Year : 2004 | Volume
| Issue : 1 | Page : 75-95
Conflicting Concepts: Contested Land Relations in North-western Vietnam
Humboldt University Berlin, Berlin, Germany
Humboldt University Berlin, Berlin
Source of Support: None, Conflict of Interest: None
|Date of Web Publication||18-Jul-2009|
| Abstract|| |
In villages of north-western Vietnam land allocation provided a window in which different conceptions of land relations came to light. Villagers resisted the implementation of key elements of the new land legislation, though the new law purported to extend people's control. Their resistance manifested a fundamental disjuncture between the exclusive and territorial concept of land rights promoted in the new land law and people's lived land relations. They refused to give up the substance of land relations that had proven useful before collectivisation, under collective agriculture and again in the initial years of decollectivisation. People's reactions highlight how post-socialist land reforms provoke their own forms of resistance. Villagers negotiate the reforms in conflicts over resources and authority as well as over the very concept of landed property. This article examines the nature of these conflicts, explores their linkages with socialist and post-socialist land legislation, and relates them to the larger literature on post-socialist property relations.
|How to cite this article:|
Sikor T. Conflicting Concepts: Contested Land Relations in North-western Vietnam. Conservat Soc 2004;2:75-95
| Introduction|| |
VIETNAM'S NATIONAL ASSEMBLY passed a new land law in 1993. Like similar laws in other post-socialist countries, the Vietnamese one promised a radical extension of private control over land.  In comparison with the previous legislation, the law accorded the same bundle of rights to all landholders, independent of their political and social status. It also augmented landholders' rights to include the right to transfer land rights to third parties and extended the duration of the rights. Furthermore, the law mandated the establishment of a new specialised agency for land administration, which for most agricultural land had been the responsibility of producer cooperatives in the past. The first task of the agency was land allocation, that is, the measurement, demarcation and registration of landholdings. The expectation was that enhanced rights and land registration would meet rural people's presumed demand for private control over land.
Resistance to land allocation comes as a surprise, then. The Black Thai of Chieng Dong, where I conducted research in 1997 and 1998, resisted land allocation vigorously.  People not only ignored and circumvented the new regulations, they also objected to some legal requirements openly. In practice, households continued to modify field sizes and exchange fields among one another, ignoring their registration and the associated legal duty to report changes in landholdings. Whole villages disregarded land classifications commanded by the state, preserving overlapping and seasonally changing uses of land for cultivation, animal husbandry and forestry. Black Thai resistance to land allocation even included open protest against the registration of some fields, forcing the local authorities to exclude them from land certificates. In sum, land allocation met fierce resistance.
In this article, I examine people's resistance to land allocation in Chieng Dong. In particular, I seek to understand why people resisted land allocation almost in its entirety. I locate people's motivations in cleavages between land relations in the villages and land legislation. My findings suggest that villagers resisted the very concept of landed property contained in the new land law. Their opposition to concrete elements of the land legislation manifested a more fundamental disjuncture between the concept promoted in the new land law and people's lived relations to the land. People were not ready to give up the substance of land relations that had proven useful before collectivisation, under collective agriculture and again in the initial years of decollectivisation. Land allocation, its implementation and the reactions it provoked, thus provided a window in which different conceptions of landed property came to light.
My examination draws upon classical anthropological approaches to property relations, as initiated by Malinowski (1935) and developed by Gluckman (1972, original in 1965). Classical approaches direct attention to the legal concepts underlying property and the wider social relations in which property is embedded. Max Gluckman's work appears particularly useful because he depicts landholdings as a series of allocation involving overlapping powers. I thus follow Hann (1998) and Verdery (1998, 1999) postulating that Gluckman's work is a useful tool by which we can look at socialist land relations, contrast them to the conceptions underlying the new land legislation and make sense of the dynamics resulting from their encounter.
My discussion of post-socialist land relations is informed by in-depth fieldwork in Black Thai villages of north-western Vietnam. The Black Thai are an ethnic group related to the Tai peoples of Thailand and Laos that migrated into the mountains of what is today north-western Vietnam in the first centuries AD. The group today includes approximately 400,000 people, accounting for a large share of the population living in the region. I spent a year in three Black Thai villages in 1997 and 1998, conducting semi-structured interviews with sixty-five households, talking to many people and observing their everyday lives. I conducted interviews with local government cadres and obtained copies of available land records. I also searched the relevant archives in Vietnam and France.
The article begins with a discussion of Gluckman's approach to property and the insights it facilitates for an understanding of post-socialist land relations. The following section introduces land relations in Black Thai villages in the first half of the twentieth century, under collective agriculture and in the initial years of decollectivisation. Discussions of the new land legislation and villagers' reactions follow, including in-depth analyses of three concrete conflicts over local land relations. The article concludes with a discussion of the theoretical implications for our understanding of post-socialist land relations.
| Gluckman's Approach to Property and Its Application to the Study of Post-Socialist Land Relations|| |
Within classical anthropology, Max Gluckman provides what I consider the most comprehensive concept of land relations in The Ideas in Barotse Jurisprudence (1972). Gluckman's concept of property relations incorporates important insights from Malinowski, that is, that land relations typically combine individual entitlements with collective control, are the object of multiple claims by various social actors, and are embedded in broader social relations (Malinowski 1935: 328-29, 376-79). Yet Gluckman goes beyond that by developing a comprehensive concept, one that I find sufficiently encompassing to describe and dissect land relations in pre-socialist, socialist and post-socialist contexts. I, therefore, want to elaborate Gluckman's concept in some more detail.
Gluckman's most fundamental conjecture is that landed property embodies a hierarchy of overlapping rights (Gluckman 1972: 89-93).  One right is to work the land, which Gluckman calls 'estate of production'. At the same time, various other claimants hold 'estates of administration', that is, nested layers of control over land (ibid.: 78, 91). Gluckman conceives of property as a 'hierarchy' in the sense of a 'series of estates' (ibid.: 91-94, 110-11). Each landholding is a series of allocations. In the case of the Barotse, the simplest series of allocation reaches from the king to village headmen, from village headmen to household heads, and from household heads to individual household members (ibid.: 85). A person's relation to property is a function of his or her socio-political status (ibid.: 93). For example, at the top of the hierarchy, the king is symbolically identified with land, receiving tributes and shares of wild animals (ibid.: 37-39, 150-57). Or common people's land claims derive from their membership in a village (ibid.: 80).
The rights enjoyed by various claimants are closely tied to obligations (ibid.: 79). In fact, rights derive from the fulfilment of obligations. They revert to the higher-level if obligations are not met or claims not enforced (ibid.: 91-92). Those holding estates of production owe harvest shares to the king and the village headman. As long as they meet the requirements, they continue to enjoy secure estates of production. Obligations and social debt thus are primary to rights. For the very same reason, the holders of estates of production cannot exercise rights of alienation. Just as rights are tied to obligations, rights are justified in reference to larger moral and social goals, not formal legal procedures. For example, the judge in a court case observed by Gluckman looks at the larger social issue and morality at hand when deciding a land dispute (Gluckman 1972: 14-18). As a consequence, the law is multi-referential in the sense that it combines reference to multiple, sometimes conflicting, principles. This allows judges an elastic conduct of adjudication, making the law apply to 'a variety of raw facts of very different kinds' and quite open to the influence of 'certain principles, presumptions, prejudices and postulates' (ibid.: 24).
Estates of production and estates of administration refer to specific resources, such as soil, wild animals and trees. Property claims, therefore, may be geographically overlapping, as one claimant may hold the right to cultivate a certain field while another one the right to plant trees on the same piece of land (ibid.: 90-91, 95). In addition, land claims may vary over time as claimants may abandon land or not meet their obligations (ibid.: 90-91). The boundaries between different claims can thus be very fluid on the ground. Property relations can also overlap in three ways. Property claims can relate to different resources on the same piece of land, they can vary over time, and they may be embedded in a series of allocations including multiple claims. Furthermore, the balance of powers between the different layers of social control may differ between places and plots. Gluckman speculates that land scarcity, as conditioned by biophysical and social factors, is the primary factor shaping the distribution of powers between the holders of estates of production and those holding estates of administration (ibid.: 97-98, 105-8). Where there is plenty of land, Gluckman observes, people are free to clear new land. In contrast, village headmen have repeatedly redistributed fields where land is scarce to safeguard that every villager works sufficient land to meet subsistence requirements and social obligations.
How does Gluckman's concept of property help us understand post-socialist land relations? Hann (1993a, 1993b, 1996), Humphrey (1983) and Verdery (this issue) suggest that socialist property relations can be usefully described in terms of Gluckman's hierarchy of estates of administration and production.  Their empirical and theoretical work makes a convincing argument for the utility of Gluckman's concept to understand land relations not only in a Soviet state farm (Humphrey) and an agricultural cooperative in Hungary (Hann), but also agrarian property relations across the socialist world. In addition, I follow Hann (1998) and Verdery (1998, 1999) when I surmise that Gluckman's concept helps us contrast socialist land relations to the conceptions underlying the notions of landed property promoted by privatisation programmes and make sense of the dynamics resulting from their encounter.
The literature notes several expressions of the conflicts between the notions underlying socialist and new land legislations. First, privatisation laws promote exclusiveness in the sense that they bundle multiple rights to land and associated resources in the hands of a single entity. In contrast, socialist land relations accommodated overlapping rights, giving different entities different rights to the same resource or to different resources associated with the same piece of land (Anderson 1998; Hann 1993a, 1996; Humphrey 1983: 6; Verdery 1998, 1999, this issue). Post-socialist land relations are thus characterised by conflicts about the balance of powers between different actors about control and use of resources. For example, local officials may grant use and management rights to individual landholders, but not the right of alienation.
Second, land reforms seek to detach land rights from their social context, separating rights and duties associated with land from the political and social status of landholders (Anderson 1998; Verdery 1996, 1998, this issue; see also Hann 1993a, 1996: 36, 43; Humphrey 1983: 4-5, 118-20). Socialist land relations, in contrast, tended to provide different legal positions for different social actors. For example, rights and obligations held by state units differed from those attributed to individual property holders. This tension between abstract and socially-embedded positions becomes evident, for example, in the frequent phenomenon that new land legislation allows outsiders (non-villagers, foreigners) to acquire land, but that people tie land claims to village membership or local state officials connect control over land with citizenship.
Third, post-socialist property relations are characterised by conflicts between references to formal legal procedures and claims derived from substantive law and notions of morality (Hann 1996: 38, 43; Verdery 1996: 163, 1998, 1999: 73- 74). Socialist states displayed a rather instrumental use of law, making it subject to the achievement of larger economic and social goals. Laws could be changed at the parliamentary level or bent at the implementation level with relative ease. This is different from the Western property law that stresses the consistent application of formal procedures. The conflict between formal and substantive laws finds expression in court rulings that ignore formal procedures in favour of decisions serving larger societal goals, such as notions of morally justified entitlements (Verdery 1996: 163). Notions of entitlements to a large part derive from labour, which was an important material and moral base for claims on land under socialism (Lampland 1995: 100-103; Verdery 1996: 163-64, 1998, 1999: 73-74).
Finally, the different notions of property turn into different geographies of physical boundaries (Hann 1993b, 1996: 37; Humphrey 1983: 132, 139, 152; Verdery 1996, this issue). The newly promoted land relations are territorial in the sense that they require clear demarcations of relatively permanent boundaries (see Sack 1983; Vandergeest and Peluso 1995). The new legislation demands the state to classify claims on resources by geographic area, demarcate impermeable territorial boundaries, designate restrictions on activities within the boundaries, and enforce those boundaries and restrictions. Boundaries become fixed in space and time, in stark contrast to the socially embedded socialist land relations, which allowed physical boundaries to shift over time and were often quite permeable and overlapping.
Thus, I suggest that Gluckman's concept of land relations provides a useful tool to describe socialist land relations, contrast them to the conceptions underlying new land laws and make sense of the dynamics resulting from their encounter.
The concept allows us to capture differences between the very concepts of landed property underlying socialist and new land laws. As Verdery (1999: 55) observes, these conceptual differences are primary sources of the fuzziness characterising post-socialist property relations. Fuzziness originates from the 'lack of routinized rules and crystallized practices around private property'. I now turn to my empirical case to examine these issues in more detail.
| The Historical Setting|| |
The Black Thai lived exclusively from subsistence agriculture in the first half of the twentieth century (Cam 1978; Dang 1972; Jacquet 1938). They derived a large part of their food needs from wet rice fields, which were the main source of their preferred food staple. Upland rice fields radiated into the surrounding slopes, as people abandoned old fields and opened up new fields every two or three years in response to declining soil fertility. The uplands also provided grass and leaves for water buffaloes and a variety of forest products for human diets and domestic needs. The sale of local products and purchase of outside goods were limited, as topographic conditions and lack of transportation infrastructure isolated the villages.
Pre-socialist Land Relations
Land relations in Black Thai villages resembled the hierarchy of overlapping powers described by Gluckman (1972).  The local lord, or phia, was considered the formal owner of the land in a muong, yet his actual powers did not go beyond symbolic ownership and associated claims for people's allegiance and shares of products and labour. At the next tier village communities exerted collective control rights over the land, but could not intervene in day-to-day land management. At the bottom of the hierarchy households held estates of production, that is, rights to cultivate wet rice fields and upland fields, graze water buffaloes and collect forest products.  In return, households owed shares of labour and products to the village collectives and phia. Since the late nineteenth century the French had added another layer at the top, when they included the Black Thai region into their overseas territory Tonkin and demanded taxes and corvee labour from the Thai leaders.
The balance of individual entitlements and collective control varied between different land uses.  Village collectives exercised strong control over the highly priced wet rice fields. They periodically adjusted individual wet rice fields or reallocated all wet rice fields in a village in response to changes in household and village demography. Collective control did not extend into the uplands, however. Households were free to clear suitable forest for new fields wherever no other household had worked the land previously. If another household had worked the land before, they negotiated directly between each other. Similarly, all households enjoyed the rights to graze their water buffaloes and collect forest products in the uplands. There were no boundaries between villages in the uplands.
Rights and obligations associated with landholdings depended on one's political status. Within villages all households had the right to work wet rice land and, in return, owed the village collective a share of output and labour duties. If households did not meet their duties, they lost the right to the land. The amount of land held by households depended on their capacity to meet associated product and labour duties. The village collectives ensured that every entitled household received a fair share of the scarce and desirable wet rice land, though the exact criterion used to determine fair shares differed from village to village.  In contrast, land in the uplands was abundant enough to enable every household to claim the desired area of land. As collective control was confined to wet rice fields, households did not incur any obligations for land use in the uplands.
Land relations were highly fluid in time and space. Periodic collective reallocations of wet rice land prevented households from claiming permanent wet rice fields. As household and village membership changed over time, so did the boundaries of household wet rice fields.  The situation was similar in the uplands, though for different reasons. The practice of shifting cultivation motivated households to rotate their fields on the abundant land. Water buffaloes roamed freely throughout the uplands, just as people collected forest products wherever they found them. What mattered to the people was that they received a share of the land but not specific plots. The major exceptions to this fluidity were the relatively rigid boundaries around village wet rice landholdings.
Land Relations under Collective Agriculture
My account of the villages in Chieng Dong begins in 1952, when Vietnamese liberation forces drove out the French administration from most of north-western Vietnam. Chieng Dong had been a small Black Thai principality located in the upper watershed of the Vat river, including a total of eight villages. The victory of the Vietnamese liberation forces caused the phia to flee to neighbouring Laos, abandoning his wet rice fields. In reaction, the village collectives of Chieng Dong redistributed the fields among households. Households were entitled to receive a share corresponding to their labour capacity. Vietnamese cadre generally encouraged the redistribution, but did not participate in it.
After independence in 1954, the Vietnamese state initiated efforts to integrate Black Thai villages into the state structure (Chu 1962; Tran 1968). The central government mandated the expansion of the state administrative structure into the mountains. Chieng Dong became a commune in Yen Chau district of Son La province. Most of the cadres recruited to staff the new authorities at the commune, district and provincial levels were Black Thai.
The Vietnamese state expanded the collectivisation drive to the mountains in 1960 (Chu 1962; Tran 1968). The people of Chieng Dong joined the newly established cooperatives without any significant resistance. They retained formal titles to village wet rice fields, as the Vietnamese state did not nationalise agricultural land. They worked most land collectively in brigades, just like their peers in other socialist economies. Yet people also resisted the complete collectivisation of agricultural production in Chieng Dong. Collective production remained limited to wet rice and some upland rice fields. Households continued to grow rice and other supplementary crops, and raise pigs and poultry on their own account. The villagers also opposed the formation of larger cooperatives including several villages for a long time. It was only in 1979 that the village cooperatives merged into only two large ones. As for relations with the state, villages had to pay a share of the output to the state as taxes, mandatory sales or in exchange for inputs. Villages also had to provide small amounts of labour to state projects, such as the construction of irrigation canals.
While the state promoted cooperatives for agricultural production, it reserved forestry for direct state management. The government nationalised forests in 1955 and put state units in charge of their management. 'Forestry land' was supposed to include all land with a slope above 20 degrees, including a large part of the uplands in Chieng Dong.  The land became subject to direct control by the authorities of Yen Chau district. Cooperatives had to seek their permission for the location and area of upland fields every year.
Though state integration and collectivisation modified land relations in Black Thai villages, they also reinforced key features. Socialist land relations preserved the hierarchy of overlapping rights. The district authorities and cooperative leadership took over the estates of administration previously held by the phia and village collectives. Brigades and households enjoyed estates of production in the valley and surrounding uplands. The state's attempt to exclude local uses of forestry land failed as people defended their overlapping rights of cultivation, grazing and forest product collection in the uplands.  Collectivisation, as much as it influenced the organisation of the labour process, had little effect on the overlapping powers characterising land relations in Chieng Dong. However, the balance of powers shifted towards the estates of administration held by the collectives and district authorities, that is, as evidenced by the brigades' new role in wet rice cultivation, the extension of collective control into the uplands and the district authorities' claims over the uplands.
Access to wet rice land remained contingent upon the fulfilment of one's obligations and membership in the village, now cooperative. The cooperative leadership allocated labour duties, hence shares in total product, to all adult villagers. Access became contingent upon the number of main labourers in a household, unifying the allocation formula across villages. Concerns for an equitable distribution of obligations and opportunities even motivated a reallocation of wet rice land between villages in the 1980s, erasing historical differences in village wet rice holdings. Tax duties and mandatory sales remained small, however, because of the low productivity of agriculture and limited use of external inputs.
Collectivisation increased the fluidity of land relations. Boundaries between wet rice fields lost significance because villagers worked the land collectively.
Upland boundaries remained fluid for the continuously changing location of upland fields. The areas negotiated with the district as well as villagers' individual fields moved from year to year in the search for fertile upland soils. Very few boundaries remained fixed in time and space, only the boundaries around the village wet rice fields continued to be impermeable for outsiders.
Villagers gradually reversed the shift in the balance of powers toward the collectives and district administration in the late 1970s and 1980s. They expanded individual upland fields up the slopes and increased their engagement in ownaccount activities. The village collectives did not stop the expansion of individual cultivation. Efforts by the local state authorities to limit people's use of the uplands failed. Control over production moved back towards households in Chieng Dong. This was a widespread phenomenon across northern Vietnam. National policy eventually followed suit, with a reform of agricultural collectives in 1981 and the call for decollectivisation in 1988 (Fforde and de Vylder 1996).
In the early 1990s land relations in Chieng Dong resembled those that had existed before collectivisation. The village collectives had allocated shares of wet rice land to households based on their labour capacity in 1988. Reallocations of wet rice holdings took place in the villages. Additionally, the collectives had begun to take away wet rice fields from those households that had not met their obligations to the collective. In the uplands access to land remained largely outside collective control. Access to new land increasingly depended on direct negotiation between households, as people had claimed most of the land suitable for cultivation. Yet collective intervention became more necessary for the solution of land disputes and the enforcement of crop protection against grazing animals.
| The New Land Legislation|| |
Vietnam's National Assembly passed a new land law in July 1993. Government officials and foreign observers heralded the new law as the cornerstone of a new strategy for rural areas and people. The law substantially widened the bundles of rights accorded to the holders of land in comparison with the previous legislation. The so-called 'use rights' not only included the right to use the land, but also to dispose of its produce, to exclude others from using the land, to use the land rights as bank collateral, to pass them on to one's heirs and to alienate 'use rights' to third parties. In the case of annual agricultural land, the rights extended over a period of twenty years, during which the state could withdraw them under narrowly circumscribed circumstances only.
Underlying the law was an exclusive concept of land rights (Carrier 1998; Verdery 1999). Land rights applied to all landholders in the same way, independent of political and social status, such as village membership. According to the law, individual persons, households and state units were all entitled to receive land rights, and their rights and duties would be the same. Additionally, for each plot of land, the law acknowledged claims by only two entities: the state and the landholder. All rights pertaining to a concrete plot of land were to be given to a single landholder, excluding the possibility that several entities may hold rights to different aspects of the land.
The law mandated that the state should classify all land by its function for housing, cultivation, animal husbandry or forestry. Once the state had classified the land, landholders were required to utilise it accordingly. Failure to use the land properly could result in the withdrawal of rights. The new legislation also entitled the state to collect taxes on allocated land. Tax loads depended on the area of allocated land and its legal classification, independent of actual usage. This was a radical break from the past practice of calculating taxes as a portion of expected or actual output. The exercise of state power over rural people, therefore, was intended to become territorial (Sack 1983; Vandergeest and Peluso 1995).
Land allocation was the vehicle to put the new legislation into practice. It included the demarcation of plots, their registration and the certification of land rights by the state in 'land use right certificates'. The land certificates were considered the basis of land administration, for which a specialised government agency was founded. The agency had the task to administer cadastral records, assist land classification, register and approve changes in landholdings, and arbitrate in land disputes. It had authority over specialised units at national, provincial and district levels, and a cadastral cadre at the commune level. The expectation was that within a few years the agency would oversee the classification of Vietnam's territory, communicate the classification to the people and connect individual plots unambiguously to the holders of land rights.
The Vietnamese government had become convinced that exclusive and territorial land rights were a vital institutional precondition for rural development. According to the view of the government, land registration and the new land administration agency would create secure tenure for rural people. Tenure security would, in turn, motivate landholders to invest labour and capital in the land. The transferable land titles would also serve as a basis for efficient allocation of land to those with better skills and more capital. In addition, the use of land titles as collateral would increase the supply of credit to landholders. Finally, land registration would also improve tax collection by creating a comprehensive database and making revenues more predictable. The passing of the land law in 1993, therefore, coincided with government decrees on agricultural credit, land taxation and agricultural extension. The far-reaching reforms met the recommendations of international advisers and were widely applauded (UNDP/FAO 1989; World Bank 1993).
The fundamental gaps between the new legislation and land relations in Chieng Dong are obvious, though. The concept of exclusive and territorial land rights informing the new legislation differed from the notions underlying people's lived land relations in Chieng Dong (see [Table 1]). At a more concrete level, land allocation had the potential to do away with three sets of village practices. Long-term allocation of exclusive rights undermined the periodic reallocation of wet rice fields, threatening the equitable distribution of rights and obligations. Field demarcation, land registration and land administration implied significant obstacles to the flexible negotiation of household upland holdings. Functional land classification also contradicted the overlapping estates of production in the uplands. In the following section, I examine how people reacted to land allocation, taking up the three discrepancies between village practice and national policy in turn.
| Resistance to Land Allocation|| |
Land allocation gained rapid momentum in the Son La province in 1994. When the programme reached Chieng Dong, local cadres asked villagers to report their landholdings and demarcated a large part of the uplands surrounding the villages as forestry land. They inspected and documented the reported fields with village leaders and affected villagers. Within a few weeks the cadres reported the successful completion of their task in Chieng Dong, as evidenced by land applications, allocation documents and maps. This claim of success is surprising considering the radical differences between the conceptions underlying the new land legislation and local land relations. Had land allocation abolished local land relations, replacing them with the exclusive and territorial conceptions contained in the new legislation?
Wet Rice Fields
Land allocation was not as successful as suggested by the district authorities' report. Most apparently, the land use right certificates did not include wet rice fields. Nor did the land allocation maps show individual wet rice fields. The district authorities had decided to exclude the fields from formal allocation. They left out the fields in order to avoid open confrontation with villagers. Local cadres had initially requested villagers to declare wet rice fields. Their request had met with protest from village leaders because long-term allocation conflicted with the practice of periodic reallocation. In response, the district leadership instructed the cadastral office to leave the fields out of the land certificates.
In the years following land allocation, all villages in Chieng Dong continued to reallocate wet rice fields. When I asked village leaders and common villagers about their motivations to continue reallocations, they unanimously referred to a concern for 'equity'.  They wanted to distribute the use of the precious wet rice fields in an equitable fashion. At the same time, the power of a household to use wet rice fields continued to depend on the fulfilment of obligations toward the collective. Village collectives took away fields from those households that had not paid the fees or contributed labour to the collective for several years. The right to work fields was connected with the obligation to contribute products and labour. 'Equity' referred to an equitable distribution of opportunities and obligations, each tied to the other. This 'equity' did not extend to outsiders, however, as they were not allowed to work village fields.
Most people did not see any trade-off between their concern for equity and tenure security, the stated objective of the new legislation. According to the rationale of the new legislation, land rights were only as secure as they were exclusive. The powers of the village collectives reduced the exclusiveness and, therefore, security of tenure. Yet in the villagers' view secure tenure on wet rice fields derived from the fulfilment of obligations in combination with village collectives' estates of administration. The collective ensured all households in a village rights to land as long as they met their obligations. Thus, the village collectives' powers were not a hindrance but the primary source of secure tenure.
Similarly, long-term allocation was not a necessary condition for investment in the land, as presumed by the new law. The rationale underlying the new land assumed that the incentives motivating investment would have to derive from the assurance of individual benefit. Only if people felt sufficiently assured of reaping the benefit of their own investment would they be ready to maintain and improve the land. In practice, villagers maintained field bunds and irrigation canals even in the absence of land certificates. They maintained the fields because those activities were obligations connected with the right to work the land. A neglect of those investments could have led to exhortations by the collective or, in the worst case, a loss of one's estate of production. Direct linkage between individual investment and individual benefit, presumably the precondition for investment, was not necessary.
Village leaders and most villagers vigorously opposed the inclusion of wet rice fields in land certificates because that would have posed a potential challenge to the estate of administration held by village collectives. Village leaders voiced their concern that households might utilise their formally accorded rights to fend off collective claims on their land. They clearly recognised a threat to the collective's estate of administration if they accepted land certificates issued by the district authorities. The formal procedures instituted by the land law were in direct conflict with village practices that had the goal to distribute opportunities in an equitable manner. Therefore, in contrast to upland fields, village leaders had openly opposed the inclusion of wet rice fields in the land certificates.
To my surprise, local state cadres justified the continuing reallocation of wet rice fields with the rationale given by villagers. They supported periodic reallocation for its contribution to 'equity'. It provided a means to create equitable opportunities for all households to meet their subsistence requirements, particularly considering the rapidly growing population. This support for reallocation was not confined to cadres in the commune and district administration. It reached up to provincial cadres, who freely admitted that periodic reallocations of wet rice fields remained a common phenomenon in Black Thai villages of Son La province.
While expressing moral support for periodic reallocation, each cadre was careful to point out that he or she did not carry the responsibility for the exclusion from formal allocation. Cadres in the provincial and district administration emphasised that it was the decision of each village to continue reallocation. The commune's cadastral officer argued that the district had left the decision about whether to allocate wet rice fields open to villages. The distribution of responsibility and blame went beyond the usual denial of responsibility and finger pointing. It mirrored the hierarchy of estates of administration held by the state and village collectives. Local state authorities pointed out that their powers did not include the right to intervene in the allocation of wet rice land within villages. That fell under the estate of administration held by the village collectives. 
Upland holdings remained flexible, despite allocation and documentation of concrete fields in the certificates. My survey of sixty-five households conducted in 1997 shows that fifty households worked larger areas than those allocated. In particular, the households that had received relatively small landholdings in 1994 had more than doubled their upland holdings. Households had expanded their landholdings in several ways. All households had expanded existing fields into forestry land or a neighbour's field. Several borrowed fields temporarily from other households, typically a close relative, or had exchanged fields with them. A few had taken up fields from the agricultural reserve land originally allocated to the village collective. One household had inherited fields from the husband's parents. None of the land transactions had been registered in the land certificates.
The presumed cadastral grid remained a fiction. Household upland holdings were much more flexible than assumed by the new land policy. Villagers gave up, acquired and exchanged fields without having the changes recorded in their land certificates. More importantly, fields expanded and contracted from year to year. Boundaries between upland holdings reflected labour capacities and negotiations between villagers. Changes in labour capacities and bargaining powers made boundaries move over time. The image of a cadastral grid, depicted in land certificates and maps, did not exist in practice. People's upland holdings and their supposed documentation in land certificates and maps increasingly differed from each other.
At the same time, households wanted to receive land certificates and carefully kept them in their wooden boxes. The certificates served as evidence of their claims on a share of village land. They formalised the estate of production in the language of the state. Yet villagers readily admitted that the individual field sizes listed in the land certificates had little meaning. They did not care about a precise geographical identification of their fields. What mattered to them was an acknowledgement of their right to a share of the land.
One may argue that the land certificates fulfilled a function that went beyond the estates of production held by villagers before land allocation. Villagers could take up small loans if they deposited their land certificates with the bank, an opportunity seized by many households. In the view of the Vietnamese government, the land certificates were an important collateral for bank loans. Land was the most significant productive asset held by most farmers. Land certificates, therefore, would drastically expand the number of rural people eligible for bank loans and boost the supply of credit to agriculture. Yet, in practice, the use of land certificates as collateral was largely symbolic, necessitated by the bank's standard procedures. Bank staff resorted to village leaders rather than land certificates in the few cases when households could not pay outstanding debts. Village leaders in turn asked the debtor's relatives and fellow villagers for help, and requested debtors to sell durable consumer goods and even their houses to raise cash. No household ever lost a land certificate. The state's estate of administration did not include the power to remove a villager's estate of production. After all, what could the bank have done with dispossessed land? Village rules allowed neither the sale of land rights nor the cultivation of village land by outsiders.
I would even suggest that local cadres viewed the significance of the land certificates in the same terms as the villagers. Just as bank staff did not attempt to foreclose on land rights, local cadres had no illusion that land certificates related to concrete fields fixed in space and time. What mattered to them was that all households had received land certificates, as evidence for their legal claim on a share of the uplands. They did not undertake serious efforts to establish verifiable links between land certificates and actual fields on the ground. Land certificates only reported rough estimates of allocated areas, not the location of allocated fields. The sketch maps of fields drawn in the field showed landholdings in relation to neighbouring ones, identified by the names of the landholders, but did not locate the fields on a grid fixed in geographical space. The formalisation of people's claims to upland fields was not necessary because those rested on social norms in the villages. 
Village boundaries became rigid in the uplands, in contrast to the boundaries between household landholdings. Local cadres had demanded the clarification and enforcement of village boundaries during land allocation. Village officials had followed suit and negotiated village boundaries. They required the district authorities to adjust some of the proposed village boundaries in order to accommodate actual landholdings. In a few cases land conflicts between neighbouring villages lingered on as significant numbers of households in both villages refused to give up fields. Yet, by 1997, a large majority of households worked fields only on the land of their own village. Several households had lost fields now located on the land of another village in the process, as rights to upland fields now depended on village membership.
Why were villagers ready to recognise village boundaries? I believe because land had become scarce in the previously abundant uplands. The demarcation of village boundaries helped village collectives exclude outsiders. The boundaries also provided the basis for the extension of the collectives' estates of administration into the uplands. The collectives' role in the allocation of wet rice fields in the valley provided a proven solution to the collective action problem in the uplands.
My explanation finds support in the collectively organised reallocation of upland fields in one village. The demarcation of village boundaries had created an inequitable distribution of upland fields in the village. A portion of households worked large fields in the more attractive area given to the village, while the other households did not have any land there and had to abandon good fields on the land given to another village. It quickly became obvious to villagers that the extent of the problem exceeded their capacity to correct it through individual negotiations. The villagers instead agreed on a process in which village leaders organised the redistribution of attractive upland fields. For the sake of 'equity', village leaders required households with large fields to pass a share on to households without any or only small fields. In the process, the village collective extended its estate of administration into the uplands, obliterating the previous allocation of upland fields done by the district cadre and documented in the land certificates.
Land allocation in 1994 included efforts to demarcate a large part of the uplands and put it under protection as forestry land. The district authorities renewed their efforts in 1995 supported by a foreign-funded watershed protection project. The project promoted the assignment of specific parcels of forestry land to households as a strategy to clarify land rights and encourage sustainable resource use. The district cadres returned to Chieng Dong to sign written protection contracts with households for specific parcels of forestry land. They had households commit to protect the parcels against use for cultivation, grazing and timber extraction, promising small cash payments in return. Livestock grazing was confined to a bare hill set aside as grazing land.
The cadres' attempt to carve up the uplands into exclusive estates of production for cultivation, grazing and forestry did not work. The expansion of agriculture into the uplands continued unabated. The land under cultivation increased by one-quarter from 1993 to 1997.  Within three years cultivation exceeded the area allocated in 1994 by more than 70 per cent. Villagers continued to use forestry land for cultivation, animal grazing and the collection of forest products. Land allocation and forest protection contracts were unable to arrest agricultural expansion. The state's claim of territorial control over the uplands, which was documented in various maps, protection contracts and a land register, could not be enforced.
Land relations in the uplands remained multi-tiered and fluid in time and space. Villagers' estates of production continued to overlap and moved back and forth across the uplands. Cultivation yielded the strongest claims to land, especially if people asserted food security problems. People with upland fields had the power to exclude other uses of the land during the cropping season. Grazing rights hills and remote areas during the season. Forestry gave claims on land little weight. Households with forest protection contracts could not stop other villagers from opening up new fields on the protected land, nor could they challenge people's rights to graze animals and collect forest products on the land. Claims to trees and forest products were strong only in the case of the individually planted bamboo groves and collectively protected spirit and bamboo forests.
The district authorities came to accept the agricultural uses of forestry land. They virtually stopped enforcing forest protection regulations after land allocation and the initiative supported by the watershed protection project had terminated. They only fined two households in 1996 and 1997, in contrast to the dozens fined in 1995.  They paid out the full amount of protection payments, though many parcels showed clearly visible evidence of their illegal use for cultivation or grazing. In addition, the local state authorities changed some of their regulations to accommodate villagers' estates of production. The district authorities expanded the permissible amount of agricultural land that could be allocated to households beyond the area specified by provincial guidelines. The provincial authorities followed suit by raising the land ceiling for the whole province. District and provincial authorities eventually gave up enforcing land ceilings completely. Their powers in the uplands after land allocation did not go beyond those before land allocation.
A brief look at the obligations associated with upland cultivation yields a similar picture. National policy now required villagers to pay taxes on all allocated land, not just land under cultivation in a given year as in the past. The change had the potential to increase overall tax loads, as villagers had claimed more land during allocation than they actually worked in any given year. Yet land allocation left overall tax duties unchanged. Villagers responded by resisting increases in their tax obligations and bargaining down the reported area of upland fields. The overall tax load of all villages in Chieng Dong after land allocation remained the same as before land allocation. Within only a year of allocation villagers bargained down the reported areas by up to one-quarter, though they continued to expand upland fields at the same time. Villagers readily admitted to me that they failed to report an average of two fields per households at a land re-survey in 1995. They explained that the state had no right to claim taxes on land that they did not actually work in the given year. In their conceptions of land relations, obligations associated with land remained tightly connected with the actual use of the land.
| Conclusion|| |
In this article I have used land allocation in Chieng Dong to look at post-socialist property relations. The Black Thai of Chieng Dong resisted the implementation of the new legislation promoting exclusive and territorial land rights. They openly protested long-term allocation of the most valuable fields because it would have undermined their 'equitable' distribution. They ignored the fixation of plot boundaries as depicted in land certificates and maps, because it would have prevented the flexible negotiation of landholdings. People also disregarded the functional classification and carving up of the uplands, and retained them under a continuously moving patchwork of overlapping rights.
The district authorities failed to put the new land legislation into practice. Yet perhaps the very failure facilitated important outcomes desired by the new legislation. Villagers maintained wet rice fields even though they did not hold land certificates. They were able to take out loans from the local bank, although the bank would not attempt to foreclose on land certificates. Villagers also continued to contribute taxes to the state coffers. Overall, agricultural products and living standards rose rapidly throughout the 1990s. Even the forest grew back in parts of the uplands as a consequence of the overall trend of agricultural intensification (Sikor and Truong 2002).
People resisted the very concept of landed property promoted by the Vietnamese government. Socialist land legislation had accommodated and even reinforced core elements of historical land relations. In contrast, the conceptions underlying the new legislation fundamentally differed from the people's lived land relations, as shaped by pre-socialist land relations and socialist land policy. People's reactions, therefore, highlight how post-socialist land reforms provoke their own forms of resistance (Burawoy and Verdery 1999: 12). Post-socialist changes in land relations are highly political as they unfold upon socialist land relations (Hann 1998; Verdery 1998).
My findings support the conflict in concepts underlying socialist land relations and post-socialist reforms identified in the literature. First, post-socialist land reforms typically promote a shift towards more exclusive land rights. Socialist land relations could better accommodate overlapping rights, such as use and control rights relating to the same piece of land held by different actors (Anderson 1998; Hann 1993a; Humphrey 1983; Verdery 1998, 1999, this issue). Second, post-socialist land reforms seek to detach land rights from their social context, separating rights and duties associated with land from the political and social status of landholders. Under socialist agriculture, land rights depended on economic and political factors, such as cooperative membership (Anderson 1998; Hann 1993a, 1996; Humphrey 1983; Verdery 1996, 1998, this issue). Third, the new legislation emphasises formal procedures as bases of claims on land. Socialist land relations stressed moral and social goals as justifications for land claims (Hann 1996; Verdery 1996, 1998, 1999). Fourth, the new conceptions presume landholdings to be relatively fixed in space and time. In contrast, socialist land relations put more emphasis on shares in land than actual fields, making boundaries much more fluid and elastic (Hann 1993b, 1996; Humphrey 1983; Verdery 1996, this issue).
I would perhaps suggest two additional loci of contestations over post-socialist land relations that originate from conflicting concepts. Land reforms shake up the previous combination of rights with obligations. People react through strategies that seek to separate rights from obligations in order to increase rights and reduce obligations. They readily seize on opportunities afforded by claims derived from new land titles, such as access to credit, and resist new liabilities associated with the very same land titles, for example, new forms of taxation. These reactions are similar to those observed in the reorganisation of state-owned industrial enterprises (Stark 1996). Enterprise managers seek to separate assets from liabilities in order to privatise assets and transform enterprise debts into public liabilities.
A final locus of contestations is the issue of how flexible the balance of powers between various layers of social control is. Land reforms challenge the variation in the distribution of powers. The new land legislation tends to unify the hierarchy of controls over different types of land and resources. Landholders are typically entitled to the same bundle of rights, while states retain the same kinds of powers in land administration. In contrast, socialist land legislation provided flexibility for different distributions of controls over land. Collective estates of administration included strong powers for some resources and types of land, yet infringed little on individual estates of production for others.
These contestations demonstrate how people actively respond to the conceptions underlying land reforms, both their concrete effects on local land relations and their more abstract justifications (Burawoy and Verdery 1999: 12). Their reactions are not merely a defence of the status quo. Instead, they incorporate elements of the new legislation, while preserving certain features of local land relations. Nor do their responses passively replicate historical land relations. People actively employ selected elements from the collective memory of pre-socialist and socialist land relations in contemporary contexts (see Cellarius, this isue). In particular, people may rely on moral and practical principles that have proven useful in the past, but modify the practices in support of the principles (Burawoy and Verdery 1999: 8; Verdery 1999: 66). People may also apply those principles to new situations as economic and political conditions change.
Differences in the very conceptions underlying people's lived land relations and new land legislation produce fuzzy land relations (Verdery 1999). Though local rights and claims follow clear practices and routines, they appear ambiguous from the point of exclusive and territorial property notions. Yet the conflicts are not just about concepts, they are also about the state of the state. My findings match those by Sturgeon (this issue) and Yeh (this issue) that divisions within the state may contribute to the persistence of conflicting concepts. The actions of local state agents may contribute to the fuzziness of post-socialist land relations. Understanding post-socialist land relations, therefore, requires opening up the post-socialist state. But this goes beyond the scope of this article and is the subject of another one (Sikor 2004).
Acknowledgements: I want to warmly thank Janet Sturgeon for her comments and encouragement. The article has also benefited from constructive comments by two anonymous readers. I am very grateful to the Ford Foundation for financial support for the fieldwork on which this article is based.
| References|| |
|1.||Anderson, David G. (1998), 'Property as a Way of Knowing on Evenki Lands in Arctic Siberia', in Christopher Hann (ed.), Property Relations: Renewing the Anthropological Tradition, pp. 64- 84. Cambridge: Cambridge University Press. |
|2.||Burawoy, Michael and Katherine Verdery (1999), 'Introduction', in Michael Burawoy and Katherine Verdery (eds), Uncertain Transition: Ethnographies of Change in the Postsocialist World, pp. 1-17. Lanham: Rowman and Littlefield Publishers. |
|3.||Cam Trong (1978), Nguoi Thai o Tay Bac Viet Nam (The Thai in North-Western Vietnam). Hanoi: Social Science Publishing House. |
|4.||Carrier, James G. (1998), 'Property and Social Relations in Melanesian Anthropology', in Christopher Hann (ed.), Property Relations: Renewing the Anthropological Tradition, pp. 85- 103. Cambridge: Cambridge University Press. |
|5.||Chu Van Tan (1962), Mot Bien Doi Cach Mang To Lon o Mien Nui (A Big Revolutionary Change in the Mountains). Hanoi: Truth Publishing House. |
|6.||Dang Nghiem Van (1972), 'An Outline of the Thai in Viet Nam', Vietnamese Studies, 32: 143-99. |
|7.||Fforde, Adam and Stefan de Vylder (1996), From Plan to Market: Economic Transition in Vietnam 1979-1994. Boulder: Westview Press. |
|8.||Gluckman, Max (1972), The Ideas in Barotse Jurisprudence. Manchester: Manchester University Press. |
|9.||Hann, Christopher (1993a), 'Property Relations in the New Eastern Europe: The Case of Specialist Co-operatives in Hungary', in Hermine de Soto and David G. Anderson (eds), The Curtain Rises: Rethinking Culture, Ideology, and the State in Eastern Europe, pp. 99-119. New Jersey: Humanities Press. |
|10.||(1993b), 'From Production to Property: Decollectivization and the Family-Land Relationship in Contemporary Hungary', Man 28: 299-320. |
|11.||(1996), 'Land Tenure and Citizenship in Tazlar', in Ray Abrahams (ed.), After Socialism: Land Reform and Social Change in Eastern Europe, pp. 85-114. Oxford: Berghan Books. |
|12.||(1998), 'Introduction: The Embeddedness of Property', in Christopher Hann (ed.), Property Relations: Renewing the Anthropological Tradition, pp. 1-47. Cambridge: Cambridge University Press. |
|13.||Humphrey, Caroline (1983), Karl Marx Collective: Economy, Society and Religion in a Siberian Collective Farm. Cambridge: Cambridge University Press. |
|14.||Jacquet, A.M. (1938), 'Enquete No. 1B Sur L'alimentation des Indigenes, Province de Sonla', unpublished report. Aix-en-Provence: Archives Ministerielles Modernes, Commission Guernot, Centre de Archives de Outre-Mer. |
|15.||Lampland, Martha (1995), The Object of Labor. Chicago: University of Chicago Press. |
|16.||Lebar, Frank M., Gerald C. Hickey and John K. Musgrave (1964), Ethnic Groups of Mainland Southeast Asia. New Haven: Human Relations Area Files Press. |
|17.||Malinowski, Bronislaw (1935), Coral Gardens and Their Magic. London: Allen and Unwin. |
|18.||Neale, Walter C. (1962), Economic Change in Rural India: Land Tenure and Reforms in Uttar Pradesh, 1800-1955. New Haven: Yale University Press. |
|19.||(1969), 'Land is to Rule', in Robert E. Frykenberg (ed.), Land Control and Social Structure in Indian History, pp. 3-15. Madison, WI: University of Wisconsin Press. |
|20.||Sack, Robert D. (1983), Human Territoriality: Its Theory and History. Cambridge: Cambridge University Press. |
|21.||Scott, James C. (1985), Weapons of the Weak: Everyday Forms of Resistance. New Haven: Yale University Press. |
|22.||Sevenier (1905), 'Commissariat du Gouvernement de Van-Bu', Revue Indo-Chinoise, 3: 149-55. |
|23.||Sikor, Thomas (2001), 'Agrarian Differentiation in (Post-)Socialist Vietnam', Development and Change, 32: 923-49. |
|24.||(2004), 'Local Government in the Exercise of State Power: The Politics of Land Allocation in Black Thai Villages', in Benedikt T. Kerkvliet and David Marr (eds), Local Government and Authority in Vietnam. Singapore: Institute of Southeast Asian Studies Press (forthcoming). |
|25.||Sikor, Thomas and Dao Minh Truong (2002), 'Agricultural Policy and Land Use Changes in Black Thai Villages of Northern Vietnam, 1952-1997', Mountain Research and Development, 22: 248-55. |
|26.||Silvestre (1918), 'Les Thai Blancs De Phong-Tho', Bulletin De L'Ecole Franfaise D'ExtremeOrient, 18: 1-56. |
|27.||Stark, David (1996), 'Recombinant Property in East European Capitalism', American Journal of Sociology, 101: 993-1027. |
|28.||Steinberg, David J. (1987), In Search of Southeast Asia. Honolulu: University of Hawaii Press. |
|29.||Tran Phuong (1968), Cach Mang Ruong Dat o Viet-Nam (Land Revolution in Vietnam). Hanoi: National Political Publishing House. |
|30.||United Nations Development Program (UNDP) / Food and Agriculture Organisation (FAO) (1989), Viet Nam: Agricultural and Food Sector Review. Hanoi: UNDP. |
|31.||Vandergeest, Peter and Nancy L. Peluso (1995), 'Territorialization and State Power in Thailand', Theory and Society, 24: 385-426. |
|32.||Verdery, Katherine (1996), What was Socialism? And What Comes Next? Princeton: Princeton University Press. |
|33.||(1998), 'Property and Power in Transylvania's Decollectivization', in Christopher Hann (ed.), Property Relations: Renewing the Anthropological Tradition, pp. 160-80. Cambridge: Cambridge University Press. |
|34.||Verdery, K. (1999), 'Fuzzy Property: Rights, Power, and Identity in Transylvania's Decollectivization', in Michael Burawoy and Katherine Verdery (eds), Uncertain Transition: Ethnographies of Change in the Postsocialist World, pp. 53-82. Lanham: Rowman and Littlefield Publishers. |
|35.||World Bank (1993), Viet Nam: Transition to the Market. Washington, DC: World Bank. |
|36.||Wyatt, David K. (1982), Thailand: A Short History. New Haven: Yale University Press. |