Skip navigation

Land and Life in Timor-Leste: Ethnographic Essays

9. Tensions of Tradition:
Making and remaking claims to land in the Oecusse enclave

Laura S. Meitzner Yoder

Introduction: Changing relations and landholding practices

This chapter explores how rural residential and agricultural land claims in Timor-Leste’s Oecusse (Ambeno) enclave are established, maintained and transferred through various means including warfare, agricultural use, allocation by customary leaders and government programs, and migration. Settlement narratives illuminate some basic principles of claim making and explain how local customary leaders and early settlers preserve their favourable positions in relation to land control. This account analyses how landownership is linked to agricultural land use and village (suco) membership; the conditions in which both landowners and non-landowners acquire, borrow and use land; and how landownership and authority are transformed through agricultural change. This discussion also demonstrates how state land-titling policies that do not account for the causes and widespread incidence of land borrowing could inadvertently misidentify landowners, and might formally deprive a significant portion of the rural population of agricultural land access.

Recent property scholarship focuses on the human-relational aspects of ownership and use of land alongside other forms of claims to physical, cultural and intellectual property. There has been much attention to how definitions of ownership reflect or exclude the relational and identity factors at the core of many land claims (Hann 1998; Kalinoe and Leach 2004; Peluso 2003; Roseman 1998; Shamir 1996; Strathern 1998, 1999; Tsing 2003; Weiner 1999; Zerner 2003b). Different actors usually have different forms and evidence of claim; rural people, urban people, migrants, the state, those with natural resource-intensive livelihoods, and others with property interests have distinct expectations about what types and expressions of resource claims are appropriate or valid (Zerner 2003a).

The stunning complexity of customary land and forest regulation—in which ownership norms can be specific to each resource, lineage or location—is understandable when we consider that the bases of claims and the possibilities of ownership are embedded in social relationships and the identity of the would-be owner(s). Customary ownership norms are no longer viewed as a static ‘tradition’ (Hutchinson 1996), and specificity, mutability and responsiveness to social transformation are understood as intrinsic qualities in land use and tenure. Studying disembodied stated rights and practices is inadequate for understanding decision making regarding land if positions of relative power and relationships define the outcomes. Where ‘claims rest on demonstrable relationships’ (Strathern 2004:9), customary ownership is tailored to each case—better visualised as a constellation of practices or cases than as a single, coherent system. As ‘ownership is a function of relations between persons with reference to things, and not between people and things’ (Leach 2004a:43), this chapter examines how ownership is narrated and expressed in ever-changing modern land claims and use. Social relations in regard to land are often made visible through tales of relocation: settlement histories, accounts of moving towards roadsides and away from highlands to the coast during the years of Indonesian rule, as well as more recent shifts related to agricultural change.

Origins and Highland Settlement: Land claimed through warfare and agricultural use

The geographic focus of this discussion is on one roughly rectangular Oecusse1 transect from the coastal lowlands to–highland interior, here named Lekot, drawing on data collected during ethnographic fieldwork conducted throughout Oecusse from May 2002 to November 2004. It outlines the development of landholding and land-use patterns contained within one administrative village (suco), covering approximately 43 sq km and inhabited by 1370 residents in 2004. Rivers largely define Lekot’s eastern and southern borders, while the western border follows streams and ridges. Residents live in three areas: a large, concentrated coastal settlement, several small inland lowland settlements and an extensive village area at an elevation of about 700 m, 7 km inland from the coast. Ascending a north–south path that transects the length of Lekot in the dry season, travellers follow flat lowlands by a rocky riverbed 2 km inland before reaching water in the river. At that point, the path climbs steeply through 2 km of closed-canopy, seasonally deciduous forest, spotted with betel groves and meticulously irrigated with ditches and bamboo pipes. The forest abruptly ends, leaving travellers on a very steep, treeless switchback gravel trail, which ends near the highland settlement. The southern half of Lekot comprises rolling hills and short climbs.

Lekot had its major origin settlements in the mountainous interior rather than on the coast.2 The story below narrates original land claims and links early settlement to local authorities’ roles in land allocation and ongoing oversight. The myth relates how the principal customary authorities, the tobe (ritual figures with specific authority over matters of land and forests, including conducting agricultural rituals, approving swidden sites, influencing natural forces that affect agricultural outcomes, and regulating sandalwood harvest) and naijuf (political figures who act as village heads within the boundaries of a suco),3 came to be established by chasing out nearly all native inhabitants, with tobe domains named after warrior ancestors.4 It describes how residential precedence is one aspect of land authority: the old naijuf (naijuf mnasi) and great tobe (tobe naek) were two of the original (re-)settlers, and the minor tobe (tobe ana) was given that land because he was the lone survivor of the original inhabitants. This narrative also explains the major family groupings for landholding and ritual responsibilities that presently exist in Lekot. The possibility of landholding and (in part) the current location of swidden fields still reflect this settlement history and tobe divisions. This story is compiled from multiple tellings of the history of Lekot settlement by different families.

Long ago, there were only a few people in Lekot, but the Ambeno king [usif]5 told our tobe-naijuf ancestors to go attack the people who were then living in the area. So the tobe and naijuf and the people attacked these enemies and won, and the former inhabitants left because a lot of them had been killed.6 When they had all fled, there was still one person left hiding in a tree, who also cooked and ate up in the tree. After our ancestors won, they went back to report to the king that the area had been cleared of enemies, but at night the king himself looked up and saw a fire burning in a tree, and asked, ‘Whose fire is that burning in the tree?’ The people answered, ‘We don’t know.’ So the king ordered the people to go again to find the person building the fire, and when they did it wasn’t a person at all, but a civet cat [makan] making the fire. The people caught it to question it, but once it was captured, it was no longer a civet cat but a person, so he answered them: ‘Do not kill me, because I guard and control this area [au es atukus ma anonot bi kuan bale].’ So the people did not kill him, and he alone was left alive after the other enemies had been killed or fled from Lekot. So that person originated from the animal, from nature, and the people made him the minor tobe [tobe ana] because he knew how to communicate very well with the spirits in the area.7 However, our ancestors conducted rituals to ensure that he would not have many descendants who might chase us out in return; indeed, the minor tobe never has more than one son, usually born when he is already old.

In this war, the people of Lekot were successful in the struggle for land, so they settled it and made gardens [swiddens]. The great tobe [tobe naek] was chosen by the king and naijuf to control the largest land area, because he had a large family and they lost the most people during the battle.8 The old naijuf [naijuf mnasi] was also there before other people came. Later, it was the naijuf and the great tobe who chose the small tobe and gave him his domain. Then both tobe received more newcomers [refugees], and those people followed the activities of each tobe in his domain. Land belonged to the person who opened a garden there, and their descendants, who divided up the various locations they inherited from their father, until Lekot was filled with fallowed garden sites. Each family may own many seimu [claimed land area used for agriculture], but separated in different locations. All landowners also received land in the tobe domains.

The area of Lekot is divided into three domains [sopu]. The first is controlled by the great tobe. The second is controlled by the minor tobe and the other tobe [tobe tonene (agriculture) or tobe ulan (rain)], who were chosen later. The names for the domain areas are those of the ancestors of each tobe, who fought their enemies and gained power over the area. The third [the region as a whole] is controlled by the old naijuf and young naijuf, but has no land of its own. Each domain has other families which belong to it, and have most of their land there. All tobe and naijuf chose assistants, who became the warriors and defenders [meob] of the realm.9

The agriculture tobe [tobe tonene] was chosen during the Japanese time when the young naijuf took a wife from Suni-Uf in western Oecusse and the naijuf learned of new agricultural rituals from that region that would improve the harvests in Lekot. Returning to Lekot with his wife, the naijuf discussed gaining and using these rituals with the Lekot elders. They agreed, and each family contributed mutin [bead necklaces] according to the groups that divide meat [sispa’]10 at the great tobe’s ritual [fua oel naek]. They went there to learn these rituals and to buy seed,11 giving a payment to the leaders there,12 and the naijuf, great tobe and the people appointed one family as the tobe tonene to carry out these planting rituals henceforth.

This account illustrates some key principles of land authorities and land claims found throughout Oecusse. Amidst the wide variability present among Oecusse suco, in all cases a tobe with a defined land domain (for example, great tobe or minor tobe) is said to be of supernatural origin and/or among the earliest settlers to an area, which lends the requisite facility in communicating with ancestors and other spiritual elements in an area. This is not necessarily so for those tobe with ritual responsibilities but no domain (for example, tobe tonene). In these accounts, authority closely follows a pattern of precedence—inherently relational:13 the first to arrive has the highest authority, followed by later arrivals. But settlers rarely encounter an uninhabited area; autochthones are eliminated or displaced to make way for the new settlers, and when they are permitted to remain (for example, in Lekot, the small tobe and the two warrior families), they only hold their present positions by virtue of the conquering settlers granting them those roles. In the story above, the remnant figure of the lone survivor who became the small tobe is feared, respected and controlled to avoid his rising to dominance.14 Superiority in warfare trumps autochthony—but never completely.15

Origins and migrations are important aspects of land-claim establishment. The current generation is the ninth or tenth since the ancestors of the two oldest-recognised resident lineages (or family names, kanaf) moved to Lekot from Biboki, West Timor. These two families are now the primary warrior (meob) lineages of Lekot. It is now the sixth or seventh generation since the great tobe’s ancestor moved from Manatuto, a district in the main body of Timor-Leste, to Lekot.16 When the great tobe lineage arrived, they gave the symbols of power that they brought with them from Manatuto (gongs and rattan) to the old naijuf lineage, because as tobe they were not permitted to keep those items.

Other settlers were accepted into the domains of the tobe and claimed land by opening gardens; the 11 landowning lineages present in Lekot all describe how their ancestors originally farmed the scattered areas now owned by members of the lineage.17 Taboos (nuni) sometimes serve as a mnemonic device for the history of family migration. Hence, asking the history of a family’s nuni array can evoke their geographic journey through various points, where events occurred and which taboos were acquired along the way.18 The possibility of being included among the original—and thus current landholding—kanaf in highland Lekot was determined by at least the late 1800s or early 1900s. All kanaf19 who moved into Lekot after the 1912 war are still considered newcomers, and are not permitted heritable land rights in the transect.

The lineages’ landholdings were established by various means: in a few cases, the tobe relegated areas to individuals or families to cultivate, but most accounts state that the tobe did not divide the land. The land became divided and claimed through the process of cultivation, without intervention or direction from any higher authority; a family selected, cleared and farmed a garden location, and, since that point, that land has been claimed by descendants of the original farmers.20 Mechanisms of inheritance and land transfer21 vary by kanaf and region, but this accounts for the island-like, patchwork pattern of land claims present today across the Lekot landscape. Clearly, the usif (ruler) and naijuf were not involved directly in dividing up large tracts of land to family groups or to individuals.

Land Transactions and the Perpetual Landlessness of Outsiders

In most of Oecusse,22 upon marriage a woman leaves her birth kanaf to join that of her husband. Accordingly, seimu land is inherited by sons and remains largely within the kanaf.23 Widows retain land rights to their husband’s kanaf land, but for an older widow the land may pass directly to her sons. Unmarried daughters retain rights to their father’s kanaf land. Women (regardless of their marital status) frequently inherit individually owned parcels of flooded rice land. In some areas of Oecusse (but not Lekot; this seems to occur in suco with more recent and numerous in-migration), a man who marries into a landowning kanaf and fully transfers his suco membership24 to his wife’s suco of residence may be able to acquire heritable25 rights to a strictly limited portion of his wife’s kanaf land after multiple years of farming a plot and subject to approval by the tobe and/or naijuf with some payment to both figures.

Land access by male in-migrants is often problematic.26 Before the new household of an external husband (not native to that suco) and local wife is allowed to settle in the suco, the local leaders and members of each landowning kanaf (atoin kua tuaf) meet to consider how the land requirements of the new family can be met. As land is almost never permanently transferred to in-migrants (referred to as atoin anaomnemat, wandering people), the new couple may be guaranteed rights to borrow—but never to own—land, sometimes for a specified number of generations (cf. McWilliam 2002:162–7).27 As this disadvantages the couple’s children, the new pair is encouraged to move back to the husband’s home suco, where future land rights are secured.28 When a couple does move to the husband’s land, the wife’s relatives might express relief that the couple and their children are ‘freed’ from their demeaning status as landless borrowers.

If the couple remains in the wife’s suco, even for a generation, it is possible that their children or grandchildren might not be received as landowning ‘kanaf insiders’ when trying to relocate to the father’s home suco. This creates the most socially undesirable status with regards to land: a perpetual borrower—rejected as a kanaf landowner in one’s paternal landowning suco and unable to acquire heritable land rights in one’s suco of residence. Some villagers describe this state as ‘atoin matustanab’, meaning someone being sent back and forth in the middle (as in a ball game), without a definitive home or stopping place. In 2004, 14 of highland Lekot’s 120 households (12 per cent) were in these circumstances.29 Twelve of these families (in two kanaf, each now with six households) are descendants of people who relocated (as ‘political refugees’) to Lekot during the 1912 and 1975 conflicts. They have attempted to return to their former suco without success, and are now permitted to remain in Lekot only because they have intermarried with local women, but they still may not make permanent claims to any land. They have no swidden garden land (seimu) of their own. They do gain permanent rights to their house yards, even if local kanaf are reluctant to give them land.

The accepted disparity between de facto and de jure land availability is important in supporting the livelihoods of these landless people. With unlimited access to free borrowing of land for swidden agriculture in the present system, ‘atoin matustanab’ are no more economically vulnerable30 than their landowning neighbours. If, however, landholding formalisation through state-based regulatory reform or agricultural change reduced their access to land borrowing, this group would be economically disadvantaged.

Occasionally, people lose land due to debt incurred in emergency situations, and may regain land only by inflated repayments.31 Families borrowing maize during hungry times might pay a high price in land lost that might take decades to regain—a debt they work steadily to cancel. In rare cases, landownership is not entirely place based, but might be linked to the actual soil of claimed land. Annual flooding rearranges landholding alongside seasonal rivers. Flooding might remove tens of metres of riverbank land for stretches of 100 m or more; where that soil is deposited further downstream, the upstream person who lost land may claim the ‘new’ land created, even on the other side of the river. This permits the highly unusual phenomenon of people owning land in an adjacent suco, when literal land transfer occurs on rivers that mark suco boundaries.

Local Authorities and Land-Use Regulation

Settlement narratives and subsequent land relations influence local political and ritual authorities’ control over land parcels. The naijuf political authorities often were selected to serve as village head (chefe do suco) under the Portuguese colonial regime and much of their preferential landholdings were acquired in that role. There are two naijuf lineages in Lekot: one ‘old’, originally from Manatuto in eastern Timor, charged with oversight of traditional ceremonies; and the other ‘young’, which came later from neighbouring Manamas, West Timor, which is concerned with suco governance including suco boundary conflicts. Regarding the naijuf landholdings, as an original settler, the old naijuf has landholdings, as do the other village member kanaf, and the lineage’s land is treated the same as any other family’s. The young naijuf, however, has very little land as a relative newcomer and, in most discussions about land, Lekot residents state that the young naijuf has no land, other than that taken from the people,32 or other than what people gave to his lineage.33

As narrated above, the great tobe came to control the entire region, and subsequently divided the land with the (autochthonous) minor tobe. As the primary land authority, the great tobe maintains a level of control over the entire domain—for example, in 2003 the great tobe denied a farming group permission to open a garden in part of a claimed agricultural region (seimu) of the minor tobe area, on the grounds that the area was too steep and should be allowed to revert to forest.34 Precedence confers special status: as a ‘landholding tobe’ with defined areas, individuals in these positions occupy them by virtue of their early residence; as such, people emphasise that they cannot be replaced with another lineage, unlike the tobe, who is responsible only for agricultural rituals concerning planting and rain/wind regulation.35

While precedence and the ordered hierarchy usif-naijuf-tobe hold for most cases, on matters relating to land and forests or within the ritual sphere, the tobe has more authority than the naijuf. Their landholding, however, reflects levels and spheres of their authority too: the naijuf may appropriate land from the people, while the tobe may not. Many say that the tobe had all the land and then distributed it to others, while the naijuf originally had no land but accumulated holdings from others. The tobe is clearly the primary land-use authority, but the naijuf’s political power enables him to usurp land and to keep it (Schulte Nordholt 1971).

There is a fundamental difference between the tobe and the naijuf with respect to landowning: while the tobe’s own land is treated just like that of other villagers, the naijuf reserves preferential tenure rights. In a 2001 example of land-tenure rearrangement when converting a swidden area to a new flooded rice field (see Meitzner Yoder 2005:218–20), land from the original landowners was redistributed in equal parts to the people who worked to build the irrigation and terraces. By contributing labour to the project, individuals previously without land in that area acquired heritable individual rights to a defined plot of land, equal in area to the land now held by the original landowners in the rice field. In this case and others, the naijuf, alone of all the original landowners (including the tobe) involved in this scheme, was exempt from dividing his land with contributing labourers. Seven people who joined the project late approached the naijuf and were granted rights to develop a given portion of the naijuf’s land, and in exchange for their labour were allowed to farm their plots for three planting seasons without any payment to the naijuf. In subsequent years they should give 10 per cent of their harvest as payment for use of the naijuf’s land.36 Unlike all other farmers in the rice field, those planting on the naijuf’s land do not have heritable rights to the land, and should they discontinue using the land, it will revert to full control by the naijuf.

Irrigation development, and subsequent conversion to fruit-tree and/or flooded rice land use, inserts the layer of cabo-oel (literally, ‘water chief/head’) into the land authority framework. A cabo-oel is not an inherited position in the traditional hierarchy, but simply any person who proposes, originates and oversees an irrigation plan for a given flooded rice field or a fruit-tree grove. Importantly, the cabo-oel is not necessarily the landowner or even a suco resident, but secures broad, long-term use rights for the duration of an irrigated system.37 Irrigation impacts on tenure in giving a measure of authority over land distribution to the cabo-oel, and in giving other non-landowners enduring rights to use the land for as long as the irrigation scheme is actively maintained and productive.38 The cabo-oel obtains use of the land closest to the water source, followed (downstream) by the landowner. Beyond that, anyone interested in joining the irrigation scheme may be given space, subject to approval by the cabo-oel (not the landowner). Significantly, for land-use decisions in irrigated areas, the cabo-oel’s authority supersedes that of the customary landowner and the tobe.

Coastward Migrations and Agricultural Change

Lekot, like much of Timor-Leste, has undergone dramatic demographic shifts during recent decades. Before World War II, most people in Lekot lived in dispersed sub-kanaf clusters in the highlands; only four households lived on the coast, including the great tobe. From 1945 to 1975, the Portuguese administration encouraged people to move to the coast for the purposes of schooling, but most were reluctant to do so for fear of physical punishments and additional labour demands made on coastal residents. A few families moved to the coast during the 1960s, but by 1975 the majority of the population was still living in family clusters in the highlands or inland lowland settlements. From 1976 to 1981, the Indonesian planned development of settlements, roads and schools attracted a few families, and there has been a steady trickle of migration to the lowlands since the mid-1980s. After every house but one in Lekot was burned during the post-referendum violence of September–October 1999, most people returned to their same places of residence to rebuild homes. By 2004, Lekot had 1370 residents, just more than half of them now in the lowlands: 120 households (640 residents) in an inland settlement at an elevation of 700 m, and 140 households (730 residents) in one peri-urban coastal and three inland lowland settlements.

Lowland Lekot residents still consider themselves mountain people who happen to live on the coast; few have lived there for more than a generation. Although some now rarely visit their former settlement area, many still regularly farm and worship in the highlands, and make weekly trips uphill to see family. Lowland land practices replicate the upland models in which the tobe and naijuf play key roles relating to land—at times unbeknownst to state authorities. In Lekot, the first hamlet head in 1976 was the minor tobe,39 and although he was illiterate and from the mountains, he was responsible for an extensive lowland land allocation program, designating each highland family grouping a specified plot of land, which very few families took as their principal residence. Ten years later, the naijuf reallocated many of these plots to outsiders, including Indonesian civil servants and church workers, following the pattern of land allocation, in the form of grants (not sale) to outsiders. Methods of claiming uncultivated land (naija sona) and restrictions on landownership in the lowlands closely resemble these practices in the highlands.

Ownership, Land Use and Changing Patterns of Rural Landholdings in Lekot

Land-use patterns are significant because they shape initial ownership claims, and are of broader significance through village membership that permits a family to become landowners. But for already claimed swidden agricultural land, ongoing use can be surprisingly unrelated to ownership or membership. This overview of everyday land-claim assertion and validation highlights how family (kanaf) identity, neighbourly relations, individual initiative, group membership and land authorities’ permission circumscribe the possibilities for different forms of ownership in Lekot’s varied agricultural landscape.

Lekot has both mountain and lowland inland settlements and peri-urban coastal areas. Land and resource tenure vary according to location and land use; each land-use category has specific patterns and norms of acquisition, purchase or temporary use, and levels of ownership and authority on decision making about that land.40 Lekot’s coast was almost unpopulated until the 1980s, leaving it with the uncommon phenomenon of unclaimed land even today. Much land has been converted from unclaimed land to individual holdings from the late 1990s onward. The significant, recent coastward migration demonstrates land claims that resemble upland patterns. As the growing coastal population addresses food needs and preferences by converting dry land to flooded rice fields, there is also conversion of land from periodic swidden to annual irrigated cultivation, with accompanying changes in landholding structures.

Much South-East Asian historiography posits a sharp divide between highland swiddening animists and lowland paddy-rice farmers who adhere to world religions, often implying a unidirectional move towards lowland livelihoods and identity (Burling 1965; Leach 1954, 1960; Wolters 1999). Scott (2009), however, notes that the migration moves in both directions—from hill to valley and valley to hill—and that there are intermediate cultural-ecological categories. Throughout this section, I highlight some of these factors that transgress these standard categories—that is, swidden farmers who start flooded rice cultivation on their own; highlanders who move to the lowlands and take their systems and norms of land claims with them, or live in the lowlands but still farm in the uplands; and how farmers’ attempts to apply new land-division schemes can function or fail based in part on prior norms of land tenure. People in Oecusse use a range of classifications for land, described below.

House Yards

In both mountain and coastal settlements, the land or yard closest to the house (poa ume, uem balef) is used for seasonal mixed gardens of root crops, maize and vegetables, as well as for fruit trees and animal pens. Individuals hold the house yard, and fenced border perimeters constructed of stone, branches, palm ribs or living fences are precisely maintained.

In highland areas (past and present), building a house and planting trees41 associated with the house form one of the most enduring possible claims. Land and its planted trees remain kanaf property for several generations where one’s ancestors were known to have lived on a given plot of land (uem balef). This convention applies even when descendants move to another suco—an action that normally diminishes claims to other (agricultural) types of land.42 Allowing members of another kanaf to build a house on one’s land usually permanently forfeits one’s claims to that land, which then transfers to the new residents’ kanaf.43 Even when the new resident moves from that site, the former owner usually has no right to reclaim the land, and any other household subsequently wishing to build a house on that land would negotiate with the most recent residents (or their kanaf descendants).44 In past times, kanaf clusters moved frequently, so there are many such residential sites for each kanaf across a suco landscape; frequent localised migration contributes to the patchwork nature of landownership in rural regions. The permanent nature of these land transfers makes landowners reluctant to release fertile or well-situated land to another kanaf, but requests for agriculturally marginal land for house building are usually still granted.

Despite the strength of the land claim resulting from residence, highland dwellers show no preference for living on their own kanaf’s land over asking for better-situated land from another kanaf. Until 1975, highland Lekot residents lived in widely scattered sub-kanaf clusters. During 1975–99, most families moved their houses two to four times to comply with changing government requirements for proximity to roads (twice rerouted) and schools (relocated three times), creating a mixed-kanaf settlement. After independence in 2002, the hamlet head initiated a program to again relocate all residents adjacent to a new road and piped water system, mandating that each family select a 25 sq m plot on which to build a house along a stretch of road that primarily runs through land of four sub-kanaf.45 As a result, most households are again relocating to another kanaf’s land. Transfers of residential land are made without any form of payment or ritual debt.

The Indonesian Government created Lekot’s coastal peri-urban housing settlements on unsettled, densely forested land in the early 1980s. As products of state planning, they were subject to some formalisation.46 Although the state oversaw the land allocation, people frequently selected a Lekot tobe to witness their formal land acquisition or subsequent transactions (following the tobe’s highland role). After the initial allocation, land that is sold follows state land administration conventions of permanent transfer. Residential land borrowing rarely happens in this region, with the exception of land used (rent free) by outside settlers during the Indonesian era that reverted to former claimants when vacated in 1999.

Swidden Fields

Seimu (land, seimu; actively swiddened fields, lele) refers to any land that has been used for swidden gardens, and therefore subject to enduring claim by a given lineage (or, as with first-generation swidden cultivation, an individual). Swidden gardens (lele) are the central food source for both highland and lowland residents; most Oecusse swiddens are cultivated for only one season, and are then left to fallow.47Seimu land can be left fallow for years or even decades, and returned to forest, without losing its status as claimed land of a given kanaf.48 As the land underpinning the subsistence livelihoods in Oecusse, a large percentage of Oecusse’s land area—approaching 90 per cent—is recognised as seimu (including both irrigated and dryland agricultural fields, fallowed and cultivated).

People usually make contiguous gardens, sharing a common perimeter fence with neighbouring farmers to reduce fencing labour.49 Making the sturdy 1.5 m-high wood/stone fences required to protect gardens from free-ranging livestock is one of the most labour-consuming activities in creating a new garden (cf. Fox 1977:34).50 Group gardening also confers yield benefits as neighbours assist in guarding ripening crops from birds and monkeys.51 Throughout Oecusse, sometimes an entire village will plant on one small mountain in a given year, on contiguous land owned by several different kanaf, with a single fence constructed around the mountain’s base. Each year, farmers form different groups that clear vegetation and burn, fence and plant a chosen area. Anyone may join a farming group, regardless of one’s status as landowner or a kin relationship to others in the group. A household may have one large or several smaller garden locations per year.

Groups form when a landowner decides that a given area is appropriate for cropping and indicates this decision by chipping the tree bark along paths near the chosen area (tae pait), usually by May. This step is usually taken only after informing the tobe with jurisdiction over that land and receiving his assent. Seeing this sign of intention to open a swidden, others approach the landowner and may join the group until the planned area is full. Farming groups of 15–40 households are common throughout Oecusse. The landowner allocates internal boundaries for each farmer, which the farmers mark in four ways.52 Each household is then responsible for managing the crops within its plot, from clearing to harvest.

While new internal boundaries on swidden fields are created and marked each time people use land, seimu ownership boundaries between kanaf are permanent and usually follow landscape features, such as ravines, streams, paths or lines between prominent stones/trees. Individually owned fields, such as those surrounding the coastal settlement, are often farmed in shorter rotations of three years’ fallow (using leguminous trees—for example, Sesbania sp., gala-gala, to improve soil fertility); these plots usually have constructed (palm rib) or planted (with living fence tree species—for example, reo) boundaries that owners maintain even while the fields are fallow.

Initial land acquisition is based on agricultural use, but on land with established kanaf claims, making a swidden garden on borrowed land confers no rights to the land user beyond that season’s harvest (cf. Vargas 1985:74 ff.). With the labour advantages of this system, people more frequently join a group and plant on borrowed land than on their own; during fieldwork, about 30 per cent of highland swidden fields were farmed by the landowners.53 There is no preference for planting annual swiddens on one’s own land over planting on someone else’s land.54 Even individuals with extensive kanaf-level landholdings borrow land for swiddens if they favour a certain location (for example, proximity to another field or their home). It is common for groups to farm land even without the landowner as a member of the group. Group composition changes annually, but during the growing season the group members have responsibility to each other—for example, if one group member clears or burns a protected area, the entire group must assist in paying any fine assessed by the tobe and/or naijuf.

When borrowing land for swiddens, residents of the landowner’s suco pay no rent or harvest share to the landowner.55 Where a swidden area is close to the suco border, members of the neighbouring suco are (reluctantly) permitted to join the farming group as they will contribute to the fence construction and help protect the swidden from grazing animals. In these cases, neighbouring suco members must pay a negotiated amount (for example, one small basket of rice and 120 ears of maize) to the landowner, even if they are members of the landowners’ own sub-kanaf;56suco boundaries are stronger than any lineage affiliation. Multiple land-use or rental arrangements commonly exist within one fence. All farmers must participate in the harvest rituals of the tobe in whose domain they make gardens; contribution to this ritual activity depends on the physical location of one’s garden, not one’s residence or belonging to a given tobe or naijuf domain.57

A seimu may be owned at a variety of social levels—by single households, among siblings or at the sub-kanaf or kanaf levels58—according to the nature and frequency of its use. These classifications are not absolute, however, and no local language terms distinguish these levels. There are very few household-held seimu in highland Lekot, other than flooded rice fields. Household-held seimu in coastal Lekot include most land outside house fences but still near settlements, which are planted to maize, dryland rice and root crops in (minimally) three-year rotations, as well as flooded rice fields. Seimu held among siblings are usually planted to maize and rice in rotations of three years (with green manures planted) or more (bush fallow), and include fields further away from settlements than household-held land. Both household and sibling-held seimu are usually farmed by the landowners themselves and are rarely borrowed by others. Much seimu held at the sub-kanaf or kanaf level is farmed with rotations of 10 years or more. When kanafseimu are farmed, it is almost always as part of a multi-kanaf farming group.

Initial claims result in household ownership of a given plot of land, which the next generation may inherit and hold as sibling-level ownership, and with the increase of descendants after several generations may be recognised as kanaf-wide land. The prevalence of individual-level ownership near the recently settled coast links to debates over whether this indicates an individualising trend in landholding, which has received much attention in ethnographies and land administration elsewhere (Healy 1971; Lea 1997; Martin 2004; Simpson 1971; ter Haar 1948; Ward and Kingdon 1995b). Observers have long reported, and legislated, a move away from (presumed) communal towards individual landholdings in Timor (Anon. 1924; Gonçalves 1937; Martinho 1943, 1945; Sumardjono et al. 1994–95; Suryosuwarno 1993; Ubbe 1995–96). In contrast, some lowland Lekot residents point out that individual claims are always the necessary first step to creating kanaf-level landholdings out of unclaimed land, and they expect that today’s first-generation coastal household claims will over time evolve—communalising—into the sibling or kanaf-level claims that fill the highlands. The earliest residents of lowland Lekot have more kanaf-level claims than recent settlers, as an individual’s claim is shared among his sons. Most Lekot people describe a communalising–individualising cycle in landownership, resulting from changes through inheritance, migration and other factors (cf. Hooper and Ward 1995; O’Meara 1995:113).

On sibling-held and kanaf seimu, different members may have unequal levels of authority over using land (for access to trees, cf. Peluso 2003). There are informal gradations of rights according to residential proximity to a given plot of land. For example, with coastward migration over the past three decades, people now living on the coast will usually ask permission or inform their highland-dwelling siblings before planning a garden on highland kanaf land. This would not necessarily be done among siblings who all live in the highlands. Land claims can diminish over time when an individual moves away from the settlement and there is competing use by other members of one’s own kanaf who live closer to the land. An individual who voluntarily moves to another suco would not expect to return to reclaim full land rights after one generation of absence; that land59 would have become controlled by the remaining members of the kanaf. Seimu held at the kanaf level are very rarely sold; most people take extensive measures to prevent land from leaving their kanaf.

Flooded Rice Fields

Irrigated rice fields (aen oek) are one class of seimu, but their tenure and use arrangements are quite different from unirrigated fields. Flooded rice fields (ane = rice, oek = water/wet) are not widespread throughout Oecusse, as most of the district highlands do not have sufficient water to support this use. There are extensive fields along the lowland rivers, especially in the centre of Oecusse’s north coast, where the main crop grown is flooded rice.60 These primary flooded rice fields occur in areas not previously settled by Oecusse natives, and in these regions the tobe-naijuf authorities are not as well defined as in the highlands. As these fields are always individually owned, people from distant suco may purchase or sharecrop land in these areas. Irrigated rice fields are surrounded by constructed wood/stone fences as for swidden gardens; internal borders are clear with the low irrigation ridges that bound each plot.

The unusually abundant water in Lekot allowed flooded rice expansion near both mountain and coastal settlements. Irrigation was initiated in one unterraced highland site during the 1960s, with the first irrigated terraced rice field constructed in 1975. In 1975, one highland Lekot farmer who observed methods of rice production in lowland central Oecusse decided to terrace a tiny plot adjacent to a spring. Despite ridicule from other villagers, he persisted and planted that area three years in a row, by which time others believed it was possible and began to make small terraces for flooding on their own land. The technology caught on and the community slowly converted several areas into flooded rice fields over the past decades.61

All of the flooded rice fields in Lekot were conceived and constructed entirely with farmer initiative; none has ever received outside financial or technical assistance.62 By 1999, there had been many hectares of rice fields constructed in six locations in Lekot, but more than half of that area was not in use, as the terraced land or the irrigation systems had been made unusable by landslides.63 Anticipating food shortages from 2000–02 onwards, existing sites were expanded and three new sites were opened; in 2004, a major new scheme was under construction in the highlands, with plans to convert more than 200 ha of former swidden areas to flooded rice fields, encompassing land held by all 11 of the landowning Lekot kanaf. People in Lekot expect that this will transform their farming systems and reduce dependency on swidden fields, which many view as more subject to the vagaries of weather that destroys their crops (especially drought and lodging in high winds). While irrigation construction and initial field terracing are labour intensive, subsequent cultivation is considered much easier—and more productive—than making swidden gardens.

Highland and lowland flooded rice fields are individually owned and may be sold for cash.64 In the extensive rice fields of central Oecusse (but not Lekot), there are no restrictions on ownership or transaction by outside-suco residents or natives of other suco. Flooded rice fields also have unique gender representation among owners: many individual Lekot women have their own plots in Lekot and in the rice-growing suco of central Oecusse.

Village Commons or Unclaimed Land

Unclaimed land (naija sona) is that which has not yet been claimed by any kanaf.65 Land remains naija sona only if it is bare, grassland, inaccessible, infertile, landslide prone or otherwise unsuitable for farming; even steep ravines, if forested, are always claimed by a kanaf. There is very little naija sona in the Oecusse highlands—mostly small landslide areas or useless patches of infertile soil. Since Oecusse’s coast was sparsely settled before the Indonesian era, some naija sona persists on mountain ridges near the coast—a relic of the former interior-centric settlement pattern, which left distant coastal lands unused while fully occupying the highlands. Naija sona is not necessarily distant from settlements. On the coast, naija sona land can be adjacent to houses, where it is gradually claimed by individuals (via farming, planting trees, building structures or fencing) and converted to seimu. In coastal Lekot itself, more than 2 sq km of naija sona has been converted to individually held seimu since 1985. Along the Oecusse coast, naija sona is receding inland as lowland settlers continue agricultural expansion. Some naija sona areas are considered dangerous—home to troublesome spirits and best left undisturbed.

Members of the suco use naija sona as common land for grazing animals and gathering firewood or construction materials; one naija sona patch in highland Lekot cannot grow annual crops but is the source for a hardwood (tas tasi) preferred in house construction. Non-village residents may not use naija sona or its products without explicit permission from local government or customary authorities. All this land requires a local authority to approve its use by those outside the village, or its conversion to seimu by village members. Aware that Timorese planting trees would later consider the planted land as their own, the Portuguese administration forbade villagers’ unauthorised tree planting along coastal areas of Lekot until the late 1960s in an effort to reduce people’s claims to lowlands areas the state wanted for plantations or other purposes. Once land has been claimed through use and thus gained the status of seimu, it does not revert to naija sona, even if left uncultivated for many years.

Conclusion: Customary rural landholding in the modern policy context

The customary features governing land claims and resource management in most of contemporary Oecusse remain more influential in everyday land-use decisions than state designations of land areas for different purposes. State land-administration efforts that do not account for existing land claims and ownership systems risk enacting programs or legislation that bear little relevance to the local situation. Recent policy studies highlight the dangers of de-contextualised land policy, and emphasise the need for land-administration structures and policies that reflect, rather than seek to supplant, existing customary patterns (Meinzen-Dick and Mwangi 2008; Mwangi 2006).

Customary landownership is inextricably linked to social identity, place, historical ties and ritual connections. These factors are not fixed, but are constantly redefined in tandem with migration and changing circumstances. Agricultural developments, including irrigation or tree planting, are significant factors in rural landholding transformations. Intra-village claims to authority and agricultural patterns have shaped landownership and land-use relationships among residents within a domain.

The layering of political and ritual authorities, in evolving forms responsive to agricultural change, is a central feature of land access and use. All rural land—claimed or unclaimed through agriculture—is under the responsibility of at least two customary and/or state authorities who mediate land use and acquisition. The tobe and/or naijuf, other members of a kanaf and irrigation overseer (cabo-oel) might all need to be included in making land-management decisions, in additional to local government officials. Even today in peri-urban areas, formal land transactions usually involve the approval and legitimation of the relevant tobe and/or naijuf. New customary authorities, whether for rituals or irrigation, are designated as needed to give oversight to land use.

Persistent land claims are established through agriculture, settlement and labour. While initial landholding patterns are determined by agricultural use, landownership is subsequently rearranged through investments such as house building and installing irrigation or planting trees. Migration histories can help distinguish whether individually held areas are individuating, or merely newly claimed and evolving towards group ownership. The validity of a given claim is contingent on evidence of a wide range of factors: the claimants’ past acquisition and inheritance, recent use, customary or government positions, political allegiances during recent conflicts, skills, initiative and labour contribution to group projects.

In Oecusse, highland land use reflects current agricultural practices more than landownership. Early settlers have favourable landowning status, and there are ownership restrictions on in-migrants. Farming in groups that form to fence swidden gardens means that landowners often do not plant on their own land, and that non-landowners presently enjoy unproblematic borrowing of land for annual cropping. The ease and frequency of temporary land borrowing contrast sharply with the near-impossible permanent acquisition of land by outsiders. Early settlers became established as eligible landowners, but the range of eligible landowning kanaf has closed. Newcomers—defined as people not native to the local naijuf’s domain—can be excluded from acquiring heritable rights to land even after four generations of residence. State land policies must account for such features: a modern land-administration program that titles land based on agricultural use at a given point in time would frequently misidentify the locally acknowledged customary landowner, and a program that simply formalises existing customary claims would exclude a significant portion of the population from ever owning land.

This case illustrates how land tenure and concepts of ownership are responsive to agricultural and demographic change. As people moved closer to urban areas and developed irrigation, they modified underlying land tenure and practices to accommodate the different requirements of permanent settlements and settled agriculture. Increasing population density also leads to agricultural change, which in turn further modifies land use and tenure across this diverse landscape. Nascent land-administration programs must be able to accommodate the substantial changes that occur if they are to have relevance among a dynamic population.


Fieldwork for this chapter was conducted between July 2001 and November 2004. I extend thanks to the Oecusse Government and customary authorities, and to multiple reviewers for helpful guidance. Research was funded by: a Yale University Doctoral Fellowship; a Fulbright-Hays Doctoral Dissertation Research Abroad Fellowship; an International Dissertation Field Research Fellowship, Social Science Research Council and the American Council of Learned Societies; a Dissertation Research Grant, Yale Center for International and Area Studies; a Mustard Seed Foundation Harvey Fellowship; a Southeast Asian Studies Doctoral Fellowship at the Research School of Pacific and Asian Studies, The Australian National University; and a Language Training Grant, Yale University Southeast Asia Program.


Anon. 1924, Provincia de Timor: Compilação da Legislação Geral e Privativa desta Provincia sobre Concessões de Terrenos (Mandada publicar por P.P. No. 40 de 8 de Março de 1924), Imprensa Nacional Timor, Dili.

Aditjondro, G. J. 1994, In the Shadow of Mount Ramelau: The impact of the occupation of East Timor, Indonesian Documentation and Information Centre, Leiden, Netherlands.

Baumann, P. C. 1997, ‘Historical evidence on the incidence and role of common property regimes in the Indian Himalayas’, Environment and History, vol. 3, pp. 323–42.

Burling, R. 1965, Hill Farms and Padi Fields: Life in mainland Southeast Asia, Prentice-Hall, Englewood Cliffs, NJ.

Castro, A. D. 1867, As Possessões Portuguezas na Oceania, Imprensa Nacional, Lisboa.

Conklin, H. C. 1954, ‘An ethnoecological approach to shifting agriculture’, Transactions of the New York Academy of Sciences, vol. 17, no. 2, pp. 133–42.

Conklin, H. C. 1957, Hanunóo Agriculture: A report on an integral system of shifting cultivation in the Philippines, Food and Agriculture Organisation of the United Nations, Rome.

Dove, M. R. 1981, Subsistence strategies in rain forest swidden agriculture: the Kantu’ at Tikul Batu, volume 1, Dissertation thesis, Stanford University, California.

Dove, M. R. 1984, ‘Man, land and game in Sumbawa: some observations on agrarian ecology and development policy in eastern Indonesia’, Singapore Journal of Tropical Geography, vol. 5, no. 2, pp. 112–24.

Dove, M. R. 1985, Swidden Agriculture in Indonesia: The subsistence strategies of the Kalimantan Kantu, Mouton, Berlin.

Fitzpatrick, D. 2002, Land Claims in East Timor, Asia Pacific Press, Canberra.

Fox, J. J. 1977, Harvest of the Palm: Ecological change in eastern Indonesia, Harvard University Press, Cambridge, Mass.

Fox, J. J. (ed.) 1980, The Flow of Life: Essays on eastern Indonesia, Harvard University Press, Cambridge, Mass.

Fox, J. J. 1996, ‘The transformation of progenitor lines of origin: patterns of precedence in eastern Indonesia’, in J. J. Fox and C. Sather (eds), Origins, Ancestry and Alliance: Explorations in Austronesian ethnography, Research School of Pacific and Asian Studies, The Australian National University, Canberra, pp. 130–53.

Fox, J. J. 1997, ‘Place and landscape in comparative Austronesian perspective’, in J. J. Fox, The Poetic Power of Place: Comparative perspectives on Austronesian ideas of locality, Research School of Pacific and Asian Studies, The Australian National University, Canberra, pp. 1–21.

Fox, J. J. 1999, ‘Precedence in practice among the Atoni Pah Meto of Timor’, in L. V. Aragon and S. D. Russell (eds), Structuralism’s Transformations: Order and revision in Indonesian and Malaysian societies. Papers written in honor of Clark E. Cunningham, Arizona State University Program for Southeast Asian Studies, Tempe.

Francillon, G. 1980, ‘Incursions upon Wehali: a modern history of an ancient empire’, in J. J. Fox (ed.), The Flow of Life: Essays on eastern Indonesia, Harvard University Press, Cambridge, Mass., pp. 248–65.

Geertz, C. 1972, ‘The wet and the dry: traditional irrigation in Bali and Morocco’, Human Ecology, vol. 1, no. 1, pp. 23–39.

Gonçalves, L. d. C. 1937, ‘Adatrecht van Portugeesch Timor. I. Direito consuetudinário dos indígenas de Timôr’, Adatrechtbundel, vol. 39, pp. 443–74.

Griffiths, T. and Robin, L. (eds) 1997, Ecology and Empire: Environmental history of settler societies, Keele University Press, Edinburgh.

Hann, C. M. (ed.) 1998, Property Relations: Renewing the anthropological tradition, Cambridge University Press, Cambridge.

Healy, A. M. 1971, ‘Land problems and land policies in Kenya and Papua New Guinea: a comparative historical perspective to 1963’, in M. W. Ward (ed.), Land Tenure and Economic Development: Problems and policies in Papua New Guinea and Kenya, New Guinea Research Bulletin No. 40, New Guinea Research Unit, The Australian National University, Canberra, pp. 63–124.

Hooper, A. and Ward, R. G. 1995, ‘Beyond the breathing space’, in R. G. Ward and E. Kingdon (eds), Land, Custom and Practice in the South Pacific, Cambridge University Press, Cambridge, pp. 250–64.

Hutchinson, S. E. 1996, Nuer Dilemmas: Coping with money, war, and the state, University of California Press, Berkeley.

Kalinoe, L. and Leach, J. (eds) 2004, Rationales of Ownership: Transactions and claims to ownership in contemporary Papua New Guinea, Sean Kingston Publishing, Wantage, UK.

Lea, D. 1997, Melanesian Land Tenure in a Contemporary and Philosophical Context, University Press of America, Lanham, Md.

Leach, E. R. 1954, Political Systems of Highland Burma: A study of Kachin social structure, Harvard University Press, Cambridge, Mass.

Leach, E. R. 1960, ‘The frontiers of “Burma”’, Comparative Studies in Society and History, vol. 3, no. 1, pp. 49–68.

Leach, J. 2004a, ‘Land, trees and history: disputes involving boundaries and identities in the context of development’, in L. Kalinoe and J. Leach (eds), Rationales of Ownership: Transactions and claims to ownership in contemporary Papua New Guinea, Sean Kingston Publishing, Wantage, UK, pp. 42–56.

Leach, J. 2004b, ‘Preface: rationales’, in L. Kalinoe and J. Leach (eds), Rationales of Ownership: Transactions and claims to ownership in contemporary Papua New Guinea, Sean Kingston Publishing, Wantage, UK, pp. x–xi.

Li, T. M. (ed.) 1999, Transforming the Indonesian Uplands: Marginality, power and production, Harwood, Amsterdam.

McWilliam, A. 1996, ‘Severed heads that germinate the state: history, politics, and headhunting in southwest Timor’, in J. Hoskins (ed.), Headhunting and the Social Imagination in Southeast Asia, Stanford University Press, Stanford, Calif., pp. 127–66.

McWilliam, A. 2002, Paths of Origin, Gates of Life: A study of place and precedence in southwest Timor, KITLV Press, Leiden, Netherlands.

Martin, K. 2004, Land, Custom and Conflict in East New Britain, Resource Management in Asia-Pacific Program, Research School of Pacific and Asian Studies, The Australian National University, Canberra.

Martinho, J. S. 1943, Timor: Quatro Séculos de Colonizacão Portuguesa, Livraria Progredior, Pôrto.

Martinho, J. S. 1945, Problemas Administrativos de Colonizacão da Provincia de Timor, Livraria Progredior, Pôrto.

Meinzen-Dick, R. and Mwangi, E. 2008, ‘Cutting the web of interests: pitfalls of formalizing property rights’, Land Use Policy, vol. 26, no. 1, pp. 36–43.

Meitzner Yoder, L. S. 2005, Custom, codification, collaboration: integrating the legacies of land and forest authorities in Oecusse enclave, East Timor, PhD dissertation, Yale University, New Haven, Conn.

Meitzner Yoder, L. S. 2007a, ‘Hybridising justice: state-customary interactions over forest crime and punishment in Oecusse, East Timor’, Asia Pacific Journal of Anthropology, vol. 8, no. 1, pp. 43–57.

Meitzner Yoder, L. S. 2007b, ‘The tobe and tara bandu: a post-independence renaissance of forest regulation authorities and practices in Oecusse, East Timor’, in R. Ellen (ed.), Modern Crises and Traditional Strategies: Local ecological knowledge in island Southeast Asia, Berghahn, New York, pp. 220–37.

Metzner, J. K. 1977, Man and Environment in Eastern Timor: A geoecological analysis of the Baucau-Viqueque area as a possible basis for regional planning, The Australian National University, Canberra.

Mwangi, E. (ed.) 2006, Land rights for African development: from knowledge to action, CAPRi Policy Briefs, CGIAR, Washington, DC.

O’Meara, J. T. 1995, ‘From corporate to individual land tenure in Western Samoa’, in R. G. Ward and E. Kingdon (eds), Land, Custom and Practice in the South Pacific, Cambridge University Press, Cambridge.

Ormeling, F. J. 1956, The Timor Problem: A geographical interpretation of an underdeveloped island, J. B. Wolters and Martinus Nijhoff, Groningen and The Hague.

Panão, A. F. 1915, ‘Timor: agricultura indigena’, Revista Colonial, vol. 3, no. 30, pp. 198–200.

Peluso, N. L. 2003, ‘Fruit trees and family trees in an anthropogenic forest: property zones, resource access, and environmental change in Indonesia’, in C. Zerner (ed.), Culture and the Question of Rights: Forests, coasts, and seas in Southeast Asia, Duke University Press, Durham, NC, pp. 184–218.

Reid, A. 1997, ‘Inside out: the colonial displacement of Sumatra’s population’, in P. Boomgaard, F. Colombijn and D. Henley (eds), Paper Landscapes: Explorations in the environmental history of Indonesia, KITLV Press, Leiden, Netherlands, pp. 61–89.

Rodman, M. C. 1987, Masters of Tradition: Consequences of customary land tenure in Longana, Vanuatu, University of British Columbia Press, Vancouver.

Roseman, M. 1998, ‘Singers of the landscape: song, history, and property rights in the Malaysian rain forest’, American Anthropologist, vol. 100, pp. 106–21.

Saldanha, J. M. and Guterres, P. 2002, Customary Property Rights and Agricultural Production: A tale of two sub-districts in East Timor, Center for Economic Studies, East Timor Study Group, Dili.

Schulte Nordholt, H. G. 1971, The Political System of the Atoni of Timor, Martinus Nijhoff, The Hague.

Scott, J. C. 2009, The Art of Not Being Governed: An anarchist history of upland Southeast Asia, Yale University Press, New Haven, Conn.

Shamir, R. 1996, ‘Suspended in space: Bedouins under the law of Israel’, Law and Society Review, vol. 30, pp. 231–57.

Silva, J. C. d. 1897, Relatorio das Operações de Guerra no Districto Autónomo de Timor no Anno de 1896, Enviado ao Ministro e Secretario D’Estado dos Negocios da Marinha e Ultramar pelo Governador do Mesmo Districto, Imprensa Nacional, Ministerio dos Negócios da Marinha e Ultramar, Lisboa.

Simpson, S. R. 1971, ‘Land problems in Papua New Guinea’, in M. W. Ward (ed.), Land Tenure and Economic Development: Problems and policies in Papua New Guinea and Kenya, New Guinea Research Bulletin No. 40, New Guinea Research Unit, The Australian National University, Canberra, pp. 1–36.

Sousa Xavier, P. d. 1997, Studi tentang Hukum Pertanahan Adat Timor Timur di Kecamatan Uato Carbau Kabupaten Viqueque, Dissertation thesis, Sekolah Tinggi Pertanahan Nasional, Yogyakarta.

Strathern, M. 1998, ‘Divisions of interest and languages of ownership’, in C. M. Hann (ed.), Property Relations: Renewing the anthropological tradition, Cambridge University Press, Cambridge, pp. 214–32.

Strathern, M. 1999, Property, Substance and Effect: Anthropological essays on persons and things, Athlone Press, London and New Brunswick, NJ.

Strathern, M. 2004, ‘Introduction: rationales of ownership’, in L. Kalinoe and J. Leach (eds), Rationales of Ownership: Transactions and claims to ownership in contemporary Papua New Guinea, Sean Kingston Publishing, Wantage, UK, pp. 1–12.

Sumardjono, M. S. W., Suyitno Iswanto, H., Widodo, W., Sudjito Ismail, N., Bosko, R. E. and Listyawati, H. 1994–95, Report on research into Adat land law of East Timor, [Unofficial translation], National Land Agency and Faculty of Law of Gajah Mada University, Yogyakarta.

Suryosuwarno, P. 1993, Pemilikan Tanah di Timor Timur (edisi II), Kantor Wilayah Badan Pertanahan Nasional, Propinsi Timor Timur, Dili.

Susanto, A. S. 1994–95, Analisa dan Evaluasi Hukum tentang Transaksi Tanah Adat di Timor Timur, Badan Pembinaan Hukum Nasional, Departemen Kehakiman Republik Indonesia, Jakarta.

ter Haar, B. 1948, Adat Law in Indonesia, Institute of Pacific Relations, New York.

Traube, E. G. 1986, Cosmology and Social Life: Ritual exchange among the Mambai of East Timor, University of Chicago Press, Chicago and London.

Tsing, A. L. 2003, ‘Cultivating the wild: honey-hunting and forest management in southeast Kalimantan’, in C. Zerner (ed.), Culture and the Question of Rights: Forests, coasts, and seas in Southeast Asia, Duke University Press, Durham, NC, pp. 24–55.

Ubbe, A. 1995–96, Laporan akhir penyusunan monografi hukum adat di Timor Timur, Badan Pembinaan Hukum Nasional, Departemen Kehakiman Republik Indonesia, Jakarta.

Van Trease, H. 1987, The Politics of Land in Vanuatu: From colony to independence, Institute of Pacific Studies of the University of the South Pacific, Suva.

van Wouden, F. A. E. 1968, Types of Social Structure in Eastern Indonesia, Martinus Nijhoff, The Hague.

Vaquinhas, J. d. S. 1883, ‘I: Timor’, Boletim da Sociedade de Geographia de Lisboa, vol. 4, no. 7, pp. 307–28.

Vaquinhas, J. d. S. 1884, ‘O Sr. Major José dos Santos Vaquinhas’, O Macaense, Macau, vol. II, no. 95, p. 210.

Vargas, D. M. 1985, The interface of customary and national land law in East Kalimantan, Indonesia, Dissertation thesis, Yale University, New Haven, Conn.

Ward, R. G. 1995, ‘Land, law and custom: diverging realities in Fiji’, in R. G. Ward and E. Kingdon (eds), Land, Custom and Practice in the South Pacific, Cambridge University Press, Cambridge, pp. 198–249.

Ward, R. G. and Kingdon, E. (eds) 1995a, Land, Custom and Practice in the South Pacific, Cambridge University Press, Cambridge.

Ward, R. G. and Kingdon, E. 1995b, ‘Land use and tenure: some comparisons’, in R. G. Ward and E. Kingdon (eds), Land, Custom and Practice in the South Pacific, Cambridge University Press, Cambridge, pp. 6–35.

Weiner, J. F. 1999, ‘Culture in a sealed envelope: the concealment of Australian Aboriginal heritage and tradition in the Hindmarsh Island Bridge affair’, Journal of the Royal Anthropological Institute, vol. 5, no. 2, pp. 193–210.

Weinstock, J. A. 1979, Land tenure practices of the swidden cultivators of Borneo, Dissertation thesis, Cornell University, Ithaca, NY.

West, H. G. 1998, ‘“This neighbor is not my uncle!”: changing relations of power and authority on the Mueda Plateau’, Journal of Southern African Studies, vol. 24, no. 1, pp. 141–60.

West, H. G. and Kloeck-Jenson, S. 1999, ‘Betwixt and between: “traditional authority” and democratic decentralization in post-war Mozambique’, African Affairs, vol. 98, pp. 455–84.

Wolters, O. W. 1999, History, Culture, and Region in Southeast Asian Perspectives, Southeast Asia Program, Cornell University, Ithaca, New York.

Zerner, C. 1994, ‘Through a green lens: the construction of customary environmental law and community in Indonesia’s Maluku Islands’, Law and Society Review, vol. 28, no. 5, pp. 1079–122.

Zerner, C. (ed.) 2003a, Culture and the Question of Rights: Forests, coasts, and seas in Southeast Asia, Duke University Press, Durham, NC.

Zerner, C. 2003b, ‘Moving translations: poetics, performance, and property in Indonesia and Malaysia’, in C. Zerner (ed.), Culture and the Question of Rights: Forests, coasts, and seas in Southeast Asia, Duke University Press, Durham, NC, pp. 1–23.

1 The primary language spoken throughout Oecusse is Uab Meto (also known as Baiqueno), which is used throughout western Timor, alongside Bahasa Indonesia. Terms given throughout are in Meto unless otherwise noted.

2 Reid (1997:61–2) proposes that for most of Timorese history, the island might have been ruled from the ‘inside out’, and ethnographies from Timor describe pre-colonial interior political systems and relatively recent relocations to the coast (Fox 1977:63; Francillon 1980; McWilliam 1996; Schulte Nordholt 1971; Traube 1986).

3 The changing roles of these figures and their interactions with modern state processes regarding land and forest regulation are discussed extensively elsewhere (Meitzner Yoder 2007a, 2007b). The named, bounded domains of each of the 18–24 Oecusse suco are delineated by the customary leadership structure referred to as tobe-naijuf. During the fieldwork period, each Oecusse suco had one or two naijuf whose control defined the domain, with each averaging three to four tobe responsible for named, bounded areas nested within the naijuf domains.

4 The name of Lekot’s suco derives from the words for ‘settlement’ and the onomatopoeic ‘to dice/chop (meat) into small pieces’, which villagers say substantiates the importance of warfare in establishing the current residents in the region.

5 For further discussion of the Ambeno and part-Portuguese kings in Oecusse, see Meitzner Yoder (2005:65–8, 96–102, 112–24). Suco-level authorities make frequent reference to their legitimation from the Ambeno king. In this case, the king commanded the people to eliminate any existing inhabitants and to settle the region from scratch; a combination of strength and precedence bolsters the great tobe’s power, exercised in selecting and appointing additional tobe, including the sole native survivor.

6 Note that the current authorities (tobe-naijuf) won their authority and do not claim to be autochthonous.

7 Note here some of the basic features of tobe authority: autochthony, supernatural origins, facility with ritual communication, and selection by the (newcomer) local residents to be the tobe. Throughout Oecusse as elsewhere in the Asia-Pacific region, settlement histories often feature total or partial displacements of autochthones.

8 One key basis for land acquisition is sacrifice or loss in warfare—a theme repeated after 1999 with auto-compensation for wrongs committed by militia families and individuals.

9 In Lekot, two sub-kanaf are known as the primary meob for the region. Their landholdings are concentrated in adjacent sections of a high mountain with a rock cliff base, originally chosen for settlement because of its inaccessibility and easy defence from approaching enemies. The meob kept their residence there long after other families settled the flat land where Lekot’s highland settlement is today. That is where their first four generations are buried, and the meob families still take sacrifices or conduct prayers. Genealogies and oral histories from all Lekot kanaf agree that the meob kanaf were resident in Lekot before the tobe-naijuf structure described in this story came to be. The majority of their agricultural land borders an adjacent suco, in keeping with their roles as guardians. The warrior families have unique authority to call the people to warfare using a buffalo horn—a responsibility exercised during conflicts of 1987 and 1999.

10 See Meitzner Yoder (2005:339–40) on sispa’. Each family’s ritual responsibilities are defined by groups visible in their participation at the great tobe’s main ceremony that closes the agricultural season.

11 The phrase used is ‘sos pin [fin] pena, fin ane’: to buy the maize seeds, the rice seeds. Like the ‘civilizing influence’ of the part-Portuguese kings (Meitzner Yoder 2005:113–17), this is another example of new ideas coming from the outside—in this case, connecting rituals and seed, knowledge and material, to the marriage transaction. This linkage of cultural and agricultural knowledge is similar to the effort of the Oecusse district Agriculture Department’s intentions regarding the (agri)cultural associations and tara bandu ceremonies (Meitzner Yoder 2007b).

12 The price paid for this knowledge was one mutin bead necklace and one horse. Strathern (2004:6) noted that for some groups in Papua New Guinea, displaying the ‘foreign, exogenous—sources from which they have purchased the rights to perform a piece of ritual’ in a festival adds value to the event, as it demonstrates the advantageous external relationship they formed.

13 Fox defines precedence as ‘an oppositional notion based on the assertion of a relational asymmetry…a socially-asserted claim to difference that generally involves an affirmation of some form of “superiority” and/or “priority”…invariably applied recursively to create a concatenation of relationships’ (1996:131).

14 Relating the origin myth of Tunbaba, just east of Oecusse, Schulte Nordholt (1971:287) also noted ‘a certain ambivalence inherent in this story, as the two founders came with Sonba’i…while on the other hand they “were already there”’—again indicative of the importance of autochthony in claims to authority.

15 There is no terra nullius in Oecusse settlement stories (cf. Griffiths and Robin 1997).

16 Several members of this lineage can name the male line extending back to the individual who left Manatuto to move to Lekot, but no further. Manatuto–Oecusse linkages of governance and military service are found in Portuguese sources as well (Silva 1897; Vaquinhas 1884).

17 Once claimed, land owned bears little relation to land farmed, as described further below.

18 Fox (1997:8) describes topogeny as ‘the recitation of an ordered sequence of place names’, which is used to transmit social knowledge and to preserve memory.

19 After this time, even in-migrants with the same kanaf name as an existing family were excluded from full membership and landownership. For example, there are multiple branches of two Lekot kanaf bearing the same name and distinguishable only by their taboos, which reflect their settlement histories; one branch of each kanaf may not own land, while the others may.

20 This is the most common means of establishing land claims (Rodman 1987; Vargas 1985; Ward and Kingdom 1995a; Zerner 2003a).

21 In general, male children inherit land, and a father’s land is divided among all sons. The father determines the pattern of equal or unequal land division; while all sons often receive equal shares of subdivided land or equal rights to use undivided land, some fathers designate a larger share of land to the eldest son or the son responsible for maintaining the family’s ritual house. Daughters rarely inherit swidden land; land from a couple without sons will be inherited by male relatives of the father.

22 An exception is the central flooded rice-growing region of Padiae (suco Cunha and Lalisuk), settled by Savunese and Rotinese in-migrants still called kaes metan (black foreigners/strangers) by Oecusse highlanders, in which most land is held and passed from mother to daughter. Until about 30 years ago, a new couple’s residence was exclusively uxorilocal, and the wife’s family took responsibility for providing the new family with land. Unlike other parts of Oecusse, Padiae inheritance frequently favours the youngest child with a larger share of family land and eventual ownership of the parents’ house.

23 It is worth comment that the subject of land tenure has been barely mentioned in the vast and detailed ethnographic literature analysing marriage, bride-wealth and kinship alliances around Timor, perhaps influenced by van Wouden’s (1968) influential study (Fox 1980).

24 In these cases, it is essential for the man to contribute animals to local ritual events, and sometimes give the local land authorities gifts of animals.

25 Even in these cases, the land retains links to the wife’s kanaf.

26 Customary restrictions on outsiders’ landownership have long been amply documented in Timor (Castro 1867; Metzner 1977; Sousa Xavier 1997; Vaquinhas 1883). Even access to unclaimed or common land can be unequal (Baumann 1997).

27 The wife’s kanaf takes responsibility for ensuring that the couple will have access to enough land to meet their housing and subsistence needs, and the couple will participate in the same system of swidden land borrowing as suco members.

28 The difficulty of outsider men obtaining land is reflected in the numbers of local and outsider marriages even in coastal, peri-urban Lekot: 57 per cent of couples have both husband and wife from the same suco; 33 per cent are a local husband and an in-migrant wife; and just 9 per cent represent a local wife and an in-migrant husband, classified as atoin anaomnemat.

29 In addition to these 14 households, in 2004 highland Lekot had three cases pending since 2001–02. This is indicative of the villagers’ reluctance to accept in-migrant males—called ‘the seeds of land borrowing’—as members of their settlement. In one pending case of outside people being rejected, a widowed mother of a local married woman (now with her husband’s kanaf) was not permitted to move in because she was accompanied by an orphaned grandson; village leaders said the woman’s move would have been unproblematic if the orphaned child had been a granddaughter or if the woman had been alone, but a landless boy must not be permitted to settle in Lekot. Village authorities, including the hamlet head and tobe, served as gatekeepers to suco land access.

30 On occasion, however, the atoin matustanab alone will clear the most marginal patch of unclaimed land (naija sona) in highland Lekot to plant low-yielding root crops—without converting that land to heritable seimu as a local kanaf would—when unable to get sufficient space in a group garden. Vargas (1985:18) demonstrated that non-landowning ‘newcomers to the village made smaller fields [than other residents], indicating that they were not free to borrow as much land as long-term village residents were able to borrow’.

31 For example, during the 1960s, one man needed immediate assistance to fulfil his tax obligation: in exchange for about 2 kg of rice, he traded 1 ha of irrigated rice land, which was regained by his son only in 2003 for the price of two large cows (US$300).

32 Various young naijuf have appropriated some of the choicest land in Lekot, taking advantage of Portuguese labour practices and Indonesian land formalisation to acquire exclusive and permanent private rights to some of the most valuable land. This included declaring the largest-existing irrigated rice field owned by several local families, stone-fenced with village labour, to be his own during the Portuguese era; acquiring a private Indonesian title (while village head) to a large coconut grove planted with community labour while he was the Portuguese-era village head; and personally claiming a former public recreation beach site after 1999 by having his resident assistant construct a fence around the area.

33 The phenomenon of the landless naijuf is heard sporadically throughout Oecusse: in four other suco, the naijuf does retain a domain of his own, though small. Naijuf sometimes state that they have no land because they divided it all out to the tobe. One naijuf gleefully tells of how his lineage gave out all the meat and was left with only the bones, handing out all the land to the tobe and the population, except a sacred mountain adjacent to his house. This is a paradoxical display of power: owning it all so he does not need to own any.

34 When the group disobeyed this prohibition and opened gardens in that location anyway, mistakenly burning a portion of a betel grove, the great tobe assessed a fine subsequently paid jointly by all members of the group.

35 Since the ritual tobe do not have particular precedence in settlement legitimating their positions of authority, their places are more tenuous and they may be subject to removal if seen to commit ritual errors that cause calamity or fail to protect harvests.

36 Villagers note that his exceptional landholding status permits flexibility in cultivating these rice fields, allowing those without land or those with extra requirements to use his land under reasonable conditions.

37 In Oecusse, individual springs not used for agricultural purposes fall under the authority of the tobe rather than a cabo-oel, differentiating the cabo-oel from the Tetun cabo-be’e (oel = be’e = water) in other districts of Timor-Leste.

38 If the scheme fails or falls into disuse, all users’ rights are fully relinquished to the original owner.

39 Villagers joked that the Indonesian Government did not know that the man they selected as hamlet head was a tobe—an appropriate person to serve as overseer for the state-sponsored land-distribution program.

40 Published land and tree tenure information for rural Timor-Leste is scant (for example, Metzner 1977; Saldanha and Guterres 2002; Sousa Xavier 1997; Susanto 1994–95; Ubbe 1995–96), and few accounts indicate the effects of migration and infrastructure development on changes in tenure patterns (Aditjondro 1994; Fitzpatrick 2002).

41 Generally, trees remain the property of their planter (or descendants) for the life of the tree. In urban and rural land transfers, the original tree/landowner may choose to retain harvesting rights to the trees or to sell the trees, separately from the land, to the new landowner. Trees and land are always transacted separately, whether land is purchased or borrowed (cf. Peluso 2003; Rodman 1987; Van Trease 1987:11; Vargas 1985).

42 Over time, kanaf members remaining in the former suco will absorb the rights to the house land, but rights to the trees remain with the original owner (or descendants) even after that point.

43 In more densely populated rural areas (for example, southern Oecusse), such land grants for residential use are reportedly becoming less common, sometimes now replaced with outright sale as practised in urban areas.

44 There are exceptions, as happened in Lekot during 2003 when Kanaf A wanted to reclaim their hillside land below a spring that had been settled in the 1930s by Kanaf B, by now home to 30 households, in order to develop a nearby water source into flooded rice fields. The decision, mediated by elders from each kanaf and the hamlet head, involved Kanaf A giving an equivalent land area just above the water source for Kanaf B to relocate their houses, as well as a share in the new rice field. This was permitted because Kanaf B’s relocation was necessary for the irrigation development that would benefit all Lekot kanaf.

45 The village head felt that formally limiting the plot size would reduce any potential complaints the original landowners would have about losing so much of their land, and the resettlement proceeded smoothly as all families received permission from the kanaf landowners for their proposed house site.

46 Of 67 households living in the coastal Lekot settlements since 1992, only 33 (48 per cent) reportedly held Indonesian land certificates, although most households (94 per cent) paid land tax annually from 1992 to 1997. Lekot highland residents report that there have never been land sales, land titling or surveying there.

47 Swidden agriculture is more prevalent in Oecusse than elsewhere on Timor. There are very detailed descriptions (cf. Conklin 1954, 1957; Dove 1981, 1985; Weinstock 1979) to complement overviews of swidden agriculture on Timor (Metzner 1977; Ormeling 1956; Panão 1915; Schulte Nordholt 1971).

48 Farmers usually establish original land claims through cultivation and maintain those claims through repeated cultivation, periodic maintenance of fallows, visits to the fallow site, establishing markers or fencing, and speaking about claimed land (or trees). The actual procedures, time to claim establishment or disappearance, and relative strength of claims are dependent on vegetation types, the status of the claimant and agricultural practices, as discussed further below (Conklin 1954, 1957; Dove 1981:88–101, 1985; Ormeling 1956; Vargas 1985; Ward and Kingdom 1995b). In Lekot, the largest contiguous forested area of kanaf-level land runs 2 km between the highland and coastal settlements, adjacent to a river. Individuals from seven kanaf farmed that land (that is, claimed and converted it to seimu) during Portuguese times, but early in the Indonesian administration the Forestry Department prohibited agricultural use by declaring the land protected forest. The area was largely left to return to closed-canopy forest after the 1970s, but despite its return to forest, the kanaf maintain their respective claims and can still indicate the boundary markers of planted and non-planted trees, rocks or ravines between kanaf sections. In some families, land claimed by the kanaf bears some measure of internal, individualised, heritable division among family members over a given land area. With today’s dense forest cover, the high population of monkeys makes agriculture difficult, so many family heads discourage family members who want to make swiddens in this region. In 2003, a customary authority (tobe) enforced the ban on farming in this area, stating that it should remain forested since gardens there were unlikely to meet with good harvests.

49 Those who do make gardens individually often take the unusual step of planting there for two or more years, to make the fencing labour worthwhile.

50 Mathematically, the labour savings in group fencing are substantial—for example, 10 people farming together must each build only 32 per cent as much fence when compared with farming (and fencing) the same land area alone. This strong preference for having a common perimeter fence contrasts with situations in which households prefer having separate, if nearby, gardens (see, for example, Dove (1981:113 ff.).

51 Other potential benefits of group gardening include pest saturation and labour exchange, as in shared fencing. Labour exchange for planting and harvesting is, however, minimal in Lekot (cf. Dove 1981:117; Vargas 1985:19, 112). In general, the planters are fully responsible for all aspects of land preparation, planting and harvesting within their own boundaries. Schulte Nordholt (1971) reported a greater degree of labour sharing in western Timor.

52 First, when clearing the land, vegetation along the boundaries is cut to 1 m high, while all other vegetation is cleared to ground level; second, after burning the field, farmers place wood or rock lines around their areas; third, farmers plant a row of maize or sorghum (different from the irregular spacing inside their plots) along the boundaries; and fourth, farmers might plant a short-season variety of maize along the borders of their area.

53 As population density increases, ease and frequency of land borrowing decrease. With 32 people/sq km (and 3.2 ha/person), Lekot’s population density was less than half the average for Oecusse. In populous southern highland Oecusse, rotating swiddens are slowly giving way to the annual cultivation of settled agriculture, including tilling, not yet practised anywhere in Lekot outside some house yards. Swidden, permanent dryland and irrigated practices coexist throughout the region (Dove 1984).

54 This point becomes important when landholdings are subject to titling. For example, Ward (1995:219) noted how this problem arose in Fiji with the Government’s ‘assumption that the registered owners of the land would also be the users’, when ‘in fact ownership and use were often not in accord’. Since most Lekot farmers practice similar land preparation and none carries out improved fallow management, landowners do not identify soil problems or damage to their fields resulting from other people’s use of the land; at this stage, no farmers were considered better than others regarding knowledge or skill in soil improvement.

55 At most, each local resident in a group will contribute a small animal towards the planting and post-harvest feasts, but a land borrower’s contribution is the same as a landowner’s contribution in these events.

56 A sub-kanaf is a subset of a large kanaf that shares additional specific taboos (nuni), described above, and is distinguishable by an additional name, appended to the kanaf name, which refers to the taboo. Full sub-kanaf names, with the appendix, are not normally used in everyday settings.

57 As members of a neighbouring suco frequently farm part of Lekot, the number of participants in Lekot’s primary harvest ritual usually exceeds the number of Lekot households.

58 In small families, sometimes all members of a kanaf or sub-kanaf are also siblings. Kanaf are small, well-defined units, limited to members of a group. It could be misleading to call this land ‘communally owned’, as all claims are within one (extended) family. See Fox (1999) for more on the intricacies of different ways to gain membership in a given kanaf. It is extremely complex for outsiders to correctly identify the socially relevant level of landownership (cf. Geertz 1972; Ward 1995; West 1998; West and Kloeck-Jenson 1999; Zerner 1994).

59 Planted trees, however, do not transfer to fellow kanaf members.

60 Significant rice-field development occurred using forced labour during the Japanese wartime occupation, with continued expansion to the present day.

61 Lekot villagers insist that they are still unskilled wet rice cultivators, and that their planting style is slow and cumbersome. They say that when working side-by-side with people from primary rice-growing areas of central Oecusse, their relative lack of familiar dexterity in transplanting the rice is humorously obvious. To them, it is still largely an outsider, lowlander (kaes metan) technology.

62 Oecusse Agriculture Department staff never visited Lekot during Indonesian times, and they were amazed to learn after independence of such extensive flooded rice development in the highlands. This lack of government intervention is notable, given Indonesia’s many programs—from transmigration to agricultural subsidies and farmer incentives—designed to promote wet rice cultivation in places with swidden cultivation or other crop staples (Aditjondro 1994; Li 1999).

63 There are many abandoned terraces in highland Lekot—by some estimates, twice as many abandoned terraces as productive terraces. One farmer likened Lekot people’s treatment of terraces to swiddens that, without purchased fertiliser, needed to be left fallow.

64 More detail on changing tenure for flooded rice fields can be found in Meitzner Yoder (2005:214–25).

65 This term is sometimes also used for infertile, unused claimed land that is left bare, so ownership status must be clarified with more detailed questions.

Land and Life in Timor-Leste: Ethnographic Essays

   by Edited by Andrew McWilliam, Elizabeth G. Traube