet. Although the distribution of genealogical ties between natural and adoptive mothers is

essentially the same as in the old hamlet, adoption does not bound and solidify female sibling sets, or other groups of women, that are residentially aggregated. Rather, since new-hamlet sibling sets tend to be dispersed over wide areas, adoption reinforces bonds between sisters and other close female enates who live far apart. New-hamlet adoptions thus differ from their old-hamlet counterparts in that they enhance the genealogical distinctiveness and autonomy of residentially proximate women. More generally, new-hamlet adoptions decrease the likelihood that the new hamlet will become internally differentiated through membership in one or another kinlike social category.

Affiliation with one of the major national-level political parties may, of course, serve some of the classificatory purposes associated with kinship elsewhere in the village. Since issues of kinship and politics are discussed elsewhere (chapter 9), I would only emphasize here that many new-hamlet residents are philosophically, if not actively, allied with the opposition party. This alliance unites many new-hamlet adults into a loose, but nonexclusive, interest group. It also sets them in sharp opposition—in ways that kinship, for instance, cannot—both to new-hamlet supporters of the dominant party (UMNO) and to Bogang's political elite, who stand behind UMNO, reside in the old hamlet, and are linked through descent or marriage with the extremely wealthy Hill lineage. The conceptual organization of a community or village neighborhood along modern political lines, however, is no substitute for social organization based on kinship. For one thing, political loyalties and alliances tend to be mercurial, and mere association with a political party does not necessarily gain one very much. Additionally, the entire domain of politics is widely held to be anathema and polluting, for it brings out the worst in people, including strife and intense hatred even among siblings. Even so, what is perhaps most striking about the new hamlet is not that political affiliation and proximate residence provide planes for sociability and cleavage, but merely that they constitute the dominant considerations in these areas, since kinship cannot.

Depending in part on the acreage of existing houseplots and the availability of suitable residential land elsewhere in the new hamlet, many of these residential compounds could conceivably become the loci of more densely populated settlements, each associated with a separate line of descent. A similar developmental course might be envisioned for all of the residential and clustering patterns discussed thus far. In actuality, however,

traditional developmental cycles of the sort outlined in chapter 2 have become increasingly uncommon owing to intravillage dispersal and, more importantly, to accelerated rates of out-migration for both males and females. Moreover, no emergent descent line or lineage branch in the entire village will ever become a jurally distinct lineage or clan, for the adat political hierarchy no longer has the institutional means to confer altered political status on descent units, the total number of which is permanently frozen in adat texts. Thus, even though traditional cycles of domestic group development are variously expressed in both of Bogang's hamlets, they are destined to be far less dominant in village social topography in the future.

Another difference between Bogang's two hamlets lies in the immediate social fields of in-marrying males. Married men residing in the new hamlet need not contend with affinal kin on a day-to-day basis, and they enjoy a measure of personal and domestic autonomy typically unknown to their old-hamlet counterparts. But as affinal relatedness also cuts both ways, neither can new-hamlet males necessarily look to men in adjacent households when seeking support for a particular cause, political campaign, or faction. Men in the old hamlet, in contrast, can turn for such purposes to residentially proximate males, that is, to their "brothers" through in-marriage (biras ) (see chapter 7).

The major analytic distinction to be gleaned from a comparison of the two hamlets, then, is in terms of structural variation of the sort noted above, together with a prevalence of attenuated enatic kinship and concentrated title rather than provisional proprietorship. The two hamlets' socio-economic contrasts, for example, are far less pronounced than is the case with certain descent compounds in the old hamlet. This in turn works against an identification between hamlet residence and economic standing, even though it also situates the disproportionately wealthy and politically well-connected households of the Hill lineage compound in conceptual opposition to virtually all other social and residential groups in the entire village.

One final point to bear in mind is that the new hamlet includes just over one-fifth of all Bogang's households. This is to say that the majority of the village population still resides in lineage compounds, or structurally comparable household clusters, whose internal dynamics and tensions continue to be informed by the long-enshrined imperatives of siblingship, descent, and other modalities of kinship. The social significance and cultural

meanings of contemporary kinship, and descent affiliation in particular, do of course vary from one area of the old hamlet to the next, but underlying uniformities are also evident, many of which are realized in the institutions of feasting and adoption. These institutions continue to highlight the structural centrality of bonds among sisters and other close female enates, and simultaneously counter the social consequences of the historically diminished importance of cooperative production and shared rights in rice land.

7—
Marriage and Affinal Bonds

This chapter focuses on contemporary marriage and affinal relatedness, including the implications of upward shifts in age at first marriage, declines in the incidence of unions between persons categorized as kin, and the marked falling off of all variants of local endogamy. I also briefly assess present-day affinal exchange and marriage celebration, declining divorce rates in Negeri Sembilan, and the distribution of divorce in contemporary Bogang.

Rates and Choices in Marriage

Bogang's resident population of 476 individuals consists of 206 persons in the "ever married" category and 270 who have yet to marry, 90 percent of whom are age nineteen and under. Virtually all the individuals in the latter group will marry when they attain the appropriate age, the few exceptions being persons considered of unsound body or mind and women whose mothers consistently demand exorbitant marital payments from potential suitors. In short, as throughout history, marriage leading to parenthood

and adult status is expected of all villagers. Likewise, everyone who has yet to marry assumes that they, too, will sooner or later be matched with, or choose, a spouse and have offspring of their own.

Marriage is viewed as socially desirable and necessary, economically advantageous, and both enjoined and sanctified by Allah and the traditions of ancestors. The institution of marriage thus retains its most basic historical features as well as its central stature in local society and culture. Similarly, childless marriages of more than a few years' duration are still essentially unheard of, since the mechanism and overall ease of informal adoption effectively counters the social consequences of sterility, infertility, and infant mortality. Childbirth in the absence of legitimizing marriage, however, does not exist. Extramarital sexual activity, a heinous, broadly tainting offense, is rare by Western standards. Marriage, then, is still, as in the past, both the expected and eminently "natural" norm and the only path to adulthood—although it should not be equated with adulthood itself, which, to emphasize a point made earlier, derives only from culturally sanctioned parenthood.

The universal expectation of marriage is succinctly expressed in formalized verbal interchange concerning relatively young villagers about whom one lacks precise social information. Here, inquiries concerning a person's marital status typically precede all other questions bearing on his or her general social standing. In attempting to discern the marital status and broader age-grade membership of adolescents and their immediate seniors, moreover, Malays do not ask, "Are you married?" but instead, "Are you already married?" In the case of someone who has yet to wed, the response is never a simple "no"; the only acceptable answer is "not yet," which expresses intent as well as unequivocal acceptance of the norm. The same "not yet," it may be noted, is the only appropriate rejoinder to questions pertaining to whether a young couple has children—which is invariably the first question put to strangers or distant relatives whose married state has just been confirmed. Concerns with affinal connectedness notwithstanding, the institution of marriage is commonly viewed in light of its reproductive functions and contributions to social continuity.

Various other attitudes toward marriage have changed over time, however, one being that girls need not be married off by age fourteen or fifteen, as occurred up until the 1930s, but should instead reach age twenty or twenty-one before they wed. Table 15 reveals a progressive rise in fe-

male age at first marriage, and a current norm of 21.1 years.^[1] Such trends prevail in part owing to contemporary convictions that early female marriage was a convenience of former and culturally backward times, and that girls today should complete their formal education and acquire some work experience, preferably of a nonagricultural sort, before taking husbands. Since the age at which young men marry has remained far more constant over time, the age differential between bride and groom at first marriage has markedly declined.

Without question, the delayed onset of initial marriage has occasioned new problems for parents and enatic guardians, who must now contend with an additional five to seven years of female adolescence and sexuality. These problems are compounded by the increasing legitimacy of Western notions of "romantic love," which, as we will see, have altered traditional male-female relations, and at the expense of conventional criteria underlying spouse choice in particular.

Table 16 reveals that approximately two-thirds of individuals in the "ever married" sample have had but one spouse. Roughly one-quarter have been married twice. Most of the individuals in the four-plus category are men with a history of serial monogamy, as opposed to polygyny, one

of whom acknowledged having had more than ten wives and as many divorces. The practice of polygyny contributes to the incidence of multiple marriages among men, but it is fairly uncommon and typically limited to two wives at any given time. (Less than 8 percent of Bogang women and 11 percent of the men in the sample have ever been partners to polygynous arrangements; each instance involved a man with two spouses.)

Throughout Bogang the general attitude is that neither divorce nor widowhood should preclude remarriage, and that in most cases the formerly wed are better off to seek out a new marital partner than to remain single. Society as a whole is also held to derive clear benefit, in the sense of being better protected, if sexually active adults are able to satisfy their physical needs in the context of legitimate unions. As a consequence, divorced and widowed women under forty-five years of age encounter strong pressures to find or accept new husbands lest they disrupt other conjugal relationships or cause community scandal. Such concerns appear far less pronounced with respect to older women owing to the local assumption that sexuality greatly declines among females after age fifty. Divorced or widowed males, in contrast, need no such encouragement, insofar as men are always eager to remarry, even if they find themselves suddenly widowed or divorced at the age of sixty or sixty-five. Indeed, table 17 reveals that less than 4 percent (3 of 88) of Bogang's "ever married" men are currently widowed or divorced, as opposed to a full 27.5 percent (35 of 127) of the women. Expressed differently, Bogang's community of formerly married adults is only 7.9 percent (3 of 38) male and 92.1 percent (35 of 38) female.

In sum, remarriage occurs frequently, especially for men, virtually all of whom are married at the time of their deaths. Various factors account for this situation, including social pressures as well as wholly pragmatic considerations concerning the advantages of economic cooperation and the sexual division of labor and resource allocation, most notably houses and land. The occurrence of multiple marriage, however, must also be examined in light of the prevalence of divorce. Polygyny aside, second and subsequent marriages occur only as a result of terminated unions, the over-whelming majority of which follow from divorce rather than death (see table 18). The broader issue here is that differential marriage rates within the village, and among persons of the same sex in particular, are best approached by focusing on the causes, distribution, and implications of

divorce. These topics will be addressed below, after we examine recent trends in spouse choice and marital exchange.

Before the turn of the century, the ideal groom-to-be was minimally an industrious man with a demonstrated competence in agricultural and trading activities, as well as Koranic recitation, adat lore, various elocutionary skills, and one or another domain of ritual knowledge (ilmu ). He was also at least a few years older than his bride-to-be, of generally comparable status and untainted pedigree, wholly free of physical disabilities, and a previously unwed bachelor, rather than a divorcé, a widower, or a man with one or more current wives. Further, he stood as a nonenatic relative, preferably a close cross-cousin such as a MBS or FZS. In any event, he should have been a man born in the village or elsewhere in the district of Rembau who was held to be of the "Malay race" and not simply a Muslim (or a Malay) by virtue of conversion, cultural assimilation, or both. Corresponding qualities were sought in brides-to-be, one principal difference being that a girl's prior marital status and virginity were of greater social concern than her potential husband's. A girl's competence in cooking, washing, sewing, and agricultural labor also counted for much more than her mastery of Koranic verse and verbal skills, since the latter, although not wholly confined to men, fell squarely within the male and public domain.

Traditional marriages were arranged, with senior kin, especially female enates, playing the principal role in spouse selection. The custom of arranged marriage coincided with a strongly sanctioned sexual segregation, such that most brides and grooms never really interacted prior to their weddings. Romantic love, as conceptualized in the West, was a nonissue. To the extent that positive enduring sentiments akin to those characteristic of Western marriage did exist, they developed long after the formation of marriage ties.

The practice of arranged marriage served to reproduce the social control enjoyed by elders, who conveyed their own standards and wishes through spouse selection (and other areas of marital management) and could in fact gain considerable prestige and utilitarian benefit from an ideal union. As discussed in chapter 2 and below, numerous advantages derived from solidifying relations and generating exchange networks with nonenatic kin residing in Bogang or elsewhere in Rembau, particularly if the eligible marriage partners met the other qualifications outlined above.

In Negeri Sembilan, we would expect declines in the incidence of ar-

ranged marriage—in conjunction with changes in the criteria for spouse selection—to transform or disrupt conventional exchange and social relations, and to engender new lines of alliance and cleavage. This process is quite clear in twentieth-century Bogang, for, along with the proliferation of "free marriages," there has been a progressive devaluation of kin connectedness and genealogical and territorial origins, as well as a trend toward purely economic and prestige criteria dominating most other issues concerned with the suitability of potential spouses. Table 19 provides a chronologically graded breakdown of marital choice in terms of husband and wife's kinship relationship prior to marriage. Viewed from the perspective of women, and contrasting the pre-1960 era with 1970–1979, one can see critical reductions in the incidence of FZS and other patrilateral unions, as well as those involving MBS and other matrilateral kin. (It should be noted, though, that marriages with FZS and MBS were rather uncommon even before 1960, accounting for less than 12 percent [14 of 118] of all pre-1960 marriages for which I have data.)^[2] In addition, a far more encompassing historical process is now evident, whereby all categories of kin are increasingly passed over in favor of persons without any prior ties of kinship. More than three-quarters of all marriages in the 1970–1979 period, for instance, involved wholly unrelated individuals. For women, this upward shift represents a 25 percent change over the pre-1960 period, and for men, even larger.

Since village matchmakers and spouses-to-be nowadays commonly turn to unrelated individuals for marriage partners, we must ask where all of these "strangers" or "other people," as they are known, are coming from. If the spouses of Bogang's inhabitants hail from Bogang itself, or primarily from the (relatively) culturally homogeneous district of Rembau, one could expect some continuity in the maintenance and spatial disposition of traditional affinal networks. But this is not the case. As can be seen from table 20, the trend toward unions with strangers has coincided with a marked decline in all varieties of territorial endogamy. A comparison of the pre-1960 era with 1970–1979 reveals a 43 percent decline in intravillage marriage, and decreases of 38 percent and 22 percent for endogamy at the levels of district and state, respectively. Particularly striking is the attendant rise in the rate of marriage with persons not from Negeri Sembilan—198 percent from before 1960 to 1979.

Marriage with individuals of other states can only accentuate the social

isolation discussed in the preceding chapter. For one thing, affinal exchange and other interactions among in-laws grow more cumbersome and expensive as the distance separating the kin of husband and wife increases. In addition, and especially nowadays, a man from outside Negeri Sembilan is unlikely to hold the institutions of adat perpatih in high regard. If such a man takes a Negeri Sembilan woman as his wife—and many are notoriously reluctant to do so—he will be less than enthusiastic to settle in his wife's natal compound and live among her kin. Of major concern here are perceptions of excessively restrictive property conventions, which continue to receive bad press throughout the Peninsula, not only as remnants of backward times, but also as being out of keeping with Islamic orthodoxy and unfair to male progeny. Also relevant are widespread views that Negeri Sembilan women have long been the inheritors and merciless practitioners of sorcery and various forms of "love magic," against which men have less than a fighting chance should they appear to lose interest in or otherwise cross their wives. As a result, men from outside Negeri Sembilan commonly avoid marriage with local women, or else they settle at a comfortable distance from affinal kin if they do decide in favor of a Negeri Sembilan woman.

To further complicate the situation, a growing number of males born and wed in Negeri Sembilan share many of these same sentiments. This is readily apparent from the settlement pattern and kinship composition of Bogang's new hamlet, where numerous adult men live who managed to break from, and not simply postpone, uxorilocal residence. In one sense, however, such cases are but the tip of the iceberg, the remainder of which appears in the absence from the village of large numbers of former residents between the ages of twenty and forty. Needless to say, all of this makes for extremely dispersed lineage membership and social isolation.

Men from other states do of course marry Negeri Sembilan women and proceed to settle in their wives' natal compounds. Although the aggregation of lineage membership need not be compromised in such cases, many of these individuals are in a sense cultural liabilities, for they typically know next to nothing of the extensive lore, or the symbols and meanings, of adat perpatih. They are thus wholly unable to contribute to the reproduction of that encompassing domain of cultural knowledge—tied in with etiquette, ritual exchange, local spirits, indeed all of social life—that, through benign

neglect coupled with Islamic and other homogenizing and essentially modernizing forces, is presently languishing on the verge of extinction.

Cultural reproduction notwithstanding, the avoidance of all kin as potential spouses has pronounced implications for local social relations. This passing over of everyone with whom one has a history of collective residential or associational ties has fostered the demise of a basic mechanism of cohesion and alliance, for present-day marriages and affinal transactions tend not to join non-enatic kin from Bogang itself or from other villages within Rembau. Instead, they typically link one particular household and lineage branch with persons and groups of extralocal birth who reside far from the village. This externalizing process is consistent with twentieth-century shifts in property and productive relations and is in fact a direct consequence of basic economic transformations, including the monetization of most productive transactions and relationships, processes of economic concentration and stratification, and stepped-up out-migration by both males and females. Hence, spouses are increasingly sought from among urban co-workers, salaried government employees, and others of comparable social standing who show some economic promise. To paraphrase one man's assessment of these trends, "a potential spouse's status [pangkat ] counts for everything nowadays, whereas questions of ancestry and descent [keturunan ] amount to nothing."

The progressive dissociation and ever-widening gap between status and ancestry has undermined institutions of affinal exchange and interdependence both within the village and among communities that share an essentially uniform culture. Moreover, as concerns with prestige and status-group endogamy take precedence over all the other potential benefits and disadvantages of legitimate unions, marriage serves increasingly to stratify the village economically and to erode the basic cultural homogeneity deriving from residents' common ancestry and shared experience. All such trends have seriously alienated the middle-aged and elderly, who rightfully conclude that community youth reject many of their fundamental values and notions of propriety. Additionally, some older residents feel deeply betrayed by the newly wed and their parents—especially the wealthiest among them—since marriage with outsiders can convey the message that eligible youth of local birth are somehow "not good enough" for these individuals. As with incestuous unions, as well as specifically political be-

trayal and treason, the cardinal issue here is loyalty to collective norms, or, more precisely, what is perceived to be overt disloyalty motivated by narrow self-interest (cf. Schneider 1976, 167).

Thus far the discussion of marital choice has centered on historical changes in the conceptualization of preferred and ideal spouses, and in the frequency of marriage with genealogically defined conjugal partners. These issues pertain to one side of the marriage coin and have no direct bearing on the current status of traditional prohibitions relating to marriage or sexual behavior on the whole. Indeed, there is little evidence to suggest a marked conceptual relaxation—let alone actual transgressions—of traditional injunctions associated with marriage or sexual behavior.^[3] That is to say, the kinship-based regulation of marriage and sexual activity has not broken down but is instead undergoing certain highly specific changes.

Historical continuity in the conceptualization of incestuous unions and cross-sexual impropriety appears most strikingly in attitudes and interactions involving brothers and sisters of the same parent(s). A culturally enjoined avoidance continues to characterize brother-sister relationships throughout virtually all stages of their lives. With the onset of puberty, for example, failure to distance oneself from a sibling of the opposite sex leads to intense feelings of embarrassment and shame (malu ). Cross-sex agemates of the same lineage branch and lineage exhibit many of these basic qualities of tension and avoidance. So, too, though to a lesser extent, do "weekend villagers," most of whom are unmarried or recently wed young people who work and live in urban areas and return home on weekends. With "weekend villagers," however, something akin to joking relationships occasionally develops, in that these individuals may engage in suggestive verbal banter and repartees of a sort altogether absent from other cross-sex interactions. Unmarried males in their late teens and early twenties are permitted greater leeway in this area than their female counterparts, but they, too, are likely to be chastised for being cheeky and "itchy" (gatal ) if they act overly familiar with females of their generation.

In the end, males still are responsible for safeguarding the interests and sexuality of same-generation females within the lineage, and lineage branch especially, and are therefore strongly discouraged from any actions that might jeopardize these women's moral standing in the eyes of the community at large. Such responsibilities have grown progressively burdensome in recent decades, the more so because females are now at least twenty

years of age when they first marry, not fourteen or fifteen. This fact alone can lead to seriously strained relations between young men and their lineage sisters, since young women today enjoy much greater freedom of movement than they used to and believe that they should select their own marital partners, not accept spouses of others' choosing.

Affinal Exchange and the Celebration of Marriage

Present-day marital transactions include one major addition to customary cycles of affinal prestation. This involves mostly gifts of food and apparel, or pemberian ("gifts," "presents"), along with cash payments, or hantaran ("things that are brought, delivered, or sent"), all of which pass from the groom's to the bride's side at about the same time as the handing over of marriage gold (mas kawin ).

The pemberian consists partly of elaborately decorated cakes and other beautifully garnished food dishes destined for postwedding consumption; it may also include makeup, a handbag, shoes, and the like, which are put away for the bride's later use. The hantaran is a negotiated sum corresponding generally to the status of the bride's immediate kin and the groom's presumed earning ability, even though it also varies in accordance with the circumstances of the wedding. Thus it can exceed M$1,500, but it may amount to a nominal, embarrassingly small fee, as in the case of "shotgun weddings." This money, which originates in the groom's natal household, is narrowly conceived as a gift to the bride. Yet since it is commonly used by the bride's parents to help defray wedding expenses, effective rights of disposal lie with them rather than with the bride herself. In this sense it constitutes a structural analogue to the bride's mother's share of the adat money (wang adat ), which it has basically superseded both in monetary and prestige value.^[4] In historical terms, this sum of money can also be seen as partial compensation to the bride's mother, who nowadays neither gains a residentially proximate son-in-law on her daughter's marriage nor retains full rights over her married daughter, since most newlywed couples settle in Kuala Lumpur or other urban areas, not in the woman's natal compound.

The temporal dimension of present-day marital exchange also differs from that of the preindependence era. For one thing, three-day weddings, known locally as bergolek , have effectively been abandoned in favor of ritu-

ally streamlined and less expensive ceremonies lasting only one day. The elaborate weddings of earlier decades included displays of the Undang's sacred regalia, which were blessed and sanctified by shamanic pawang and then carried about in circumambulatory processions harking back to pre-Islamic (Hindu) times. In contrast, contemporary weddings not only dispense with these features but also require that most marital transactions be conducted within twenty-four hours. These shifts are seen as having numerous advantages, both in reducing the sponsors' wedding costs and in greatly simplifying the logistical problems of nonresident kin. Yet a clear trade-off exists, for certain exchanges and wedding rituals must be performed in a rather hurried, haphazard fashion—or left out entirely—lest the most basic transactions and events (or what are currently defined as such) not be concluded within the allotted time frame.

This inattentiveness to the adat of marriage and affinal exchange is a source of great concern to middle-aged and elderly villagers. The problem is not merely that local youth know next to nothing of adat, although this is patently true. Rather—and in some ways far worse—too few adults seem to care about the proper orchestration of adat rituals or whether the cultural knowledge underlying these rituals will endure. Even those who do show a sincere commitment in such matters often do not know the proper meanings and order of the ritual sequences in question. Further aggravating this situation is the fact that the highest-ranked clan officer in all of Bogang (the clan subchief of the prestigious Lelahmaharaja clan, who belongs to the Hill lineage) has himself transgressed a basic tenet of adat and is thus barred from overseeing the engagement and marriage rituals for most persons in his clan. As will be discussed in chapter 9, the circumstances surrounding this transgression have precipitated the villagewide schism known as the adat crisis (krisis adat ). Owing to the lines of cleavage and alliance engendered by this profound moral dilemma, marital exchange, together with the institutions of feasting and prayer, has become highly politicized.

The other major temporal shift in marital prestation derives from the same economic considerations responsible for weddings being compressed into a single day. Interestingly, this second trend moves in the opposite direction, insofar as the creation and full sanctioning of conjugal bonds occur over a period of months, even a year or more, rather than three or four days. I refer to the increasing prevalence of conditional or deferred unions, known as nikah gantung , or "hanging marriages." In these in-

stances the specifically Islamic dimensions of marriage ritual, the akad nikah, assume preeminence, and are performed much as in a standard wedding. As a result, the marriage is wholly acceptable in the eyes of Allah and Islamic religious law. The critical difference is that the hantaran is not presented to the bride's relatives or spokesmen until well after the union is religiously solemnized. Instead, the two groups of kin involved agree to postpone its payment; they then reconvene over a series of ritual feasts (known as bersatu , "to make one") to effect all outstanding transactions and thereby fully legitimize the marriage. During the intervening period, which may range from a few months to a year or more, the couple joined in conditional wedlock must refrain from all physical contact, not to mention sexual activity and cohabitation. Any breach of these injunctions results in severe censure, for, however acceptable and final the marriage is from a strictly Islamic perspective, so far as the community at large is concerned the couple do not yet enjoy the full status of husband and wife.

One advantage of these unions is that an eligible bachelor who can raise some but not all of the capital required for marriage payments and feasting expenses can still publicize his desire to wed a particular girl and simultaneously establish provisional but nonetheless exclusive conjugal rights over her. A deferred marriage may be of even greater benefit to the girl's relatives, who can thereby disperse marriage costs, the brunt of which they shoulder, over a period of months. Issues of conjugal maintenance are also relevant, since the man is expected to support his wife even though she resides in separate quarters (typically with her own parents). In the case of a husband who serves in the army or police or has salaried civil-service employment, this financial assistance may be an appreciable and much-welcomed addition to the wife's household coffers (Lewis 1962, 180), especially because her parents will have major expenses when they sponsor the rituals and feasting that finalize the marriage.

Another incentive for deferred unions is to minimize the shame and embarrassment of cross-sex impropriety. If an unmarried youth is known or presumed to have been overly intimate with someone of the opposite sex, one possible solution is to plan immediately for a deferred marriage. Such an alternative exists only if sexual activity is assumed not to have occurred; if, however, this assumption is not widely held, the only acceptable recourse is a "shotgun wedding," or bidan terjun —a term that calls to mind the unexpected, and hence hurried, calling of a midwife, as happens when a pregnant woman experiences labor contractions sooner than anticipated.

Before turning to some other hallmarks of contemporary marital prestations and weddings, I should stress that affinal exchanges and marriage feasts provide the principal occasions for the households of bride and groom to publicize and render concrete their most zealously treasured claims to status and prestige. Stated differently, a household stands to gain (or lose) considerable esteem by dint of the weddings it sponsors for its children. Aside from the issue of cash outlays, the ritually demonstrated commitment to things modern (and Western) bears heavily on one's prestige standing in the eyes of the community. Expressions of such cultural affinities may be advanced in myriad ways, such as in the attire of bride and groom, the display of marital payments, and the music accompanying the wedding; indeed, rivalries among wedding sponsors can become extremely intense, with one-upmanship knowing no bounds in terms of extravagance and creativity. Further, since all but shotgun weddings are usually broadly advertised up to a year in advance, and are discussed for months or even years afterward, these competitions are of extensive duration and are forever being revived by pointed references to the quantities of food and cash consumed in the course of marriage festivities.

But while everyone eagerly recalls the most lavish and joyous weddings, most people prefer to forget those hastily arranged unions effected for the purpose of containing community scandal. Such marriages, which cluster among low-income households, bring enduring shame and embarrassment to the parents and lineages of the two individuals concerned. Most relevant in the present context, all such weddings deprive parents of the opportunity both to display their household wealth and urbanity and to substantiate their claims to impressive social networks (as would, under normal circumstances, be revealed by the volume of guests and the number of local notables among them). When a forced marriage involves a couple's only child, a once-in-a-lifetime chance is effectively lost. The personal and social honor at issue in such a loss accounts for the morose atmosphere permeating shotgun weddings, for everyone can appreciate what is at stake. More generally, a proper wedding, let alone one that bespeaks a certain amount of prestige, requires months, even years, of capital accumulation and planning. Thus, both on logical and empirical grounds, the wealthy are ideally situated to retain or upgrade their positions of esteem through the weddings they sponsor, just as the poor are the most likely to slide (further) down the prestige ladder.

Among the most sought after and costly additions to contemporary wedding festivities are energetic performances by amateur "rock 'n' roll" bands. Unmarried males embrace these innovations with great enthusiasm, if only because the entertainment attracts large crowds and thus allows them a rare opportunity to gaze at unwed females. It also offers them a chance to engage in Western-style dance should they get up enough courage to ask a single girl to accompany them to the wooden dance platform. Few village girls risk soiling their reputations (or incurring the wrath of their parents) by accepting such invitations; but there are usually a few "out-of-town" girls who accompany the musicians and tirelessly accept most requests to dance—thereby providing excellent negative examples of ideal female behavior so far as most adults are concerned.

Even the adults who pay for such modern entertainment, however, sense that it is at sharp variance with Islamic codes of conduct, which are generally interpreted as prohibiting dancing, associated forms of hedonism, and all carefree abandon. Here, then, the contemporary requirements of advancing or simply maintaining one's prestige within the community entail an explicit violation of God's will. This problem has in turn spawned what might emerge as a further innovation in the cycle of marriage rituals sponsored by the rich—that is, postwedding feasting aimed at expiating the sins that attend weddings celebrated with Western-style music and dance. I attended one such evening feast, which the wealthy mother of the groom described as essentially twofold in objective: first (though not necessarily of greater importance), to obtain Allah's blessing for the newlywed couple; and second, to effect an across-the-board cleansing (kasih bersih semua ). The groom's mother, who sponsored the affair along with her husband, elaborated on the "cleansing" as follows (which I paraphrase from field notes):

All of the young people wanted a modern, Western-sounding band to play at the wedding, like the ones we had at the marriages of our other children. So we made provisions for the music and dancing, even though Allah doesn't approve of these sinful doings. Since it is especially wrong for those of us who have made the pilgrimage to Mecca to sponsor or otherwise participate in such things, this feast is to ask forgiveness or pardon [ampun ] from God.

Various issues strike us when we stop to consider the symbolic importance at weddings of Western-style music and "pop rock" dancing. Perhaps

most obvious is the explicit clash in ethos symbolized by the wane of long-established musical forms of the stylized gong and rebana drum variety, and their progressive displacement by modern music built around the eminently suggestive lyrics and gyrations of long-haired young men sporting bell-bottoms and dark glasses. These clashes sometimes take on a literal "battle of the bands" aspect, as when men seated on the verandah of a wedding sponsor's house commence a gong and drum sequence accompanied by rhythmic chanting, only to find their Sufistic praise of the Prophet drowned out by overamplified electric instruments. Such competitions may occur repeatedly over the course of a single evening, since the hire of both types of musicians is increasingly seen as the ideal, however beyond the financial reach of many households it may be. Such clashes can, of course, be less fierce, as when guests flock to the evening wedding festivities expecting to see and hear conventional music but are greeted instead by a modern band running through Malaysia's current pop hits.

For adult villagers, the most disconcerting feature in all this is youthful disregard of sanctified bodies of religious knowledge and normative constraints. I was told on numerous occasions, for example, that many adolescents living in Bogang cannot even speak Malay properly, as indicated by their difficulties in passing secondary-school language examinations. Worse yet, they act as if they were immortal, which is to say they evince no concern with—or fear of—God or the afterlife. The elderly find a ready explanation for these errant ways: things Western and profane, especially the contemporary fads and fashions touted in the mass media and urban centers, have seduced youth away from substance of true value.

In summary, shifts in marital exchanges and wedding festivities attest to social processes generated largely by economic transformation and Western-oriented modernization. So, too, do major changes in such areas as age at first marriage, cross-cousin unions, and endogamy. It remains to consider how these broadly ramifying trends are realized in marriage durability and in the nexus of affinal institutions.

The Distribution and Decline of Divorce

In chapter 2 we noted that nineteenth-century divorce resulted in the immediate and permanent rupture of affinal exchange between the kin of estranged husband and wife. A man's divorce effectively canceled his rights

both to the house he likely built and provisioned himself and to all land that he and his wife prepared for residential or agricultural purposes. Additionally, upon divorce the husband had to leave his wife's compound and also sever his relations with his own children. Divorce was perhaps less disruptive for women, for they neither relocated nor lost contact with their offspring as a consequence. Nonetheless, even though divorcées bore no particular stigma and could remarry easily enough, a woman's divorce could also be traumatic and regrettable.^[5] Most of these generalizations hold at present as well. Why, then, are divorce rates so high, especially since Malays strive for harmony with everyone in their social universe?

Divorce accounts for over 65 percent of all terminated unions in which current female residents of the village have been involved, and 73 percent in the case of males (see table 18, above). Such data constitute incontrovertible evidence as to the prevalence of this phenomenon. As we shall see presently, the ease of obtaining divorce clearly contributes to its high incidence. So, too, do numerous other factors, including relatively mild social and economic implications of divorce as compared with other societies. In the final analysis, however, divorce rates in Bogang and Negeri Sembilan as a whole are best approached by focusing on the problematic position of married men in relation to their affines, and on affinal demands on male productivity in particular.

Although I am concerned here with divorce (cerai [hidup ]) in its most final form, I should point out that divorce as a process comprises three distinct stages. Each stage is effected by a husband reciting the divorce formula, "I repudiate you," to his wife. One such repudiation, or talak , is not irreversible so long as it is revoked within one hundred days. The same holds for the second talak. Three repudiations, however, constitute an irrevocable divorce; persons thus separated are no longer considered husband and wife. Should they resolve their differences and decide to resume cohabitation or sexual relations, they are required to marry again, which can in turn occur only if the woman has been married and irrevocably divorced in the meantime, and has in each case also observed the one-hundred-day-long edah restrictions (see Taylor [1929] 1970, 113–114; Djamour 1959, 110–113).

Divorce is primarily a male prerogative so far as Islamic doctrine and village ideology are concerned, especially since males alone are entitled to utter formal repudiations.^[6] A woman may, however, given proper grounds,

request a decree of divorce from the local kadi. If she convinces the kadi that her husband has abused her or failed to provide her with financial maintenance (nafkah ) for over four months, for example, her marriage will be judged irrevocably terminated. But since female initiative of this sort occurs relatively infrequently, the majority of those who appear before the kadi to pursue or simply register divorce are males.

Even though women enjoy fewer jural prerogatives with regard to divorce than do men, other, more cathartic strategies exist that are virtually guaranteed to result in divorce should they so desire. One such strategy involves a woman leaving her husband and returning to her mother's home. Aside from conveying the unambiguous message that she no longer seeks her spouse's company or partnership, this action causes the husband extreme embarrassment and shame. As a general rule, the only face-saving alternative available to a man so abandoned and humiliated is to fully repudiate his wife. Thus, tactics of shaming are both widespread and highly effective.

It should be evident that divorces are rather easily procured. Equally important, the cash expenditures involved in divorce are usually negligible—a minor registration fee. For that matter, the financial responsibilities of divorcés are frequently nominal, since it is virtually impossible for women to obtain child support from former husbands who fail to cooperate and in any case typically relocate to other villages. The legal and economic requirements of divorce, then, do not necessarily hinder its occurrence. But neither do they actually encourage divorce, for they are really relevant only in situations where marriage bonds or affinal roles are strained or marked by antagonism. We are left, then, with questions concerning the manifestations and structural origins of these tensions.

Throughout Bogang I encountered remarkable consensus regarding the divisive issues in marriage and the assignment of "fault" in divorce—that is, that husbands feel the tensions of marriage more acutely than wives, and that the husband's, rather than the wife's, behavior is usually the major cause of divorce. More generally, marital difficulties and divorce derive from a man's being insufficiently attentive to his obligations as husband and son-in-law. These points are well illustrated by the comments of two current residents of Bogang.

Musa, a man in his early forties, said that divorce arose because many married men simply "don't follow the established rules" (tak ikut per -

aturan ): they're lazy, too quick to sell off household possessions, and basically expect to "eat for free." Male pride and self-righteousness figure in as well, Musa observed, particularly since men commonly refuse to acknowledge any of their faults or poor decisions.

A slightly older man named Abdul Ghani, who served for twenty-five years as a clerk at the office of Rembau's Islamic magistrate, which certifies all local marriages and divorces and attempts conjugal reconciliations wherever possible, expressed similar views. In Abdul Ghani's experience, men are usually to blame in situations of divorce because they are not straightforward in their dealings with their wives. The basic problem is that too many men like "the good life"; they enjoy gambling and alcohol^[7] and are overly inclined to purchase on credit. Aside from wasting their earnings instead of handing them over to their wives, men often run up debts and then falsely tell their wives that the outstanding debts have been paid off.

The other side of the coin, in Abdul Ghani's view, is interference by the wife's mother and her other close kin. By virtue of residential proximity alone, a woman's immediate relatives usually know what goes on within her household and how her husband behaves both in and outside of the home. Fundamentally, the pressures from a man's in-laws can be overbearing at times, especially since affinal kin nowadays often goad in-marrying males into buying their wives Western-style consumer goods, such as radios, televisions, and refrigerators, that are frequently beyond the men's means.

Of course, such affinal demands are nothing new. Contested rights over the productivity of in-marrying males have always marked affinal relations and social reproduction, as well as been a major cause of divorce. What is new is that today's married men share a much narrower range of social and economic interests with their enates. Accordingly, the contested rights no longer involve two groups of enates linked through marriage; instead, they typically involve a solitary man's strivings to maintain his personal and economic autonomy, and to a lesser extent the autonomy of his household, in the face of countervailing pressures from his wife's kin.

A more general theme running throughout the data on divorce is that a man's prerogatives and obligations as a father are rarely at issue in divorce, whereas his role as a husband and in-law is frequently the source of major contention. Essentially, then, a man's ties with his children are not felt to

be especially problematic, from either his own perspective or that of his wife or her kin. (This should come as no surprise, since a man's wife and immediate affines assume the bulk of childrearing and socialization responsibilities, and since a man's role as father is not nearly so socially necessary or culturally valued as that of a woman as mother.) Interests and privileges regarding children, then, do not seem to exert any destabilizing effects on marriage. But neither do they work in the opposite direction, for if they did we would expect far fewer divorces subsequent to the birth or adoption of a couple's first child than in fact occur; we would also encounter more concern with rights over the children of individuals contemplating or actually party to divorce.

Instead, we find a situation in which tensions focusing on married men's productivity as husbands and in-laws are the dynamic and determining factors in the relative durability of marriage bonds. Such tensions can, if present in any scope or force, go a long way toward destroying a marriage; if relatively absent, however, they can generate the conditions under which a marital union will likely endure until one spouse dies. My references to degrees or levels of tension, antagonism, and the like, are admittedly rather imprecise, being based on villagers' after-the-fact reconstructions of their own subjective experiences. Nonetheless, as we shall see, the data suggest that such tensions vary significantly, and are at the highest pitch—or at least most acutely felt—in the case of in-marrying males of low economic standing or promise. In these men's lives, the discrepancies between affinal demands and their own successes as household entrepreneurs are most pronounced. For that matter, because married men judged to be poor providers usually lack proprietary rights to rubber land, concerns with property loss through divorce cannot exert any stabilizing influence on their marriages.

The situation is strikingly different for village men who are considered "good providers," men with impressive incomes from rubber acreage they themselves acquired (and work or lease out), from government pensions, or both. A wealthy man, or one who displays economic promise, is best situated to satisfy affinal demands for household commodities and other emblems of prestige. Moreover, a man with substantial capital investments either in his wife's home or in land within her village has much to lose as a result of divorce. So, too, does his wife. On both sides, then, the incentives and satisfactions of economics and prestige contribute to the durability of the marriage.

These arguments are consistent with a striking feature of divorce data that has eluded previous observers, that is, the inverse correlation between monthly household income and the number of divorces per ever-married resident. Specifically, we find a marked concentration of divorce among households at the lowest end of the income scale, and a generally progressive decline in the divorce rate as income increases (see table 21). These associations exist for persons of both sexes, although they are more uniform for women. Among households earning less than M$100 per month, for instance, the female divorce rate is 83.3 percent (15 of 18). This figure diminishes to 53.3 percent (16 of 30) for women of households with monthly incomes in the M$100–199 range and continues to decrease at a moderate rate as monthly household earnings rise from M$200 to M$499. In the next income bracket the rate drops to 0.0 percent (0 of 15); there-

after it is a mere 8.3 percent (1 of 12). Questions of causality aside, this means that women living in the poorest households are likely to have experienced at least one divorce, whereas their female counterparts in the richest segment of the community have effectively avoided divorce altogether. Since I have indicated that the same directional trends can be discerned for the males in the sample, these latter generalizations may be applied to Bogang's population of ever-married men as well.

The data in table 21 reveal associations rather than causality, but they do support my argument that conjugal and affinal tensions are at their highest pitch—or the least likely to be resolved within the confines of marriage—among households with the lowest incomes. As for Bogang's wealthiest residents, and the ever-married men among them especially, such tensions either are less severe or are dealt with in the context of marriage, rather than at its expense. One partial explanation for this circumstance is that most male members of Bogang's socioeconomic elite enjoy positions of privilege thanks to their long careers in government employ. Their salaries and pensions then enabled them to purchase commercially valuable rubber acreage, which in turn generated additional wealth over the years. Also, civil service employment has usually involved considerable geographic mobility, as in the Federated Malay States Railways or a branch of Singapore's armed forces, which could easily postpone a newly wed government servant's uxorilocal residence for up to twenty years, after which time he would retire with a pension and settle in his wife's natal compound. This is not to imply that these men had no contact with their wives or affines during the course of government service. It is to suggest that their domestic circumstances largely precluded the progressive buildup of antagonisms and hostilities between married men and their residentially proximate in-laws, tensions that are especially disruptive during the first six or so years of marriage.

Of further interest here is the fact that many of Bogang's socioeconomic elite are married to women of the Hill lineage of the Lelahmaharaja clan, which is the descent unit at the top of Bogang's status hierarchy as well as the core around which the local political structure revolves. The privileged economic standing of the households associated with this lineage, together with the lineage's low rate of female divorce (the lowest in the village, with 0.18 divorces per ever-married female, as compared with the village mean of 0.44), derives from productive resources obtained not

through inheritance but through the efforts of in-marrying males still alive, many of whom draw handsome government pensions. Simply stated, the wives of these men married very well. That they did so can be explained largely by the prestige of their own lineage as measured by its residential origins and untainted pedigree, and by its having among its members a number of highly placed political luminaries. For these women and their husbands, the prospect of property and prestige loss upon divorce works effectively to control any conjugal and affinal antagonisms that may arise in the course of marriage. Also, such stresses would be less prevalent in the upper-income brackets in any case, since in-marrying males of impressive economic standing can satisfy affinal demands on their productivity. Historically increased concerns with status-group endogamy and the resultant concentrations of wealth within the elite stratum, then, lead me to expect that low rates of divorce will prevail within this lineage in the future.

Not surprisingly, the exceedingly marginal Biduanda Dagang clan records the highest divorce rate in the village (0.76 divorces per ever-married female). This clan, having always occupied the bottom rung of the prestige ladder owing to its comparatively recent migration to the village and to its ancestors including Meccan-born slaves and other non-Malays, has been extremely limited in its ability to obtain productive or high-ranking males as husbands. It has therefore had to settle for itinerant Javanese or unrelated men affiliated with other lines of the (non-exogamous) clan, who tend to be of inferior economic standing and potential—otherwise they would find wives elsewhere. Thus, when conjugal or affinal tensions arise, these men have relatively little incentive to maintain the bonds with their wives, and vice versa. This is especially true given the usual absence here of jointly acquired land over which each partner to a union would stand to lose rights (that is, a half-claim) in consequence of divorce. The constraints working against divorce are thus very weak; moreover, this situation will likely prevail into the future, for the dynamics in operation here tend to reproduce themselves from one generation to the next.

The interpretations presented thus far have dealt with marriage durability more or less synchronically, and in terms of a single nexus of structural entailments whose refractions differ according to socioeconomic and prestige variables. This approach allows, I believe, insights far more powerful than those offered by earlier observers such as Swift (1958; 1965, 133

et passim), who explained the distribution of divorce in the Jelebu district of Negeri Sembilan with reference to factors of psychology and personality.^[8] I would also contend that the analyses presented here are relevant to historical trends that have been misconstrued by Swift (1965, 132–134) and other investigators (e.g., DeMoubray 1931, 205; de Josselin de Jong 1951, 162).

It has been widely assumed that the weakening of enatic bonds would cause divorce rates to rise, since disgruntled spouses would be less susceptible to pressures from enates and affines, who, according to this perspective, generally strive to effect reconciliation and in any event exert an ultimately stabilizing influence on marriage. In my view, this position rests on a fundamental misunderstanding of enatic and affinal ties, and also be-speaks a failure to grasp certain of the most critical dynamics and tensions in the system of property and social relations. Of more immediate concern, though, is the fact that divorce rates have actually diminished quite radically in recent decades, a change that is consistent with historical processes discussed here and in previous chapters.

The phenomenal decline in state- and district-level rates of divorce (measured in terms of divorces recorded each year as a percentage of marriages registered over the same period of time) can be seen from table 22. The earliest figures, which pertain to the 1940s, indicate a state-level divorce rate of 65 percent. The corresponding rate for the 1950s is much lower, and is approximately 51.5 percent. Statewide data for the 1960s are not available, but the figures for the 1970s point to a further decline, and to a state-level rate of divorce of 14.7 percent.^[9]

This trend, I would argue, is by no means confined to the post-1940 era; rather, it probably dates from the earliest decades of colonial rule. Moreover, the same set of factors that accounts for divorce patterns in contemporary Bogang is at work here: the historically variable, but nearly always problematic, structural ties between in-marrying males and their affines, and the tensions arising from affinal demands on men's productivity. In other words, the issue is a married man's relative autonomy among his wife's kin. And it is precisely in this area that married men have made such enormous advances in this century, and during the postindependence era especially, even though these gains—if they are in fact to be regarded as such—are quite unevenly distributed throughout present-day villages.

More generally, enatic kin have lost much of the social control they once

exercised over in-marrying males and their labor power, owing to the infusion of cash into the village and other processes of economic development. Hence, significant reductions in divorce rates would be expected, insofar as the more encompassing trend—the emergence of progressively individualistic atomized household units—has occurred at the expense of siblingship and other matrilineal ties linking spouses with their respective enates. These ties appear increasingly incapable of seriously disrupting marriages, even though they still do cause stress and antagonism and, of course, continue to contribute to divorce.

Married Men and Their Affines

The preceding discussions elucidated various aspects of continuity and change in the relationships established through marriage, but dealt largely with affinal bonds in the aggregate. A more comprehensive assessment of persistence and transformation in the domain of affinity requires that we examine some of these relationships in greater detail.

Twentieth-century trends in affinal relations are well illustrated in the settlement pattern and kinship composition of Bogang's new hamlet. To oversimplify, married men of the new hamlet do not reside among their

affinal (or enatic) kin, and have consciously chosen to distance themselves from their affines owing to tensions and demands that they can now literally afford to do without.

The situations that new-hamlet men escaped are structurally identical to those currently confronting the adult men of Bogang's much larger old hamlet. Affinal relations in the old hamlet continue to be characterized by tensions, even though they differ significantly from their counterparts of earlier decades. Indeed, although old-hamlet men enjoy greater economic and jural autonomy vis-à-vis affines, they are still morally and socially bound to abide their affines' wishes as long as they reside among them. This is especially true for men who live in houses built on ancestral land over which their wives' enates enjoy residual, if not provisional, rights.

I was told, for example, that the verandah of a house built on ancestral land belongs to the wife's kin, who may use it as a meeting place or hall (dewan ), and that a man has legitimate powers only in the central region of his wife's house that is bounded by the "four thresholds" (bendul yang empat ). Thus, a man's affines may enter his garden and climb up to his verandah on almost any pretext, but they may not enter the area marked off by the four thresholds without appropriate cause and good intentions. This authority evidently does not obtain in houses situated on nonancestral land, for there a man may refuse his affines permission to enter his garden as well as his home. Both instances involve a paradox, however, since the interior of a home is still defined as the women's domain, just as the verandah continues to be strongly associated with men and maleness.

In the case of homes on ancestral property, then, in-marrying males must continue to defer to their affines with respect to activities occurring outside the four thresholds, and as long as they are within these architectural boundaries they are beholden only to their wives. As for houses on nonancestral property, husbands need not entertain or host their affines at all, even though they must of course consider their wives' sensibilities.

All this would seem to suggest significant variation between Bogang's two hamlets, since the newer of these residential enclaves is composed entirely of households built on nonancestral acreage. In actuality, however, the social and cultural realization of such variation is minimal at best. For even if a man is not subject to pressures or meddling from his affines, he remains constrained by basic constructs of gender that are shared by his wife, children, and neighbors, as well as village society as a whole (see

chapter 8). While these constructs shape behavior in a multitude of diverse contexts, their relevance here lies in their contributing to what men jokingly refer to at times as "queen control," or kwin kontrol .^[10] This term, denoting situations in which women either dominate their husbands or assume the greater role in decision making, aptly summarizes an important dimension of domestic reality in both of Bogang's hamlets. Thus, what would appear to be a major contrast between the new and old hamlets—the differential standing of married men in their own homes—dissolves. The significant contrast is that new-hamlet men need not be bothered with their affines and that they are unable to turn to them should they seek support for some personal reason or political cause.

Let us move now beyond the domestic realm and look at a married man's other affinal relationships. The old hamlet is an appropriate focus for discussion, insofar as most of Bogang's inhabitants live there and also because the married men of this hamlet generally reside among their immediate affines.

A married man must contend first and foremost with his wife's parents, and his wife's mother in particular. These ties are characterized by latent tension and basic avoidance; when the son-in-law encounters or otherwise interacts with either of his wife's parents, let alone the two of them together, formal deference is the rule. In general, his parents-in-law will call on him—to pay a social visit or request favors or services—not vice versa. Although the generational difference by itself is an important factor here, it is less significant than the man's standing as guest in relation to his parents-in-law, who enjoy the status of hosts, having provided him with his wife, land to cultivate and on which to live, and perhaps a house as well (as when his wife is their youngest daughter). Interestingly, it is precisely because his parents-in-law bestowed the property rights from which he and his wife derive benefit and obligation that they are potentially his best allies, at least in their capacity to forestall and, if necessary, mediate conflicts over land and other property rights involving his wife and her sisters as future co-heirs. In these disputes, it may be noted, the mother need not assume an especially active role as mediator, for her mere existence exerts a moderating influence on the sisters' relationships.

A married man's ties with his wife's sisters are likewise rigidly structured and usually marked by avoidance. They neither seek one another out nor converse freely when they do interact, such as during preparations for

feasts held in the lineage compound. As far as the villagers are concerned, this is as it should be. Physical intimacy with a sister of one's wife is still among the most reprehensible of all moral offenses (even though "anticipatory sororate" and sororal polygyny no longer meet with capital punishment).^[11] Sororatic marriage with the sister of a deceased wife, however, though rare, is still permitted, and continues to be seen as an ideal remarriage for a widowed man with young children.^[12]

A married man's relationships with his wife's brothers and her mother's brothers are profitably viewed in the light of the extremely formal bonds between women and their male siblings, and in terms of cross-sex siblingship in general. A woman's adult brothers usually live some distance from their natal compound, and even when these men marry women from their own community, they rarely visit or interact with their sisters. Stated simply, a woman's brothers and mother's brothers influence her affairs but little once she marries. True, these men do have rights to the produce of fruit trees planted on residential land classed as ancestral, and also to ancestral sawah should they participate in its cultivation and harvest. Such prerogatives, however, are not usually exercised, and in most areas adult cross-sex siblings lack common interests, in land or anything else. In and of themselves these facts need not contribute to harmonious relationships—witness competitive ties among brothers—but they do effect a divergence of life concerns as well as relatively extreme geographic dispersion. Consequently, men who reside uxorilocally upon marriage tend not to have major problems with their wives' male (as opposed to female) enates.

A married man does encounter difficulties, however, in his dealings with other men who have married into his wife's lineage. These individuals include his father-in-law, as well as the men married to his wife's lineage sisters, that is, his biras , or "brothers" through in-marriage. Because they—and he—are in-marrying males (orang semenda ), they are all identically situated with respect to the wife's lineage and clan (tempat semenda ); they are also usually total strangers to one another prior to marriage, deriving from separate lines of descent and dispersed geographic origins.

Village men linked as biras nonetheless generally live in aggregated households. The behavioral imperatives of residential proximity further structure and intensify their relationships of mutual interdependence and social commonality. Owing largely to the feasting complex, for example,

these men eat, pray, and engage in other ceremonial activities together far more than with anyone else, for almost every ritual feast includes, at the very least, all socially active adults living in the compound(s) of the wife's lineage branch, if not lineage. The ritual celebration of marriage, moreover, being typically on a scale that demands the construction of large wooden or bamboo platforms with thatch roofs to be used both in the preparation and serving of food and to shelter participants from rain, requires extensive cooperation and labor exchange among men linked as biras. Ad hoc work groups formed for fishing expeditions, too, usually have as their core closely related female enates and their husbands. Although an analogous situation once prevailed in the context of traditional rice production, neither economic cooperation and exchange nor residential proximity generated the resentments and antagonisms among biras with which we are concerned. Rather, the problems that did arise—and still prevail—were a consequence of the structurally enforced interdependence among biras that exists in the absence of any centripetal mechanisms guaranteeing a man unconditional political backing from the others in this group. Demands for political loyalty and support are in fact the hallmark of biras ties, and bonds among adult males in general. Needless to say, such demands can be far more frequent and intense—and responses to them much further from the ideal—among residentially proximate biras whose relationships may be put to the test on a regular basis.

As seen by the rest of the community, "brothers" through marriage with the female collaterals of a single lineage constitute an interest group in the most basic sense; and to a lesser extent the same holds for all men married to the women of any one lineage, regardless of their generation. These men, as noted above, are enjoined to support and further one another's concerns, if only because their personal strivings and social identities are merged with those of their own wives, who themselves are sisters or close enates, as well as immediate neighbors. As compared with their wives and offspring, however, the married men of a lineage compound have relatively weak emotional and social involvements, both with one another and with the enatically related residents of other compound households. As a general rule, there is little visiting or genuine friendship among them, despite their residential aggregation. Nor do these men share rights in one another's offspring, for such rights are distributed and reallocated among natural and adoptive mothers, not their husbands.

Especially at present, moreover, a man's financial standing—and hence, that of his household—is only minimally bound up with the economic well-being of his biras and their households, who at any rate are far more likely to constitute a drain on his labor and capital resources than to further his chances of economic advance. Perhaps more to the point, biras are in competition by virtue of the status rivalries among their wives, who are almost certain to view one another's successes with marked ambivalence. Thus, although the community at large sees married men as acting on behalf of their spouses and the latter's immediate kin, the very households to whom they pledge their loyalty may view them in a wholly different light. This in turn means that while the backing a man receives from his biras may be conditional at times, it is generally far more substantial and reliable than that forthcoming from men who are not his principal allies, since these men's loyalties are defined by their ties with their own wives.

Of broader concern is that the conceptualization and force of biras bonds undercut and contradict the logic of enatic morality that prevails among males affiliated by descent with a particular lineage or clan. More specifically, since actual brothers invariably marry women of different lineages, if not altogether separate clans, they are constrained to honor their respective biras relationships over those with their own brothers.

There is yet another level at which the institution of biras contradicts normative theories of brotherhood, simultaneously ensuring their highly qualified realization in social action. I refer to the siblingship said to obtain among all Muslims, and most emphatically—both according to formal theory and in village practice—among men. Islamic ritual and ideology argue quite pointedly for a united community of Muslim believers with no internal subdivisions or opposed allegiances, for all who follow the teachings of the Prophet Muhammad are held to be identically situated with respect to God and therefore of equal standing in relation to one another. These sacred bonds of equality and commonality, in short, supersede the particularistic and cross-cutting ties of kinship and residence. Here I may cite al-Ghazali, the eminent Muslim philosopher, who explored the nature of Islamic brotherhood in some depth.^[13] Commencing with a quote from the scriptures, al-Ghazali wrote:

The Muslim is brother to the Muslim. He does not wrong him, does not forsake him, does not betray him. What treachery and desertion to abandon him to

the rending of his honor! It is like abandoning him to the rending of his flesh . . . which is why God . . . compared it with the eating of carrion meat. For He said: "Would any of you like to eat the flesh of his brother's corpse?"

Elsewhere in the essays, al-Ghazali emphasized that

[Islamic] fellowship is a bond of flesh, like the bond of blood-kinship. . . . [Indeed,] brotherhood is a contract on the same footing as kinship; once it is contracted the duty is confirmed, and that which the contract entails must be fulfilled. . . . Brotherhood is provision for the vicissitudes and accidents of time. . . . Brotherhood in religion is firmer than brotherhood in kinship.

Here lies the essential contradiction for villagers of present-day Bogang, for brotherhood in religion would seem to be even more elusive and unobtainable now than at any point in the past. Cross-cutting affinal ties are not the only factors accounting for this situation, but they certainly constitute an appreciable obstacle to the realization of a broad siblingship commonality. More generally, the ambivalences characteristic of relationships among biras continue to derive from structurally induced tensions that have always permeated this domain of affinal relatedness. When one recalls as well that most other variants of affinal ties are marked by analogous structural tensions, it becomes clear that the realm of affinity in contemporary villages such as Bogang shares a good many similarities with its counterparts of earlier decades and the nineteenth century as a whole.

8—
Property and Inheritance

Anthropological inquiries into twentieth-century changes in Negeri Sembilan's system of property and inheritance relations have been primarily synchronic in approach and have invariably focused on some aspect of filiation or descent; the most common conclusion is that "the system has broken down" (or is well along the road to decay) and has given rise to a system characterized by bilaterality or patriliny.

In this chapter I start with a clean slate and the most basic of questions: What types of shifts have occurred in property and inheritance institutions, and what continuities, if any, can be discerned? As we will see, the demise of divided title and most other twentieth-century developments in property and inheritance concern principles and bonds of siblingship more than descent constructs or the conceptual units or social groupings associated therewith. Despite disjunctions in property relationships among various categories of siblings, however, continuities in basic modalities of siblingship have also played a critical role in maintaining traditional property bonds that remain central to the overall nexus of social relations. The reproduction of long-established concepts of gender has worked similarly,

such that daughters continue to be strongly favored over sons with respect to the devolution of rights over houses, residential plots, and most other categories of land. In sum, even though the emergence of proprietorship based on full rather than divided title has excluded enatically related female collaterals from the inheritance of rights to houses and land, property ties between successive generations of lineally related women still constitute the central links through which such rights flow.

General Attitudes toward Property

To ascertain twentieth-century shifts in villagers' attitudes toward property, we would best focus our attention on noncustomary property, which tends not to be encrusted by legal definitions and restrictions that obscure villagers' preferences and attitudes concerning the use and conveyance of property rights. The question thus becomes, How is noncustomary property dealt with, and what types of rights are associated with it? Further, in what areas can proprietors exercise autonomy in terms of loan, mortgage, sale, or other forms of right conveyance, and to what degree is partial or complete autonomy actually manifest? Finally, how does the contemporary taxonomy of property and property rights relate to basic nineteenth-century distinctions among economic sectors or spheres, such as those associated with subsistence and reciprocal exchange, prestige goods and luxury items, and the like? Since issues of this latter sort are in some ways the most fundamental of all, I will first assess the degree to which historical consistency prevails in this area.

Throughout the subsistence sector we find considerable historical continuity in modes of acquiring, utilizing, and conveying property rights. As regards land and houses, to take examples of corporeal property of the utmost utilitarian and symbolic value, property rights are still most honorably obtained through inheritance or purchase, and most honorably conveyed through inheritance, postmortem inheritance in particular—that is, freely given to junior kin. Land and houses remain earmarked for basic needs of subsistence and shelter and for the domestic maintenance of progeny, even though they may confer prestige on their proprietors as well. Rights over these items tend not to be bartered, sold, or otherwise made available on the market; in brief, they have not become commodities for exchange.

The mortgage or sale of land tends to occur only in situations of acute economic necessity and even then is a source of embarrassment and shame, causing others to "view one thinly." Land sale for the purpose of financing a trip to Mecca is also inappropriate and is experienced by kin and nonkin alike as somewhat unsettling. This is less a function of the history of specific land properties than a manifestation of the view that land is to be preserved and maintained both for household subsistence requirements and for the future needs of children who are eminently deserving, jurally entitled to shares of inheritance, or both. This same view shapes attitudes toward houses, which are never rented, pawned, mortgaged, or sold.^[1]

All such views as to the proper uses and conveyance of land and houses are of great antiquity, and have been neither undermined nor significantly transformed by economic processes or other forces of modernization. To gloss over such basic facts when considering the monetization of various segments of the economy would be to lose sight of a crucial thread of historical continuity in property concepts and relations.

Most types of agricultural (as distinct from primary forest) produce, too, are still used for domestic consumption and have not become exchangeable commodities. Liquid and coagulated latex are the obvious exceptions, but only partially rather than across the board. As discussed in chapter 5, the inhabitants of Bogang do not market rubber in the community or to other Malays. The same is true for rice. This situation seems to have obtained for decades, and even when locally grown rice was sold it was apparently done so at prices fixed by custom as opposed to market supply and demand (NSAR 1888, 6; R. Hill 1977, 126, 137–138).

The fruit of coconuts, papaya, and other trees grown on residential plots is likewise consumed by the household, or offered as gifts to neighbors and guests, rather than sold. In short, there is a basic distinction between eating (makan ) and selling (jual ), and assuming that agricultural or forest produce is edible or can be processed for eventual home consumption, it should be eaten and shared rather than sold. The overriding concern is that village relationships be mediated by reciprocity in exchange—and hence be warm, personal, and continuous—rather than by barter or the transfer of cash or commodities through hired labor. Thus, even when durian and other orchard produce is marketed along the roadside, it is typically sold to unrelated Malays of other communities or to Chinese. Moreover, some durian sellers acknowledge they could command a much

higher price for their fruits but feel constrained to charge below what the market will bear, since the produce derives from trees painstakingly planted and tended by ancestors who would disapprove of any profiteering by descendants.

Such generalized sentiments recall the rights enjoyed by males in relation to the ancestral sawah vested in their sisters. These rights are known as makan hasil , which means literally "to eat the yield." This particular expression and category of privilege once stood as a basic metaphor for traditional rights over land on the whole. In the case at hand it amounted to males claiming a share of the harvest of their sisters' land in proportion to their contributed labor. If they did not participate in producing the crop, however, such claims could not be lodged. Clearly, then, the maintenance of property rights hinged on the productive use of property over time. While this situation still applies to most land, it does not apply to acreage planted in rubber, insofar as rights over rubber plots have always been relatively autonomous vis-à-vis other categories and taxonomic schemes of property. For obvious reasons relating to the local importance of rubber acreage and the cash earnings derived therefrom, this relative autonomy helped foster the demise of major components of the property system, simultaneously severing it from its cultural basis in notions of origin point and diachrony.

Domesticated fowl, for their part, are generally raised for household subsistence needs and not with an eye toward future exchange or sale. This is all too apparent to those who find themselves forced by the prospect of a sudden or unexpectedly large ritual feast to attempt the purchase of domesticated chicken from neighbors or friends. Typically only those individuals who are owed debts can claim any success in this area. Such is the case with the village headman of Bogang, who is also a renowned curer specializing in the treatment of poisoning and sorcery. As his wife put it, "He has cured quite a few people and is very clever in finding people willing to sell him their chickens."

Goats, sheep, cattle, and oxen also serve as subsistence guarantees; the owners of bovines derive prestige as well from having such readily apparent stores of wealth at their disposal. Although livestock are easily enough sold outside the village and thereby converted into hard cash, villagers usually raise them either as a hedge against hard times or for the weddings of their children, when, rather than being sold to help defray feasting and

other expenses, they are instead slaughtered to feed, honor, and impress the guests.^[2] Although appreciable historical continuity might be supposed here, it is well to bear in mind that domesticated animals, especially bovines, were an important source of cash for luxury items and the pilgrimage to Mecca before the local economy was transformed and cash cropping became widespread. Hence, the economic functions that these animals once served have not only been assumed by other property in the form of rubber land, but they have also been conceptually recast and situated more squarely within the domain of subsistence as commercially valued acreage has spiraled in importance, both in the village economy as a whole and in the local taxonomy of property and property rights.

This overview of villagers' attitudes and behavior concerning the use and conveyance of rights over land, houses, agricultural produce, and domesticated animals illustrates how little subsistence property has been redefined as a commodity to be freely exchanged or sold. Labor, of course, has been commoditized—most notably the labor of males engaged in rubber tapping. Land itself, however, has largely escaped this process (but see Fett 1983, 79–80).

Prestige goods and luxury items display an equally pronounced consistency over time. I am not speaking here merely of ancestral items, such as inherited bridal ornamentation, but refer also to newly purchased gold given to women in the course of engagement and wedding rituals and during the hair-clipping ceremonies performed when they are three or four years of age. As with land and houses, in other words, gold is best acquired through purchase, gift, or inheritance, and thereafter conveyed as a gift, typically as a postmortem bequest. In the interim it should be safeguarded, added to if possible, and never parted with on account of subsistence or accumulative concerns, however important these may be. Thus, while many women over fifty years old boast finely crafted and extremely valuable bracelets, necklaces, earrings, and other ornaments of high-quality gold, none of them would ever pawn or sell any of these possessions unless forced to do so by otherwise unmanageable funerary expenses.

These traditional attitudes encompass the phenomenon of conversion, whereby prestige items, for example, or the monetary returns therefrom, are channeled into the subsistence domain, as well as the related process of conveyance, which may involve the retaining of goods or wealth within a particular economic sector while utilizing them for purposes other than

those originally intended.^[3] Thus, however strongly a woman yearns to make the pilgrimage to Mecca and thereby attain exalted spiritual standing and social prestige, she would never sell off her gold to help finance the journey—this despite the fact that rising to haji status is perhaps the most cherished dream of all villagers, and clearly within the reach of many couples should they choose to liquidate their gold assets.

Issues of conversion and conveyance aside, it is also true that gold with ancestral standing has a higher prestige value than gold with a less distant origin point. Just as the value of such gold is enhanced over the years (since the passage of time leads ultimately to death and inheritance), so does the improper use and transfer of ancestral gold occasion greater moral opprobrium.

The other side of the coin in all of this is the modern-day ease with which household earnings from the sale of agricultural produce and labor can be converted into luxury items and mass-produced consumer goods, the possession and display of which may bring considerable prestige. As we have seen, the availability and distribution of such items is strongly connected with the progressive centrality of cash in the local economy and attendant processes of economic stratification; they are, moreover, profoundly significant in how villagers perceive their own and one another's socioeconomic standing. I would only add here that the influx of Westernstyle consumer goods and fashions—such as kerosene and gas stoves, refrigerators, radios, motorcycles, televisions, automobiles, and ready-made clothes—has not only generated new and cross-cutting indices of status but also served to profane traditional categories of prestige and luxury.

It merits mention as well that many of the more valued prestige symbols in today's village are not of enduring quality and are thus unlikely to be conveyed through a gift or inheritance transaction. Televisions, for example, are frequently leased by the month from Chinese in nearby towns. Rights of proprietorship over these commodities are qualitatively different from those over such prestige items as gold jewelry, being essentially divided between urban Chinese and a household's adult residents. In one important sense this particular property relationship is a paradigm for twentieth-century change in property and social relations, for here individual household units are not only disconnected from one another—rather than linked—by virtue of socially valued property ties, but they are also bound to urban Chinese and the encompassing market networks they

so effectively mediate. To appreciate the point we need only recall that all residents of Bogang who sell latex do so to Chinese middlemen, and that the practice of hiring Chinese carpenters to build village houses has supplanted the collective "barn raising" traditions of earlier times, as well as largely displaced Malay carpenters.

In sum, even though household earnings are easily converted into emblems of prestige, the boundaries of traditional economic sectors remain more or less intact, with the partial but noteworthy exception of labor (agricultural and otherwise). This exception deserves special emphasis in light of the transformation of productive relations discussed in chapter 5.

Nineteenth-century labor exchanges were couched in a nexus of idioms drawn from kinship, reciprocity, and essentially disinterested service, and they typically occurred in one of two forms. In the first, glossed tolong-menolong, or "mutual assistance," groups of female proprietors, aided at times by their spouses and other male kin, joined together for the more labor-intensive stages of cultivation, proceeding from one woman's fields to the next in accordance with the varying labor and production requirements of each plot. Here, notions of sisterhood, common lineage membership, and egalitarianism and interdependence among cultivators provided the basic idioms for the provisioning and exchange of labor, and for the realization of domestic subsistence guarantees on the whole. The second institution, seraya, differed in concept from tolong-menolong in that it could also draw directly on debt repayments from beneficiaries of certain specialized services, particularly various indispensable and otherwise largely unobtainable ritual services. The ranks of those who called in their debts through seraya were thus composed of persons who both knew and performed the sacred arts, namely midwives (bidan ), curers and exorcists (dukun ), and the elite pawang. These repositories of esoteric lore apparently delegated much of their agricultural work to villagers who stood essentially as their clients. More explicitly political and asymmetric, though not necessarily culturally defined as such, were the seraya generated at the bidding of clan chiefs and Rembau's Undang. These seraya especially entailed service to ruling patrons whose mandates called for pursuing public affairs in the best interests of their adherents.

Significantly, both of these institutions have all but disappeared owing to political and economic processes that have not only progressively narrowed the field of social obligation but also generated the view that specialized

practitioners are most appropriately compensated by on-the-spot cash payments. This is not to suggest that clan chiefs or midwives are placed in the same social category as fishmongers or local dry-good peddlers. It is simply to underscore that labor provisioning of the sort considered here is not forthcoming on a voluntary (or any other) basis and that debts for services rendered are of exceedingly short-term duration. It is to emphasize as well that individuals vested with traditional titles no longer enjoy the privileged positions they once did and are thus incapable of commanding extrahousehold labor solely on the basis of their unique skills or regalia.

The individuals who provide these specialized services have thus witnessed an appreciable narrowing of their legitimate claims over both the labor and the loyalties of those who benefit from their expertise. Such developments are not merely by-products of a situation in which the sanctions, authorities, and functions once concentrated among clan leaders and ritual practitioners are nowadays diffused among officials associated with Islamic administrative hierarchies and secular bureaucracies. Rather, they are a logical and empirically evident outcome of social and cultural trends precipitated primarily by economic processes that encourage individual and household autonomy at the expense of vertical ties as well as interdependence among households and larger groupings comprising social equals.

Viewed in terms of the village majority, the marked contraction of social and moral obligations linking domestic units is but one manifestation of the progressive concentration within individual households of rights over labor and other forms of property. The long-range manifestations of such trends appear in varied guises, one of which is symbolized by the decline of the agricultural rituals discouraged by Dato Abdullah and like-minded members of the Malay elite beginning in the 1920s, and by foreign-born Muslims during the second half of the nineteenth century. The performance of berpuar and related sacred ceremonies associated with local shrines testified to the protective powers of the spirits of pioneering culture heroes credited with transforming Bogang and much of the rest of Rembau into a habitable region conducive to social reproduction and cultural florescence. All such rituals were predicated on the sanctified axiom that these spirits not only shared rights of proprietorship over all the land in and around the village, but were in fact responsible for the very creation

of these rights. The broader theme, which local elders occasionally address, is that rituals like berpuar amount to gestures of reciprocity by the beneficiaries of the local spirits' protection, based on their common bonds of proprietorship. To raise this issue, however, is also to draw attention to the fact that the decline of berpuar and other local rituals constitutes a symbolic and public negation of those bonds of proprietorship that link long-deceased ancestral figures with the living. Of course, the beliefs underlying the performance of such rituals persist among most villagers over forty years of age (see chapter 9). Yet the symbolic disjunction does exist; it is, moreover, clearly recognized by elders, who, as discussed earlier, attribute diminishing rice yields to the actions—or, more precisely, the nonactions—of guardian (but nonetheless potentially malevolent) spirits gravely incensed by collective behavioral denials of their roles in proprietorship and community affairs in general.

The narrowing field of social and moral obligation thus impinges on relationships not only among the living but also with the realms of the spirits and the dead. The concentration of property rights among the living, and among individuals and their household groups in particular, also finds a structural analogue in the fact that Allah is increasingly viewed as monopolizing all of the spiritual energy and power once seen as concentrated in Him as well as in sacred shrines, traditional curers and exorcists, and various local spirits (see chapter 9).^[4]

Shifts in Property Rights Conveyed through Inheritance

I have already noted that property rights are most commonly transferred through inheritance. Here I will examine continuity and disjunction both in the nature of such rights and in the direction of their transmission. These issues are analytically distinct and will be dealt with accordingly. I will first consider various issues bearing on the rights at stake in inheritance; thereafter, and also in the following section, I address historical and contemporary ethnographic data pertaining to the genealogical links involved in the intergenerational flow of rights over property, particularly houses and land.

One salient feature of precolonial property institutions was the explicit differentiation between "acquired" and "ancestral" holdings and the rights thereover. Underlying this distinction was the longitudinal dimension, or

diachronic grading, of the taxonomic scheme, through which land was reclassified over time such that all acreage acquired through individual or conjugal effort was ultimately redefined as the ancestral property of enatic segments. This reclassification occurred more or less automatically upon inheritance. Hence, any previously unalienated land cleared by a woman and her husband came to be provisionally vested in their daughters, subject to the residual control of the daughters' enates—that is, it became ancestral. The daughters and all future recipients of these provisional rights were thus constrained in their dealings with the property in ways the original proprietors never were.

Over the course of the past century, however, an exceedingly selective juridical bolstering of the property system occurred. As discussed in chapters 3 and 4, certain features of the system were formally reinforced by the colonial state, whereas others were either left intact or eviscerated through administrative fiat. More specifically, whereas traditional restrictions on the transfer of rights over ancestral holdings were codified and frozen in various colonial enactments (many of which remain in force even at present), there was no corresponding attempt to preserve the diachrony that was both inherent in the system and central to its reproduction. Thus, while most land titles first registered before 1910 were designated as "customary"—which was intended as a synonym for ancestral—holdings not so labeled were never formally reclassified upon inheritance. This failure to record the necessary changes in cadastral surveys and village titles had no immediate implications so long as consensus obtained among potential heirs and those individuals empowered to validate the conveyance of postmortem property rights. If, however, disagreement arose as to the precise shares of intestate property due this or that individual or class of heirs, satisfactory resolutions could be extremely problematic, if not impossible, to effect.

Further complicating this picture were historically variable attitudes among state officials, the traditional elite, and villagers themselves concerning the proper relationship between long-established property traditions and Islamic religious codes pertaining to inheritance and property in general. Islamic property law recognizes no distinction between ancestral and nonancestral property; nor does it allow for the diachrony with which we are concerned. Without question, the existence of an alternative set of property and inheritance codes bearing the stamp of Islam provided vil-

lagers and state officials alike with an ideological precedent and rationale for gradually relaxing numerous constraints associated with precolonial property institutions.

We should also bear in mind that we are dealing with well-financed religious hierarchies backed by the state and empowered to act independently of the wishes of those persons most directly involved in the devolution of rights over land. The distinction is important if only because present-day settlement officers adhere to a policy that expressly disavows the history of all land and focuses only on its official status (as customary or noncustomary) in deciding whether it falls within the domain of Islamic law or should be partitioned instead according to the relevant adat conventions as interpreted by clan leaders. So far as the contemporary state is concerned, there are still two categories of land and land rights but no mechanism whereby noncustomary property is redefined upon inheritance.

With the juridical elimination of diachrony from property relations, divided title persists, at least in a technical sense, only in the case of sawah, residential acreage, and other lands formally designated as customary. Virtually all such holdings date from before 1910, for beginning in the second decade of the twentieth century, by which time commercially valued plots of rubber had been well integrated into the local economy, there were hardly any new lands registered as customary. Although many of these holdings were subject to various clan restrictions through the end of the colonial period, this situation no longer prevails due to the augmented charters of Islamic bodies and their stepped-up involvement in the property, marital, and other affairs of village inhabitants. Contemporary notions of land proprietorship are thus markedly different from those of earlier decades, let alone those of the nineteenth century, primarily on account of the emergence and spread of undivided title (or fee-simple interest), which has gradually displaced conventions of divided title.

In sum, even if we assume for the sake of argument that females are nowadays the sole recipients of land rights dating from after 1910, we must recognize that because such rights fall outside the system of divided title, they are wholly concentrated among the individuals whose names actually appear on cadastral surveys. From this perspective, the effective demise of the system's diachrony bears most directly on the classificatory siblingship of collateral enates. Shifts in these and other modalities of

siblingship in turn provide the analytic key to continuity and disjunction in the nexus of property relations.

Aside from the state's position regarding the classification of land, it remains to emphasize that villagers themselves do not usually operate with the same set of conceptual categories invoked by settlement officers and Islamic magistrates. Numerous factors account for this situation, including disagreements about whether a particular plot of land ought to be subject to Islamic property codes and lack of consensus as to the actual nature of these codes as well as their application in any given circumstance.

Proprietors, moreover, often view all plots (and categories) of inherited land as ancestral and therefore feel morally bound, at least in certain contexts, to treat them as such. Because one's land rights are not usually mortgaged, sold, or otherwise formally alienated before death, the moral imperatives at issue typically come into play only when a proprietor informs children and other immediate kin of his or her wishes concerning the future devolution of rights over that property. Such information is conveyed to close relatives in a highly variable, extremely casual way, and rarely in the guise of a formal will. More important, just as villagers are generally loath to transgress even the implicit desires of deceased kin, and mothers especially, so too do they tend to view all inherited land as most appropriately passed on to daughters, regardless of Islamic provisions highlighting the jural prerogatives of sons in this area. Assuming formal agreement on the part of males who might otherwise receive shares of their parents' acreage, such sentiments are easily enough accommodated by religious officials, who commonly aim only to ensure that all potential heirs, as defined by Islam, are advised of their privileges with respect to inheritable property. Even when sons might well derive substantial material benefit by exercising these privileges, they too generally agree that women have a greater need for subsistence guarantees in the form of land than do men. In short, it is by no means unusual for men to relinquish such claims in favor of their sisters.

One might thus contend that the elimination of diachrony from the system is relative rather than absolute, since villagers themselves still hold to an implicit reclassification of property over time in consequence of inheritance even though the state does not. There is some merit in this contention, although the resolution of contrasting policy positions invariably

favors the state. More precisely, the state always has the upper hand both in determining property distribution and in the analytically prior domain of defining the rights associated with inheritable property. Having recently decreed that all land not designated as customary is to be dealt with in accordance with Islamic conventions, the postcolonial state further undermined the system of divided title as well as the jural dimensions of the diachrony central to its reproduction. The extent to which a reclassification of land and land rights still occurs as a result of inheritance may be seen, then, as a conceptual or cultural phenomenon severed from its earlier juridical foundations.

After having considered various state, regional, and national developments, both economic and otherwise, we turn now to certain modalities of nineteenth-century property transmission through which subsequent change in the property system as a whole was channeled. The marriage payments and inheritance practices in question not only allowed for structural continuity but also simultaneously provided the template for the partial systemic collapse analyzed above.

Of greatest relevance here are the assets a groom brought to his marriage (harta pembawa ), which could—and ideally did—include rights to cash, livestock, and other movables that he obtained from his parents or sisters (hence from a narrow range of his kin) through premortem inheritance. The conveyance of such properties to a groom diminished the conjugal or familial fund over which his parents and sisters enjoyed rights, and in much the same fashion as occurred with the engagement rings, marriage gold (mas kawin ), and other marriage payments.

The harta pembawa complex is of particular significance insofar as a married man could legitimately designate his children as heirs to this property, which his enates could otherwise claim upon his decease. Since such paternal provisioning (tentukan ) could include rights over the nonancestral portion of the harta pembawa property he brought to the union, it functioned in essence as an extension of harta pembawa. More generally, paternal provisioning can be seen as an extension of harta pembawa even when the rights at issue centered entirely on the husband's share of conjugal earnings. This is to say that the institution of harta pembawa embodied the structural logic informing all variants of paternal provisioning and that the narrow provisioning of a man at marriage prefigured the distribution of his property to his children (that is, just as his mother's sisters and her

other collateral enates did not contribute to his harta pembawa or other marriage payments, so too might his own sisters and other collateral enates be excluded from inheriting rights to the property he left at his death). Viewed from this perspective, the harta pembawa complex provided the structural precedent not only for paternal provisioning but also for the colonial-era erosion of property rights grounded in the ties of siblingship and descent that linked married or formerly married men with their sisters and other collateral enates.

The phenomenon of paternal provisioning merits consideration apart from its connections with harta pembawa as well. Its mere ideological existence testified to a man's prerogative to favor his own children over enatic kin in matters of inheritance. This privilege served primarily to qualify the rights of sisters and other immediate enates with respect to the yields of a married man's labor, and to highlight the rights of his own offspring, who belonged to an altogether separate clan.

Of course, the aforementioned rights of a man's sisters and other immediate enates need not have been compromised to any significant degree if his son had married his father's sister's daughter. In such cases the property conveyed via paternal provisioning might be brought to the son's marriage as harta pembawa, utilized to help sustain the domestic unit established with the father's sister's daughter, and thenceforth conveyed to the son's own children, who would be associated with the same descent line and residential compound as his (the son's) father's sister. As noted earlier, however, unions of this sort were probably fairly uncommon in the nineteenth century (even though they may have been seen as ideal) and in any event constitute a very small percentage of the marriages in the villagewide sample discussed in chapter 7. These cases aside, it is clear that the practice of paternal provisioning obstructed the flow of property rights from a deceased man to his sister for the ultimate benefit of her offspring and thus impinged directly on the traditional entailments of cross-sex sibling bonds. Of far broader concern is that a woman's resultant inability to lay claim to the fruits of a deceased brother's labor foreshadowed both the historical linealization of inheritance and the concentration of title among individual households.

These trends are profitably viewed in relation to historical shifts in the significance of paternal provisioning, most of which concern the property rights at stake after a married man's death. Rembau men, it may be re-

called, first acquired commercially valued acreage in their own names beginning in 1888 or 1889; by that time they had already realized extremely lucrative returns from trading activities, all of which greatly enhanced the financial stakes at issue upon their decease. As a result, a reassessment and partial collapse of the cultural distinction between movable and immovable property occurred. This realignment of property concepts is evident by 1890, by which time Rembau's indigenous leaders had decided that a dead man's rights over jointly acquired movables (cash, livestock, and so forth) should thenceforth devolve on his children rather than his enatic survivors, and that his rights over conjugally acquired land could revert to his enates. Rembau's leaders were thus not only equating land with movable property that had always been divisible between spouses or their enatic representatives on the termination of a marriage, but they were also effectively upgrading the stakes at issue in paternal provisioning, since its occurrence or nonoccurrence could now effect the devolution of rights over jointly obtained lands as well.

Beginning around 1890, then, the legitimacy of a child's claims to his or her father's share of conjugally acquired movable goods was no longer contingent on the prior occurrence of paternal provisioning. As the traditional sine qua non for children to advance legitimate claims in this area was, in effect, written out of existence, their rights over certain of their father's properties were strengthened at the direct expense of competing jural interests vested in his sisters and other collateral enates.

Congruent developments date from the early part of this century, as in the ways the notion of capital interest (bunga ) was accommodated within the conceptual framework of inheritance. Before 1900, since little if any capital interest usually accrued to the property a man brought to his marriage, no indigenous policy was needed for the distribution of such interest on the dissolution of a marriage. Owing to the widespread purchase of rubber land by males, however, this facet of inheritance soon assumed critical importance. A man might possess potentially valuable acreage suitable for rubber cultivation before his wedding, or he might inherit rights over such property in the course of his marriage. Assuming the land was developed, what was to become of the increase in value, which could be phenomenal? Historical reports vary, but by the late 1920s a general consensus seems to have emerged that the enates of the decedent could claim as his personal estate only the original value of the land, and not the in-

crease or the land itself, both of which fell within the category of jointly acquired earnings with its attendant restrictions. In such cases the practice of paternal provisioning clearly resulted in further losses to enatic kin, who might otherwise have obtained up to one-half of the capital interest at issue as well as a half-share of all conjugal earnings.

A recurrent theme in many of the nineteenth- and twentieth-century changes examined here is that the interests of a man's wife and children came to receive priority over those of his sisters and their offspring. Such being the case, to what extent did these and attendant shifts represent departures from the precolonial state of property and inheritance affairs? To answer this question we must assess the genealogical ties and categories of property implicated in precolonial paternal provisioning, which provided the structural basis for later changes in inheritance, including those reinforced by the state or by developments within Islam.

First off I might stress the obvious fact that paternal provisioning emphasized links of patrifiliation at the expense of rights and interests inhering in the cross-sex siblingship and other enatic ties that joined married men with their matrilineal kin. Further, the enatic prerogatives initially undercut by paternal provisioning of jointly acquired earnings centered on movable goods, not land. Likewise, whereas children might inherit shares of their father's conjugally obtained cash, livestock, and weapons through paternal provisioning, they did not gain land rights in this way. These same generalizations apply to paternal provisioning involving a married man's personal estate, for here too we are dealing exclusively with movables (of a nonancestral sort). Significantly, though, the conveyance of property rights from men directly to their children seems not to have caused that property to be recategorized as ancestral. Although such a recategorization would occur if the property passed thereafter to the children's enatic kin, there is an additional structural precedent here for subsequent transformations in property institutions. I refer to the existence of a category of property which not only passed from a man to his offspring but which also retained its nonancestral status despite inheritance.

In sum, to point up that previously inherited land is no longer formally defined as ancestral is to emphasize once again that processes of change over the past century have resulted less in a breakdown of property institutions generally than in a realignment of the relationship between movable goods and certain categories of land. It is to underscore as well the emer-

gence of a conceptual disjunction between commercially valued rubber holdings and all other types of land (that is, residential acreage, sawah, and fruit orchards), and an attendant process whereby rubber plots and rights over them came to be associated with the property category centered on movable goods. In time, of course, this likening of rights over rubber plots to those pertaining to livestock, weapons, and such became subject to a process analogous to what Flannery (1972, 413) has referred to as evolutionary promotion—in short, an elevation within the conceptual system that ultimately rendered all land subject to generally comparable jural definitions. The emergence and spread of undivided title at the expense of earlier divided-title forms of proprietorship is but one facet of this process. Such developments were, moreover, effectively anticipated in the practice of paternal provisioning, as well as in the analytically prior domain of harta pembawa. Generally, then, there is considerable justification for viewing all shifts in the social links and categories of inheritable property rights as involving a progressive realignment and reweighting of traditional structural principles and property constructs, rather than the creation ex nihilo of a wholly distinct nexus of property or social constructs.

Female Heirship and the Reproduction of Matriliny

To what extent do the shifts outlined above constitute significant departures from precolonial patterns, whereby rights over previously inherited land typically devolved on a female proprietor's enates? More specifically, how has enatic heirship fared over the past seventy or so years, and how have female enates in particular made out in light of property, inheritance, and other economic changes?

A summary answer to these questions is fourfold. First, female heirs continue to be favored over males with respect to all categories of land titles that have been previously conveyed through inheritance. Second, a consistent and positive association exists between the temporal origin point of land rights on the one hand and the likelihood that such rights will be vested in women on the other; in other words, the more distant the origin point of a particular land title, the greater the chances that women will hold the rights to it (see also Fett 1983; Stivens 1985). Third, one long-range correlate of such trends is the maintenance and reproduction of a generally unbroken chain of property ties among matrilineally related women. Fourth, and perhaps of greatest theoretical relevance, patterns of

this nature derive largely from the interplay of principles of siblingship and gender, which thus perpetuate the conceptual significance and social entailments of matrifiliation and enatic descent and inheritance.

The implications of this situation force a major revision of earlier thinking about many features of Negeri Sembilan society and culture. The anthropological literature on Negeri Sembilan is replete with predictions and actual claims about the supposed transformation of inheritance relationships in favor of explicitly patrilineal or bilateral arrangements. De Josselin de Jong (1960, 165, 190), for instance, points to a shift whereby Islamic law governs the devolution of rights over jointly acquired acreage. A. Wahab Alwee (1967, 40–41) goes even further, proposing that "newly opened land [has] follow[ed] the Islamic laws of inheritance" ever since the mid 1920s, which decade also witnessed the beginning of a shift from matriliny to patriliny in the inheritance of rights over all noncustomary plots.^[5] Swift (1965, 172) asserts that sons tend to receive rubber lands, while daughters get residential and rice acreage, and that these patterns prevail regardless of the "legal status" of the land—by which he apparently means its cultural categorization as customary, noncustomary but ancestral, or noncustomary and nonancestral. Lewis (1962, 192–193) posits a generally comparable situation, contending that male children acquire larger shares of noncustomary rubber land through inheritance than do their sisters but are almost invariably excluded from inheriting noncustomary and customary rice and residential acreage. Here, too, the previous history of proprietorship over noncustomary land is regarded as a nonissue, and we are left with the impression that the devolution of rights over such land is characterized by bilaterality and, in the case of rubber holdings, a bias toward sons.

The basic problem with all such views is that they are products of essentially synchronic inquiries. Since they do not build on historical data obtained from cadastral surveys and household interviews, they fail to grasp the significance of the diachrony within the inheritance system. As a result, these views are wholly inconsistent with the fact that females are far more likely than males to inherit rights over previously inherited land in all agricultural categories, and that this dynamic serves to reproduce a form of matriliny.

Some of the data supporting my assertions are summarized in table 23. These data derive from a detailed study of all District Land Office records for the village of Bogang, which span the period 1888–1980 and thus

extend back to the very first year that land titles were recorded in written form. The collection and analysis of this material represents the first systematic attempt to examine female heirship in Negeri Sembilan from a diachronic perspective capable of illuminating the temporal dynamics of the system of inheritance itself.^[6] This material reveals that females still enjoy a privileged position in heirship, even in the case of land first registered in the past few decades. The tendency holds during the post-1909 era on the whole—that is, throughout the period in which lands were no longer formally redefined as ancestral in consequence of inheritance or designated legally as customary.

The fact that females still enjoy a privileged position in heirship is clear from table 23, where the percentages of land titles registered to females increase markedly over time in six of the eight columns. The exceptions are the first and last columns, which show no significant change. In the case of the last column, too little time has passed for any significant change to have occurred. In the first column, showing land titles first registered in 1888–1889, 93.2 percent were under female proprietorship at the time of initial registration, and fully 92.3 percent of these same titles were still under female proprietorship in 1980. The second column shows that, in 1890–1909, a smaller percentage of the land titles were under female proprietorship at initial registration than was the case in the preceding interval (66.3 percent and 93.2 percent, respectively). This relative decline in female proprietorship at the point of initial registration has nothing to do with inheritance; it merely reflects the increasingly pronounced impact of British colonial policies that encouraged males to purchase and register previously unalienated state lands in their own names. The second column as a whole, however, reveals significant historical increases in female proprietorship between initial registration and later decades, which are of course a result of females being favored in inheritance.^[7] The data summarized in the remaining columns of table 23 are generally consistent with both these trends, reflecting greater male involvement in the acquisition and registration of previously unalienated land on the one hand and increases in female proprietorship over time through inheritance on the other.

The second trend just noted can be seen in summary form by reading the 1980 row of table 23 from right to left. I should emphasize that the increase in female proprietorship of land through inheritance is

not confined to titles first registered before 1909, nearly all of which are designated legally as customary. The trend can also be discerned, though less strikingly, for land titles first registered after 1909, over 95 percent of which are, and always have been, designated as noncustomary. Particularly significant, too, is that this trend obtains with respect to all categories of land—including rubber.

What, then, of the contentions of earlier observers? Regarding the suggestion that the devolution of rights over newly opened lands, such as rubber holdings purchased from the state in the course of marriage, tends to be governed by Islamic law, it is certainly true that sons are occasionally awarded shares of the land that their parents acquired through conjugal effort. If this were not the case, increases in female proprietorship over time would be even more pronounced. The frequent implication by other scholars that, owing to Islamic influences, male shares are usually twice the size or value of those allotted to female offspring, however, is entirely unfounded. Even at present the property laws of Islam are commonly interpreted by villagers to mean only that sons merit inclusion in heirship, not that they are entitled to shares of inheritance larger than their sisters. Throughout most of the twentieth century, furthermore, secular and religious officials with authority over the distribution of intestate property rights seem to have held the same basic view, and in any event sought primarily to inform potential heirs (as defined by Islam) of their jural prerogatives in this area. Male children, I might reiterate, commonly choose to forego their legitimate claims in favor of their sisters.

Nor is there any evidence for a mid-1920s shift toward patrilineal inheritance of noncustomary land. The inclusion of sons as heirs to land jointly acquired by a married couple dos not generate patrilineal inheritance. For even if sons retain such land until they die, it typically then passes to their sisters, wives, or daughters, and subsequently from mother to daughter. Moreover, a great many of the heirs one would expect to encounter under patrilineal inheritance (F, FB, FBS, FBD, BS, BD, and SS, among others) have never been incorporated into the theory or practice of Negeri Sembilan inheritance. Comparable arguments could be advanced with respect to the system's putative evolution toward bilaterality (cf. von Benda-Beckmann 1979, 373–375).

The evidence given here is highly consistent with a complementary corpus of data, obtained from structured interviews at each of Bogang's 106 households, pertaining to proprietors' intentions or wishes concerning the

future devolution of rights over their houses, house plots, and agricultural acreage. Let us first consider the anticipated inheritance of rights over currently occupied houses, ninety-nine cases of which (93.4 percent of the village total) are suitable for analysis (see table 24). In 84.9 percent of households, the intended heirs are the natural or adopted daughters of current household heads (defined in most instances as the eldest resident female). The future heirs of another 11.1 percent are sisters, sisters' daughters, or other lineage females; in these cases proprietors have no daughters of their own and live on land registered as customary. Although some of these people could conceivably disassemble and relocate their houses with an eye to bequeathing them to their sons, most proprietors feel morally bound to pass over sons and all male relatives in favor of female collaterals in the matriline, regardless of who financed the construction of the house.

The question of why the vast majority of village houses will pass to daughters and other females in the matriline might be seen as a nonissue, since more than three-quarters of Bogang's residential structures rest on customary land. In other words, many owners who might wish to bequeath their homes to their sons could not do so without relocating onto noncustomary land. Such a possibility, moreover, does not really exist in the case of houses inherited, as opposed to constructed, by current owners. Although one could argue that houses pass to daughters or other female enates in the lineage owing to jural conventions originating in centuries past, which do not correspond with contemporary notions of propriety and inheritance, such arguments gloss over a critical issue: people of both sexes clearly prefer to bestow rights over houses on their daughters rather than—and to the exclusion of—their sons, regardless of whether the rights entail full or provisional title. This is so even in the few instances where houses are owned partly or fully by men. In fact, 86 percent (19 of 22) of the houses situated on noncustomary holdings are expected to pass to daughters.

The future devolution of rights to the homestead plots on which Bogang's currently occupied houses rest is much the same as for houses. As table 25 indicates, daughters and other females sharing lineage membership with the current household heads will be the heirs in approximately 92 percent of the cases at issue.^[8]

Data pertaining to the anticipated inheritance of rights over Bogang's agricultural acreage are summarized in tables 26, 27, and 28, revealing patterns similar to those pertaining to houses and houseplots. This data,

however, is far from comprehensive, partly because interviewees did not always divulge the full extent of their holdings.

In any event, table 26 suggests that rights over all sawah will pass to daughters or other female enates of household heads, regardless of the nature of the property rights at issue or whether the land is designated as customary or noncustomary. As for fruit orchards (table 27), the female enates of household heads will be favored overwhelmingly in the case of customary land, and to some extent in the case of noncustomary acreage as well. The least likelihood of female heirship as regards orchards is in noncustomary holdings purchased by current proprietors—resembling the case of residential acreage. Viewed conversely, male progeny are far more likely to receive shares of their parents' purchased rather than inherited noncustomary orchards and homestead plots.

The devolution of rights over land planted in rubber (table 28) is subject to the same set of temporal dynamics. Thus, we need not follow earlier observers in invoking a separate set of genealogical links or structural principles to account for the inheritance of rubber land or its relationship to other modalities of inheritance. At the same time, if we consider only residential plots and all types of agricultural holdings that have never before been inherited, the highest estimates for future male heirship are clearly those pertaining to rubber. This situation certainly derives in part from the fact that rubber land is utilized solely to generate cash income and is therefore strongly associated with domestic entrepreneurs, hence males. Yet the conceptual link between males and rubber land does not preclude females from inheriting purchased noncustomary rubber lands. For that matter, the conceptual link is far less pronounced with regard to noncustomary rubber lands that have already been inherited at least once, as evidenced by the projection that females will be the exclusive heirs in 60 percent of cases, joining their brothers in heirship in the remaining 40 percent.

The larger explanation for the extent of male inclusion in rubber acreage heirship lies with notions of siblingship and the pervasive view that brothers and sisters ought to be accorded equivalent shares of their parents' estate. The point is frequently made that, since daughters will inherit rights to houses, sawah, and residential acreage, sons should get at least some rubber land. Providing sons with access to rubber holdings is all the more important in light of the much-reduced economic and political significance of males' traditional inheritance shares, such as weapons and clan titles.

Beneath this concern with equivalence, however, is a more fundamental concern to provide daughters with property that will enable them to fend for themselves and their children if they are abandoned by or outlive their husbands. Rubber lands have become important components of such inheritance shares owing to declines in the economic value of women's rice plots. These factors help explain why so many women are expected to inherit rubber plots as well as rice and residential land, and why many, if not most, men will probably be bequeathed smaller shares than their sisters.

Siblingship and Gender in Inheritance

To account for the fact that female heirs have been favored over males throughout the past century, it is useful to examine local constructs of siblingship and gender. The social and especially the moral implications of these constructs demonstrate the uselessness of arguments resting on vacuous notions of "cultural lag" or "the dead hand of tradition" to explain this favoring of females; they also allow for powerful insights into continuity and change in property transactions and myriad other dimensions of social relations.

As I have stated, in recent decades neither the clan structure nor its associated sanctions—or those of other political or administrative bodies—have been instrumental in promoting villagers' adherence to practices that tended to exclude males from the inheritance of rights to noncustomary land or the houses thereon. In point of fact quite the opposite has occurred, inasmuch as twentieth-century forces of Islamic nationalism and reform have underscored Koranic perspectives on the privileged position of sons vis-à-vis daughters. We might also recall villagers' historically stepped-up strivings to orient their activities to Islamic idioms, and the fact that individuals with extralocal religious training perceive much of adat as incongruent with Islamic orthodoxy.

A crucial factor in the continuity at issue here is that moral obligations are generally felt to be of far greater scope and force with regard to daughters than to sons. This attitude is manifested in proprietors' explanations of why sons are or are not likely to inherit rights to the noncustomary land vested in a household member. Many individuals who desire or expect such holdings to pass partly or wholly to their sons referred to the religious and jural entailments of being Muslim. The question of sons desir-

ing, needing, or deserving shares of their parents' land, however, rarely arose. In marked contrast, those who intended effectively to exclude male offspring from such inheritance did not refer to formal duties codified in sacred texts, but spoke instead of affection for their daughters, who were far more needy and deserving of land than sons.

The broader theme is realized in the local view that adult males should bestow land rights on certain female relatives—including their sisters and daughters, and of course their wives, but not mothers or any other female kin—so as to contribute to the current and future well-being of these women. Such views attest to a pervasive concern that women not be left unprovided for owing to the death or desertion of husbands, who should assume protective and "elder-brotherly" (abang ) roles in relation to their wives but commonly fall short of the ideal. The comportment of husbands and fathers, and of in-marrying males generally, is implicitly judged by the ideals pertaining to the treatment their wives and children should receive from enatically related elder brothers. Just as disjunctions between theory and practice in this area account strongly for villagers' frequently negative appraisals of the behavior of in-marrying males, so too do these same appraisals underwrite those aspects of the gender ideology that portray males as inherently different from females and ultimately far less responsible and reliable. In short, notions of siblingship inform gender constructs and likewise conduce toward a favoring of female heirs.

Present-day gender ideology appears highly congruent with its counterpart of centuries past and quite likely to be reproduced in its contemporary form for some time to come. This ideology specifies and encodes models for the proper—that is, the God-given and "natural"—activities, roles, and relationships of men and women, and does so in accordance with local premises concerning the "innate differences" between male and female personality and behavior patterns. Here women are depicted as the more sacrificing and reliable of the two sexes, and far more deserving and in need of assistance and support than men. The personal sacrifices of women are in fact enshrined in myths of origin and both encountered and recalled in diverse contexts of contemporary village experience and activity. Women's emotional reliability and sense of social responsibility are also the subject of considerable cultural elaboration, particularly given the relative absence of such with regard to men. In the eyes of males and females alike, women are less given to mistreating or abandoning kin—be they same-sex siblings,

spouses, children, or parents—than are men. Similarly, they are much quicker than men to patch up differences and heal the occasionally strained bonds of cooperation and friendship linking them with relatives and other villagers. All this is reflected in the notion that women are "soft-hearted" (hati lembut ), whereas men have "hard hearts" (hati keras ).

Local perceptions of the innate differences between males and females figure into the bestowal of rights over houses and the overall thrust of property transactions at two critical points, both of which villagers commonly cite when they explain their adat to outsiders. First, since females are seen as the weaker and more vulnerable of the two sexes and more likely than males to assume a dominant role in childrearing, they require substantial protection and assistance from their parents and kin. It is universally assumed, for example, that all females will eventually marry and become mothers. Equally widespread is the conviction that having a house, along with part of a house plot and a stretch of cultivable land, will satisfy at least some of a mother's material needs and will also enable her to provide her offspring with shelter and subsistence. A mother may well have to fend for herself and her children for months at a time, since married men often take leave of their wives' communities for extended periods. Given the age differential at marriage, women also tend to outlive their husbands and therefore must have resources to fall back on if they are to survive, let alone live somewhat comfortably.

But women must be prepared not merely for the temporary absence or death of the household entrepreneur. Marriage itself is seen as a tenuous arrangement, a man's loyalties to his wife and children provisional, even capricious. The theme of desertion by husbands and fathers frequently dominates in villagers' explanations of why adequately providing for their daughters is so important. In this connection villagers also refer to the circumstances of outmigration (merantau ) and the demonstrated ability of males to eke out a living and make do whether they find themselves in their natal communities, among affinal kin, or in distant and culturally foreign locales such as Kuala Lumpur or Singapore. According to this perspective, men are simply more adaptable and therefore do not require institutionalized subsistence guarantees of the sort usually extended to women. In the final analysis, however, it is on account of male infidelity and whim that women require sustained assistance from their kin. Property in the form of houses and land has long been central to such assistance.

The other major reason why daughters tend to receive most, and sometimes all, of a household's land lies in the understanding that proprietors may someday depend on their heirs for nurturance and support. The unquestioned assumption that reciprocity should underlie the relationship between parents and their children is of critical importance in this regard, for children incur lifelong debts to their parents, which can never be repaid in full. Such debts of gratitude devolve on heirs and nonheirs alike, but children singled out for heirship stand especially beholden to their parents. At the same time, of course, children's support and aid are neither fully guaranteed in advance nor necessarily forthcoming when most needed.

One way of increasing the likelihood that a child will respond appropriately during hard times is to forestall the formal transmission of household property rights as long as possible, even if this means—as it usually does—going to the grave with unallotted holdings. This approach is straightforward in its logic and appeal: children who look after their mothers or parents during sickness and old age will receive the larger share or even all of the household property; those who do not will be left less, and perhaps none at all. Both contemporary experience and traditional lore testify to the value, though not the infallibility, of this approach. Similarly, villagers take as a given fact of nature that daughters will respond more compassionately than sons to their parents' hardships, whether or not future inheritance figures into the picture. Promises to sons concerning future heirship would not likely alter this fact appreciably. In any case, there remains the overriding concern to provide fully for daughters who might be neglected or abandoned by their husbands. Giving sons a share in the land that would otherwise pass exclusively to daughters might undercut the women's subsistence guarantees, leaving them less able to cope on their own. Worse yet, it could lead to the forcible mortgage or sale of that property owing to mismanagement, failure to pay taxes, or debts on the part of male co-heirs (that is, the women's brothers).

Issues relating to women's sexuality, particularly that of sisters, enter in here as well in that villagers argue a direct correlation between female landlessness and prostitution. Adults of both sexes boast that very few Negeri Sembilan women become prostitutes, referring by way of explanation to the subsistence guarantees extended to them through inheritance. Such guarantees are said not to exist among Malays elsewhere in the Peninsula owing to what some villagers regard as the "more Islamic" adat

temenggong. In local lore the Malay women of Kelantan serve as the negative case par excellence, for it is widely believed that female landlessness from male biases in inheritance forces many Kelantanese women into petty trade as well as prostitution. Bequeathing land to sons thus makes these men indirectly responsible for their sisters' highly inappropriate sexual behavior. This could occur even if the actions of male co-heirs did not result in the squandering of resources in which their sisters also held interests. In almost all cases we are speaking of rather limited quantities of productive land conveyed through inheritance; hence, the inclusion of male progeny as co-heirs could easily diminish their sisters' shares by just enough to preclude year-round satisfaction of domestic needs.

The continued biases against male heirs might also be seen as evidence of an attempt to preserve complementarity among cross-sex siblings, and this in the face of strong historical pressures to mute (or transform) this complementarity and promote a more pronounced (or simply restructured) sibling equivalence. After all, many villagers interpret Islamic property codes to mean only that sons are entitled to inheritance shares equal to those of their sisters. Yet in many cases even this view is resisted, for it enhances the likelihood of arguments and antagonisms between brothers and sisters, thereby undercutting the protective role and disinterestedness that should characterize brothers' behavior toward their sisters.^[9]

Gender Ideology, Female Heirship, and the Autonomy of Women

The gender ideology that I have outlined here raises important theoretical issues regarding the comparative study of women in society and culture. For one thing, much of the foregoing would seem to suggest a local perception of men as somehow "closer to nature" than women, since so often men are seen as the more likely to act on their base or animal impulses (nafsu ) and ipso facto as less bound than women by social and cultural constraints of the sort separating humanity from the animal domain and nature in general. In certain respects such a view does exist and is shared by males and females alike. This fact would thus appear to pose problems for Ortner's (1974) thesis that women are everywhere held to be "closer to nature" than males and therefore the "less cultural" of the two sexes (as well as culturally devalued and socially secondary as a group).

In fact, however, no such problems exist, for in the final analysis men and women alike feel that women are the "closer to nature," both in the general sense of having more natural affinities and in the more limited sense of being more able to threaten the established order. To appreciate this point and the apparent contradiction with perceptions of males as the more capricious, self-interested, and libidinally driven of the two sexes, we need only look to metaphors of biology and reproduction. These provide the key not only to numerous features of the gender ideology but also to many of the mechanisms by which this ideology promotes continuity in the favoring of female heirs.

More broadly, I suggest that it is precisely because this ideology portrays females as the weaker and in some sense the "less cultural" sex that they are accorded the larger shares of inheritance and subsistence guarantees, which in turn provide them with a material base (although not the only one) allowing them considerable autonomy and social control relative to males.

These contentions are profitably viewed in light of Malay notions of human nature and the fact that villagers frequently refer to the existence of two opposed elements or forces forever struggling within the individual. These forces are rationality or intellect (akal ) on the one hand and desire, lust, or passion (nafsu ) on the other. The point is often made that, much as proper actions testify to the dominance, however temporary, of rationality over passion, a "good person" (orang baik ) is one whose general behavior appears governed by the intellect rather than desire or lust. Conversely, improper actions and comportment that is less than refined bespeak an inability or lack of concern to control the baser impulses. Behavior seen as contravening social codes is not merely aesthetically offensive (tak sedap ) and unrefined (kurang halus ), but it also reflects faulty or incomplete socialization. Hence, individuals whose comportment is considered seriously improper are sometimes referred to as "less than fully taught" (kurang ajar ); they are thus accorded an intermediate standing between the world of animals and nature—where moral codes do not exist and thus need not be learned—and the rule-governed realm of humanity. We must ask, then, why females emerge in the final analysis as the more passionate and lustful of the two sexes, and therefore the less rational and cultural.

One explanation I encountered for the "greater animality" of females was rooted in procreative behavior and interpretations of the ease with

which males achieve sexual satisfaction (orgasm) as compared with their wives. According to this perspective, men are the more quickly gratified in intercourse because their sexual longings are simply less intense. That women are possessed by greater lust is evidenced by their desire to continue lovemaking even after their husbands have collapsed in satisfaction. I should remark, however, that these comments surfaced in discussions with male villagers and that I was unable to broach this particular subject with females. Yet since I found little disagreement between men and women on any issues related to procreation, gender, and the like, I assume that women interpret such things similarly. In any event, women clearly regard menstrual blood as extremely dirty and polluting, as do men, and both take very seriously all Islamic proscriptions concerning menstruating women's physical contact with the Koran as well as their involvement in prayer, fasting, and mosque activities. Menstrual blood is in fact referred to, by women and men alike, as "dirty blood" (darah kotor ), just as menstruation is likened to the moon's arrival (kedatangan bulan ) and a flood (banjir ). Since dirt or dirty blood, following Douglas (1966), may be viewed as matter out of place, hence disorder, the conceptual linkage between menstruation, lunar cycles, and flooding argues even more strongly for the association of women with nature and natural threats to established orders.

Numerous other areas would support the contention that local culture fosters a more intimate association with nature on the part of females than males.^[10] I wish here, however, merely to suggest that the social implications of gender ideologies that portray females as "closer to nature" than males are far more variegated than commonly recognized. Specifically, I would emphasize that these gender constructs may go far toward guaranteeing women rights over houses, land, and other forms of property and thereby relatively privileged positions of autonomy and social control vis-à-vis brothers, husbands, and men in other kinship and social roles.

In sum, females in Bogang and elsewhere in Rembau and Negeri Sembilan continue to be strongly favored over males in the inheritance of rights to houses, residential acreage, and most other categories of land, despite the colonial-era demise of divided title and numerous other precolonial property and inheritance conventions. Women's privileged position with respect to rights over houses and homestead plots contributes to the residential clustering of female kin and likewise promotes the dispersal

of their brothers and other male relatives. Accordingly, enatically related women continue to comprise the stable and solidary membership of contiguous households, lineage compounds, and residential settlements generally. Most men spend the better part of their married lives residing among their wives' immediate kin, where they are inevitably defined as inmarrying males beholden to their affines—which situation necessarily limits their authority over their wives and children and otherwise constrains their behavior both within the domestic domain and beyond. Thus men are typically unable to dominate or coerce women, either as individuals or in groups.

The continued favoring of female heirs is generally congruent with present-day gender ideology, which has remained largely intact from the nineteenth century to the present. This ideology also informs gender-based distinctions in production and exchange. Thus, it is a major force in ensuring that residentially aggregated female kin still assume most of the labor associated with rice cultivation, in addition to much of the responsibility for the maintenance and reproduction of socially valued exchange. Female predominance in these spheres affords women numerous opportunities to congregate together and to distance themselves, both physically and conceptually, from the domestic domain and the myriad entailments of caring for their husbands and children. In light of the recent decline of exchanges associated with rice cultivation and processing and villagers' much-reduced dependence on rice production, female exchanges keyed to ritual feasting and the creation and maintenance of affinal bonds are of particular analytic significance. Women's activities in these public arenas are both indispensable and a source of considerable prestige in the eyes of all villagers. They also feed into local views that women are more responsible than men when it comes to honoring kinship and social obligations, especially those involving the nurturance and provisioning of spouses and children. These views are, as we have seen, integral to a largely implicit but widely redounding ideology of gender that depicts men as more prone than women to capricious behavior and perfectly willing to abandon their wives and children when difficulties arise. This ideology, moreover, helps to explain why females are seen as requiring greater subsistence guarantees than males, and why so many females still end up with the lion's shares of their parents' estates.

More generally, the reproduction of this ideology has played a crucial

role in guaranteeing Negeri Sembilan women many of their precolonial prerogatives—including their relatively privileged positions of autonomy and social control. The data examined here should thus encourage a critical reassessment of the widely held view, espoused by Engels ([1884] 1972), Boserup (1970), Goody (1973, 1976), Sacks (1974), and others, that colonialism and modern market forces necessarily undermine the position of women in agrarian societies. Despite the general validity of many of these analysts' propositions, their universal applicability cannot be empirically justified (cf. Stoler [1977, 76], Peletz [1987b]).

9—
Social Change and Cultural Continuity

This chapter focuses on an intraclan feud in Bogang known as the "adat crisis" (krisis adat ). The events and ramifications of this social drama provide a critical window on continuity and change in local society and culture, elucidating general features of social process as well. So, too, do the entailments of ritual change and cultural rationalization, which are taken up in the final section of the chapter. To summarize, I argue that numerous dynamics and schisms continue to be framed in the symbols and imagery of adat and that these latter cultural phenomena still serve as the foci around which local sentiment is galvanized. I also contend that the twentieth century has witnessed a narrowing of adat's semantic range, which trend, ironically, attests strongly to the overwhelming cultural centrality of adat in village society.

Bogang's Adat Crisis

Bogang's adat crisis comprises a series of events that occurred in the mid 1960s. Both the incidents and their present-day repercussions merit careful

consideration for a number of reasons, particularly for what they reveal about personal and kin-based loyalties as well as about the scope and force of adat idioms in the rapidly modernizing Negeri Sembilan of today. It should be noted, however, that certain details of the events in question are hazy, even contradictory, reflecting discrepant contemporary accounts as well as after-the-fact embellishments and divergent reconstructions.

In major outline, the sequence of events occurred as follows. A young boy named Isa stole some sugarcane from a garden belonging to Hamzah, a fellow clansman (Lelahmaharaja) affiliated with Bogang's wealthiest and most prestigious lineage (Hill). Incensed by the theft, Hamzah struck Isa, thereby seriously alienating the boy's kin and violating a fundamental adat injunction against assaulting persons of the same clan. Isa's close relatives considered taking the case to court but later decided on an adat resolution to the problem. Shortly thereafter local clan leaders met to weigh the evidence and to determine the most appropriate course of action. After pronouncing Hamzah guilty, they ordered him to host Isa's kin at a ritual feast. Since Hamzah's older brother, Ibrahim, happened to be the local clan's highest-ranking official (that is, the clan subchief), it was his responsibility to see that Hamzah paid his retribution to young Isa's kin. But for various reasons—some quite obvious, others less so—Hamzah's brother did not pursue the matter with vigor. Almost everyone in the community saw his failure to enforce justice as a dereliction of duty, the principal exception being persons associated through descent or marriage with his lineage, who had little choice but to support him. Other villagers brought the matter to the attention of the clan chief and finally to the Undang himself. These officials concurred with earlier decisions regarding Hamzah's culpability and ordered Hamzah's brother to oversee the sponsoring of the feast and other retribution. Owing to a multitude of contemporary political considerations, however, they did not exercise their authority to remove Ibrahim from office, even though it was soon clear that the feast would never take place.

To this day the matter has not been resolved, and Hamzah's brother still enjoys the title of clan subchief, the most prestigious adat title in Bogang. This situation lies at the heart of what community residents refer to as the adat crisis, which is manifested both in the general boycotting of the clan subchief and in the sharp antagonisms between his wealthy and politically well-connected lineage and the other two lineages of the clan. As one ex-

ample, the clan subchief is not called on to preside over the engagement or marital exchanges involving persons from other lineages of his clan. Consequently, all such exchanges are in certain critical respects out of keeping with "the way things should be," as ordained by the traditions of ancestors. Individuals associated with his lineage, moreover, are commonly excluded from feasts sponsored by the other lineages, and vice versa. Further reifying this cleavage is the villagewide schism defined in terms of support for one or the other of the two national political parties. The clan subchief and his backers are allied with the ruling party, UMNO, whereas their principal adversaries in the adat crisis stand firmly behind the opposition group, PAS. Marriage rituals and the entire institution of feasting have thus become highly politicized. So, too, has prayer itself, for feasting centers around collective prayer and the sharing of food, and many members of the opposed political parties prefer not to pray—or eat—together. I will return to these issues in a more detailed analysis of the events leading up to and perpetuating the crisis.

I have already noted that the crisis dates from about 1965, which is when the eight- or nine-year-old Isa stole one or more stalks of sugar cane from another villager's garden. Although such petty thefts by youngsters were by no means unheard of, they typically reflected badly both on the children's upbringing and on their parents and immediate collaterals. These actions might, however, be written off simply as childhood or teenage shenanigans, especially if the offender's prior behavior had never drawn criticism from fellow villagers or if his kin enjoyed relatively high status in the community. Regardless of anyone's verdict concerning such a young thief's character, it would be extremely inappropriate to strike a child or young teenager caught stealing garden produce. Here, though, the owner of the garden, a married man named Hamzah, did strike the young Isa, who by almost every account was plainly guilty of the theft.

Unfortunately, it was never clear to me if Hamzah caught Isa red-handed and hit him then or caught up with him later. What's more, there was little consensus as to the severity of the retaliation, or even as to whether Hamzah struck Isa with an open hand or a clenched fist. The more reliable accounts indicate that Hamzah did not witness the actual theft but that he definitely overreacted when he later encountered Isa walking along the railroad tracks, for he grabbed the boy by his shirt and struck him about the head with enough force to bloody his lip. The inci-

dent generated considerable commotion—especially since any form of physical assault is extremely rare—and a sizable crowd of onlookers soon gathered. Some of those present decided Isa's condition required medical attention and so he was taken to the local hospital for treatment. The doctor in charge apparently felt that his injuries were not all that serious and did not merit Isa's remaining in the hospital.

When Isa's kin heard of the altercation they considered taking Hamzah to court on charges of assault. They reportedly went through the preliminary stages necessary for the initiation of judiciary action, although this may well have been a bluff and nothing more. At any rate, this approach was probably all the more appealing because one of the boy's close relatives worked at either the local or state-level court. As such, he may have been able to expedite the proceedings or, better yet, to contribute toward a court settlement in Isa's favor. As it turned out, however, a short-lived reconciliation was effected, and Isa's kin decided not to pursue the matter in court but to resolve the conflict within the village, in accordance with the appropriate adat traditions. It was (and still is) generally preferred to find a village-level solution to disputes of this sort rather than call on outside authorities, and this case was no exception. Indeed, it is safe to assume that virtually everyone involved experienced great relief—and a renewed sense of faith in the workings and relevance of adat—when the boy's side signaled their willingness to drop the idea of court litigation in favor of an adat resolution.

To appreciate the workings of adat justice we need to recall that the kinship relationship of the principal parties in a dispute is very important, both with respect to the person responsible for investigating the background and details of the conflict and in terms of the restitution or punishment and its enforcement. In the altercation at hand, the most salient feature of Hamzah's errant behavior was that the boy he struck belonged to one of his own clan's lineages. (Had Isa been affiliated with Hamzah's lineage, or with another clan, the incident and its ramifications would have assumed altogether different dimensions.) In theory, then, whoever was head of Hamzah's lineage bore full responsibility for convening local lineage elders to obtain accurate information concerning exactly what happened. In this case the de facto head of the lineage, holding the superior title of clan subchief, was also Hamzah's brother, Ibrahim (see figure 7).^[1] By virtue of office, Ibrahim had the constitutional duty not only to see that

Figure 7.
Principal Actors in Bogang's Adat Crisis

lineage elders agreed on the compensation owed the boy's kin, but also to ensure its prompt payment. The head of the boy's lineage had a comparable jural mandate, which included investigating the incident in the first place; calling a meeting of his constituents; helping them to reach a consensus on the restitution due them; and finally, seeing that the agreed-upon payment was in fact forthcoming. The two leaders were also expected to negotiate a settlement with each other if necessary and to cooperate with other lineage heads of the local clan as well as any higher ranking adat figures (such as the clan chief) brought in on the matter.

It was apparently in this latter connection that major problems first arose, for when villagers trace the roots of the feuding and factionalism that still dominate political dynamics and loyalties, they frequently cite conflicting loyalties and obligations, highly partisan behavior, and dereliction of duty, all on the part of Hamzah's brother Ibrahim. The basic issue was that although Ibrahim met with the three lineage heads of the clan and

seemed to concur with their decision on patching up relations between the two lineages, he did not proceed to enforce the payment of retribution—this despite the fact that as clan subchief he should have played the dominant role in settling the affair to the satisfaction of all concerned. His jural and moral responsibilities were not only to the members of his lineage but also, and more importantly, to the entire local clan; thus, loyalties to his lineage and close kin should have yielded to the broader issue of clan solidarity and the upholding of adat.

The verdict was that Hamzah should slaughter a goat and sponsor a ritual feast (kenduri ) at one of the village shrines (keramat ) to which Isa's kin, and presumably Isa himself, would be invited. Additionally, the lineage heads and clan subchief reportedly all agreed on the fine for which Hamzah stood liable—to wit, one bahara (M$7.20), this being a standard unit for the calculation of adat fines. This money, if paid, would have been claimed by the clan chief (Dato Perba ) and then partially redistributed among the four clan officers in the village. Hamzah, however, never paid the fine, nor—and far more serious—did he ever sponsor the ritual feast that would have symbolized both an acknowledgment of his wrongdoing and his remorse and goodwill toward the boy's lineage. Of perhaps greater historical relevance, these acts of contrition would have contributed both to the viability of concepts and processes of adat justice and to a general sense that the dominant values grounded in adat still held meaning for the conduct of daily affairs. Since nothing of this sort occurred, and in fact the subsequent stalemate worked in the opposite direction, the question emerges as to what forces led to the nonpayment of retribution.

The most general answer to this question lies in the nature of historical change over the past century and, more specifically, in the imposition and consolidation of modern governmental bureaucracies that deprived adat leaders of their control over effective sanctions. These trends fostered conditions conducive to the nonpayment of retribution and the ensuing crisis, all of which would have been inconceivable before the twentieth century or at least would have been of temporary duration and relatively minor significance in village history. This is not to suggest that traditional sanctions were always effective or that intraclan or other political feuding rarely occurred before the events leading to Bogang's adat crisis. It is to point out instead that in earlier times, if retribution were not paid, the affair

would have been settled by force of arms resulting in the impeachment or murder of the clan subchief, resettlement by the boy's lineage, or some combination thereof. In the contemporary era, however, such alternative solutions do not exist owing to the state's consolidation of power and its monopoly on the exercise of legitimate force, in conjunction with the current realities of demography and land acquisition, which render any sort of resettlement virtually impossible.

Yet history is certainly not the sole explanation here, since the loyalties and actions of particular individuals—however much a product and barometer of historical change—both precipitated and drew out the crisis. Here the plot thickens and we must focus on relations of kinship and social control within Hamzah's lineage, which seem to have precluded the payment of retribution. I refer to an essential fact not yet mentioned, namely, that when the incident occurred Hamzah and Ibrahim's mother not only was alive and residing in the village but also was the eldest, thus the most esteemed and influential, woman in the entire lineage. The mother, Wan (see chapter 6), apparently felt Hamzah's reaction to the theft of his sugarcane justified and not deserving of any censure. She refused to abide the supposedly unanimous verdict of the adat leaders and ordered both of her sons to ignore it.

Her position of loyalty to Hamzah created serious problems for Ibrahim in his capacity as subchief. Were he to insist that Hamzah pay the fine and sponsor the feast at the village shrine, he would so anger and alienate his mother that she might well disown him. (Indeed, she reportedly told him as much.) Nor could Ibrahim sidestep the issue by paying the fine and sponsoring the feast himself, for this would have amounted to an acknowledgment of Hamzah's guilt and a breach of fraternal and filial loyalty in his mother's eyes. And yet if Ibrahim adhered to his mother's wishes he clearly ran the risk of being discredited by Isa's lineage and possibly by other segments of the clan as well. In any case, Ibrahim chose to defer to his mother, ostensibly so as to remain on good terms with her and avoid being cast in the role of disobedient and ungrateful son, even though such deference meant dereliction of official duty and abrogation of moral responsibility. His overriding filial allegiance was not immediately apparent to members of the other lineages, however, for, as already noted, he had seemingly agreed to the particulars of the retribution during a meeting with the three

lineage heads. On that occasion, and for a short while thereafter, it was assumed that he would effect the payment of the fine and ensure Hamzah's holding the feast.

As the weeks passed it became increasingly obvious that Ibrahim was not seriously committed to compelling his brother to make the prescribed amends. The three lineage heads soon recognized that their only recourse lay in soliciting the intervention of the Dato Perba, Ibrahim's political superior and chief of the entire clan. This man lived in a village a few miles from Bogang and, like all his predecessors, was the second-highest-ranked adat figure in Rembau, being junior only to the Undang. By dint of office and regalia alone, his wisdom and authority in clan matters were above question, at least in theory. The three lineage heads were therefore confident that the Dato Perba would acknowledge the validity of their claims against the clan subchief, and would also insist on the latter's compliance with their decision concerning Hamzah's restitution. Indeed, the Dato Perba reportedly sided with them to the extent of concurring on Hamzah's guilt, and he ordered Ibrahim to insist that Hamzah honor his debt to the boy's lineage. Unfortunately for almost everyone involved, however, the Dato Perba's intervention proved ineffective. Consequently, and as a final resort, the matter was taken to the Undang. Here again the boy's kin looked forward to a binding decision involving some censure of Ibrahim's comportment as clan subchief, and an eventual end to the stalemate.

It may have been at this stage that a precise timetable for the payment of retribution was established (although some such details were probably considered at an earlier phase in the mounting crisis). The decisive—and ultimately divisive—issue is that the deadline passed, as did the subsequent one-hundred-day grace period. In fact, by mid 1980 the situation remained essentially unchanged, the critical differences being the progressive buildup of antagonisms and resentment between Ibrahim as clan subchief on the one hand and the three lineage heads on the other, and the attendant solidification of alliances and oppositions centering on the principal actors and their most loyal supporters.

Before turning to the various ramifications of the crisis, we might consider why neither the clan chief nor the Undang demanded Ibrahim's resignation once it became clear he had no intention of executing his responsibilities in the matter. They could have done so, of course, and in the view of many Bogang residents they should have, given their mandate to up-

hold justice and maintain the integrity of traditional offices. That they adopted instead a position of benign neglect as regards Ibrahim's dereliction of duty is one of the more distressing aspects of the entire crisis, for villagers tend to interpret this action (or nonaction) as colored, if not determined, by considerations of power, status, money, and politics—all of which should have been irrelevant in the case at hand.

To elaborate, both at the time of his assault on young Isa and throughout the ensuing crisis, Hamzah held a federal government appointment of considerable importance. Owing to his prominent position—not to mention a salary of astronomical dimensions, by local standards—Hamzah enjoyed almost unparalleled prestige in Rembau and the rest of Negeri Sembilan. In addition, because of his presumed close connections with state and other federal government officials, villagers regarded Hamzah as capable of exercising inordinate influence over the clan chief and the Undang. Villagers therefore assumed that these latter figures responded accordingly, especially to Hamzah's strategic placement within the corridors of power. In brief, the stock explanation as to why neither of these leaders called for Ibrahim's resignation is that they either bowed to deftly applied pressures or anticipated heavy political fallout if they pursued the issue to its logical conclusion. Expressed in one of the many village euphemisms, "they were playing politics" (dia main politik ). Worse yet—although villagers define this as an integral dimension of virtually all political behavior—their actions constituted a complete disregard of sanctified concepts of local justice and retribution, as well as the moral order in its entirety. In an era of all-consuming secularization, rapid erosion of long sacrosanct rural institutions, and pervasive uncertainty as to the viability of established local custom, these events assumed a profound and disconcerting symbolic significance. For, to paraphrase a comment I heard on numerous occasions, "if those at the top don't adhere to adat, why should anyone else?"

Let us now look at the subsequent ramifications and manifestations of this collective dilemma in greater detail. I should underscore that Ibrahim and Hamzah's entire lineage appears to have endorsed Ibrahim's general position throughout the whole affair. So too, it seems, did all of the men married to Ibrahim's lineage sisters, for in general the males who marry into a particular lineage do band together in political and most other broadly social affairs (and hence tend to be classified in terms of their

affinal connections rather than their own descent affiliations). Whether Ibrahim's affinal supporters were spurred by loyalty to Ibrahim, his brother Hamzah, or the two men's mother, they did stand basically united behind Ibrahim's decision not to force Hamzah to pay the adat fine or host the prescribed feast. Such at any rate was the public stance these men adopted, and I suspect that they also advised Ibrahim not to go ahead and sponsor the feast himself. Here again some motivational elements elude us, and a critical factor may well have been the long history of strained relations between Ibrahim's lineage and that of Isa. In any case, the prior existence of bad blood between the two groups definitely figured into the perpetuation of the crisis and its subsequent transformation into a villagewide affair with polarizing implications for group prayer, public animal slaughter, and officiation at adat rituals, as well as village leadership and affiliation with national political parties.^[2]

The circumstances responsible for the preexisting antagonisms between the two groups are probably far more complex than most contemporary villagers (or I) realize, and could easily date to early in the century, if not before. Nevertheless, dynamic factors of more recent origin occasionally assume dominance in villagers' accounts of these strained relations, which would seem significant in the generation and sustenance of the tensions in question. I refer to the existence within each camp of a few men regarded as both well versed in the art of public, persuasive speech and resourceful in mobilizing sentiment and resources for "getting things done"—in brief, men with charisma and leadership qualities. Spoken of at times as "big men" (orang besar ), these individuals bear certain similarities to Melanesian "men of renown," whether or not they hold clan office, a seat on a village council, or any other formal status. That is, in exchange for pledges of loyalty and support, they act to further the needs and interests of persons who either are less politically adept or simply prefer not to get involved in the strife-torn arenas of public life. Competition between big men is occasionally quite fierce, although it rarely erupts in physical violence or direct confrontations, verbal or otherwise. This rivalry may revolve around contests for the loyalty of lineage mates or affines, but it need not be constrained by distinctions based on descent or affinal ties. Thus, big men associated with a single lineage through descent or marriage may compete for the support of the same individuals. Such rivalry among full brothers is in fact a common theme in myth, history, and contemporary village politics. Con-

versely, big men with different descent and marriage affiliations frequently find themselves similarly pitted, but here the loyalties sought are those of men associated with various lineages.

In the Bogang case, Ibrahim, the clan subchief, is one such big man by virtue of education, literacy, verbal skills, and prior government employment. Also relevant is his brother Hamzah's access to state and federal officials and resources, and his own orientation toward "getting things done" for close kin, other allies, and the village community as a whole. Ibrahim's successes in this latter capacity and his stature as a big man are further enhanced by the titled positions he holds on the village councils (the UMNO Committee, the Village Development Committee, and the Mosque Committee), which nowadays constitute the principal organs of local government in Bogang and throughout the Peninsula. Finally, there is Ibrahim's close relationship with the village headman, who is married to Ibrahim's lineage sister (his MMZDD) and lives with affines who are Ibrahim's most immediate enates. This in-law bond makes the two leaders natural allies and mutes any potential antagonisms and rivalries between them—for the village headman is also a "big man," and he derives much of his support from individuals who, in the event of an intraclan dispute, would usually back Ibrahim. The larger issue, though, is not the muted competition between the two leaders but rather their reputation for furthering the material and other interests of the same relatively small group of people, most notably the members of Ibrahim's lineage (Hill) and the men who, like the village headman, have married women of that lineage. As mentioned before, this group is far and away the wealthiest in the entire community.

As for the broader matter of local coalition, Ibrahim and the village headman are allied with the half-dozen or so other prominent figures in Bogang, most of whom sit on one or another village council and support the ruling political party, UMNO. The principal exception lies with the big men who were—and still are—the most incensed by Ibrahim's betrayal of office. Interestingly, these two individuals are not associated through descent or marriage with either young Isa's lineage (River) or that of Ibrahim (Hill); rather, they represent the third lineage, Valley. They are, moreover, full brothers, and more important, they stand as half-brothers to Isa by virtue of a common father. For the past few years and perhaps longer, neither of these men has served on the more powerful of the village councils; and by numerous accounts they have steadfastly endeavored to

situate themselves and their followers in an adversarial position to the clan subchief, the village headman, and the power bloc those men represent. Facilitating this objective is their affiliation, and that of a disproportionate number of their lineage members, with the opposition political party, PAS.

Additionally, the junior and less influential of young Isa's half-brothers has been the head of the Valley lineage for the past two decades or so. Questions of patrilateral half-siblingship, big-man rivalry, and politicking aside, this man was fully required by the mandate of adat office and the duties of clan brotherhood to make sure that Hamzah paid the compensation to Isa's lineage or, failing that, to effect a mutually satisfactory compromise between the two lineages. More to the point, as we will see, his loyalties and responsibilities cast him as the principal adversary in relation to the clan subchief, especially once it became evident that Ibrahim had no intention of transgressing his mother's wishes by forcing Hamzah to effect retribution.

Further bolstering the younger man's position in the affair is the support he has received from his brother, Zakariah, who was—and still is—the only other big man of the lineage. Owing to Zakariah's standing as one of the community's most successful entrepreneurs, and perhaps the most charismatic of all Bogang residents, he has long constituted an appreciable threat to the established powers in the village. Aside from his expertise in syncretic ritual knowledge (ilmu ) and the formal teachings of Islam, Zakariah possesses widely admired verbal and social skills, an awe-inspiring bravery (berani ), and great determination in the face of potential obstacles to the realization of his objectives. He therefore enjoys an impressive following composed largely of lineage mates, male affines married to his lineage sisters, and others who are dissatisfied with Bogang's political elite, are allied with the opposition party, or, since the two tend to go hand in hand, both.

In sum, were it not for the prior existence of strained relations between Hamzah's lineage and the allies of Isa's half-brothers, both the incident and the ensuing antagonisms caused by nonpayment of the retribution might have faded into insignificance some years ago. For that matter, were it not for the bad blood between the two groups, Ibrahim's mother might have allowed, or even forced, one of her sons to go ahead with the restitution.

A related, but far more significant, theme is that the lines of cleavage and alliance engendered—or at least reified—by this feud are defined by variants of siblingship; hence, they cannot be interpreted in terms of idioms of descent, or with respect to hostilities between the two lineages of Isa and Hamzah. If the entailments of descent were the primary idioms informing the behavior of the principal social actors in the Valley lineage, for example, we would expect that these individuals would eventually have assumed a mediating role vis-à-vis the other two lineages of the clan, to which they stand equally related by descent. Instead, these men, like most other people associated with their lineage, have not only continued to boycott Ibrahim as clan subchief, but have also been far more actively involved than members of young Isa's lineage both in criticizing Ibrahim and in perpetuating the crisis—even though their lineage's honor was less seriously besmirched by the events following Hamzah's overreaction to the theft of his sugarcane. The dominant idiom underlying the alignment of these two lineages and their affines, and unifying them against the clan subchief's lineage and its affines, is siblingship; specifically, the tie of brotherhood that crosscuts lines of descent and links young Isa's lineage with that of his patrilateral half-brothers. More generally, just as the whole affair has revolved around various issues and categories of male siblingship from the beginning, its most disturbing implications concern the bonds of brotherhood that are supposed to obtain among all Bogang men by virtue of common village citizenship and equality in the eyes of Allah.

Isa's theft, to reiterate, was especially reprehensible insofar as Hamzah stood as a clan brother. Owing to this same tie of clan brotherhood, Hamzah should not have struck Isa—all the more so since he was endowed with the rationality (akal ) of an adult, whereas Isa was still a child. The clan subchief, in turn, failed to discharge his jurally mandated duties, which would have involved a breach of fraternal loyalty as well as seriously alienated his mother. Thus he, too, compromised the broad spirit of siblingship underlying clanship while honoring a narrowly conceived variant of sibling relatedness. The same could be said of young Isa's half-brothers, whose indignant response to the assault on the boy, and to the events that followed, was equally compromising to the same spirit of siblingship. Noteworthy as well, but of much wider concern, is that all of the men who have married into the clan are required not merely to take sides in the dis-

pute but to adopt the position of their wives' enates, as defined by the most influential males in the lineage. Those who are biras on account of their unions with women of a single lineage are therefore allied with one another and defined in opposition to the men married to women of different lineages within the clan, despite the fact that they themselves may be actual brothers or classificatory siblings owing to common enatic origin.

The question of taking sides in the dispute (since a middle ground seems never to have existed) emerges as a factor of critical importance both socially and politically in Bogang. This situation is quite explicit in the domain of ritual, where issues of sides and loyalties—not only to individuals, affines, and lineally related kin, but also to notions of propriety, due process, and justice—receive their most succinct expression. The orchestration of certain rituals necessarily involves making unambiguous ("either/or") public statements of allegiance with respect to the clan subchief's comportment and the entire crisis. I refer most directly to the ritual performances marking the engagements and weddings of villagers belonging to the Lelahmaharaja clan. On these occasions especially, adat conventions call for the beleaguered clan subchief to be present and to participate actively, both on grounds of protocol and so as to render the ritual efficacious. Further underscoring the clan subchief's pivotal role in these contexts is the fact that his immediate superior, the head of the entire clan, need not attend such rites of passage unless ceremonial displays of the Undang's sacred regalia (flags, swords, kris, and so forth) are incorporated into the ritual sequence. (Such embellishments are encountered nowadays only on the extremely rare occasions of three-day weddings.) For that matter, lineage head participation in engagements and weddings is customarily secondary, almost peripheral in relation to that of the clan subchief. In short, the rituals in question both revolve around and fully require the officiation and validation of the clan subchief.

Herein lies what is perhaps the most visible and disturbing element of the adat crisis: two of the three lineages of the Lelahmaharaja clan refuse to allow the clan subchief to discharge his ritual duties during their engagements and weddings. Viewed from the other side, the only members of the clan who request and sanction his participation in the betel, bridewealth, and other exchanges that are indispensable to both of the aforementioned sets of rituals are persons of his own lineage. Their actions in this area stand as clear testimony to their backing of his position in the

crisis. Conversely, the stance of the other two lineages is both intended and interpreted as a direct and searing indictment of the subchief's partisan behavior toward his brother Hamzah and in the management of clan affairs more generally. It amounts, in sum, to a denial of his supremacy within the local clan; in fact, as far as these two lineages are concerned, the Lelahmaharaja clan of Bogang has no legitimate head.

This boycotting of Ibrahim as clan subchief would not necessarily entail profound dilemmas for the clan and much of the rest of the village were it not for the simple fact, noted above, that proper performance of the rituals in question hinges on his direct involvement in them. As we have seen, rituals of engagement and marriage are not limited to effecting or publicizing a particular stage in the formation of conjugal bonds; rather, they are keyed to the attainment of far more encompassing objectives, such as the creation or perpetuation of symbolic ties between entire clans and lineage branches thereof. Related to these eminently valued goals is the lofty and increasingly elusive ideal of unity and solidarity among all kin and village residents. The attainment of all such culturally desired states is especially problematic at present owing to the adat crisis and the historical erosion of all social and cultural mechanisms capable of resolving it. For aside from the boycott of Ibrahim as clan subchief, most individuals tied by descent or marriage to his lineage are not only tainted by the affair but also excluded from many of the marriage feasts sponsored by the lineage of young Isa's half-brothers. Similarly, Ibrahim's lineage does not usually extend feasting invitations to persons of the Valley lineage, and even when it does, few Valley members actually attend. To do so is to incur the wrath and lasting vengeance of Isa's half-brothers, influential men in the lineage and in Bogang as a whole.

A good deal more might be said of Bogang's adat crisis and the schisms it has generated, but I shall confine my remarks to a few of its implications for ritual and politics. Especially noteworthy are the ramifications of the crisis in the domain of public animal sacrifice. A major rift occurred when young Isa's half-brother, Zakariah, was informed that allies of the clan subchief had decided the water buffaloes Zakariah procured for communitywide Islamic calendrical celebrations were not halal , or ritually acceptable to Muslims. This statement embodied a searing indictment of the dealings leading up to Zakariah's obtaining the oxen, which if unscrupulous, as alleged, would indeed render the animals unfit for Muslim consumption.

Gravely incensed by this attack on his religious credentials and personal integrity, Zakariah vowed to forego the future slaughter of oxen at the village mosque and to look elsewhere for a suitable sacrifice site. He settled on a pleasantly shaded spot near his wife's home, proceeding in addition to construct a small prayer house (surau ) in the same vicinity that allowed him and his supporters to minimize the time spent praying among and otherwise mixing with their rivals and enemies. Because the clan subchief and his principal backers still slaughter oxen near the mosque, every Islamic holiday marked by public animal slaughter sees the two factions separating into their mutually antagonistic camps. Furthermore, since these camps are simultaneously defined in terms of two opposed political coalitions, both the sacrifice of oxen and the entire adat crisis have come to be infused with the symbols and idioms of national politics. So too, of course, have the institutions of marriage, feasting, and prayer, as well as those of secular village administration and the adat polity, up to and including Rembau's Undang.

The politicization of public animal slaughter, as well as the adat crisis as a whole, is profitably viewed in relation to a controversy that emerged in Rembau in the early 1900s, when four disaffected clan chiefs refused to attend the installation ceremony of the newly elected Undang (NSSSF [1905] 2960/05). In consequence of so publicizing their disapproval of the candidate, they lost their titles. Nonetheless, and in a clear contravention of adat, they proceeded to sanction a feast at which a water buffalo was slaughtered. This comportment provoked the ire of British officials, who promptly informed them that all such behavior constituted "a great lack of respect [toward] His Excellency the Governor in whose hands the selection of the Penghulu [Undang] had by all Chiefs in Rembau been placed." Particularly interesting here are the parallels between these turn-of-the-century expressions of political opposition and those manifest at collective animal sacrifices in present-day Bogang.

Far more encompassing patterns of historical continuity emerge from the fact that the two lineages most directly involved in Bogang's adat crisis include among their respective members and spouses many of the principal luminaries and supporters of the two opposed political parties. This situation indicates that lineage-defined factions still predominate in local political arenas, and that association, whether by descent or marriage, with one or another lineage tends to dictate modern-day party affiliation. In

other words, traditional structures continue to define and empower social cleavages, even while they help to legitimize new divisions capable of transcending them.

The involvement of in-marrying males in this long-stalemated affair is especially relevant to these themes of structural continuity, and provides further support for my earlier point that the lines of cleavage and alliance expressed in Bogang's adat crisis ramify throughout the institution of marriage, being by no means confined to matters of ritual exchange. To appreciate this, as well as the quintessential politics of contemporary affinal alliance, we need only consider that the men who have married into the Lelahmaharaja clan are embroiled in the schism far more passionately and inextricably than their wives or, for that matter, than most male clan members—if only because so many of the latter do not reside in Bogang. Far from being a function of personalities or other adventitious circumstances, this situation indicates considerable continuity in the social identities of married men, especially those born outside the village. Even at present, such identities derive largely from a man's close affines, that is, from individuals affiliated with his wife's lineage. Thus, no mater how unencumbered these identities may be by jural and economic sanctions from affinal quarters, they are still informed by the moral and broadly social bonds obtaining between married men and their wives' closest enates. Bogang's adat crisis certainly stands as testimony to this theme.

Ritual Change and Cultural Rationalization

Bogang's adat crisis reveals that numerous dynamics and schisms continue to be framed in the symbols of adat. It indicates in addition that symbols of adat still have appreciable moral force and are clearly capable of galvanizing local sentiment.

Nonetheless, the twentieth century has witnessed a progressive narrowing of adat's semantic range. To appreciate this fact we need only examine ritual change and cultural rationalization. These processes highlight the increasing attenuation of adat, just as they suggest a concurrent conceptual trend entailing the progressive monopolization of sacred power and associations by Allah. While all such developments might be interpreted as evidence of the partial devaluation of adat, I would argue that they also attest to the ethical and overall cultural centrality of adat in village society.

Historical continuities and transformations in the ritual domain reveal numerous features and implications of cultural rationalization processes. It is patently evident, for instance, that virtually all sacred rites perceived by villagers to be Islamic in origin—such as those associated with circumcision, daily and Friday prayer, the Prophet's birthday, and the pilgrimage to Mecca—are approached with at least as much sincerity and devotion now as ever before. And yet there has also occurred a radical falling off, and in some cases the total abandonment, of previously sacrosanct ceremonies that today are seen, at least by the comparatively cosmopolitan segment of the community, as decidedly pre-Islamic in origin. These include sacred performances linked with ear boring, teeth filing, and the clearing of house sites and forest lands, along with other essentially pre-Islamic rituals such as berpuar.

The week-long berpuar complex constitutes the most elaborate of the rituals that are no longer performed locally; it also merits privileged consideration in light of Dato Abdullah's 1920s proclamations that this tradition testifies to the "residues of pre-Islamic Days of Ignorance." Notwithstanding the disdain of Dato Abdullah and like-minded contemporaries among the Malay elite, berpuar continued to be held throughout Rembau until the mid to late 1950s in accordance with the directives of traditionally highly esteemed and much-feared shamanic pawang. The performance of this triennial ritual involved representatives of all village households and was aimed most directly at propitiating the life force of cultivated rice (semangat padi ) and simultaneously ensuring both the safety of the fields and a bountiful harvest. Centering around offerings of incense, rice, fresh animal blood and meat, and a spectacular, celebrated phase of mock battle, the ritual was geared generally toward placating beneficent ancestral beings, like those of the earliest Undang, credited with the ability to drive away potential pests and demons (soil or earth spirits, or jembalang , for example). Interestingly, most villagers over forty years of age believe strongly in these spirits and attribute unsatisfactory rice yields to the cessation of this ritual. Even so, there is no interest at present in reinstating its performance, and consequently the reciprocal and nurturant bonds linking the community with heroic ancestral figures have been all but severed. As a result, the ancestors have not only ceased to protect the living but have also taken to venting their anger on them in myriad unpredictable ways—and in certain generally expected ones as well.

Parallel developments have occurred in the ritual repayment of vows (bayar niat ). Because these rituals contain Islamic elements, they can be addressed solely to Allah. Villagers point out, though, that they were most frequently performed as an obligatory gesture of gratitude to spirit beings and mythical figures for the assistance they had rendered individuals seeking good health, safe journeys, offspring, and economic prosperity. Until quite recently bayar niat were most often held at one of the keramat shrines that serve as the homes and resting places of dominant spirits (who occasionally take the form of weretigers under the partial control of shamanic pawang). In contrast even to the situation of ten or fifteen years ago, however, today persons participating in bayar niat usually do so in their own homes and also tend to neglect those aspects of the ceremony concerned with spirits. Both the focus and meaning of bayar niat have thus shifted considerably, and the once highly venerated institution of local shrines, like the office of pawang, has waned in relevance almost to the point of extinction. Owing to the languishing of traditions associated with shrines and pawang, it is increasingly difficult to ceremonially invoke Rembau's mythic past in the present and, in this way, to maintain an unbroken spiritual link between the region's legendary heroes and earliest settlers and its contemporary inhabitants.

Additional changes in the realm of ritual can be discerned in practices honoring deceased villagers. Since the late 1970s, for example, the annual feasts held for the local dead a week or so after the end of Ramadan have not included graveyard offerings of food, prayer, and incense and have been moved from the cemetery to the mosque. This change also originated with Islamic men of learning (ulama ), who visit communities like Bogang fairly regularly and seek to dissuade villagers from believing that the living can petition the dead for blessings and other aid. Despite villagers' willingness to have these feasts in the mosque rather than the graveyard, elders are unwaveringly convinced that their departed kin can indeed assist them in their strivings to attain good health, successful harvests, and prosperity, and they understandably lament the dwindling ties of reciprocity uniting them with their deceased kinsmen. They are also struck at times by the painful contrast between their own actions and those of their children, who, on moving out of Bogang, fail to maintain contact with even their closest relatives.

Ritual transitions of the sort outlined here involve far more than mere

shifts in the frequency, locale, or components of long-established sacred practices. Indeed, what has changed—or, more accurately, what is in the process of transformation—is the very nature of villagers' conceptions of the supernatural order. Also being transformed—since historical persistence here, too, depends on continued ritual validation of sacred axioms—are villagers' views concerning sin (dosa ), divine praise and reward (puji and pahala , respectively), and the extent to which adat is compatible with Islam and relevant to social action.

To elaborate, an appreciable and rapidly growing number of Rembau inhabitants—composed primarily of village youth and their better traveled, better educated, more politically active, and usually male elders—do not believe that local spirit beings can aid them in any way. Most of these individuals lean toward the position that spiritual assistance derives solely from Allah and that it is to Allah alone that prayers, sacrifices, and requests for blessing should be directed. In the view of these relatively cosmopolitan and religiously informed villagers, spiritual power and energy as well as sacred power tend to be wholly concentrated in God, rather than simply held by Him for the most part but also diffused among sacred shrines, graveyards, traditional curers and exorcists, and various local spirits.

This reform-minded but not yet actively reformist segment of the village population also feels that pre-Islamic rituals, long subsumed under the rubric of adat, should not be so regarded but should be seen instead as the behavioral outgrowth of kepercayaan . This recent conceptual shift is of particular significance in that the term kepercayaan not only denotes beliefs of a non-Islamic sort—and, more precisely, "superstition"—but is also used as a gloss for the "heathen" beliefs and practices of non-Muslim aborigines. The low standing of kepercayaan in relation to Islamic beliefs and activities makes clear that many of the previously central and sacrosanct features of adat have been relegated, by some villagers at least, to a much-reduced, even denigrated status. Equally apparent is that the once unitary and all-embracing concept of adat—a concept fragmented in the first instance owing to culturally divisive British colonial policies—has been further shattered such that its scope is both more restricted and more secular now than ever before.

The adat concept is thus being stripped of its most overtly non-Islamic elements, through the reclassification of these features as kepercayaan, and

recast as more congruent with the teachings of Islam. This paring down and rationalizing of adat by certain villagers may be interpreted as a highly creative (though not necessarily conscious) cultural accommodation to a historically specific context in which loyalty and adherence to the adat order, particularly its ritual dimensions, has come under attack as being inconsonant with Islamic orthodoxy. Through this conceptual readjustment, reform-minded villagers who have ceased participating in doctrinally controversial rituals but remain committed to the social and moral codes of adat can strive to maintain the integrity of the relationship between adat and Islam, as expressed in the customary saying "Adat hinges on law, law hinges on the Book of God" (Adat bersendi hukum, hukum bersendi Kitab'ullah ).

All such developments are of further historical significance given that the reform-minded segment of the community includes the village elite, and the adult males among them especially. These individuals are mostly former government employees with long records of military or bureaucratic service far beyond the confines of the village and Negeri Sembilan. As the "new elite," these individuals do of course differ from their nineteenth-century predecessors in terms of the trappings of prestige and power. More importantly though, they are identically situated both as cultural brokers and in their privileged capacity to influence social and cultural change. (Much of the reformist energy over the past century, it may be recalled, emanated from the local elite—be they 1880s clan chiefs, English-educated leaders of the 1920s and 1930s like Dato Abdullah, or members of the Religious Affairs Section of the Rembau branch of UMNO, who sparked the 1951 adat controversy.) In this regard, then, relatively little has changed, even though the cultural accommodation of varied historical events and processes has entailed a profound recasting of both central cultural constructs and their long-enshrined interrelationships.

The ritual declines noted above are in any event of relatively recent occurrence, however, and most of Rembau's adult inhabitants have taken part in innumerable ritual performances once widely categorized as belonging to adat. Through participation in the syncretic, although somewhat toned down, ritual feasts that still accompany marriage, circumcision, funerals, and the like, these people have directly supported and reaffirmed the sacred axioms embodied in the customary saying cited above. These axioms sanctify the entire realm of adat, however broadly or

narrowly that realm is defined, by asserting its inherent compatibility with Islamic doctrine, including religious law. It is therefore eminently comprehensible that adat, in either its rationalized or its earlier historic form, continues to dominate the cultural identity and social orientations of most adult villagers, regardless of their critical attitudes toward or breaks with specific adat conventions, and despite the differentiation and polarization of adat and Islam at various institutional levels.

Yet at the same time, a radically dissimilar orientation prevails among village youth and some of the younger, more mobile members of the adult community. Their understandings of the adat construct differ in important ways from those of their parents' generation, owing primarily to the demise of many adat-based social and political arrangements and rituals and to the institutional segregation of adat and Islam. Local youth especially know very little about even the most fundamental features and principles of adat, partly because customary sayings and other oral texts are no longer commonly recited as a form of recreation, to display linguistic virtuosity and erudition, or at marriage, land, and other comparable proceedings. Since these traditions, like the entire domain of sacred knowledge associated with pawang and keramat shrines, are only tenuously maintained at present, local youth see the adat concept as relating primarily to the non-Islamic aspects of marriage ceremonies, the clan structure, and the system of female proprietorship and divided title. Unlike the village majority of previous decades and even their reform-conscious elders, then, local youth do not view the construct as referring to an all-embracing set of beliefs, assumptions, and behavioral guidelines. Moreover, and in contrast to persons of their parents' generation, contemporary youth tend to perceive adat as neither integrally nor for the most part harmoniously related to the realm of Islam. In this view, adat constitutes a rather superfluous supplement to Islam, such that conformity to adat will have only marginal, if any, bearing on divine punishments and rewards encountered in the afterlife. Indeed, in the opinion of village youth, the proposition "Adat bersendi hukum, hukum bersendi Kitab'ullah" has little meaning; rather, it represents an anachronism that in times past legitimized a wide array of non-Islamic traditions and that today serves to hinder modernization and religious development especially.

The coming years will probably see a dramatic increase in the scope and force of these perspectives on adat and its relation to Islam. Much will de-

pend on the relative successes of groups of urban-based, and predominantly middle-class, Malays who seek the creation of a new social order congruent with their interpretations of the will of Allah as revealed to the Prophet. These groups spearhead the recent pan-Malaysian Islamic revitalization movements (glossed dakwah ), which are part of a more encompassing historical process involving much of the Islamic world (see Nagata 1984). Throughout the 1970s Malaysia's dakwah groups targeted many of the conventions and axioms central to village religion and culture, and also gained considerable urban support and national attention, especially from the media and the government. As of 1980, however, their impact on the rural inhabitants of Negeri Sembilan did not appear to be all that significant. Discussions with villagers in Bogang suggest that this situation prevailed largely because of the widely publicized (and frequently exaggerated) antimaterialistic orientation of certain dakwah groups. This orientation has had the effect of muting dakwah's influence in the countryside, since villagers throughout Negeri Sembilan remain committed to expanding their typically meager holdings and acquiring Western-style consumer goods along with other emblems of status and prestige.

Whatever the fate and rural impact of the dakwah movement in coming decades, it is clear that these and earlier currents of Islamic revivalism have served to elevate Islamic symbols and idioms in all domains of social activity and daily existence. It is highly ironic that these trends bear much of the responsibility for the increasing attenuation of adat, as expressed in part by the decline of pawang and spirit cults and by Allah's progressive monopolization of sacred power and associations. For although the irrevocable narrowing of adat's semantic range is matched in the demise of important features of clan structure, proprietorship, and exchange, such developments have not necessarily brought villagers any closer to what is certainly one of the most critical objectives of Muslims everywhere—namely, a sociological realization of the seamless brotherhood enjoined upon all believers by virtue of their identical standing under Allah. In fact, many villagers lament that the achievement of this goal is more elusive now than ever before.