Pragmatic Mobilities and Uncertain Lives: Agency and the Everyday Mobility of South Sudanese Refugees in Uganda

This article investigates the pragmatic, everyday journeys of South Sudanese refugees in northern Uganda’s Palabek Refugee Settlement through a mobilities-focused analytical

regional historical perspectives, we demonstrate that the movements of South Sudanese refugees disrupt simplistic discourses of return and repatriation. Using the context of ongoing movements among people displaced by the Second Sudanese War  and especially the recently concluded South Sudanese Civil War (2013-18), we argue that the complexity and diversity of our interlocutors' varying mobility practices are particularly powerful manifestations of agency amid unpredictability. That is, having personal histories of 'enforced domination, lived uncertainty, and extreme collective suffering' (Finnstro¨m 2008: 189) and now living in a situation of 'existential uncertainty bordering on crisis' (O'Byrne 2017: 93), they seek to use their-often fragmented-mobilities to at least allow for the possibility of greater existential control (and therefore certainty). We consequently understand these sometimes rather commonplace movements to be essential parts of everyday-yet crucial-practices to gain and maintain personal and collective control in otherwise uncertain contexts.
For the United Nations High Commissioner for Refugees (UNHCR)-the UN body mandated to protect, manage, and assist the world's growing refugee populace-'repatriation' refers to the movement of refugees from their country of exile to that of their origin. However, in many instances, it is used as a form of discursive shorthand to specifically mean 'voluntary repatriation': UNHCR's preferred, primary (and often unproblematized) 'durable solution', in which a refugee willingly and permanently returns to their country of origin, ending their experience of 'the refugee cycle'. Nonetheless, despite widespread repatriation across the globe in the last several decades, little is known about refugees' experiences of multiple displacements or the diversity of their later movements and mobilities (cf. Monsutti 2008;Grabska 2014: 205).
Mobility has become an essential part of the everyday and life-long trajectories of many South Sudanese over the last several decades. This is especially true for those living close to neighbouring countries or in fragile climate-effected or conflict-ridden environments and was most apparent immediately following the Comprehensive Peace Agreement (CPA) that ended the Second Sudanese War in January 2005. The repatriations which followed did not stop returnees from moving back and forth across the region, however, and South Sudanese continued to move throughout both the brief period of peace and over the course of the recently concluded conflict. Recognition of the complexity surrounding these movements is important, lest conflations between return and repatriation continue to inform local manifestations of the (inter)national refugee response. In fact, as noted by Long (2010: 36;cf. Monsutti 2008: 59), not only do many returned refugees continue to be mobility after their displacement ends but such continuities should even be expected. After all, as Ramadan (2013: 70) has argued, 'in the absence of a durable solution to refugee status, migration and transnational networks may represent an "enduring" and effective livelihood strategy'.
This article explicitly builds upon previous work among (South) Sudanese refugees in Uganda during or immediately following the Second Sudanese War (cf. Allen 1996;Kaiser 2006Kaiser , 2010Hovil 2010). Almost all our interlocutors have been refugees at least once before and, despite the recent violence in South Sudan, 748 Ryan Joseph O'Byrne and Charles Ogeno many continued to move across the border as they had throughout their lives. For many, 'the leaving and coming back has been constant' (Monsutti 2008: 59) and, like the Afghanis with whom Monsutti worked as well as the South Sudanese discussed by these earlier scholars, many of our interlocutors have moved much the same during times of war as they did in periods of peace. Given such complexity, knowledge about the current and former lives of refugee individuals is crucial.
The factors impacting movement in this region have always been multifaceted, determined by a diverse array of personal, familial, and communal concerns and local, regional, and international contexts (cf. Kaiser 2010: 54-55). The dynamics of many contemporary South Sudanese's movements manifests 'a continuation of the mobility practices of earlier generations' (Bjarnesen 2016: 61) as well as individuals' own prior journeys (cf. Hovil 2010: 12-14;IRRI 2018b: 4). Although some of these are no doubt economic-based or livelihoods-focused, such rationales do not account for the full range of mobilities (cf. Monsutti 2008: 58-59).
This does not mean livelihoods are not important; clearly they are, and many of the back-and-forth movements in this region are intimately connected with obtaining not only food-without a doubt the most important consideration for most of our interlocutors-but other aspects of individual and familial continuity: jobs, money, healthcare, schooling, spouses, and so on. However, as this short list already demonstrates, the everyday political economy for many refugees is wider than standard humanitarian conceptualizations of 'livelihoods' allow, encompassing an extensive range of sociocultural and relational connections.
As Allen (1996: 7) has argued, failure to account for diversity of movement would be to 'give a false . . . impression that one is dealing with a simple and wellcircumscribed event rather than with an untidy process, involving multiple, and sometimes overlapping migrations in both directions'. Nonetheless, although many of the people we spoke with undeniably take part in multiple, interconnected movements-across the border, throughout Uganda, or beyond-those undertaken by many refugees frequently remain irregular and unpredictable, largely because of a multitude of factors over which they have little control or certainty.
We suggest that the irregular and uncertain mobilities so characteristic of South Sudanese refugees in Uganda not only problematize the one-dimensional direction and duration of the term 'repatriation', but also many more nuanced concepts such as circular' (Kaiser 2010: 52), 'oscillating' (Allen 1996), or 'pendular' (REACH 2018a. Although these more nuanced concepts do correctly highlight spatial and temporal multiplicity, they tend to assume one or more underlying regularities, as well as a certain amount of existential certainty and predictability. Accordingly, we define the 'pragmatic mobilities' of our interlocutors as 'the experience and practice of multiple, distinct yet interconnected mobilities, despite trying times and unknowable circumstances', thereby attending to the sometimes fractured (dis)junctures between journeys as well as to their full temporal and geographical scope. We argue that this conceptualization highlights the inherent complexity of refugees' mobilities, allowing for movements which take place over Pragmatic Mobilities and Uncertain Lives 749 weeks or seasons as much as lives, as well as to additional places beyond either an assumed final destination or a supposed point of origin. Significantly, it emphasizes that in even the most trying existential circumstances, some refugees can still engage in directed if sometimes spontaneous and sporadic movements. Moreover, it allows that these movements are illustrative of refugees' agentive capacities more generally, as well as connecting the underlying imperatives of these movements to that more general human endeavour to act within and upon the essential uncertainties of one's life, no matter its surroundings (cf. Finnstro¨m 2008).

Background: History, Location, Methods
Uganda and Sudan/South Sudan have been hosting each other's refugees since at least the 1950s, with large numbers of people from both countries spending significant periods in the other (Akol 1994: 77-78;Allen 1996: 226-228). The first significant repatriation in southern Sudan took place when thousands of refugees returned after the Addis Ababa Agreement ended the First Sudanese War in 1972. At the time, this was 'one of the largest repatriation operations on the African continent' (Akol 1994: 78). Other large repatriations followed the signing of the CPA in 2005 and the creation of the independent country of South Sudan on 9 July 2011 (Hovil 2010: 12-14;Kaiser 2010: 45-47). These returns were by no means temporally or spatially simple, however: some South Sudanese refugees refused to repatriate while many who did soon fled again, becoming refugees once more.
Most currently displaced South Sudanese derive from the ongoing conflict which erupted in December 2013. This conflict has killed hundreds of thousands (Checci et al., 2018) while nearly 4 million people have been displaced, half of whom are children (OCHA 2016: 2). By 2016, South Sudan was the world's third largest refugee crisis and the largest in Africa (OCHA 2016: 1). There are now over 2.2 million South Sudanese in exile, 860,000 of whom are hosted in Uganda (UNHCR 2019b), the world's third largest refugee-hosting nation (UNHCR 2019a: 6). Nearly all Uganda's refugees live in one of 30 refugee settlements in eleven districts (REACH 2018b: 2, 7; UNHCR 2019a: 6), although there are small numbers of self-settled, urban refugees.
Underpinned by the rights enshrined in the 2006 Refugees Act and the 2010 Refugee Regulations, Uganda's legislation has garnered significant attention for its allegedly refugee-friendly nature. Under these provisions, all South Sudanese in Uganda have prima facie refugee status (UNHCR 2019a: 7) and are entitled to the same basic services as citizens, as well as some freedom of movement and rights to employment and business ownership (UN and WBG 2017: 2). Such rights are often practically unavailable, however, either due to a lack of settlement-based services (Kaiser 2006: 601;UNHCR 2019a: 6); because basic humanitarian assistance targets rural settlements to the detriment of urban refugees (IRRI 2018b: 7); because of the local interpretation and implementation of law (IRRI 2018a: 4); or because many settlements' rural surroundings suffer from chronic underdevelopment, effectively encamping residents (IRRI 2018a: 4;Kaiser 2006: 601, 620).

Ryan Joseph O'Byrne and Charles Ogeno
However, we will argue that these problems may not be quite so relevant for residents of PRS, in the northern District of Lamwo (Figure 1), as those in some of Uganda's other settlements. The newest of Uganda's refugee settlements, PRS was opened on 12 April 2017 following an unexpected influx of South Sudanese refugees east of the Nile from late 2016 (UNHCR 2017: 1). As PRS chanced to open almost immediately after a Sudan Peoples' Liberation Army/In Government (SPLA/IG) attack on the Acholi-speaking community of Pajok in Magwi County on 3 April 2017, Pajok's refugees became PRS's first residents, with over 5,000 settling in PRS by 15 April (OPM and UNHCR 2017: 1). Although it is planned that PRS will eventually house 100,000 people (UNHCR 2017: 1), the population at the end of fieldwork in November 2018 it was approximately 34,000 while it was 52,022 by the end of 2019 (UNHCR 2019b). All residents of PRS are South Sudanese, primarily from Acholi-and Lotuko-speaking areas (UNHCR 2018). Especially prevalent are those from the Acholi-speaking community of Pajok.
This article is grounded in 12 months ethnographic fieldwork undertaken over 2017 and 2018. Our research proceeded from the methodological premise that 'direct dialogue with others, afford[s] opportunities to explore knowledge . . . as an intersubjective process of sharing experience, comparing notes, exchanging ideas, and finding common ground' (Jackson 1996: 9). As Finnstro¨m (2015: S224) has argued, 'we build our ethnography by way of the relationships that we establish in the process'. Rather than just asking questions, we therefore spent time and effort building close interpersonal relationships in an attempt to 'minimize the distance between the researcher and the researched' (Grabska 2014: 15). Our findings derive from the strength and quality of these relationships.
To achieve the relationships required, we prioritized the parts of life our interlocutors deemed most essential and spent significant time engaging in the type of everyday activities James Clifford (1997: 56) might have called 'deep hanging out': compound maintenance, the collection and preparation of food, sitting and talking, and eating and drinking. Moreover, each author has a longer personal history of living and working with South Sudanese. O'Byrne's doctoral research was conducted in Pajok during 2013-14, where he was adopted into the Bobi subclan and became 'a Bobi boy' and 'a son of Agola Kapuk' (a Pajok-specific mwoc or 'praise name'). His son was later named by Bobi elders according to Acholi custom. Ogeno originates from Pajok, and he spent his childhood living in Kyangwali Refugee Settlement in Hoima District ( Figure 1). There he lived alongside some of the same people now living in PRS, many of whom are close friends or family. Thus, both authors have ('real' or 'fictive') kin and other connections to PRS residents or their families, and a major impetus for this study were the stories we heard before research began.
These relationships allowed privileged access to much sensitive information. These same relationships also introduce a potential weakness, however: an overreliance on Acholi voices to the neglect of other groups. In our defence, at the time of fieldwork around 70% of PRS's residents were from Acholi-speaking South Sudan, while refugee leadership was closer to 80-85%. The other main ethnic groups in PRS (the predominantly agro-pastoralist Langi and Lotuko) come from a much larger, famine and drought stuck but relatively conflict-safe area further east in Eastern Equatoria, beginning in the eastern foothills of the Imatong ranges and extending as far as the Ethiopian and Kenyan borders. These refugees often paid significant amounts for safe transport to the Ugandan border point and, due to the basic lack of income generating activities within PRS, over the period of our research generally could not afford the transportation fees required to engage in temporary or irregular returns to their former communities.
Given these demographic variations, differences in ethnic mobility were more matters of geography and available money than ethnic-based identities, relations, 752 Ryan Joseph O'Byrne and Charles Ogeno and expectations. Even so, to rectify any problems with representation, we spent significant effort diversifying the ethnic base of our primary interlocutors in midlate 2018. Nonetheless, the majority of our interlocutors remained Acholi South Sudanese, more specifically from Ayaci and Magwi Counties, and especially Pajok. Moreover, given the location of PRS in an Acholi-speaking area of Uganda, PRS itself has a decidedly Acholi feel: we were at many formal events where Acholi was the only language spoken, especially beyond the gates of the main UNHCR/NGO compound, and both the surrounding Ugandan community and the majority of NGO staff are all Acholi. It is also the native tongue of Ogeno. When needing non-Acholi language translation, we simply did what UNHCR does, too: approach refugee leadership and request use of their own preferred translator(s).
Examples of the types of events we observed include: food aid delivery across multiple Food Distribution Cycles (FDCs) both before and after the UNHCR biometric verification exercise which ran from April to June 2018 (BVE, for more information see below); a wide array of stakeholder meetings; the Refugee Welfare Council elections in July and August 2018; and a number of church, clan, family, school, or other gatherings. Beyond ethnography, numerous formal, semi-formal, and informal interviews were also conducted, with NGO, OPM, and UNHCR employees as well as across both the refugee and host communities. Further, a total of 50 open-ended questionnaires on individual mobility practices was administered to Acholi and Lotuko refugees across each of PRS's eight zones. Nonetheless, deeper investigation is needed into the connections between various types of refugee mobility and their direction, duration, and rationale. Further studies could benefit from undertaking more longitudinal perspectives, especially life history methodologies or cohort-based forms of observational study.

The Everyday Pragmatics of Mobility Among South Sudanese Refugees in Uganda
During our research, it was apparent that cross-border movements between Uganda and South Sudan continued throughout the recent conflict, despite South Sudan being unsafe. Our findings suggest that a combination of regional variation in South Sudan's wider conflict dynamics and the specific location and demographic composition of PRS were the primary factors allowing many of these cross-border mobilities undertaken. Thus, throughout the period of our fieldwork, and although much of the rest of South Sudan remained unsafe, one obvious distinction between PRS and many of Uganda's other refugee receiving locations is that the majority of PRS's residents originated from the Ugandan/ South Sudan borderlands east of the Equatorian Nile. In fact, except for the one major SPLA/IG attack upon the community of Pajok in April 2017, these areas were generally safer from large-scale violence than most other refugee-producing regions. Return was therefore easier for these refugees than those whose origins were located further away or who had to pass through more conflict-affected regions. This is especially true for those areas directly north of Lamwo District Pragmatic Mobilities and Uncertain Lives 753 from which PRS's numerically dominant Acholi-speaking residents derived. This finding affirms similar patterns among Equatorian refugees in Uganda's northwest settlements (REACH 2018a: 5).
This combination of (relative) peace and proximity meant that, during our research, cross-border mobility was definitely possible, if neither predictable nor entirely normal (although as we note below, it became increasingly common with the progress of the peace process). As with Afghani refugees in central Asia (Monsutti 2008: 59), South Sudanese might also return across the border for reasons ranging across the full gamut of personal, familial, sociocultural, political, and economic registers. In no particular order, the most common reasons among our interlocutors were: to visit friends and family; to collect objects left behind (usually for small business enterprises); to engage in cultivation (often on land close to the border so access was cheaper and easier); to gather information about possible future repatriation; or perhaps most commonly, to attend a funeral or accompany a body returning for burial (cf. Hovil 2010: 6;Kaiser 2010: 52;REACH 2018a: 4).
Often the reasons underlying seemingly divergent mobilities are interlinked, as for one young woman we spoke with in November 2018. Amy-like all names used, this is a pseudonym-had just returned from accompanying the body of her uncle back to Acholi South Sudan and, knowing our interest in cross-border movement, was happy to talk with us. She said that, if they can afford it, refugees originating from close to the border prefer to take family 'back home to be buried' rather than intern them in the settlement cemetery. There were several reasons for this, she said, but not only was the settlement cemetery 'too full and the remaining pieces too close to the river [a stream running east-west through PRS]' but host community 'land lords do not allow refugees to bury the dead ones within the residential plots' as they would in South Sudan. She also told us that some families were denied burial 'if they did not first pay the landlord for the body', citing the example of her neighbour, whose landlord demanded UGX 2,000,000 (£420/$540) for the polluting presence of a deceased person on his land. 'Better to pay to take that body back than give the money to someone else', she said.
People's attitudes to movement also changed frequently, often within the space of a single week. One notable example of attitudinal change took place not long after early progress in the Revitalized Agreement on the Resolution of the Conflict in South Sudan (R-ARCSS) peace initiative, during the period of 2018's June and July 'short dry' ('oro matidi'), when communities of the region begin preparations for the year's second agricultural season. Reminiscent of exactly those type of seasonally-based, agriculturally-defined movements that had distinguished their pre-refugee lives-when entire families might go and stay in their bush gardens for a period of several weeks to try and best ensure a productive growing seasonmany PRS households continued to engage in similar seasonal movements, despite becoming refugees, and many families told us their sons had 'gone back to South Sudan to open up new gardens ' (cf. REACH 2018a: 4-5).
This included Bosco, a paternal cousin of Ogeno. In July 2018, his mother informed us he had left PRS to begin preparations on the family's sesame fields 754 Ryan Joseph O'Byrne and Charles Ogeno in Pajok. She said this would not only bring an important source of income but would provide some much needed dietary diversity. The family's hope was shortlived, however: Bosco soon returned with the news that the SPLA/IG had deployed new regiments to the area, composed of unknown and ethnicallydifferent soldiers. Recent experiences had taught the population that such deployments came with increased civilian predation. Not long afterwards the other youth who had travelled to South Sudan also returned to PRS, bringing word of an escalation in army-related theft and violence. Moreover, their cultivation efforts were largely unsuccessful, significantly affecting dietary diversity, livelihood success, and basic cash flow in the settlement and beyond.
As in any community, some people in PRS travel more frequently or for longer periods than others, and in Palabek this demonstrated very obvious class dimensions. Indeed, many useful indicators of social differentiation can be found within variations in who chooses to move, to where, how often, and for how long (cf. Kaiser 2010;Grabska 2014). For Acholi residents of PRS during 2017 and early 2018 especially, cross-border movement was more common among those located at both extremes of the class spectrum and, beyond the seemingly unpredictable nature of more irregular forms of mobility, had a distinctive class profile: on the one hand, while some are involved in international business and have dependable access to vehicular transport and a variety of sought after trade goods, most people move out of sheer desperation, their mobility induced by uncertainties around the predictable provision of services in the settlement and the inadequacies of available resources.
Those on the first side of this class dichotomy are those who exploit the opportunities of conflict and resettlement for their own advantage. They encapsulate the positive dimensions of the ambiguities inherent within any refugee situation, demonstrating not only that mobility is possible but that it can be leveraged for one's benefit. For these men and women, mobility and wealth help continue access to the other. Although those few individuals able to engage in repeated, profitable border crossings were communally glorified, however, the number of people undertaking such movements were actually very few. Thus, we were told abundant stories about 'Owot the Driver', a man who seemingly lived life on the road, transporting goods and people-living and dead-back and forth between Pajok and Palabek. According to these stories, Owot had several wives, multiple vehicles, and a large disposable income that allowed him to reliably eat meat and drink beer. Owot was said to be demonstrable proof that mobility was not only possible but profitable. However, as Owot was able to mobilize a level of personal wealth not accessible to most PRS residents, the experience of him and others like him does not greatly represent the majority of the community.
Nonetheless, those at the other end of the class spectrum were always more common, and more commonly discussed. This was especially true from November 2017 through to the completion of the UNHCR biometric verification exercise (BVE) in April 2018. Indeed, the majority seemed among the most marginalized and peripheral of the refugee community: near destitution; perhaps suffering from the many physical or psychological ailments associated with conflict and post-Pragmatic Mobilities and Uncertain Lives 755 conflict life; possibly a widow, orphan, or some other UNHCR-designated PSN (Person with Special Needs). For many of these, perhaps best called 'quasi-refugees', life in a refugee settlement was simply too fragile to be bearable: despite their prima facie status, for various personal or political reasons they had either failed to officially register with the OPM or could not afford the bribe money necessary to do so (cf. Ogeno and O'Byrne 2018). Unable to afford life in the settlement, and without food, money, health services, or other humanitarian assistance, sheer desperation drove these forgotten refugees back to the uncertainties of life in South Sudan. When our interlocutors told us about these returnees, mobility was portrayed as dangerous rather than glorious, often resulting in serious setbacks or even death for those in pursuit of life's basic necessities. Their stories always noted these people had only 'returned home' because of the deplorable conditions of their settlement-based lives, the pain they individually experienced a form of symbolic analogue for the existential anxiety of the refugee collective.
Given the basic 'protection' parameters of UNHCR's mandate and operation, ensuring timely and problem-free food distribution should be fundamental. In PRS, however, this was not always the case: at least until the BVE stabilized humanitarian assistance in April 2018, lack of dependable food provision was the single greatest concern of most residents. Although the World Food Programme's (WFP) food distribution seemed taken for granted by most camp authorities and humanitarian actors, including UNHCR and the OPM, it certainly was not among refugees. In fact, missing or delayed food aid seemed one of the defining features of settlement life and we were repeatedly told it was the primary reason someone would leave relative safety in Uganda for uncertainty and danger in South Sudan. Various refugee leaders also showed us multiple abandoned compounds, telling us their owners had been among those denied food and other humanitarian assistance and, given the circumstances, had decided their best chance of survival was to return to living in a country beset by a volatile conflict involving gross, civilian-directed violence.
Regrettably, as with most informal migrations, evidence for this humanitarian failure-induced mobility is largely anecdotal. Nonetheless, we heard variations of this phenomenon from refugee leaders within six of the settlement's eight zones at various stages through late 2017 and early 2018, as well as from multiple refugees themselves. All these repeated near-identical situations and outcomes and, unfortunately, these stories nearly always ended badly as well, with the migrant's untimely death (cf. REACH 2018a: 4-5). These dangerous informal repatriations were not unknown to UNHCR or settlement officials, either, as we were present at several large meetings during which refugee leaders complained about the effects of unpredictable food delivery. Furthermore, one unusually forthright UNHCR representative told us: These refugees move a lot, sometimes back to South Sudan or Kiryandongo (Refugee Settlement) . . . Our refugee policy says once someone crosses the border back into their country, they lose the status of being a refugee. But people do return back without informing the authorities. You see, it is true that, because of the 756 Ryan Joseph O'Byrne and Charles Ogeno problems getting food, the refugees find life difficult in the settlement. So they go back to South Sudan to do farming and business, to get money and food (UNHCR Official, August 2018).
Although some NGO staff and refugee leaders may have portrayed the diverse mobilities of PRS residents in rather simplistic and generalizing terms, most refugees moved because of a combination of factors. Poverty, insecurity, and general existential uncertainty were among the most common of a range of intersecting rationales, and certainly a lack of dependable cash flow meant that even relatively unreliable opportunities to access money were seized upon. In understanding the complex interplay between competing needs and rationale, the story of a Lotuko woman called Agnes proves particularly instructive, highlighting the ways that personal rationalizations of everyday and future needs intersect with individual and household uncertainties and (in)securities.
We first met Agnes in late 2017, when she approached us after an NGO livelihoods workshop, presuming us to be NGO staff and seeking our support with the non-delivery of food aid. Agnes became one of our primary interlocutors, but towards the end of 2018, she suddenly returned to South Sudan for a period of several weeks. When we finally located her again, she told us she had been away due to a new business venture: as the local go-between for a probably illegal international logging business, one that took trees from areas of southern South Sudan depopulated by violence to sell to foreign speculators in Kampala. When we asked her why she had done what she did, her answer was telling: I went looking for a better life because life here is too difficult. No food, no resources, nothing. Most of these organisations are not willing to help us out of our situation-they come and ask questions and they end there, they do not want to assist [. . .] This is why I got involved in this business, because if I do not do something nobody will help my children (South Sudanese refugee woman, October 2018).
However, despite spending several weeks living away from her children in the South Sudanese bush, an objective assessment of this venture might say it was largely unsuccessful. As well as bribes paid to police, soldiers, border guards, and government officials on both sides of the border, Agnes' Kampala-based business associates demanded all costs were paid from her own profits. Furthermore, although she had not died, her experience was distinctly unpleasant: as well as being generally sick and hungry for much of the time, she implied she had been forced into non-consensual sexual activity. Despite her best efforts, she barely broke even, hardly covering the money she owed her neighbours for feeding her children while away.
Not all refugee movements involve crossing borders, however, and the majority of those undertaken by PRS residents meant travelling to different places within Uganda, often for extended periods. For example, many teenage boys (especially, although not exclusively) are undertaking secondary education in northern Ugandan boarding schools, and the seasonal variation within their mobility patterns is that of the Ugandan education system rather than the agricultural cycle.

Pragmatic Mobilities and Uncertain Lives 757
Likewise, any number of individuals may transit backwards and forwards between PRS and the primary northern towns of Kitgum or Gulu for health or business-related purposes.
Moreover, due to their experience as refugees during the Second Sudanese War, most Acholi-speaking adults in PRS have previously lived in at least one Ugandan refugee settlement. Indeed, a substantial number continue to maintain houses, farms, families, or businesses in or nearby these settlements. This is particularly true for the area around Kiryandongo Refugee Settlement, home to significant numbers of Acholi South Sudanese during the 1990s and 2000s (Kaiser 2006(Kaiser , 2010. Although most repatriated following the CPA, some self-settled in the area around nearby Bweyale town, and visits to friends and family remaining in that region make up a significant portion of the Ugandan-based movements of PRS's Acholi-speaking refugees. In October 2017, one elderly Acholi man from Pajok told us 'because I still have a wife living on our land down there [in the Bweyale area] from before [during their time as refugees in the Second Sudanese War], I will try to go down there every two or three months, if I can afford it'. Others repeated similar patterns (cf. REACH 2018a: 5).
However, these movements were severely curtailed over the first half of 2018: a scandal involving the systematic inflation of refugee numbers rocked the Ugandan refugee industry in early 2018, resulting in substantial international fallout and threats of reduced funding from donor organizations and countries. In response, between March and September 2018 UNHCR instituted the organization's largest ever biometric registration and verification programme, the Biometric Verification Exercise or BVE. Seeking to quantify the true number of Ugandan-based refugees, the BVE used iris scanning and finger printing technology to count each refugee once and only once, reducing the possibility of corruption and theft on the part of either refugees or humanitarian staff, and tying the allocation and distribution of all food and other humanitarian assistance to the final outcome.
One result of the BVE was a series of significant changes in how food was processed, distributed, and accounted for. In PRS, this meant that from June 2018, camp authorities (the OPM, UNHCR, and WFP) began insisting all refugees could only collect food aid from a single specified collection point within their own zone on one particular day per monthly delivery cycle. As well as the continuing irregularity of distribution days-as before the BVE, it could still be the start of the month during one cycle but the end or middle of the month during another-food collection was suddenly now only available to persons older than 14 who could provide valid biometric data on that specific day, with the precise day for distribution usually made public less than a week before food delivery began. Because of these changes, it was no longer possible for friends, extended kin, or even refugee leaders to collect food for absentees, as they had been able to under the previous card-based system. The BVE therefore not only had a negative effect on those who had not been able to correctly register (most of whom now lost any ability to access food or other humanitarian services) but also upon the general intra-Ugandan mobility of settlement residents, significantly reducing the duration and frequency of previously common inter-settlement movements. After the new food distribution rules came into play in June 2018, the same elderly Acholi man quoted above told us that 'because of this thing [the BVE], now I can no longer go down [to Bweyale] to see my wife there, because if I do I may miss out on the distribution here, and then I will not eat'.
Refugees do not only think about their present lives and movements, however (cf. Hovil 2010; Kaiser 2010): they also consider those in the future, whether back to their place of origin, a location from an earlier refugee experience, or to visit or live close to friends and family elsewhere in the same receiving country. Indeed, many South Sudanese continue to move to and from various Ugandan locations for a variety of often mundane reasons during the course of their everyday lives, and the same is true of those who remain in South Sudan. As noted earlier, many residents of PRS have visited compatriots across Uganda multiple times during their most recent flight, some of whom live in the same settlements as during a previous displacement. In all cases, and much like movements taking place throughout the recent past and present, individual and communal analyses about peace, proximity, and (un)predictability seem to be the most important considerations involved in planning potential future movements.
Concerns about security are especially significant, as many plans for the future depend upon the success of R-ARCSS, South Sudan's current but fragile and unpredictable peace process. Although several attempts were made to end the recent war, from the outset PRS residents deemed that R-ARCSS was somehow qualitatively different. Early progress only confirmed this, and it is undeniable that the number of those making temporary visits across the border increased substantially following its signing in September 2018. In fact, over the November 2018 to March 2019 dry season, many of those making the crossing began remaining in South Sudan for longer periods. Some of these visits may even have adopted a more semi-permanent form, although changes in patterns of mobility were apparent even before this. Nonetheless, even those who seemed to be most proactive about undertaking longer or multiple cross-border journeys generally still remained cautious about the future.
Take, for example, Paul. One of our closest interlocutors, Paul was a newly married man in his early thirties who had worked closely with O'Byrne during his PhD fieldwork. He also owned his own motorcycle and used this to return to Pajok several times over the last half of 2018. The first time he crossed the border, in August 2018, he told us he went 'just to go for a visit, to see what was happening, and return back [to Uganda]'. This was followed not long after by an overnight trip for a burial, and then another trip in October 2018, to help a friend visit their family. Just before we finished fieldwork he went again, this time to clear his gardens for planting in early 2019, and he said he hoped to go back and forth several times during the 2018-19 dry season. He would not take his family back, however, because: In Pajok there are still too many soldiers. And it is scaring us to return, because we fear the soldiers. They are still sometimes raping women. Some men are returning, but they are living with fear, because of the soldiers. People there do not talk about There is significant risk attached to these and any returns, however: as everyone recognizes, the peace process is (and remains) fragile. And, as almost all our interlocutors acknowledge, there have been other ceasefires, some of which they also thought might 'stick'. Consequently, as with Paul, most people took a pragmatic approach to R-ARCSS's fragility and whether or not they would repatriate. These feelings were widespread, and continued for months after R-ARCSS was initially signed. For example, during an unrelated conversation in late January 2019, the Refugee Welfare Council Chairman (the refugees' primary elected representative or RWC) told Ogeno that: Many people are returning to South Sudan illegally, without informing UNHCR or the OPM. They return illegally as they want to keep their refugee status, because they don't know what will happen in South Sudan, even after this peace agreement. On the other side, many who left earlier are crossing back into Uganda. The driving force is famine in South Sudan and access to better social services in Uganda, like education and health. There are also more new arrivals coming and they are fleeing because for the same reasons. And now people are starting to become unsure about the current peace deal (Refugee Welfare Council Chairman, January 2019).
As the RWC astutely noted, fear of renewed violence in South Sudan was not the only reason driving people to seek refuge in Uganda, nor was it the only issue which worried returnees: they were also concerned about losing their refugee status (cf. Hovil 2010: 12;Kaiser 2010: 54). As Kaiser (2010: 57) noted following the Second Sudanese War, such fears are intimately interconnected. Given refugees' very real concerns about how the lives of themselves and their families might be effected by unknown or unpredictable elements beyond their immediate control, it is little wonder they were reluctant to give up the benefits of being a refugee. After all, beyond personal or physical security, being a legally enrolled and verified refugee in Uganda brings not only food and security but health services and 'free' primary-level education, none of which are readily available in rural southern South Sudan. This is why people told us that, for the great majority, permanent return was between three and five years away, if R-ARCSS continues to hold. Moreover, repatriation would begin by sending young men to open gardens and build homes first, while the rest of the family-especially students and the sick, weak, and aged-remained in Uganda to maintain households, compounds and refugee status (cf. Hovil 2010: 16;Kaiser 2010: 55). Nonetheless, even in this circumstance, nearly everyone we spoke with said repatriation would comprise much cross-border back-and-forth. Thus, even in the case of future UNHCRassisted repatriations, PRS residents will generally maintain the refugee status of some family members for as long as possible, allowing a return to Palabek and a certain amount of certainty should life in South Sudan prove-perhaps predictably, given recent experiences-too violent or difficult.

Conclusion: Pragmatic Mobilities, Uncertainty, and Refugee Agency
This article has outlined some of the various and diverse everyday mobilities of South Sudanese in Uganda, predominantly through the lens of Acholi-speaking refugees in PRS in the country's northern Lamwo District. Cross-border movements between Uganda and South Sudan have continued throughout the current conflict, despite South Sudan being unsafe, and we have demonstrated that the complex interactions between refugees' seemingly distinct and divergent movements have important implications for how displacement-based mobilities are conceptualized and understood. In an attempt to capture the temporal, geographical and existential scope of these journeys, as well as their fractured (dis)junctures, we have defined these journeys as 'pragmatic mobilities', that is, as 'the experience and practice of multiple, distinct yet interconnected journeys, despite trying times and unknowable circumstances'. We have also argued that these pragmatic mobilities practices are particularly powerful manifestations of agency, seeking to at least allow for the possibility of greater personal and collective control in otherwise uncertain contexts.
In the case of refugees in PRS, at least, the complex rationales through which most refugees try to bring a certain element of predictability to their uncertain mobilities are diverse and involve a combination of factors, with the most significant including: regional variation in conflict-related violence, with movement to areas of relatively safety likely to occur more frequently; the precise proximity of the locations being left and returned to and some closeness of connection to the place or the people who reside there; and specific refugee community demographics, with more permanent movements more common among young men and refugees at either end of the class spectrum. Thus, unless it were for the purposes of a funeral and burial, most of PRS's mobile population either generally returned to their most recently inhabited location in rural southern South Sudan during the November to March dry season, or went to visit friends or family in another Ugandan refugee settlement, with most journeys taking place for a period of only a few days.
We have also argued that the complexity and diversity of PRS residents' 'uncertain' multiple mobilities have more or less 'irregular' or 'rhythmic' forms. Although the great majority of movements undertaken were irregular or onceoff visits to one place or another, some definitely had more 'rhythmic' dimensions, and some of the former even became the latter over time and through repetition. Alongside their expected differences around the frequency and length of movements undertaken, 'irregular' and 'rhythmic' forms of 'uncertain' mobility share many similarities, including a variety of geographical (location returned to), temporal (duration of the visit), and seasonal (time of year) dimensions. The major differences between 'irregular' and more regular or 'rhythmic' movements were simply that the latter tended to be of a more enduring nature and practiced by a smaller number of individuals for more explicitly economic rationales. Moreover, 'rhythmic' returns are more likely to include a lengthier stay at the place of arrival, perhaps demonstrating some element of longer-term commitment to both the Pragmatic Mobilities and Uncertain Lives 761 place being returned to as well as the very act or promise of mobility itself. Thus, although irregular or once-off movements might be more common in terms of the number of people practising them, they are almost necessarily uncertain in nature, while the more regular nature of 'rhythmic' mobilities that frequently and perhaps even repeatedly follow similar patterns or pathways lend them a certain predictability.
We end by providing the following, empirically-based conclusions and recommendations: Firstly, it is important to note that the contemporary cross-border mobilities of PRS residents seem intimately connected to the unique location of PRS vis a vis South Sudan and must not be conflated with either voluntary repatriation or any greater sense of personal security. This is due to several reasons, many of which do not easily or directly transfer to lives of those living in some other Ugandan refugee settlements. After all, most PRS residents originate in areas directly north of Lamwo District, and so the border area and its legal and less legal crossing routes are well-known from residents' pre-exile lives. Furthermore, most residents share a language with their Ugandan hosts and are able to pass themselves off as a Ugandan if required. Moreover, as well as these returns being relatively quick, easy, and cheap, the 'home' areas being returned to are at least relatively free of localized violence. This makes them safe to visit, if not entirely safe for extended durations. In other words, they are more predictable, less uncertain. The mobility patterns found in PRS should not necessarily be expected to be repeated elsewhere, especially when these basic dimensions are not shared, and must remain a matter of empirical discovery. Such considerations should also be taken into account when planning and implementing future policy or research initiatives.
Similarly of note is that some of the most experientially significant movements in a refugee's life are not at first obvious or perhaps even expected, for example, those relating to life's end. Indeed, death-related mobility was so common in PRS that 'returning' a body to its native soil, to be buried where it 'belongs', is an important yet underappreciated facet of refugees' movements. These mobilities are also important manifestations of refugees' agency and place-making practices: by acting where possible to return their deceased, they take advantage of the opportunities and paradoxes within local governance regimes to temporarily visit areas of origin and belonging. In this way, continuity of life, meaning, and connection are maintained, despite other uncertainties. Further research into refugees mortuary and funeral practices is therefore suggested, especially death's migratory or transnational dimensions.
Moreover, the ways in which residents of PRS speak about and practice returns to and from South Sudan are often framed through the positive and negative experiences of uncertainty and unpredictability within life in exile. As Grabska (2014: 6) has noted, 'the visions for the future and the imagined homes that women and men long for are shaped according to their experiences in the specific framework of refugee camps'. Thus, despite the somewhat unique proximity to point of origin and return of PRS in the Ugandan refugee context, many who repatriated did not do so because they wanted to 'return home' at the precise time they left. Rather, they did so because of the problems of resettlement life. This is especially true for those who experienced difficulties accessing the basic food aid to which they should have been entitled. UNHCR and other humanitarian providers and governance actors in the refugee industry must therefore ensure their registration, verification, and distribution practices are transparent, accountable, and humane, with resources actually reaching the designated refugees.
Likewise, there is a problematic dissonance in the fact someone can be South Sudanese and legally holding prima facie refugee status and yet not be able to claim either that status or the protections it affords, simply because the Ugandan refugee authorities seem to find it logistically easier to insist all refugees are 'properly' biometrically-enrolled either at a border or during a time-limited and arbitrarily defined period within one of the settlements. Again, like many of the failings of the Ugandan refugee industry, in not disputing the legality of these requirements, UNHCR seems to have chosen a smooth operational environment over either international human rights legislation or its own mandates around protection. Similarly, from a human rights perspective, the changes brought about by the BVE has meant that UNHCR and WFP have effectively removed the same refugees' rights to freedom of movement enshrined in both Ugandan law and international human rights legislation and about which they have been so vocally supportive of the Ugandan government.
Finally, transformations in how, when, and why displaced people move are instructional: along with the existential difficulties of settlement life, perhaps the most important parameters affecting South Sudanese refugees' cross-border mobilities were a reduction in localized violence at the destination and the uncertain institution of a perhaps temporary peace. The likelihood of any large-scale, future repatriation therefore depends upon the stability and success of this peace, with a return to either widespread or extreme violence limiting future returns. This demonstrates the continuing importance of the international community in South Sudan's peacebuilding efforts and we recommend that international resources are directed not only towards the provision of security, justice, or the rule of law but also in developing infrastructure within the war-effected rural areas that many refugees left and to which they might return. After all, without significant, localized rural investment, repatriation may ultimately prove unsustainable. And without the predictability of a safe and sustainable return, will mobility not only continue to 'be seen as a key livelihood strategy' (Monsutti 2008: 58) but as a fundamental means of ensuring existential certainty.