(Un)making the upland: resettlement, rubber and land use planning in Namai village, Laos

ABSTRACT This paper highlights how farmers in a northern Lao village transformed their customary land rights – in the face of incoherent overlapping state territorialization attempts – into a territorial strategy to secure their land tenure. By planting rubber, some villagers have engaged in a crop boom to lay claim to land which has recently been zoned for upland rice cultivation (and conservation) as part of a state-led land use planning initiative. We show how internal resettlement, ethnic division and the influx of commercial agriculture in the Lao uplands intersect in a novel land use planning process and predetermine the plan's actual significance.


Introduction
State territorialization, or the ways in which the state seeks to exert control over people and resources through 'the creation of systems of resource control, rights, authorities, jurisdictions, and their spatial representations' (Rasmussen and Lund 2018, 388) is not a new phenomenon in many developing countries in the Global South. Vandergeest and Peluso (1995) show how processes of state territorialization translate into state defined policies and institutions, to justify, legitimize and enforce state's control, while excluding or including people within particular geographic boundaries. While some have emphasized external territorialization and the spatial development of state's sovereign territory (Elden 2013;Sassen 2008), others have focused on the internal territorial strategies of states to establish control inside their boundaries. In Laos, internal territorialization has been pursued since pre-colonial time (Rigg 2005). Only recently growing attention has been given to non-state actors and their role in making territory (Baird 2009;Corson 2011;Lukas and Peluso 2019;Peluso 2009), in some cases, when working against state territorialization, also referred to as counter-territorialization (Bryant 2002;Isager and Ivarsson 2002;Lestrelin 2011).
This paper explores a case in which non-state actors 'make territory'. We use the term territorialization from the ground up to show how different groups of farm households responded to a recent land use planning initiative which aims to demarcate farmers' farmlands and forest areas. We look at how these responses are derived from their respective status as original settlers, recent and late comers as shaped by the government's earlier resettlement policies, and how it predetermines the actual significance of the defined land use plan. We argue that even though villagers' actions may appear as to be directly opposed to the government-led land use planning initiative they feed into an overarching project of state control. Villagers' actions likewise aim to demarcate a new territorial configuration, albeit from the ground up. In particular, the paper looks at: (1) how farm households perceive state's territorial projects centered on internal resettlement and in relation to the latter land use planning processes; (2) how these views are closely interlinked with their access to land and how such access is (re)negotiated at the village level through various local institutional arrangements; and (3) how it reflects back on and influences farmers' land tenure security and their strategy to engage in a crop boom (by planting rubber), while relying on their customary land rights.
Drawing on literature that views political authority as contested and in the making (Lund and Rachman 2018;Rasmussen and Lund 2018), the paper shows how the process of state formation is contested and remade through a power interplay shaping and reshaping local institutional arrangements at the village level. We examine this interplay in our case study of Namai village, hereafter Ban Namai, as 'it allows us to see space's enduring state of being under construction' (Lund and Rachman 2018, 419). Our case illustrates how internal resettlement, ethnic divisions and the influx of commercial agriculture manifested in a booming rubber plantationintersect in a state land use planning process and predetermine the plan's actual significance in the Namai uplands. Resettlement of three upland communities into the low-lying Namai valley has resulted in a rupture of people's social and spatial organization that has affected villagers in the resettlement village differently based on their resources, social network and ethnic identity. Even though the Government of Laos (GoL) aimed to install institutional control over people and resources in newly resettled communities it did not manage to fully do so, leaving room for local actors (e.g. original settlers, recent and late comers) to exercise their power in (re)shaping their access to land. The presented case of Ban Namai shows how (partly successful) resettlement and resulting power asymmetries between original settlers, recent and late comers have predetermined further government interventions that seek to establish spatial control through land use planning and allocation.
We unpack existing customary tenure arrangements of the two ethnic groups in the village by looking at interactions through a web of power relations that connects the different, partly overlapping, property rights regimes. Asymmetrical power relations in Ban Namai have manifested in land rental arrangements and in conflicts over boundaries. We illuminate villagers' strategies to secure land tenure and access to land, while unpacking the role of less and more powerful actors within the local community pertaining to land use planning and other state territorialization attempts as well as frontier dynamics.
The paper contributes to the current discourse on state territorialization and frontier dynamics in two ways. First, it brings to light the power dynamics at work (re)shaping farm households' access to land, and how the latter contributes to both the shaping of state territorialization and frontier dynamics. It argues that while common assumptions tend to suggest that customary land rights would get swept away or disregarded in the face of crop booms, the first could also serve as a tool for farmers to strengthen their land tenure (Baird 2008) while mimicking private sector actors' strategy to gain access to land. When farmers changed their upland rice fields into a rubber plantation area to defend their access to and control over land from the government's policy interventions and private sector actors' land concessionwhile at the same time benefitting from the rubber boomthey transformed their customary land rights from merely a local institutional arrangement into a territorial strategy.
Second, it presents the concept of territorialization from the ground up as a theoretical underpinning to unpack how farmers harness frontier dynamics while exploiting the incoherence of overlapping state territorialization attempts. It also brings to light the prevalence of customary land rights and tenure systems, as the latter forms the center piece of farmers' strategy to defend their access to and control over land. As stated by Fitzpatrick (1983, 50): 'Power is not simply based on prohibition but also on the positive formation of norms and shaping of individuals to fit these norms'. Our hypothesis is that as farm households view rubber planting both as a market but more importantly as a territorial strategy to secure their land tenure, it becomes a tool for farmers to apply territorialization from the ground up, while contesting the state's territorialization strategy to (re)form state space (Brenner et al. 2003) through land use planning processes. Similarly, as farm households mimic private sector actors' strategy to gain access to land through rubber planting, such strategy allows them to oppose the state's territorialization project, albeit informally and indirectly, and within state spaces (Kenney-Lazar, Suhardiman, and Dwyer 2018b).
The line of analysis and arguments presented in this paper are derived from in-depth case study research (Burawoy et al. 1991;Yin 2009) conducted by the first and third authors over the period of one year beginning in October 2017. The latest field visit in October 2018 looked at customary land rights of Khmu villagers and Tai-Lue original settlers previous to and after resettlement as well as at the underpinning power relations between the two groups that structure access to land in Ban Namai. We focus on three elements: (1) how customary land rights have evolved over time (2) the role of customary land rights in shaping farmers' strategies to secure their present land holdings; and (3) the implications of such strategies for land use planning and other state territorialization efforts.
To unpack present customary land rights in Ban Namai and how they have evolved over time we conducted a series of focus group discussions with representatives of the elderly council of both ethnic groups and the Village Development Committee (VDC) and facilitated smaller discussions among villagers. This was complemented by in-depth semistructured interviews with 11 households of each ethnic group to understand how the local community perceives their capabilities to gain and secure access to land as well as their perceptions of existing local institutional arrangements regarding access to land. We moreover collected perceptions of Khmu and Tai-Lue on how villagers' customary land use rights have been influenced by (and have influenced) state-led initiatives such as resettlement, a rubber concession and land use planning. As part of this field research the first and third authors conducted key informant interviews with representatives of the Provincial Agriculture and Forestry Office (PAFO), the District Agriculture and Forestry Office (DAFO), as well as with staff from The Agro Biodiversity Initiative (TABI) in Vientiane. TABI is a project funded by Swiss Development Cooperation (SDC) and led by the Ministry of Agriculture and Forestry (MAF). Putting a greater emphasis on agrobiodiversity in land use planning processes, TABI has formulated and implemented the participatory Forest and Agricultural Land Use Planning, Allocation and Management (pFALUPAM) in 315 villages, in 13 provinces, throughout the country since 2011. An ethnographic approach was additionally employed in regard to information collected through observations, informal discussions and a visit to a former Khmu settlement.
2. Theorizing state-society relations: from territorialization to frontier dynamics and public authority from below Rasmussen and Lund (2018) have brought to light the importance of applying a bifocal perspective of territorialization and frontier dynamics, making visible their interplay in (un)making space and its social orders of property rights, political subjectivities and institutional arrangements (Rasmussen and Lund 2018). The frontier has been invoked by many scholars to illuminate the transformations of upland areas across the region (Barney 2009;Diepart and Sem 2018;Dressler, Fletcher, and Fabinyi 2018;Hall, Hirsch, and Li 2011;Tsing 2005). We underline this argument, as in Laos too rural society and especially upland communities are increasingly transformed not only due to territorial strategies but also through market integration and a confrontation of the mainly subsistence-based peasant economy with a growing internationalized market economy (Tappe 2010). Frontier spaces have mushroomed across the country and rather than delineating a geo-political separation of physical space, they construct a contact zone or epistemological and political distinction between civilization and the wild (Rasmussen and Lund 2018). Anna Tsing has famously described this phenomenon in the Indonesian context of Kalimantan: A frontier is an edge of space and time: a zone of not yetnot yet mapped, not yet regulated. It is a zone of unmapping: even in its planning, a frontier is imagined as unplanned. Frontiers aren't just discovered at the edge; they are projects in making geographical and temporal experience. (Tsing 2005, 32) Frontier spaces indicate where new resources and commodities come into being and may thus be described in their basic form as: 'epistemological, discursive and political operations [that enable] powerful actors to turn nature into economic commodities' (Rasmussen and Lund 2018, 391). In the Lao upland, the government's 1 reforestation policy, for example, aiming to increase forest cover by 2020-70% (MRLG 2019), paired with framing swidden agriculture as destructive, have worked among other factors as frontier dynamics conjuring an image of a desired future situation in which 'backward' swidden agriculture is stabilized and production is fixed. This legitimizes the demarcation of swidden fallows as state-owned forest land ('degraded forest') rendering them unfit for production under shifting cultivation, i.e. appropriating land and resources for environmental ends (Fairhead, Leach, and Scoones 2012). Moreover, recent economic policies have produced frontier spaces in the Lao upland through concessions and encouragement of foreign investment under the slogan of 'turning land into capital' 2 and under the rationale of progress and rural development (Kenney-Lazar et al. 2018c).
Aligned with Tsing's (2005) understanding of frontiers, Barney (2009) describes the Lao upland frontier as unevenly distributed ('striated and patchworked') and produced by legal and extralegal means. Drawing on Rasmussen and Lund (2018), we argue that frontier dynamics in Laos and beyond have not only transformed property relations of uplanders but also dissolve other aspects of existing social orderse.g. customary rights, political jurisdictions and subjectivities. Hence, these two sets of dynamics are co-existent and -constitutive as frontier dynamics unmake and territorialization establishes spatial control (Rasmussen and Lund 2018).
State formation processes through territorialization and frontier dynamics are anchored in the local struggle for recognition of customary land (use) rights and membership to the community. In Laos, Dwyer (2013) highlights the need to better understand the local territorial relations, in relation to transnational land deals, while incorporating internal resettlement and the rubber boom as key decisive factors shaping these relations. As stated by Dwyer (2013:): 'reveal[ing] the specific political and social relations that shape the ways transnational land deals unfold on the ground [would] help to "resolve" the global land grab in finer scale geographic terms'. Hence, understanding public authority as in the making and institutional arrangements that compete for it as transitional allows for a perspective that brings to light the 'unfinished' character of the state, and the role of non-state actors in (re)shaping state space. According to Lund '[t]he state, [presented as an assemblage of actors with diverse interests and goals, rather than a single entity with unified objective], is always in the making' (Lund 2016(Lund , 1200. This analysis leaves room for a closer examination of those fragments of political authority (Lund 2011) which have operated outside the formal state realms, may not coherently corresponded with government institutions, and which might have been in opposition to them. Particularly, in those spaces where state's power and authority are contested, with no easily identifiable unequivocal domination (Lund 2016). To better understand how institutional arrangements and social order in the Namai upland are contested, we follow Dwyer (2013) and Lund (2016) in studying the specific political and social territorial relations that (re)shape socio-political production of property and citizenship rights, as looking at property (what we have) and citizenship (who we are) enables a concrete understanding of the dynamics of authority and the process of state formation (Lund 2016).
Understanding political authority as (re)produced through defining and enforcing rights, brings to the fore the dialectic relation of rights and authority. Lund notes that authority and rights are co-constitutive, contemporaneous and are conceptually tied together by recognition. Thus, 'when institutions recognize claims to property or citizenship, they themselves become recognized by the claimants of these rights' (Eilenberg and Lund 2017). Mutual recognition, then, constitutes a contract that links property and citizenship rights to authority. Looking at the dialectic relation of authority and rights and the production of social contracts of recognition directs our attention to the agency of institutional actors and their capacity to shape the process of state formation. We argue that local arrangements over access to land and customary land rights regimes in Laos illuminate the complexity of state formationas the latter is shaped by territorialization and the production of frontier space, which are themselves shaped and reshaped by various dispossession, whereby companies acquired farmers' land and communal forest as part of their land concession agreement (Kenney-Lazar, Suhardiman, and Dwyer 2018b;Schumann et al. 2006;Suhardiman et al. 2015).
actors' (i.e. state actors, private sector actors (e.g. rubber company), and farmers) multiple goals and interests, manifested in local strategies and practices, occurring at the intersection of resettlement, ethnic division and commercial agricultural production. How these interactions (re)shape farmers' strategies to transform their customary land rights as a means to secure their land tenure in relation to the government's land use planning and allocation processes and rubber boom, and how these strategies in turn predetermine the government's land use planning and allocation processes and outcomes are primary research questions explored in this paper.

Upland-lowland dynamics and state territorialization through resettlement and land use planning
The interrelationship between upland and lowland regions in Laos has a long history in which civilizational dynamics have connected the 'civilized' Buddhist lowland population and the 'barbaric' upland tribes in the unruly hinterlands. In such a dichotomized understanding upland and lowland might appear to be homogenous in themselves, omitting that they are subjects of complex ethnic relations (Pholsena 2006;Tappe 2018). The civilizational continuum between up-and lowland is rather 'characterized by the interplay between sociocultural difference and tentative rapprochement, by different levels of civilizational interpenetration, by transfers of knowledge and ideas, by political frictions and hierarchies, and by permanent movement of people' (Tappe 2018, 211). These entanglements have in recent Lao history been in many places drastically transformed by policy interventions that have sought to establish territorial control over the unruly hinterlands, most notably by resettling upland populations into the valley. The notion of rupture, described by Lund (2016) as an 'open moment' 'when opportunities and risks multiply, when the scope of outcomes widens, and when new structural scaffolding is erected' (Lund 2016(Lund , 1202 is therefore the entry point of this paper and its analysis of how different modes of territorial governance overlap and intertwine with frontier dynamics resulting in a contestation and fragmentation of political authority. Internal resettlement in Laos occurred as early as during the Laotian civil war (1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975) but was only widely implemented in 1975 (Baird and Shoemaker 2005;Evrard and Goudineau 2004). Shortly after Laos' liberation in 1975 the newly established Government of Laos (GoL) engaged in the resettlement of upland villages into the lower lying valleys often in response to armed rebel groups and associated security concerns (Baird and Shoemaker 2007). The rearrangement of population has been justified by the GoL as to tackle poverty (lout phone khvam nyak chon) and to foster rural development (phathana xonabot). Here, the idea of resettlement is presented as a means to bring villagers closer to public infrastructure and services (e.g. roads, markets, schools, hospitals), while also transitioning their livelihoods away from upland, swidden cultivation toward lowland, paddy cultivation (Baird and Shoemaker 2005;Cunnington 2011;Lestrelin, Castella, and Bourgoin 2012). However, researchers have also demonstrated that an underlying goal of resettlement was to increase state control over rebellious and mobile upland communities, many of whom were associated with the American-backed royalists during the Second Indochina War, by making them more visible and culturally integrated into the national development project (Baird and Shoemaker 2007;Evrard and Goudineau 2004). Baird and Shoemaker (2007, 870) categorize the GoL's motivations and justifying factors for internal resettlement into: 'opium eradication, security concerns, access and service delivery, cultural integration and nation building, and swidden agriculture reduction' which depending on location in Laos have different relevance (e.g. opium eradication in northern Laos) (Lu 2017). On a national scale, accessibility for providing public services and swidden eradication appear to be the most important (Baird and Shoemaker 2007). While the central Lao government sets the overall policies, interpretation, to some degree, and implementation of resettlement related policies on the ground was done by the provincial and district authorities Shoemaker 2005, 2007).
Internal resettlement by the GoL was pursued through three major policy instruments: Focal Sites, Village Consolidation, and Land and Forest Allocation (Baird and Shoemaker 2007). The 'focal site strategy' began in 1994 under the National Development Programme seeking to cluster villages based on criteria such as poverty alleviation, potential for development, etc. Villages were selected by the Central Leading Committee for Rural Development 3 and provincial authorities (Thomas 2005). The second policy tool, village consolidation, can be seen as a repackaging of the 'focal site strategy' similarly leading to the resettlement of rural population. Officially formulated in 1989 the Village Relocation and Consolidation strategy sought to improve villagers' access to public infrastructure and services by merging villages with less than 50 households into larger villages (with a minimum of 200 people in upland areas and 500 people in lowland areas), while also relocating remote upland villages to lowland areas (Baird and Shoemaker 2005;Cunnington 2011;Lestrelin, Castella, and Bourgoin 2012). 4 Lastly, the Land and Forest Allocation (LFA) policy aimed to secure resource tenure and sought to reduce environmental degradation by controlling swidden agriculture (Fujita and Phanvilay 2008). In its pilot phase, supported by international donors such as the Swedish International Development Agency (SIDA), Asian Development Bank (ADB) and the Food and Agriculture Organisation (FAO), LFA aimed to develop a system of land classification (Baird and Shoemaker 2007). The LFA policy targeted swidden agriculture by imposing restrictions on this practice, consequently leading to food shortages which in sum resulted into a 'push' for upland villagers to resettle into the lower lying areas (Baird and Shoemaker 2007). While the LFA policy has led to spatially reorganizing people, it has done so by delineating resource use boundaries through spatial planning.
State territorialization in Laos has in this regard not only manifested in policies leading to internal resettlement but also through the establishment of spatial control delineating agricultural and forest lands aiming to stop the degradation of natural resources (Evrard and Goudineau 2004;Goldman 2001;Lestrelin, Castella, and Bourgoin 2012). The GoL has therefore positioned land use planning as a key element in the country's land governance (Suhardiman, Keovilignavong, and Kenney-Lazar 2019). Land use planning is a powerful tool for states to make territory, as they 'divide their territories into complex and overlapping political and economic zones, rearrange people and resources within these units, and create regulations delineating how and by whom these areas can be used' (Vandergeest and Peluso 1995, 387).
In practice, however, land use planning efforts have remained contested since the early stages of the LFA program. Land Use Planning and Allocation (LUPLA) 5 is since the 1990s one of the main elements of land use planning in Laos. As an area-based approach to development it has been used to identify village boundaries as well as to demarcate different types of land uses for an improved management of natural resources. With regard to the national policy goal to reach 70% forest cover by 2020 (MRLG 2019) a categorization was decided that declared fallow land used for shifting cultivation as 'degraded' forest land ('unstocked forest') implying an absence of agricultural activity and hence opening the door for industrial tree plantations (e.g. rubber). The enclosure of swidden land has had negative effects on shifting cultivation communities, as concessions, as well as forest zoning based on the new categories (conservation forest, protection forest and production forests), exclude people from their previously used land (Barney 2008;Heinimann et al. 2017). The threefold forest categorization in Laos has according to Rigg (2005) become a major instrument of LUPLA and has favored forest conservation over agricultural land allocation. For example, LUPLA limited every household to three plots and rendered the remaining land that was laying fallow as forest under one of the three categories and hence as 'empty' state land (Lu and Schönweger 2017). The fallow length of agricultural land was moreover reduced to two years, turning shifting cultivation into an unsustainable venture (Fujita and Phanvilay 2008;Heinimann et al. 2017). Responding to critiques of LUPLA, as a top-down planning exercise that overlooked farmers' interests in using rural lands, the program was reformulated in 2009 as Participatory Land Use Planning and Land Allocation (PLUPLA). 6 However, a more recent (2008-2020) land use planning approach, carried out by The Agro-Biodiversity Initiative (TABI) has included shifting cultivation as an integral part of upland farming system, and into the land use planning processes. In the next section we discuss how this land use planning process unfolded in Ban Namai.

Land use planning in the Namai upland
Ban Namai, a low-lying village in the northern province of Luang Prabang, located in Nambak district, lies approximately 150 km northeast of the town Luang Prabang (see Figure 1). The Namai valley stretches in-between two mountain ranges limiting the available lowland area which was originally settled by a group of ethnic Tai-Lue. Around 1986, following the government's policies that led to resettlement (Baird and Shoemaker 2007), three ethnic Khmu villages (Ban Luk, Ban Klok and Ban Huana) were resettled from the surrounding upland into the valley. The rapid population growth of the village among other factors co-produced a scarcity of available farmland which was exacerbated in 2006, by a land concession granted to the Sino-Lao Chilan Rubber Development Company over the period of 50 years and covering 255 ha. The village is part of the Nayang village development cluster (Kumban Nayang), one of the GoL's priority development zones to exercise its Sam-Sang strategy. 7 Among the many development objectives that are addressed, one of the priorities of this rural development strategy is land use planning, for example to improve natural resource management, delineate village boundaries, and clarify land for concessions.
In Ban Namai land use planning was taken up in 2015 by The Agro-Biodiversity Initiative (TABI). Land use planning provides an entry point for the government to ensure sustainable agricultural development and forest protection through developing an agricultural and forest land zonation system (MAF 2017). TABI has been implementing the so-called Participatory Forest and Agricultural Land Use Planning, Allocation and Management (pFALUPAM) process at the village and village cluster levels, since the program started in 2008. The proclaimed goal of this community-based approach is to be more participatory than previous land use planning efforts in Laos by including the local population into the pFALUPAM process. This includes delineating village boundaries, developing agricultural land use and forest cover maps, developing the village's plan for Forest Land and Agriculture Management Zonation (FLUMZ) as well as rules and regulations surrounding the different land use in these zones.
TABI's objectives are to reduce deforestation and to reframe shifting cultivation as a sustainable practice, if done accordingly, by showing its feasibility on the local level (e.g. longer rotation). In Ban Namai implementation is done through the provincial and Nambak district authority (i.e. PAFO and DAFO) in collaboration with technical staff from TABI. The rotation schedule in Ban Namai is based on a four-year rotation in four larger zones which are subdivided into two (see Table 1 and Figure 2). Each year a total of two smaller zones become active and eligible for a limited number of farmers that can use the land for cultivation according to the agreed upon terms.
Even though the pFALUPAM process is theoretically designed to be participatory, by including villagers in the overall discussion and shaping of the plan it did not, in practice, adequately capture the local reality before it perpetuated a new land use system through delineation, mapping and other technical exercises. While PAFO/TABI 8 staff stated that they were aware of previous land use of villagers they did not include insights about villagers' claims to land in the planning process. Here, discussion on land use planning processes  8 There is a considerable overlap between PAFO/DAFO and TABI as the project staff for the pFALUPAM process is recruited by TABI from the PAFO and DAFO offices. In cases when TABI project staff is not part of either agency people in the village often perceive them as 'government staff'.
were focused on farmers' actual land use, but overlooked the customary land rights systems, how the latter evolve over time in relation to the government's previous policy on internal resettlement, and how it in turn predetermines farmers' access to land. This omission is reflected in the poor uptake of the plan in its first year as only 29 households (PAFO 2017) used the first zone for upland cultivation whereas 43 households practiced shifting cultivation outside of it. In the following year only ten households participated which resulted in PAFO and DAFO urging the village chief to hold another meeting to explain the system in detail and to encourage villagers to cultivate inside the appropriate zone.

Unpacking the 'black box' of local land governance and filling the institutional void
State territorialization in the Namai upland has been framed as inclusive through the participatory design of the pFALUPAM process. Involving different groups in the village, the land use planning processes was very much driven by the idea that everyone would be open to discuss existing rules and mechanisms pertaining to land use allocation, even when the latter is rooted in asymmetrical power relations between Tai-Lue and Khmu villagers as well as original settlers, early and late comers. Similarly, villagers did not always reveal their views on the proposed land use plan (as discussed and facilitated by TABI), especially when doing so would also mean confronting the idea for more equal land allocation between the different groups in the village. 9 In the next sub-sections, we present the brief history of land use in Ban Namai, while highlighting the importance of internal resettlement as a key determinant for shaping the land use planning processes and its outcomes.

Tai-Lue side of the storyearly history of the settlement and new Khmu settlers
The original settlers of Ban Namai, a group of seven ethnic Tai-Lue households from nearby Ban Nayang, founded this village at around 1969 during the period of the Laotian civil war when their families faced a food shortage. To be able to farm, our informants stated, you would have to do this in secrecy, out of security reasons, so the enemy could not spot and attack settlements easily. Tai-Lue farmers of Ban Nayang consequently moved to the surrounding areas and hid out close-by their scattered small plots in places where some had already cleared land in the past for paddy rice cultivation. The low-lying part of the smaller and slightly uneven Namai valley used to be a wetland which, in a communal effort, was transformed into paddy fields and divided amongst those who had helped. The seven households stayed nearby the stream using the upland to cultivate upland rice to diversify their food production. After 1975 and the end of unrest the new settlers expanded their fields near the stream but also on the other end of the valley were some families had cultivated paddy rice before the war. In 1986, Nambak district officials informed the Tai-Lue village chief about the impending resettlement of three ethnic Khmu upland villages: Ban Luk, Ban Klok and Ban Huana (see Figure 2). Relationships between Ban Namai and these villages had already existed, then through trade, volunteering work or upland activities. Due to the proximity of Ban Huana for example, some Khmu had the chance to visit Namai to make sure they would have land to build a house and to do agriculture. During this period the Tai-Lue community was split between those who welcomed the idea of new settlers in order for Ban Namai to grow and develop and those who opposed the idea of strangers settling on their land. Nevertheless, the first resettled Khmu arrived in 1987 and were given land to build their houses on what had previously been the garden area of the Tai-Lue, i.e. the reserved land for future generations. This was decided in a meeting between Tai-Lue and Khmu under the condition that Khmu should not expand the given area and expand onto the fields of Tai-Lue. Agricultural land in the valley was not available for the vast majority of Khmu and still today only a few Khmu households were able to acquire small parcels for paddy production. After more resettled Khmu arrived in Ban Namai in the following years, Tai-Lue families were not willing to provide more of their land and handed the responsibility to distribute land to the Khmu settlers who had arrived earlier.
In response to the growing influx of Khmu migrants Tai-Lue farmers who cleared upland plots and had paid the land tax were allowed by the district authority to reserve the land but were not provided any legal document. Only in 2007 the first land records were issued in the form of family land books which delineate residential area and register the measurements of low-lying paddy fields. This formalization process however did not record upland use rendering these areas as state land.
Migration in Ban Namai led to a spatial separation of the two ethnic groups in which the Tai-Lue maintained their old village structures (e.g. houses, town hall, etc.) whereas Khmu migrants formed clusters of affiliation to previous village, kinship and according to time of arrival in the area provided to them. Besides the existing spatial segregation, village governance is also divided as each ethnic group delegated decision-making power to representatives, so that: 'Lue look after Lue and Khmu look after Khmu' (Interview with a Tai-Lue original settler, October 2018). This decision was made by both groups on the grounds that their cultures were too different to decide on matters of the other group. Subsequently, this governance structure was integrated into the formal arrangement through elections of deputy chiefs carving out space for local power. 10 Moreover, the two ethnic groups make decisions through informal meetings within their communities and only when outside state institutions enter Ban Namai the whole village gathers formally.

Khmu side of the storyupland communities and arrival in Ban Namai
In their previous villages, our informants reported that Khmu villagers had access to abundant land from which they could choose a plot fit for shifting cultivation. Customary tenure systems across the former Khmu villages differ as some described that land was held as a common pool resource which was freely available to everyone in the community and only had to be cleared for cultivation. Most of our informants however described that customary land rights were held inside a family structure organizing access to a number of shifting cultivation plots in the surroundings. Land rights thus remained inside the family, through inheritance, and the next generation was familiarized by the older generation with the 10 Villagers had decided beforehand to vote two people of each group however three Khmu were elected and only one Tai-Lue representative for which reason another Tai-Lue was appointed as an unofficial assistant. families farming practices. In this customary tenure system every family knew the location of their land and of the land of the others without written documentation. The remaining land in the surroundings of the village which did not belong to any family was considered a common pool resource and use rights were acquired by clearing it. Due to the small population size in the previous villages and their vast territories, conflicts over land allegedly did not occur and if there was a misunderstanding it was settled between the heads of the two families who had the disagreement. If this failed the elderly people of the village were involved to settle the dispute. One of our informants stated that to avoid conflicts families would gather before the season to discuss their plan and the location of their selected plots. Shifting cultivation was described as a communal effort in which families help one another and work together (samaki) in reciprocal relationships. Evrard (2007, 129) accordingly describes the Khmu upland social system based on an egalitarian tenor with 'no real, absolute differences in status between lineages of a locality'. However, differences between Khmu come to the fore increasingly through differing economic resources and changing ways of life (e.g. paddy rice cultivation).
Resettlement of the former villages began in 1986 when the district authorities pushed people to move into the Namai valley where they would have access to a primary school, to a lower secondary school in the neighboring village and road access to the district center Nambak with a hospital, market, and upper secondary school. In some cases, Khmu villagers paid visits to Ban Namai to discuss with Tai-Lue residents where they could acquire residential and farmland before leaving their old village. Other Khmu households migrated without ever having been to Ban Namai. The relocation of the three villages took place over a period of time in which a larger number settled around 2004 and 2005 with a continuing migration flow of Khmu guided by kinship relations. The outflow of Khmu had a domino effect on these villagesin Ban Luk for example which has now disappearedas no one wanted to stay behind. Over the following years the number of newly settled Khmu families surpassed the number of original Tai-Lue settlers. Today there are 146 Khmu and 100 Tai-Lue families in Ban Namai. The uncontrolled movement of people led to a rapid population growth and smaller settlements near the stream of the village in which new arrivals stayed temporarily until either district authorities or relatives could provide them with a residential plot to build their house. The insufficient amount of available land for agriculture nearby the village moreover forced many Khmu to use their old plots which in some cases are located near Ban Namai (2-3 hours walk) but more generally are further away in about half to one day walking distance. Many Khmu have thus expressed their lack of understanding about the government induced resettlement. Furthermore, de-and reterritorialization in these cases are often not clear-cut but have resulted in hybrid forms in which families partly or temporarily relocate and maintain old landholdings near which they camp out to practice agriculture. 11 5.3. After the rupture: social structure and a layered system of 'fuzzy' land rights Understanding resettlement as a rupture in the spatial order and social structure of Ban Namai and thus in villagers' property relations, opens up room to analyze how power is 11 See also Baird's (2013) explanation of why people subverted government rules and how people returned to their old swidden land, as well as Baird and Shoemaker (2005) and Evrard and Goudineau (2004) for other examples.
(re)established, and how new social contracts of property and citizenship are formed between original settlers, recent, late comers, and the government. In this contested field of political authority, territory is co-produced between claimants and those who authorize their claims as rights.
Territorialization through resettlement has only been partially successful in establishing state authority in Ban Namai but has nevertheless led to further state interventions, i.e. through land concession and land use planning, aiming to establish spatial control not only in the Namai valley but also in the upland region where fragments of local society hold control over land. The upland area near Ban Namai can be described as a layered property system in which customary and formal property rights partly overlap and intertwine. Customary land rights in the upland persist until today but have been subject to change by multiple dynamics that have unfolded in the landscape.
The fringes of upland near Ban Namai are predominantly controlled by Tai-Lue who hold land rights based on customary tenure. Even though government authorities allowed these households to reserve land it did not provide them with a formal right and hence little security. Therefore, Tai-Lue upland farmers have increasingly converted their plots over the past years into small cash crop plantationsmostly rubber. This way, they transformed their plots from upland rice cultivation to commercial cash crops, more in line with the government's policy to promote market integration. In many cases, Tai-Lue farmers hired Khmu laborers who would be granted use rights to another upland plot in exchange for clearing and preparing land for nurseries. Other land rental arrangements between Tai-Lue upland holders and Khmu include land use rights in exchange for money depending on soil quality and size of the plot or in exchange for part of the rice yield (around 2-3 bags). Upland lands have also been sold by Tai-Lue families, predominantly to Khmu. Tai-Lue farmers' secured land is in high demand due to the proximity of its location to the village leaving Tai-Lue farmers in a powerful position.
Khmu similarly rent out or share land to fellow Khmu mostly to those belonging to the group of late arrivals coming from further away places. Access to upland lands among Khmu is regulated by kinship and/or is based on land rental arrangements similar to those between Tai-Lue and Khmu. In cases where land is located far away family members camp out in huts during the rice growing season. Village affiliation and kinship play a key role in being eligible to use land and to obtain customary land use rights.
Focusing on the social structure of Ban Namai directs attention to the patchy web of Khmu social networks torn from their previous fabric and reassembled in the new setting. Migrants of the same village have often built their houses close to one another likely out of reasons to rebuild old village structures but possibly also due to assigned and limited space. Resettlement has resulted in a spatial separation of both ethnic groups leading to fragmented social networks in the early period of the resettled village. Over the past years these fragments have increasingly established links, for example, through invitations to social events (e.g. marriages, funerals, etc.), neighborhood in the village or upland, provision of work or services, etc. Overall, villagers of the Tai-Lue group take on higher positions in the village's social structure, for example, due to resource access, as evidenced by their more privileged access to land (e.g. closer to village) and ethnic identity, as the original settlers in the village. Relations between Tai-Lue and Khmu are to some degree characterized by patronage relationships exemplified in rental agreements. For Khmu to be able to rent land from Tai-Lue upland holders it is mandatory to establish a social relation beforehand which may be described by sung hak sang peng (to build a good relationship/friendship). Rental agreements can thus be understood as social contracts of mutual recognition as Khmu recognize Tai-Lue upland holders as an authority providing rights to them and Tai-Lue landholders recognize Khmu farmers' property use rights and admit them to the community. The conditions of each social contract vary according to the terms of the rental agreement and are beyond that terminable, lasting from one season to indefinite use rights. These different sources of authority or power poles (Bierschenk and Olivier de Sardan, 1997) form a complex network of constant change, regarding new arrangements and their conditions. Asymmetric power relations paired with overlapping claims to land have led upland farmers to secure their 'fuzzy' (use) rights to land by adopting several strategies as we will show in the following sections.

Unclear boundaries and security through tax receipts and cash crops
In the context of Ban Namai, kinship relations work as a form of trust and assurance that give some degree of security to lessor and lessee. Rental arrangements which are not based on kinship relations and trust pose more risk to the party seeking land. One upland holder thus stated: 'before doing a land deal you should always do an investigation and talk to the neighbors so that the boundaries of the field are clear' (Interview with a Tai-Lue original settler, October 2018). Some informants furthermore stated that unclear boundaries are also used to an advantage by those renting out their land in exchange for labor to clear a designated upland area which can after clearing be partly used for upland rice by the hired laborer. Those clearing the land in exchange for use rights have experienced that land holders planted on the cleared land and only provided them another area for cultivation after they had confronted them. Boundaries in the upland tenure regimes are therefore 'fuzzy' and might overlap wherefore villagers employ different tactics to secure their upland such as clearing the land, cultivating it, fencing it and by paying the land tax. The 'fuzziness' of customary land rights in the Namai upland shows how intricately power relations are linked to property and how diffuse power works in this contested and fragmented terrain of political authority.
Overlapping rights regimes and the blurring of formal and informal legal spaces moreover reflect in villagers' land tax payments. By paying the tax for their upland holding they perceive it to be secure and view the tax receipt as a proof of ownership. Linked to this strategy of using different forums to secure their access to land, villagers moreover lodge indirect claims to citizenship in direct claims to property, i.e. through tax payments. Lund and Rachman (2018) refer to this key mechanism in acquiring rights as indirect recognition. Tax payments for upland holdings thus directly aim to solidify property rights by recognizing the authority of government institutions, while at the same time indirectly entail a legal visibility of villagers as right bearing subjects. Vice versa, Khmu villagers lodge claims to property rights in their role as 'exemplary citizens' or in their past role as soldiers. Khmu villagers for example demonstrate that they are good citizens by actively involving themselves in the community and in its different committees hoping to reap potential rewards through recognition. 'So, in addition to the overarching absolute claims, an understory of partial, pragmatic, and indirect claims plays a role in the everyday struggle over space, resources, and the recognition of rights to them' (Lund and Rachman 2018, 418).
This complex situation of local land governance in which actors employ a wide variety of strategies to secure access and rights to land has since 2004 come under pressure by two state interventions that have operated under the presumption of an institutional vacuum seeking to establish territory. The next section therefore grounds territorialization and frontier dynamics and subsequently links these processes to farmers' strategies that navigate an uneven terrain.
6. Making territory, frontier dynamics and farmers' strategies to secure access to land Until 2006 the relationship between villagers and government institutions (i.e. PAFO and DAFO) has been marked by two territorial projects, that of resettlement and secondly, through the conversion of 255 ha of state landin the Namai upland ( Figure 2)into a concession for the Sino-Lao Chilan Rubber Development Company. The latter move is part of a larger project in the districts of Nambak and Pak Ou in Luang Prabang province using multiple smaller locations for rubber plantations aiming to convert a total of 20,000 ha (McAllister 2015). In Nambak the Sino-Lao Chilan Rubber Development Company was granted a 50-year concession in 2004, arriving in Ban Namai in 2006 to inform villagers and to start converting the leased upland. Existing customary rights were erased in this process as both Tai-Lue but mostly Khmu villagers were dispossessed of their upland holdings. Villagers stated that they wanted to file a complaint with the authorities but were too afraid of doing so and instead silently bore the loss of their land. Regarding their future, one of our informants stated that he was not confident that people would get their land back after the end of the concession term. A study by Friis et al. (2016) in the neighboring village of Ban Nayang similarly found that the enclosure of land for rubber plantations further contributed to the negative effects of resettlement policies on people's access to natural resources, livelihood activities and income opportunities.
The upland frontier space that emerged and freed up resources set into motion a dynamic that has resulted in an ongoing conversion of upland plots into small-scale rubber plantations. 12 In the Namai frontier customary tenure is already imagined as unmade, as farmers gradually transformed their swidden plots into rubber garden, either by contract farming with the Sino-Lao Chilan Rubber Development Company or independently. The former practice can be characterized as plantation-sharing schemes in which 70% of the profits go to the farmer and 30% to the company which provides seedlings, fertilizer and knowledge. Smallholder non-contract farmers invest in and cultivate rubber independently but sell their produce to the Sino-Lao Chilan Rubber Development Company for a rate below the price that contract farmers receive from the company. Farmers' decision to transform their swidden fields into rubber garden in order to secure their land holdings can also be viewed as a strategy to mimic the rubber company's land concession practices. 13 As said by a number of farmers: 'If the company can claim villagers' farmland through its rubber plantation, villagers can also plant rubber trees to secure their 12 Farmers refer to these plantations as rubber gardens.

13
See below. land tenure while also benefiting from the crop production' (Interview with farmers, October 2018).
In 2015, subsequent land use planning led by TABI attracted only those who were in need of land (but did not hold upland plots themselves in the designated zones) (Figure 2). While TABI included the historical dimension in land use in its consultation processes, it overlooked the importance of internal resettlement as one of the key factors shaping villagers' access to land, which later predetermined the actual land use planning processes and outcomes. TABI's land use planning processes did not explicitly link villagers' access to upland plots (or the lack thereof) with their respective status of original settlers, recent and late comers, and how they view the proposed/discussed land use plan in relation to these statuses. Consequently, TABI's spatial planning project has not come into full effect as it conflicts with customary land rights of both ethnic groups and the interests of powerful local landholders. Or as our Khmu and Tai-Lue informants put it: Tai-Lue original settlers disagree with the plan and insisted that Khmu recent settlers have to continue paying land rental to them. Otherwise, they would not allow us to clear and cultivate their land, even when this is stated in the defined plan. (Interview with a farmer from Khmu recent settlers, November 2017) Over time, we have managed to rent our reserved land to others as additional means to increase household income. However, the proposed land use plan will erase this rental arrangement when it allows Khmu recent settlers to access the land without paying any rent (in cash and kind). (Interview with a Tai-Lue original settler, November 2017) Namai's upland frontier space has created an internal momentum for rubber expansion on a smaller-scale, exacerbating land scarcity, while also reducing the actual significance of land use planning processes. Over a period of 4-5 decades (from the early 1970s until recently), territorial aspirations and frontier dynamics have come together in shaping the upland landscape in a contemporaneous effort. However, there are some inconsistencies in the making of state territory by government institutions, such as continuing land tax collection even though land has been rearranged formally through TABI's new land use plan. State territorialization in Namai has in that sense been contradictory and has been hampered by villagers' unwillingness to give up their land. Customary land rights occupy in this regard a central role in villagers' decision-making as they are used as a means to secure land holdings and in consequence work to defy state territorialization. The conversion of upland land into rubber gardens feeds on the other hand into the broad scale project of turning land into capital (i.e. specifically the promotion of rubber as a modern cash crop and a form of forest cover), while also reaffirming the government's misrecognition of claims to swidden land. Even though such conversions are primarily made to secure access to land locally 14 , they reinforce state authority over land by contributing to, rather than challenging, overarching structures. Moreover, by producing rubber independently or as contract farmers villagers enter into a relationship of mutual recognition in which villagers 'breathe life' (Lund 2016) into the Sino-Lao Chilan Rubber Development Company by recognizing its political authority while indirectly lodging a claim (to be recognized by the government) to their land. 14 Baird (2008) similarly showed how ethnic minorities in northeastern Cambodia planted cashew trees to protect land from external intervention.
Apart from turning land into rubber gardens, some Tai-Lue and Khmu upland holders have also engaged in making territory, for example by clearing upland plots, planting crops on them (e.g. upland rice, maize, mangos, mak kha, etc.), fencing them or by providing use rights to other people. In the local scramble for land, such efforts impeding the defined land use plan of TABI (and its associated mechanism to unmake spatial control of villagers) are often based on subsistence ethics 15 , aiming to maintain customary land (use) rights that ensure access to farmland for upland rice and the production of other crops for consumption. Customary tenure, based on reciprocity, family (and/or former village) ties and traditionalism, in that sense works as a sort of homeostasis in the spatial organization of the Namai upland. However, customary land tenure in the immediate proximities of the village has been increasingly transformed by powerful upland holders who preserve land rental arrangements and plant rubber as a territorial strategy with implications for social as well as labor relations, land use and land control.
The co-production of territories by the assemblage of state and non-state institutional actors in Ban Namai has produced a complex patchwork of territorial projects. In this fragmented landscape uncolonized interstitial spaces have been strategically exploited by villagers to secure their customary land rights, for example, by starting small-scale rubber plantations. TABI does not differentiate on their maps between different categories of rubber production such as: agro-industrial rubber plantation, contract farms and independent smallholdings (Figure 2). The general category 'Rubber' used by TABI thus does not make explicit and thereby does not 'expose' which land belongs to independent smallholders, contract farmers or the company. TABI's generalization has in that sense allowed those villagers with enough resources to secure their landholdings as rubber gardens, as stabilized rubber cultivation opposed to rotating swidden farming has been tolerated and has met no opposition by government authorities or TABI so far. Rubber expansion and TABI's land use planning process in this regard overlap as rubber spills over into the reserved area for swidden cultivation 16 , in TABI's land use plan.
Territorial strategies of both newly settled and original settlers operate in response to the expanding pressure on land triggered by resettlement and deepened by the concession project, taking shape for example in non-compliance regarding land use planning. Furthermore, we have shown that the commodification of natural resources is not only part of state territorialization but has become a strategy to make territory on both ends. Frontier dynamics enable farmers to exploit the interstices between land use planning and concession project, offering a strategy to those with enough resources to bridge the maturation period of their rubber gardens.

15
Drawing on Scott's (1976) idea of subsistence ethics, Rehbein (2007, 26) writes, [p]easants' interests are focused on having enough until the next harvest, not on having as much as possible. They achieve this by mutual aid (reciprocity) and by reinforcing family ties and traditionalism. They aim at survival and security, not at affluence and profit. Reciprocity, family orientation and traditionalism, subsumed under the term subsistence ethics, characterize village society in Laos. 16 PAFO and TABI staff underlined that one of the primary goals of the joint venture is to promote food crops such as rice and livestock farming to tackle food insufficiency.

Conclusions
By understanding the internal resettlement of three Khmu communities into the Namai valley as a rupture ('open moment'), we have illustrated that state territorialization, aiming to establish spatial control in Ban Namai, has been rather unsuccessful in the surrounding upland, as Tai-Lue and Khmu villagers began governing their disrupted access to land through informal institutional arrangements. Even though this open moment allowed villagers to maintain upland control, it affected them differently, based on their resources, social ties and ethnicity. Moreover, the engendered uneven access to land and the resulting power asymmetry in Ban Namai have predetermined subsequent state territorialization attempts. Today, the spatio-temporal effects of internal resettlement, in that sense, continue to haunt local land governance in the Namai upland.
Against this backdrop, we have argued that customary land rights are a key decisive factor in farmers' strategies to secure their land tenure and serve as a means to defend their landholdings against the government's land use planning and land allocation initiative. Villagers have in this regard used the state's incoherent territorial ambitions to their advantage, as villagers exploit the interstices within and between different government interventions and harness frontier dynamics to secure their land tenure, while also relying on their customary land rights. When farmers planted rubber as a means to secure their land tenure, they transformed their customary land rights into a territorial strategy. In other words, farmers planted rubber not only as a market strategy to increase farm households' income, but primarily as a territorial means to protect their land from an external intervention. Hence, villagers take part in the production of territory, by refusal and/or by mimicking the nearby rubber concession, while ensuring that such refusal occurred within state spaces (Kenney-Lazar, Suhardiman, and Dwyer 2018b).
Positioning farmers centrally reveals their multiple strategies to secure customary land (use) rights. The case study has illuminated how power relations, different legal orders, and accumulation processes intersect and are shaped by villagers in order to secure their access to land. By adopting a 'bifocal perspective' of territorialization and frontier dynamics, bringing to light the dynamic structural processes that farmers navigate, we have also shown how a rubber frontier space has unmade local land tenure, but more recently has been used as a means to secure farmers' upland plots. Integrating frontier dynamics with a territorialization lens thus allowed us not only to better understand farmers' strategies to secure their access to land, but also to unpack the rationales behind these strategies, and how farmers' reasoning is rooted in views of their customary land rights vis-à-vis external policy interventions. Frontier dynamics, then, take an ambiguous role in the overall production of authority and rights as the commodification of upland space has become a strategy by both local cash crop farmers, to secure their land tenure, and by the Government of Laos, in its plan to 'turn land into capital'. Resources such as land or rubber and the dynamics surrounding them have in this regard profound consequences on the governance of people and places, as they reshape space and its state-society relations. Baird (2019) has in this regard recently pointed to the GoL's changed understanding of land that is nowadays treated also as an economic asset, thus broadening the positioning of land as merely a building block of national sovereignty. Within the GoL's increasingly neoliberal economic agenda, land concessions have become a crucial component and have, as we have shown, provided a channel for smallholders to turn their land into capital and their customary rights over it into a territorial strategy to secure their land tenure. This process of neoliberalizing nature (Bakker 2005) took shape within a larger spatial-territorial-configuration (Barney 2012;Dwyer 2013) involving other modes of governance (e.g. coercive resettlement, controlling swidden agriculture by delineating resource use, patronage and rent seeking of villagers, and subsistence ethics). What is interesting in this development is that farmers who have historically faced difficulties in getting government institutions to recognize their land tenure were able to secure the latter through rubber cultivation. Local rent seeking of villagers in their scramble for land has in that sense worked together with a neoliberal mode of governance that has established a market for rubber and has allowed for an increasing influence of the Sino-Lao Chilan Rubber Development Company over local resource use. Looking at the process of state formation and the intensification of capitalist relations from below has shown how new social contracts among non-state actors have crystalized, giving rise to new configurations of recognition on the local level.
Subsistence ethics and the existing institutional setup have in this process undergone a capitalist transformation with implications for the village's social structure, prompting an increasing social differentiation between villagers. Ethnic minorities take in this development a disadvantaged position due to their ethnic, educational, political, social and geographical resources. Social inequality of ethnic minority groups within the Laotian process of state formation has in this regard become more pronounced as the reconfiguration of spatial relations engendered exclusionary effects. Land use planning interventions seeking to transform upland regions therefore need to take into account more carefully local power relations and customary land rights, as our case has shown. We assert that while land use planning processes should make visible and integrate locals' claims to land, uncovering these claims might also mean unraveling unequal power relationships within the different groups of local communities. Hence, recognizing such power asymmetries is key to make land use planning more meaningful, if it is to provide alternative measures to deal with the issue at stake.