Assembling land control after displacement: some reflections from rural Southern Sudan

In the years prior to independence, large numbers of displaced people returned to Southern Sudan. Returnees were seen to pose particular challenges in relation to land rights. But in many areas, returnees had no difficulty regaining access to land. I recount how returnees successfully assembled land for inhabitation and productive use through autochthonous modes of governance, legitimation and inscription. My study argues that the quotidian practices of inhabitance and production are as critical to assembling of land rights as more institutionalized processes of legal regulation or policy developments, the promotion of markets, or the organized use of violence or force.


Introduction
Displacement in Southern Sudan during Sudan's Second Civil War was generalized and occurred on a massive scale. Precise numbers are difficult to verify, but according to Elnur (2009) approximately 80 percent of the population of Southern Sudan was displaced in 2005, when the Comprehensive Peace Agreement (CPA) that ended the conflict entered into force. 1 Three years later, more than two million people had returned to their communities of origin to begin the process of reconstructing their lives, livelihoods and country (Duffield, Diagne, and Tennant 2008), with returns continuing throughout the transitional period (2005)(2006)(2007)(2008)(2009)(2010)(2011) and after independence (UNOCHA 2011;Reeve 2012). Returns took place in the context of tense, contested and shifting political relations between the rebel lead autonomous government in Southern Sudan and the national government based in Khartoum as well as among ethnic communities within the South (Schomerus 2008;Rolandsen 2009;Hovil 2010;Schomerus and Allen 2010). Given the fragile peace, returnees were critically important to political hopes for Southern Sudanese independence, while at the same time the number and pace of returns challenged the newly formed government's capacity to project its authority to many of the remote rural areas it purportedly governed (Duffield, Located in the agriculturally productive Didinga Hills, historically residents of the area practice agro-pastoralism, with climatic conditions, animal health and political developments shaping how extended families strike a balance between agricultural and pastoral production. More recently, those with the requisite skills and capital have added waged labor in civil service and the Non-Governmental Organization (NGO) sector or petty trading to their mix of livelihood strategies. Both the town and the ethnic group are comparatively marginal in terms of economic and political clout in the national arena.
The data I present here are drawn from a broader project examining the shifting sociospatial practices that accompanied return migration from Kakuma Refugee Camp in Kenya to Chukudum. The research combined traditional ethnographic approaches such as participant observation and informal conversation with over 130 semi-structured interviews and a detailed migration history and livelihood survey (n = 79). 4 The majority of the research took place during the transitional period 5 between November 2009 and September 2010, with return visits in 2011 before and after Southern Sudan gained its formal independence.

Frontier assemblages
Within the field of political ecology, the notion of the frontier has gained currency as a way to understand the renegotiations of space and authority that surround moments of migration and political, economic and ecological transition (Tsing 2005;Barney 2009;Curry and Koczberski 2009;Fold and Hirsch 2009;Hirsch 2009;Peluso 2011;Vellema, Borras, and Lara 2011). Originally associated with settler colonialism of North America, the frontier has been understood as a material and ideological space productive of self-reliance, unchecked individualism and the freeing up of new lands and resources for exploitation and development (Turner 1920, 212-14). Despite its colonial legacy, the concept of the frontier has since been reworked in ways that allow for a more grounded analysis of the dynamics that unfold as places come to be dis/articulated through violence (Peluso and Lund 2011). Frontiers operate in the grey zones between licit and illicit modes of power, in the shifting overlap of competing regimes of rule. State and extralegal violence often play a central role in the creation of frontier-like spaces, where violence is used to free up land, resources and labor for profit generation (Woods 2011;Grajales 2011;Peluso and Vandergeest 2011;Watts 2012;Leonardi 2013). As a conceptual tool, the notion of the frontier encompasses both the imaginative and material processes that come together to assemble land and resources as being there for the taking (Tsing 2005;Li 2014).
Often frontier-creating deregulations emanate, at least in part, from weak political authority that is selectively or intentionally employed (Mbembe 2001;Ferguson 2006). But deregulation can also describe how spaces are set free from embedded social and cultural multitude of smaller ethnic groups, the Equatorians, at approximately 31 percent of the population (SSCCSE 2011), have historically worked together as a political bloc as a way to balance the influence of the country's two largest ethnic groupsthe Dinka and the Nuer, at approximately 35 percent and 15 percent of the population, respectively (CIA World Factbook 2014). 4 Participants for the survey were selected by means of a geographically stratified sampling frame, which was developed using Google Earth satellite images. However, because no reliable data on the population of the town exists and the satellite data was several months out of date, it is unclear if the sample selected was representative of the population. For this reason, data from the survey is descriptive of the sample, rather than the population as a whole. 5 The CPA mandated a 6-year transitional period from the signing of the accord in January 2005, to the referendum in January 2011 when southerners voted on whether to stay with a unified Sudan or to secede and form an independent state. regulations of space and practice. In much of Africa, legal pluralism and the subsumption of land issues under the authority of customary law remains one of the enduring legacies of colonial strategies of indirect rule (Mamdani 1996;Boone 2013;Lentz 2013;Leonardi 2013). As a result, local social and cultural practices continue to play a central role in legitimating authority structures, as well as in shaping norms of land access and resource use (Utas 2012). While land control practices are sometimes recognized and reinforced by local and national state action (itself often highly contested; see Moore 2005;Badiey 2013;Boone 2013), much of the contestation around access to productive resources in the African contextincluding landrelies on a shared idiom of autochthony (Geschiere 2009;Chauveau and Colin 2010;Woodhouse 2012;Lentz 2013) and takes place through the everyday negotiations of embodied material practices (Sikor 2012;Lund and Boone 2013). The alienation of land and other resources that occur in frontier zones requires the active undermining of these other forms of socially embedded authority. Whether or not these authorities operate outside state authorities or are incorporated into the formal bureaucratic structure through customary law, the ability to sustain autochthonous claims to land and resources practically plays a key role in determining which places are incorporated into a given frontier space.
A focus on land control through the lens of social practice brings into focus the ways in which local communities and individuals lose, retain or reclaim authority over land. The practices may work broadly in collaboration with the state and, indeed, play an integral role in state formation, or, as in this case, they may be effected outside of the state, or in direct opposition to state attempts to assert authority over and through the legitimation of land claims (Sikor and Lund 2009;Peluso and Lund 2011;Sikor 2012;Badiey 2013;Lund and Boone 2013). My study suggests that the quotidian practices of inhabitance and production are as critical to assembling of land rights as more institutionalized processes of legal regulation or policy developments, the promotion of markets, or the organized use of violence or force . I document the ways in which 'exclusion is a double-edged sword that can be wielded by the weak against the powerful with the right strategies and under the right conditions' (Kenney-Lazar 2012, 1021. In Chukudum, this meant that returnees successfully asserted their authority over land through an assemblage that combined acting as if their claims were legitimate with broader legal and cultural recognition of autochthonous land rights as well as the potential that more violent forms of contestation (on the part of a militarily connected commercial elite) might destabilize the newly forming state in the critical run-up to the referendum on independence. The importance of everyday practices and informal modes of authority in the negotiation of land rights also sheds light on the tenuous role of formal political institutions in day-to-day governance in rural areas and their limited scope for mediating intraand intercommunity tensions.

Dis/assembling autochthonous authority in Chukudum
In Southern Sudan, state penetration has always been quite weak, dating back to the colonial period, with the projection of state power largely restricted to the major towns (Johnson 2007;Leonardi 2007Leonardi , 2011Leonardi , 2013Rolandsen 2009). After independence in 1956, national law in Sudan stipulated that land belonged to the state; in the southern region, the government lacked both the interest and the capacity to enforce this policy, especially in rural areas (Collins 2008;Rolandsen 2009;Johnson 2010). State authority in rural Southern Sudanto the extent that it existedwas largely administered by appointed chiefs and traditional leaders, who acted more as intermediaries between the government and local populations than they did as instruments of rule (Leonardi 2011). With respect to land, authority was vested in a variety of communal institutions of land management and resource governanceincluding, depending on location, local chieftaincies, land priests and clan or lineage leaders (Johnson 2007;Rolandsen 2009;Leonardi 2011). These institutions administered land and grazing rights and mediated conflict both internally and with neighboring tribes, clans and/or sectionsoften in a flexible manner that recognized both primary and secondary rights to particular areas for grazing, agriculture and transit (Johnson 2010).
During the nearly three-decades-long second civil war in Southern Sudan, the pattern of limited state presence in rural areas was exacerbated. By very early in the conflict, the rebel SPLA exerted control over much of the region's rural areas, while control over many major towns remained with the Sudan Armed Forces (SAF) (Rolandsen 2009). However, the SPLA's attempts to establish civil administration (as opposed to military control) in the liberated areas were limited and often not effective (de Waal 1997;Branch and Mampilly 2005;Rolandsen 2005;Badiey 2010). The conflict also produced considerable displacement in the south, particularly after the 1991 split in the SPLA. 6 During this period, civilians were increasingly targeted by competing factions of the SPLA as well as the SAF and fled in large numbers both across international borders and within the south (Johnson 2007). The civil war and the factional fighting within the SPLA produced openings for new militarized regimes of land and resource control to take root (Schomerus 2008;Walraet 2008;Badiey 2010;Sluga 2011). The war substantively transformed local political authority and economic relations in Southern Sudan, particularly in the territories controlled by the SPLA rebels (Hutchinson 1996;Rolandsen 2005;Johnson 2007;Leonardi 2007;Collins 2008). Massive population displacement in combination with the subsumption of local authority to military command during this period considerably eroded the capacity and authority of many local regimes and institutions of land control.
Among the Didinga, historically land was governed through a complex web of relations between immediate family members, wider lineages groups, and clans (Driberg 1939). Rather than an understanding of land ownership based on exclusive property rights, rights to use land for cultivation, pasturage and resource extraction have been more relationally defined (Driberg 1939;Allen 1996). In a national context characterized by cultural values that celebrate transhumant pastoral production, these kinds of relational property rights regimes are both commonplace and generally accepted by the government and across the diverse population (Rolandsen 2009). Formal land documentation procedures such as a registry or cadaster did not exist outside of the major towns (Deng 2011). As agro-pastoral people practicing shifting cultivation, Didinga rights to land are negotiated through a set of cultural practices that associate particular clans with loosely defined geographical territories. 7 According to anthropologist and colonial administrator J.H. Driberg (1939), historically, ritual and governance functions were split, with one kind of clan official having authority over settlements or inhabitance, and another over soil or production (Driberg 1939). In my discussions with residents, land claims were still legitimated through oral accounts that detailed personal and lineage migration histories that brought people and families to the comparatively new settlement of Chukudum over the course of the past century. 8 Productive land near town remained largely under the authority of those who had previously cultivated there, with the descendants of early settlers and prominent past leaders most likely to retain access to larger plots of land. As I will later show, while years of violence and displacement have impacted ways in which land rights are obtained and distributed in Budi County, these culturally mediated regulations of both inhabitance and production have been surprisingly resilient. 9 However, these modes of authority and resource governance faltered when confronted with the overwhelming militarization that occurred in the area in the 1990s. The significant and longstanding presence of Dr. John Garang's faction of the SPLA created an economy of extraction where military commanders exerted both political and economic control of the area (HRW/Africa 1994;Rolandsen 2005;Walraet 2008). Rebel and military operations slowly pushed many Chukudum town residents to abandon their homes and gardens, seeking refuge farther in the mountains and over international borders. 10 As with many other places, provisions were irregular for fighters stationed in the area and SPLA soldiers were not regularly paid for military service. This placed considerable stress on local populations who were expected to contribute food and other material support to the liberation effort across Southern Sudan. 11 Some of the fighters arriving in Chukudum set up trading operations taking advantage of the area's proximity to the border with both Kenya and Uganda. In addition, a number of internally displaced people from other conflict-affected regions (notably Dinka from Bor) decided to settle in Chukudum rather than to become refugees. Those who chose to stay were either former rebel soldiers or those with family connections to people high up in the SPLA military hierarchy (Walraet 2008). 12 These militarily connected commercial and political elite relied heavily on the extortion of local producers and the expropriation of local resources while discursively constructing their activities as part and parcel of the struggle for liberation (Rolandsen 2005;Collins 2008;Walraet 2008). The newcomersboth the internally displaced and those in the SPLAassembled land, environmental resources and control over mobility and the border into a profitable frontier economy, adversely affecting the livelihoods of residents of the area.
The consolidation of economic and political authority under the SPLA military leadership elicited broad dissatisfaction among residents of Chukudum. These resentments surfaced in 1999 during what came to be called the 'Chukudum Crisis', an internal conflict sparked by the assassination of the ranking SPLA Commander, a Dinka Bor named Deng Agwang, by Peter Lorot, a Didinga SPLA officer from the local area (Johnson 8 This echoes what Leonardi (2013) has documented in relation to settlement and the legitimation of authority in Juba, Yei and Rumbek, though Chukudum is much smaller and more marginal to national political developments. 9 The governance of both cultivation and settlement rights continued to be mediated through cultural norms relating to family and clan membership. And within families, land rights were further mediated through marriage as well as gendered and generational relationships. 10 This account of the history of the SPLA in the area is based on informal conversations and interview data that have been triangulated through an examination of documents and records from the period, in addition to the scholarly historical research cited. 11 It is important to note that many people were very willing to support the rebels in this way, at least initially. To this day, people in the region speak with pride about their role in feeding the SPLA. 12 The IDPs arriving in Chukudum were largely ethnic Dinka from Bor, who had been targeted by a rival faction of the SPLA because Bor was the hometown of SPLA commander John Garang (Johnson 2007;Collins 2008). 2007; Walraet 2008). The assassination prompted reprisal attacks by SPLA soldiers on Chukudum residents. 13 When speaking to me of that time, Didinga residents reported increased levels of harassment and expropriation by the SPLA, lest they provide any material support to Lorot's militia. The remaining civilian population of the town fled to the mountains and to refugee camps across the borders in both Kenya and Uganda. 14 Their flight further freed up land and other resources for the taking, as it meant that residents were not present to contest the military's appropriations. This left the soldiers stationed in Chukudum in the role of an occupying force rather than a liberation army. But the SPLA's control over movement and resources was limited to the lowland areas. Lorot and his militia holed up in the mountains and fiercely defended the uplands from encroachment by SPLA soldiers. The ongoing fighting resulted in a tense standoff, with the Didigna militia in control of the mountains, and SPLA soldiers the valley.
The arrival of the military in the area following the 1991 split in the SPLA, then, was a critical moment when ideas of property, ownership rights and access were reworked, and when the question of which social actors had the authority to legitimize claims shifted dramatically. As Sikor and Lund (2009) argue, the question of who has authority to define and legitimate property and access rights is better understood as a site of contestation and negotiation: '[w]hat is legitimate varies between and within cultures and over time, and is continuously (re-) established through conflict and negotiation' (Sikor and Lund 2009, 7). And as Li (2014, 592) notes, 'assembling land as a resource available for some purposes to the exclusion of others requires a great deal of complex cultural work'. In Chukudum during the war, the SPLA put aside the cultural work of persuasion in favor of the more expedient use of force. But persuasion is a critical factor in determining which assemblages of land cohere and stabilize over time (Blomley 2007;Li 2014). While residents supported the struggle for national liberation, the SPLA was not able to sustain the persuasive element to legitimize their control over land, either in the short term (hence the crisis) or, as I will show in the next section, over the longer term.

Reassembling autochthonous authority
Two interconnected formal political developments set the stage for the return of residents to Chukudum town in the mid 2000s. These were the negotiated resolution of the 'Chukudum Crisis' in 2002 and the entering into force of the CPA in 2005. In the early 2000s, both the SPLA and the Khartoum government came under increased pressure to resolve the decadeslong civil war between the two sides (Rogier 2005;Johnson 2007;Collins 2008). In order to project themselves as the only political force in Southern Sudan capable of governance, the SPLA needed to resolve internal conflicts with local populations and competing armed groups. They did so through a series of local peace conferences and negotiations with rival military factions (Rogier 2005;Rolandsen 2005). 13 The crisis, which many local people argue was fundamentally about resources, quickly took on ethnic overtones, as the assassinated commander was of the Dinka Bor ethnic group that dominated the political leadership of Garang's faction at that time. This caused spillover violence to erupt between the two groups in Kakuma Refugee Camp across the border in Kenya (Crisp 1999;Human Rights Watch 2002). Whether Lorot and his militia were supported by Khartoum beforehand or only turned to them for money and arms after the assassination is still hotly debated.
The crisis in Chukudum was one of several visible and public conflicts at that time, and the SPLA leadership made considerable effort to come to some sort of resolution with the local population, culminating in a formal peace conference convened under the guidance of religious leaders associated with the New Sudan Council of Churches and the Catholic Diocese of Torit. Delegates from all sides attended including civil Sudan People's Liberation Movement (SPLM) authorities, SPLA military authorities, traditional leaders of Didinga and neighboring communities, and representatives of Dinka Bor IDPs resident in the area (SPLM/Didinga Peace Statement 2002). The re-establishment of security was the main focus of the conference; however, commitments made by the SPLA during the conference show that the Didinga participants were very concerned with the economic effects of hosting the rebel military for so long. Many of the provisions related to the material conditions that contributed to the violence, including (1) requiring the section commander to provide food for recruits, (2) a prohibition on trading activities on the part of military personnel, (3) new regulations on road blocks that reduced their number and mandated that receipts be given for any tax collected (SPLM/Didinga Peace Statement 2002). Each of these provisions aimed to limit the necessity, incentives and opportunities for members of the military to appropriate local resources and wealth. 15 The commitments made by the SPLA leadership in the Chukudum agreement gained further momentum as the SPLM/A entered into serious negotiations with the government in Khartoum to end the conflict shortly afterwards, including agreeing to a cessation of hostilities a few months later (Collins 2008). The negotiations culminated in the entering into force of the CPA, signed in Naivasha in January 2005. The agreement set up an autonomous civilian government in the south, and made provisions for a referendum on independence for Southern Sudan in 2011. For many of the displaced from Chukudum, the signing of the CPA was the final signal that there would not be a return to war, and that they could safely return to the area.
However, as a newly created entity, the semi-autonomous Government of Southern Sudan (GOSS) had to build political institutions and a new legal framework from the ground up. There was a 'high degree of legal ambiguity' with regard to basic questions of governance (Deng 2011, 13), particularly in the first few years of the transitional period when returns were high. There were no formal legal provisions for land governance until early 2009 when the Land Act passed in the legislature. While the Land Act recognized customary tenure and granted significant power to traditional authorities with respect to land governance, the creation and dissemination of concrete policy recommendations did not occur until 2011, on the eve of independence (Deng 2011). In relation to land, then, there was considerable scope for action on the part of competing political actors at various scales of government during the transitional period (Badiey 2013).
The formal political developments discussed above created the conditions of possibility for the re-socialization of space in and around Chukudum and the ability of local people to reclaim authority over land from the soldier-traders that had been dominant during the conflict. Yet the success of this re-assertion of local authority cannot be read off from political agreements and new imperatives that emanate from the national and international scale of governance. Instead, it was in the realm of practice that Didinga people challenged the dominant role of the military in relation to land and resource rights. Despite significant wartime land and resource alienation at the hands of armed actors, social norms relating 15 The use of SPLMthe Sudan People's Liberation Movementgenerally refers to the political (rather than the military) arm of the SPLA.
to the legitimacy of autochthonous claims and the responsibility of landholders towards coethnics persisted, smoothing the way for the accommodation of returnees. In this context, the reassertion of returnee authority over land was assembled out of autochthonous modes of legitimation, material practices of inscription on the land (including inhabitance, agricultural and pastoral production, the collection of environmental resources) and the production of insecurity.

Practices of inhabitance
One way that local people were able to assert their rights relating to settlement and land use was through their practices of inhabitance. For an illustration of how this unfolded, I turn to Nadai, 16 an elder in her 60s or 70s who I spoke with in January 2010 outside of her home. Nadai had left Chukudum after one of the aerial bombardments by the SAF in the mid 1990s. Having fled to Uganda during the Sudan's first civil war (1954)(1955)(1956)(1957)(1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972), when the conflict returned to the area in the 1990s, she chose to move to the mountains rather than to flee across an international border. She returned to Chukudum from the mountains shortly after the signing of the CPA in 2005. She recounted: 'So when I came down there weren't many here. Just some soldiers, so I chased the soldiers [away]. Those that are here now, they came with me from the mountain'. Nadai, who was widowed and whose children were now living in Khartoum, Juba and the United States, made her return accompanied by others from the mountains. Nadai's family had claim over a large garden (mana) for cultivation near the central area of town. When she returned from the mountains, she had been able to successfully evict soldiers who had been living on her land. However, part of the reason that she was successful in this effort was because she was accompanied in her return by others who were interested in re/settling in the area: Nadai: I used to live just there, behind that fence. This whole area here is my plot. All the people who are staying here, they ask me for permission to build. Yes, they ask me. Author: Do you ask anything from them in return, like rent? Nadai: No, I give it to them for free. They ask, and I am to give it to them. Author: Are they relatives, then? Nadai: There are some of them that are relatives and there are others that need a place to stay.
As she spoke Nadai gestured to a large area, comprising several distinct family compounds (halang, p. halan'ngiook), indicating the boundaries with reference to features in the landscape, pointing to specific trees, fenced yards and the nearby streambed. It is unlikely that Nadai's efforts to reclaim her family's land from the soldiers living there would have met with success if she, an elderly woman without a husband, had attempted to do so on her own. Nadai's authority over the land was thus tied to her willingness to accommodate those without prior land claims in town on her own plot. By allotting parcels of her own productive land to distant relatives and neighbors from the mountains for habitation, Nadai catalyzed the building of several residential compounds, complete with tukuls, cooking shelters, cleanly swept yards and, as she mentions, fences. Inhabitance did not just require the permission to erect shelters on a particular plot but also set into motion a whole set of practical and material relations with the broader landscapefrom collecting thatching grass and digging clay for bricks or plastering, to harvesting bamboo and 16 All names and identifying information have been changed to preserve confidentiality.
hardwood poles for construction. In performing these mundane actsin acting as if Nadai's land rights were validthose given permission to settle on the land materially manifested her authority. At the same time, these manifestations were also more legible to the soldiers occupying her lands than the traces of other forms of inscription that remained in the landscape, and which were nevertheless immediately recognizable to Didinga (e.g. cleared land, living fences, the graves of ancestors, mango trees planted in the 1970s). Nadai's authority to permit and assign residence rights then emerges through the practices of inhabitance of the new neighbors she settled with on her lands. Notably, Nadai is not a chief or a clan authority but rather only a land rights holder; however, in practice her permission was sought and secured as a way to legitimize new households' claim on lands for settlement in the center of town, thus at the same time legitimating Nadai's own authority over the area. Following early movers like Nadai, the pace of returns picked up through time, peaking with formally organized voluntary assisted repatriation (VAR) for refugees and assisted voluntary return (AVR) for IDPs in 2007-2008 from both Uganda and Kenya, and then continuing at a steady, but lower, pace through the study period. 17 In contrast to expectations of both the United Nations Refugee Agency (UNHCR) and the encouragement of newly established county-level civil authorities which envisioned returnees settling together in a newly designated area on the southwestern outskirts of town, returnees instead made their own arrangements about where to settle and build their homes. 18 When repatriating refugees arrived in town, their relatives by and large welcomed them home.
Many return migrants mirrored Nadai's stories, recounting their own negotiations with kin and friends to allow them to erect a home on land close to the central area of town. Cultural practices in the area favor patrilocal, multifamily residence, so returnees considered building their home on land near parents, brothers or in-laws both appropriate and desirable. Lotiman, the headmaster of one of Chukudum's primary schools, explained that when he returned from Kakuma in 2009, he had no problem finding a place to build his home: 'I got them [my family] when they were already settling. So I squeezed to put my house on the same plot' (Interview, Chukudum, 13 January 2010). Another young man returning from Uganda with United Nations (UN) assistance had a similar experience. He told me that he had no difficulties in finding a place to settle in Chukudum: 'I have family land. When I came back, my mom was here. So I was able to go back to my family's lands' (Interview, 08 January 2010, Chukudum). Having family land rights that pre-dated the conflict, and family members who had returned early in the process, then, was clearly an advantage to those returning to the town.
Yet the lack of historical claims did not bar other Didinga returnees from gaining access to land for housing within the town. Returnees originating from outlying villages first sought out 17 Reliable numbers on returnee flows are difficult to find, as many who returned to this community were not included in official counts by assisting agencies or the local arm of the South Sudan Relief and Rehabilitation Commission, the responsible government agency (but for an overview of returns across Southern Sudan during this period see IOM 2010). In a survey I conducted of local livelihood practices and migration history in June 2010, 75 of 79 people reported that they were displaced from the town at some point during the second civil war. Of the remaining four people, two reported that they were IDPs from other parts of Southern Sudan who settled in Chukudum in the1990s. 18 Prior to beginning of the VAR program, the UNHCR had funded the drilling of several new boreholes and the construction of both a primary and secondary school in this area, which had previously been the bush. But the area was viewed as insecure because it was sparsely populated and on the western outskirts which were more susceptible to raiding from across the valley. Despite the investments in infrastructure, very few returnees chose to settle in this area because they had other options closer to town. close family members for land for housing; when that failed they looked to more distant relatives who had returned previously to town. However, access to land for housing was also mediated through social ties that extended beyond family and clan-based affinities, to friendly and neighborly relationships developed in the refugee camps and in the areas to which the internally displaced migrated. Most returnees and Didinga migrants to Chukudum who did not have historical land rights in the immediate area could obtain access to land without any costs through informal channels. While these kinds of ties were not typically the first recourse of returnees, they could be called upon when people lacked or were unable to capitalize on family ties, including (some) widows and orphans. This was the case for Natemo, a young man who told me 'I just went to see where I could stay. So like I said, I am living with a friend. It is not in the same place as I lived before, it is some meters from there' (Interview, 11 January 2010, Chukudum). In this case, Natemo's friend offered him hospitality, allowing him to erect a small house in their compound.
In contrast to many other instances of population growth due to return migration, residents of Chukudum by and large viewed the growth of the town in a positive light. 19 The growth of the area from a large village into a more densely populated town signaled, at least for some, the possibilities of developmentin the form of concrete material improvements and improved access to services. Lokibe, another elder whose family has long lived in Chukudum, explained: Lokibe: People have come back and now there are many. So it is hard to own a big plot, and even more so for cultivation. Author: Do you give land to others? Lokibe: I do not give the land, but maybe if some one wants to build a small tukul [traditional thatched roofed cottage] then yes I allow it … . We are happy when we see many people in town. We have some development, and even some organizations are here to help. We need more people to come to town so we can get development. Even if you are white from America or black from South Africa, we need you to come. Because this one is a town and we need development.
Lokibe expresses a willingness to accommodate Didinga return migrants on land over which he had authority. Beyond that, he also expressed a welcome to any person from anywhere who might come and contribute to development in the town. Like many others I interviewed, Lokibe links population growth directly to development, and sees this change contributing to attracting NGO programs to the town. 20 And as the quotation 19 For a discussion of more conflictual relations relating to town growth in South Sudan, see Leonardi (2013) and Badiey (2010Badiey ( , 2013. Chukudum's small size and economic and political marginality may account for the differences in attitudes toward population growth compared to these other examples. 20 The positive attitude toward population growth could be traced to three main factors, one political and the others based on past experiences. First, in his development agenda, John Garang emphasized 'taking the town to the people' instead of bringing the people to the town (Copnall 2014;Motasim 2014). The slogan, often echoed by people in Chukudum, was a commitment to bringing the hallmarks of developmentroads, schools, medical servicesto rural areas that had not previously had such services. However, the statement is also a clear manifestation of the strong association between 'the town' and development in the minds of many Southerners. On a more practical level, being a town meant attracting infrastructure projects as well as NGO programs that provided critical support in the education, agriculture, health and sanitation sectors. Second, many of the elders and political leaders had lived through a previous return migration after the first civil war in the 1970s, and had witnessed or been part of changes (especially in terms of new agricultural techniques) that had a lasting positive effect food security and wellbeing in the area. Finally, for some of the suggests, there was a recognition that towns inherently draw people from other areas and that space should be made for them as well.
According to interviews with county officials and town residents, SPLA soldiers stationed in Chukudum, faced with returnees claiming the lands they lived on, initially retreated to the market area, but relations between the civilian and military population remained tense. 21 In response to ongoing harassment in the market, the new civil authorities at the county level representing the autochthonous Didinga successfully pressured the military to move the barracks area even farther from the town center. Furthermore, the cycling through of new regiments after the CPA meant that many of the soldiers who were 'chased away' by returnees did not have a longstanding presence in the area. And, with the transfer of the barracks to a location on the outskirts of town, traders who had merely a loose association with the military had less to gain through physical proximity with the military. Many chose not to move to the barracks. With this clearer distinction between the military and traders operating in the area, the autochthonous population was more willing to allow traders to retain their stalls, storehouses and residences in the market. 22 Tania Li (2014, 591-92) draws attention to the 'complex cultural work' that is necessary to stabilize assemblages of land control and the ways in which this work so often depends on moral arguments about the affordances of land and the propriety of certain categories of people deriving profits from it. In Chukudum, land for housing was generally seen as a moral obligation, something that should be afforded to allco-ethnics, traders and IDPs alikethough the terms of this affordance differed for each group. 23 When asked, both those on the giving and receiving end of these arrangements denied that land rights holders demanded any payment for land used for housing. Rather, there was an expectation that, at least with respect to housing, land rights holders should make land available to anyone settling in Chukudum, even those without historical ties to the area. However, this picture changed considerably when land was put to productive use. Thus, it was in relation to the productive uses of land that autochthonous modes of legitimating land claims were most exclusionary. The return to autochthonous modes of legitimation in the county hinged on the question of who could rightfully derive profit from land and land-based resources. This meant that while soldiers, traders and internally displaced people, and even new co-ethnic migrants living in Chukudum, were able to gain access to land for settlement, they were largely not able to make similar claims on land for subsistence or commercial agricultural use, pastoral production, or for the profitable disposal of environmental resources. younger generation, time spent during the conflict in cities, and densely populated refugee camps, shifted the norms around settlement patterns towards greater concentration. 21 It was not feasible to interview rank and file soldiers or the military leadership at the time of my research as all matters relating to the army, its movements and location, and its relationship with local communities were extremely sensitive during the transitional period. 22 Traders I spoke with were open in their discussions with me and emphasized their primary concern was that they be allowed to continue to run their businesses and maintain control over the plots on which their shops, stores and homes were built. The traders had also organized an association which regularly liaised with Didinga county authorities on concerns related to commerce and interethnic tensions (when they arose). 23 Land rights holders still had a ranked preference in terms of who they would allow to settle on their own lands, with close family members and in-laws coming first, clan members and neighbors from displacement somewhere in the middle, and unknown co-ethnics and traders from other ethnic groups or countries coming last. Soldiers were not part of this ranking, as the community and autochthonous civil authorities ceded land in the bush for a barracks, which was where soldiers therefore belonged.

Practices of production
As I have shown, collective practices of inhabitance helped to give weight to informal autochthonous authorities and produced new assemblages of land governance within the town. But habitation meant more than just settling on a plot in town. In going about the business of settling in, town residents engaged in a host of productive practices on the landcollecting grass for thatching and firewood, cutting bamboo and hardwood poles for constructing their homes, clearing and cultivating land for their subsistence, and burning grasslands to maintain pasturagethat had even broader impacts on land and resource governance in the immediate vicinity of the town. Through socially regulated practices of cultivation, pastoral production and natural resource extraction, Didinga residents were able to extend their own authority over land to many areas that had formerly been under the control of the military forces, heavily mined or perceived as no-go areas due to insecurity. However, productive use of the land was heavily restricted, only granted to those with historical ties to the area based on family and clan membership. 24 As previously mentioned, rights to use land for cultivation, pasturage and resource extraction in much of Southern Sudan are relationally defined (Driberg 1939;Allen 1996;Rolandsen 2009;Johnson 2010). These kinds of relational property rights regimes vary in the specific modes of regulation and forms of authority they rely uponbased on cultural identity and geography. But regardless of this variation they are, nevertheless, widely recognized and respected by the highly diverse population. Where land rights are contested in Southern Sudan, fault lines fall along ethnic boundary zones and center around the questions of which collectivity's claims should prevail (Rolandsen 2009), or of in-migrants' abuse of local hospitality through, for example, profiting from the resale of lands to which they have only usufruct rights (Badiey 2013;Leonardi 2013;Motasim 2014). Broadly shared ideas relating to the rights afforded through autochthony and the corresponding moral obligations and limits of hospitality towards strangers/guests undergird these kinds of conflict.
As hinted by Lokibe (quoted previously), gaining or retaining access to land for cultivation in or near the town was more difficult, even for those with prior claims, due to the increased pressure for residential lands that accompanied return. But outside of the town, Didinga residents were able to effectively reassert control over land and other resources through a variety of productive practices on the land. If not everyone was able to gain access to garden lands near town, Didinga returnees did have the potential to lay claim to land somewhere for cultivation through existing clan or familial ties. 25 In contrast, it was very difficult to acquire lands for extensive cultivation for in-migrant IDPs and tradersthat is, for those who lacked the requisite ethnic affiliation to claim productive land through an appeal to autochthony.
At the time of my research, only a single family among the Dinka Bor IDPs living in town had been given permission (by a Didinga neighbor) to use land for exclusively agricultural purposes. Ayen, a Dinka Bor woman of 40, had arrived in Chukudum, by way of 24 These claims were mediated through cultural norms relating to family and clan membership, in particular those around marriage, gender and generational relationships that largely determine land distribution and rights to other kinds of resources. For example, women marrying in to a Didinga family were accorded land rights through their husband's clan. 25 Traveling to distant gardens was not uncommon, as many families maintain multiple fields or gardens in different agro-ecological zonessome a day's walk awayto maximize crop diversity and spread risk. Kakuma, in 1997. As a woman raising seven children alone, she worked for many years as a cleaner at the hospital for a small wage. Her income, however, was not sufficient to support her children and she had asked to borrow land from one of her Didinga neighbors to supplement her salary. Her access to a garden was a recent development (when we spoke it was her second year cultivating), which she credited to her friendly relations with her neighbors. Of the Dinka Bor IDPs that I interviewed, Ayen was the only one to report that she hosted and participated in collective work parties, an important practice through which Didinga mobilized labor at key moments in the agricultural season. Her widow status and close personal relationship with neighbors, as well as her willingness to engage in the social and labor exchanges common to Didigna culture, made her access to land possible. As noted, Ayen's situation was an exception. The remaining Dinka Bor residents of Chukudum were generally limited to planting food crops on their residential plots, in the fairly limited space directly adjacent to their homes. While this was helpful as a supplement to purchased grain and vegetables, the small size of the plots meant they were generally not productive enough to sustain a family exclusively. The restriction on agricultural production extended to migrants from other backgrounds as well. With the exception of small home gardens, it was uncommon for migrants to build the kind of confidence and trust with local landholders to enter into these kinds of land-borrowing relations.
However authoritative the culturally mediated claims over gardens and fields were, it must also be noted that the extensification of cultivation into areas of prior use was also deeply tied to pastoral practices. 26 Pastoral production is central to social life in much of Southern Sudan, including in Budi County. Though not many cattle were kept in Chukudum town itself, the lowland areas in the vicinity were productive grazing areas that were protected by well-armed young men, mostly from neighboring villages. The social and physical landscape showed traces of pastoral production throughout. Cattle were routinely run through the town center when pastured nearby, accompanied by cohorts of young men armed with bows, rifles and the occasional Kalashnikov. Further traces could be seen in the patches of grasslands burned at certain points in the season to regenerate new growth and rebalance species make-up for superior nutritional composition.
But perhaps the most powerful mode of inscribing grazing rights on the landscape was through violence, specifically through cattle raiding and the production of other forms of insecurity. In Southern Sudan, tending herds of cattle and other livestock is a very risky endeavor. Cattle keeping in the area is fraught with violence, as inter-community cattle raiding is a common and deadly practice that regularly claims lives (Schomerus 2008). Owing to the prevalence of raiding, Southern Sudanese cowherds are well-armed bands of young men who do not fear to defend both their cattle and their grazing areas. Given the potential for violence, non-Didinga town residents did not even consider raising livestock in the area. To do so risked prompting a cycle of attack and counter-attack that might escalate to communal violence that, once unleashed, could threaten the stability of the nascent state (Schomerus 2008). Additionally, given the prior conflict between Didinga residents and the largely Dinka Bor SPLA soldiers during the Chukudum Crisis, both groups recognized that the entrance into the pastoral sector by Dinka Bor town 26 Persistent raiding in the months following the failure of the 2009 rainy seasons limited the willingness of people to live and engage in cultivation in lowland areas of the county during the next rainy season. This area was known to be fertile, but its proximity to the western border areas where raiders from neighboring tribes were particularly active meant that several families chose not to cultivate there that season. However, when the security situation had improved, many lowland areas were again open to cultivation. residents would likely provoke outright violence. It was clear from conversation with both the Dinka Bor IDP population and Didinga residents of town that members of the Dinka community were effectively barred from raising livestock in the county. The Dinka displaced to Chukudum in 1991 had arrived without their substantial herds, and this set the groundwork for contemporary restriction on livestock keeping by in-migrants. 27 That said, this did not mean that Dina Bor traders living in town did not invest in livestock; many used profits earned in their shops to add to family herds in their home areas.
The risk and violence associated with cattle keeping had knock-on effects that impacted access to areas for other productive activitiesfor cultivation and the collection of environmental resources including fuel wood, construction materials, thatching materials and wild foods. While long-time Dinka Bor residents were not prevented from gathering fuel wood and thatching for their own use, collection for re-sale in the market was not considered appropriate. Additionally, the persistent insecurity associated with cattle raiding contributed to the IDPs' hesitancy to risk spending time in the 'bush' collecting environmental resources such as fuel wood, thatching grass and poles for construction. As they themselves were not engaged in pastoral production, they were further removed from the day-to-day information sharing about where cattle might be pasturing, and the level of risk of a raid or counter-raid based on ongoing events elsewhere in the county. Fear, rather than outright prohibition or targeted attacks, kept migrants from gathering environmental resources.
In this context, land control emerged in the flux of the embodied productive practices of residentsin their labor upon garden plots, their transit through forests and woodlands in search of fuel and other environmental resources, their protection of livestock, and the burning of pastures to encourage new growth for their animals. Didinga residents often had superior knowledge of the physical and social landscape in the region that allowed them to better judge the risks associated with venturing to, and laboring upon, particular areas, whether for grazing, cultivation or fuel wood collection. In particular, they knew which areas were open to each activity, whose claims were widely accepted or hotly contested, and which areas were prone to raiding from neighboring communities. Their knowledge and embeddedness in the dense social networks through which new information traveled helped them to better navigate the endemic insecurity and, perhaps at times, even leverage it to their advantage. These dynamics worked in their favor in reclaiming land and authority over resource use in the area. These same dynamics also worked against soldiers, traders and other non-Didinga migrants living in the area, as they often lacked local knowledge of risks and had more limited access to the networks through which this information typically flowed. 28 27 This stands in contrast to Yei, another area (in Central Equatoria) displaced Dinka Bor settled after the 1991 split, where many arrived with their cattle. There, the struggles over land and grazing rights have become quite contentious, and occasionally violent (Sluga 2011). 28 While some of the Dinka Bor traders had lived in the town for more than a decade, much of that time had been characterized by both low-level and open conflict with the autochthonous Didinga community who retained control of the nearby mountains. The tensions meant that when they traveled outside of the town, they stuck to well-worn transportation corridors. Additionally, as per the terms of the agreement relating to the Chukudum Crisis, soldiers who had a history of expropriation and abuse of civilians were transferred to other areas (SPLM/Didinga Peace Statement 2002). And after the signing of the CPA in 2005, a significant minority of soldiers 'self-demobilized', making the decision to return to their home communities (Munive 2013, 21). This cycling of personnel meant that soldiers living in the town at the time of the research had limited familiarity with both the physical and social landscape in the vicinity. The reclamation of authority by Didinga returnees in Chukudum in the late 2000s relied on an understanding of ethnically defined rights to land and resources that is common throughout Southern Sudan (Rolandsen 2009). The specific sets of norms governing land use vary across locations, but in rural areas the legitimacy of autochthonous authorities (whether embodied in government chiefs, clan elders or respected family heads) to determine how and by whom land is used is not typically challenged. In Chukudum this has taken the form of closing off a set of spaces, resources and practices to those perceived as outsidersincluding agricultural and pastoral production, and the productive use or re-sale of environmental resourcesechoing traditional distinctions made between authorities with responsibilities over settlement versus soil (see Driberg 1939). Because the reassertion of authority over the productive spaces of Budi county is based on norms relating to territory and ethnic belonging that are widely held in Southern Sudan, the exclusion from productive land and resources was not a point of overt tension at the time of my research. Yet while the authority of Didinga to determine settlement in their own areas was seen as legitimate (within limits), the recognition of this authority cannot be extracted from the history of past violence and the unstated threat of ongoing violence, which operated as an often-unspoken foundation for the success of this process.
While even outsiders were granted the affordance of habitation and of the rights to engage in trade, these were contingent on respecting limitations on the productive and profitable uses of land. As I have noted, local practices of inscription on the land were part of a broader assemblage that included shifts in SPLA political priorities towards governance and away from military rule, the transfer of abusive commanders, and more generally the cycling of troops in the post-conflict period. These factors also articulated with the history of fierce resistance of the local population to the exercise of arbitrary authority by the military during the Chukudum Crisis, which played a part in the relative willingness for soldier-traders to give way in land disputes with former residents. For traders it was more important to maintain the ability to conduct business in the area than to farm, or raise cattlean affordance that might be threatened if they appeared to push too hard against local authority over land rights.
Rather than government policy, the attitudes around access to land within the town were governed by underlying cultural understandings about who had the legitimate authority to allow or disallow settlement, and also about the moral responsibility of land rights holders to give room in town for housing new arrivals. Land for both housing and cultivation was assembled through the embodied material practices of returnees, who reclaimed authority (or reconstituted it through their recognition of the legitimacy of others' claims) in part through building homes and fences, and encouraging or allowing other returnees to do the same. Indeed, the reassertion of historical rights by Didinga returnees against those of the military was effected not through a family's ability to exclude all others but through their ability to determine which others should be allowed to use the land, and for what purpose.

Epilogue: emergent fault lines
An assemblage is the product of multiple determinations that are not reducible to a single logic. The temporality of an assemblage is emergent. It does not always involve new forms, but forms that are shifting, in formation or at stake … assemblage implies heterogeneous, contingent, unstable, partial and situated. (Collier and Ong 2005, 12) Autochthonous claims based on clan membership could only be pushed so far. Just as clan-and family-based authority to determine settlement and production in Chukudum stabilized, new political entrepreneurs drawing on emerging urban sensibilities and diasporic rationalities threatened to unravel the assemblage. In 2009, the land law was passed by the national legislature, recognizing the collective rights of local populations to their historic lands. But while the law was being debated and finalized in Juba, local administrative actors pursued other modes of governing land, taking inspiration from their more urban neighbors (notably the state capital, Torit).
A land survey was planned, and partially executed. A month or so of considerable activity and even more considerable concern followed. Rank and file members of the SPLM (all but one of them Didinga) moved through the town, bisecting family compounds to stake out the location of broad avenues, future roundabouts and individual plots, with bamboo pegs. They spray-painted a red X on every structure that violated the new order. Registration forms were printed and sold, for a profit. Unsure of the future, some town residents raced to inscribe their ownership of ancestral properties in the register, others ignored the effort as yet another government scheme aimed at 'scooping' up the wealth of the people and still more worried about what would happen if/when they couldn't scrape together the required funds. The county commissioner hijacked a public meeting to defend the efforts, justifying this new mode of land governance through appeals to development ('Chukudum will become like Nairobi') and references to the liberation struggle ('your ancestors did not liberate this land from the Arabs'). In response, chiefs and elders stood up to offer subtle speeches, voicing their unqualified support for the registration in the same breath as they made scathing critiques of the poor timing, unclear process and high fees associated with the effort.
A few months passed, and no substantive change to settlement patterns and land use occurred. The pegs marking out the grid of future roads had been eaten by termites or plowed under for the new season's planting, and the red Xs hidden under newly smeared walls. The governor sacked the commissioner, for unrelated reasons. The surveyor who had directed it all left town with his maps because he had never been fully paid for his services. No one knew where the money collected went. It has now been more than five years since the land survey controversy took place. The question of land registration for the town continues to surface regularly, always amidst some level of worry and a great deal of contestation.
I mention the failed survey here to draw attention to the contingent nature of the assemblage through which autochthonous authority over settlement and production on the land was instantiated in Chukudum, and to recognize its inherent instability. This is perhaps not so surprising in the context of ongoing and significant population movements, processes of state consolidation, contentious ethno-political balancing acts and the widespread proliferation of weaponry. Yet in the midst of such significant shifts, returnees were able to push back against the military's economic control of their land and resources by acting as if their claims mattered. These autochthonous claims gained traction as a critical mass of returnees arrived in town through a collectivity of embodied practicesthat is to say, they were recognized because they involved the material shaping of the land through the construction of homes, the sweeping of compounds, and the clearing, planting and weeding of gardens.
In this paper I have drawn attention to the practices that produce land control as one element in assembling autochthonous authority and pushing back against military appropriations. This approach offers a way to see how land is drawn into particular moral orders and modes of projecting authority. It also shows how everyday acts can be an effective counter to the attempts at accumulation and dispossession driven by more powerful actors. It can be tempting to read such acts as silent resistance (Moreda 2015) or a Polanyian re-embedding of the economy within social life. But, as Hall et al. (2015) note, 'what happens on the ground is more varied and complex'. In this case the ability to create new facts on the ground stemmed from a particular assemblage of political imperatives, autochthonous norms and practices which were for the most part accepted by the various actors, and took place largely without open conflict. The land assemblage that stabilized in the transitional period combined broadly held social norms generally viewed as legitimate (the rights of first comers, the moral responsibility to grant all the affordance of land for inhabitance) and local land governance traditions (where discretion relating to land distribution is ceded to families) with material practices anchored in subsistence. This assemblage consolidated within a weak institutional context characterized by a considerable lack of legal clarity that was emblematic of the transitional period. But each of the elements contributing to the assemblage was undercut (or buttressed) by the potential of violence. While the territorialization of identity and authority through control over resources did, in this instance, protect the claims of comparatively marginal people, this same process also points to the entrenchment of an exclusionary politics of ethnic mistrust in the Southern Sudanese context. The result was a tense and unstable accommodation with the potential to breakdown as elements shifted and new power arrangements consolidated.
The arrival of large numbers of returnees to Southern Sudan during the transitional period did not generate the level of conflict and instability that was expected with respect to access to land. Instead, individuals, families and communities in rural areas drew on existing customary norms to allocate lands for the returnees, often with little conflict. This suggests that community-led reintegration of the displaced can succeed, and that such ad-hoc arrangements may be more practicable and long-lived than formal policies and plans. But the lessons drawn from the Chukudum case are by no means unequivocally positive. The evolution of land governance in Chukudum points to fundamental questions regarding the nature of governance and political authority in rural South Sudan. In the day-to-day management of rural lives and livelihoods, formal state structures and institutions remain remote, mistrusted and sometimes ignored. Instead, rural people have turned to more informal ways of governing social life, based on historical practice or (neo)traditional modes of authority, enforced through a combination of social sanction and vigilante violence. These modes of governance strongly favor autochthonous claims, disadvantaging those considered marginal in their own communities, internally displaced people and migrants from other ethnic groups, as well as decommissioned soldiers and their families. In the transitional period, the collective goal of achieving independence helped to hold in check the potential for violence bound up in this assemblage, as diversely situated actors feared that inter-ethnic conflict would derail the referendum. But with the achievement of an independent state, the checks on violence weakened. As we have seen in the turbulent years since 2011, tensions over governance based on territorialized notions of ethnic belonging remain close to surface, offering fertile grounds for mobilization by elites in their struggles over political power.