From Exile to Homeland Return: Ethnographic Mapping to Inform Peacebuilding from Afar

When violent conflict flares up, forced migration often follows. Ethnographic data shows that forced migrants remain attached to their places of origin and often express a desire to return once conflict has abated, be it after weeks, months, or years. Conversely, peacebuilders in the homeland have not effectively integrated displaced persons within their strategic programming. This is cause for concern considering the literature connecting the collapse of fragile peace to ‘refugee spoilers.’ There is a critical gap in peacebuilders’ commitment to understanding refugees’ needs and claims, and the implications these pose on peace stability following repatriation. This article argues that ethnography of refugees still living in exile can generate rich datasets useful to the development of peacebuilding programming. More than this, it proposes a methodology — ethnographic mapping — that can collect both spatial (maps) and narrative (descriptions) information in tandem and across cultural groups living in refugee camps.


Introduction
When refugees repatriate once conflict has abated, they frequently face new challenges that disappoint their long-awaited hope for stability and a return to normalcy. They often find that the landscape they left has undergone radical changes through the process of war and violence. Secondary occupants may be living in their childhood home or working their land. These ground-level realities routinely result in the emergence of new forms of grievance that potentially threaten peace that is already fragile and uncertain (Bradley 2013;Unruh, Frank, and Pritchard 2017). With sometimes tragic outcomes, the worst being a relapse in conflict and violence, refugees and other displaced persons bear the burden of being characterised as 'spoilers' to peace (Loescher et al. 2007;Milner 2008;Perera 2013). As many examples in recent history attest to the scale and frequency of this phenomenon, we must ask ourselves why peacebuilding programmes do not take forcefully displaced persons into greater account when working through conflict environments and towards landscapes of peace. Contrary to sedentarist assumptions that associate dislocation with absence of interest, capacity, or legitimacy in having a voice on important matters of the homeland, ethnographic evidence shows that refugees remain deeply connected to their country of origin (Betts and Jones 2016;Chatty 2010;Hammond 1999). While in exile, they need opportunities to communicate memories, hopes, desires, claims, and perceptions on peace and conflict. Particularly in considering the prospect of large-scale return migration, this information can be instrumental and formative in developing comprehensive peacebuilding programmes that include the refugee perspective.
This article focuses on the importance of refugee voices in peacebuilding efforts. While an ideal circumstance would directly involve refugees in peacebuilding programming, the first step towards this is recognising and providing opportunities for those in exile to share their experiences and perspectives on conflict, displacement, peace, and return with peacebuilders. A methodology to gather and organise this information, based on approaches in geography and anthropology, will be presented. As will be shown, ethnographic mapping in refugee contexts has the potential to generate rich spatio-narrative visualisations and descriptions. Data generated through this approach can, in turn, inform peacebuilding initiatives and help pre-empt potential conflicts resulting from voluntary repatriation.
The first section of this article connects the dots between the literatures on return migration, conflict, and peacebuilding. The second section outlines ethnographic evidence showing that refugees are politically capable and culturally invested in their homeland, therefore establishing their aptitude and importance in having a place at the 'peacebuilding table.' Next, a methodological outline for ethnographic mapping is provided, followed by a model for its use in refugee contexts. Before concluding, the article explores the risks and ethical considerations of ethnographic mapping.

The Peace and Return Migration Nexus
Voluntary repatriation, or voluntary return migration, is one part of a trifecta of ' durable solutions' set out by the United Nations High Commissioner for Refugees (UNHCR); local integration and resettlement being the other two. Although UNHCR affirms that no hierarchy exists between the three, that they are all part of an integrated approach (UNHCR 2012: 186), data on refugee flows and state policies clearly shows that 'repatriation and return will likely continue to be the most favoured durable solution, not only for those hosting refugees but for many refugees themselves' (Hammond 2014: 508). Acknowledging that the reasons for this are quite broad, much of the literature suggests that while some refugees may embody a desire to return home, political processes that go far beyond their reach consistently undermine their agencyand sometimes choice -in whether or not they repatriate (Tegenbos and Vlassenroot 2018;Bradley 2013). Within this context, Long (2013: 2) problematises the identityemplacement nexus, explaining that: Too often, repatriation has been presented as a 'natural' best solution to refugees' exile, when in fact the assumptions that underpin this claim -that people belong in a particular, fixed place -reflect the interests of the politically powerful in retaining the status quo international order, rather than the capacity of return to realize the rights of the displaced.
The now-widespread assumption that repatriation represents the ultimate route to restoring refugee livelihoods arguably developed as a result of two intertwined paradigms emerging in the late 1980s and 1990s. The first is the result of changing migration policies following Cold War-era politics, where states veered away from the humanitarian philosophy that underpinned sensibilities for refugee protection following World War II. Where states used to react based on obligation and expectation (Goodwin-Gill 2001), they now 'introduced a wide range of measures designed to keep asylum seekers from reaching national borders' (Hansen 2014: 258). Moving from asylum to containment, Loescher (2014: 219) notes that 'Western states have largely limited the asylum they offer to refugees and have focused on efforts to contain refugees in their region of origin.' Repatriation, then, proved to be most representative of this political environment, as other ' durable solutions'resettlement or local integration -involved refugees remaining outside of their country of origin and under the care of a host country. This coincided with a second paradigm within the broader international community, where multilateral agencies consistently asserted that voluntary repatriation was the most logical, humane, and 'natural' way forward in addressing the 'refugee issue.' Following former UN High Commissioner for Refugees Sadako Ogata's designation of the 1990s as the ' decade of repatriation,' scholars warned that this attempt to universalise responses to displacement was predicated on false assumptions rather than a fully formed policy programme based on evidence and best practice (Harrell-Bond 1989;Bascom 1994;Black and Koser 1999). However, with the political landscape now tilting towards containment, securitisation, and criminalisation of migrants and refugees, this paradigm quickly took ground within policy, development, diplomacy, and peacebuilding circles. Reactions within academia widely condemned the extent to which voluntary repatriation had been embraced by the international community. Scholars such as Mertus (1998: 345) went as far as considering voluntary repatriation as a direct externality of states' irresponsibility towards displaced persons: 'the primary cause of so-called "voluntary repatriation" is the host State's unwillingness to provide for and protect refugees.' From the late twentieth century until now, several examples have shown what many scholars had forewarned as this paradigm emerged. A striking example is that of repatriating Bosnians following the Dayton Agreement, where thousands were forcefully evicted while others found their homes inhabited by secondary occupants. For many, however, motivations to repatriate were overshadowed by a fear that their return would reignite ethnic tensions, and that those involved in the already haphazard, ad hoc, and unjust post-conflict repatriation strategy would fail to protect them under these circumstances. Despite this reality on the ground, the UN remained confident that its strategy was a success (Kleck 2006). It is surprising, or perhaps not, that the UN had held itself to such a low standard of success, considering that previously mismanaged repatriation protocols as part of the UN-supervised Arusha Accords had demonstrably exacerbated local rifts, leading up to the 1994 Rwandan Genocide (Jones 2001;United States Holocaust Memorial Museum [USHMM] 2014). In 2002, return migration of Burundians was jointly facilitated by UNHCR and the governments of Burundi and Tanzania. This effort followed the signing of the Arusha Peace and Reconciliation Agreement in 2000, an agreement that established returnees' rights to property restitution or compensation (Government of Burundi 2000). However, as with previous projects of large-scale return, the responsibility to deliver on the agreement's promises remained in the hands of state actors. Almost two decades later, the vast majority of returnees still remain landless and uncompensated, all the while experiencing food insecurity and contributing to local resource scarcity (Fransen 2017). Evidence from recent years continues to show that the complexity of post-conflict return migration is still underestimated and unaccounted for. In the context of disarmament, demobilisation, and reintegration (DDR) programmes, Knight (2008) argues that the 'R' is the most problematic component because it typically constitutes a long-term process; a timeline familiar to development actors, but out of the operational scope of peacebuilders. The complexity of the latter component of DDR is exemplified in the many iterations of 'R,' where reintegration is often substituted with rehabilitation, reinsertion, repatriation, and return. In the case of the Democratic Republic of the Congo (DR Congo), for instance, return of demobilised soldiers (including children), internally displaced persons (IDPs), and refugees to their home villages is highly problematic and lacks foresight from peacebuilders in these areas. Through interviews with demobilised Congolese, Richards (2016) demonstrates that ex-combatants fear -and often face -capture by not-yet-demobilised groups following their return to native villages. In the case of child soldiers, where DDR programmes are focused on a return to the way it was before the war, Haer (2017) explains that this strategy often further entrenches historical and structural gender roles that harm girl ex-soldiers upon return. For IDPs and refugees, lands and homes left vacant following displacement are often occupied by demobilised ex-combatants, leading to new grievances that can renew community rifts and conflict (Paddon and Lacaille 2011). Where UN peace operations are the leading partner in the implementation of a dozen DDR initiatives globally (World Bank, United Nations Peacekeeping and Social Science Research Council 2018), there is evidence that the implications of return migration are not thoroughly considered within these programmes.
While problems with return migration persist, so does the desire of refugees to return home. That said, economic, political, and social conditions within the country of origin are strong determinants of whether or not a refugee will undertake a return movement. When these conditions are not clearly understood, or when refugees are incentivisedor sometimes coerced -into return, their arrival can threaten post-conflict environments. Bradley (2013) outlines three unique scenarios that commonly play out following repatriation. First, tensions may occur when returnees lay claims to lost housing, land, and properties that are now in the hands of occupants who may have participated in their dislocation and dispossession. Unruh, Frank, and Pritchard (2017: 1) share this observation, confirming that: Among the most problematic aspects of any return process is the reclaiming, reoccupation and reconstruction of housing, land and property (HLP) for the hundreds of thousands to millions that will attempt this in today's dislocation scenarios. Equally difficult for those unable or unwilling to return to areas of origin, are the forms of redress for their HLP loss -due to destruction, secondary occupation, ethnic cleansing, etc. -that could put to rest the acute grievances that unresolved HLP dispossession produces. 1 Second, return migration may lead to a scarcity of natural and social resources that are already under pressure. Third, disorganised and large return movements can put pressure on fragile post-conflict institutions, potentially risking destabilising peace and triggering new displacement movements. As such, Bradley (2013: 6) suggests that 'repatriation has not been a boon for peace processes.' From a peacebuilding perspective, Milner (2008) argues that refugee return can negatively impact peace prospects in two ways. First, referring to the existence of 'refugee spoilers' (or 'refugee warriors'), some of which inhabit militarised refugee camps, these persons or groups plan their return with an opportunistic gaze to gain from or cause outright destabilisation of fragile peace (see also Loescher et al. 2007;Perera 2013). Second, and reaffirming other themes explored above, is the push for early and unsustainable repatriation by host states with shifting donor interest. In these situations, there is usually an absence of ground-level preconditions for sustainable and peaceful return, and the voluntariness of 'voluntary' return is suspect, even when return is assisted by the UN International Organization for Migration (IOM) (Bendixsen 2019;Flegar 2018;Gerver 2018). Milner (2008Milner ( , 2011Milner ( , 2015 confirms what other scholars have claimed, that refugees can play a positive role in building peace (Duthie 2012;Grace and Mooney 2009;Johansson 2009;Pope 2008;Sommers 2001), but has also recognised that targeted and programmed opportunities for this to take place are few and far between.
With what we know from research work and past experience, why does assisted voluntary repatriation continue to be recklessly advocated for and facilitated? Why does peacebuilding not incorporate refugees -or, more generally, displaced persons -more effectively in their approach if we know that return may undo peace efforts? Looking at the UN organisational culture can help us answer these questions. In her widely acclaimed and prize-winning book The Trouble with the Congo: Local Violence and the Failure of International Peacebuilding (2010), Séverine Autesserre provides an erudite diagnosis explaining why a ' one-size-fits-all' approach to peacebuilding persists despite overwhelming evidence that it is inadequate and dysfunctional. Drawing on observations made when working in conflict and post-conflict environments in Afghanistan, the Balkans, and DR Congo, she attributes this failure to four dimensions of the dominant international peacebuilding culture: 'The conception of the UN and diplomatic staff's role as "naturally" focused on the regional and national realms, the belief that specific strategies are appropriate for "postconflict" environments, the veneration of elections, and the view of humanitarian and development aid as an ideal solution to local conflict' (Autesserre 2010: 85). Peace and conflict scholars have often pointed to these same issues in the context of the Bosnian war (Merlinger and Ostraukaite 2005;Richmond and Franks 2009), Colombia (Berents 2018;Naucke 2017), and African conflicts more generally (Castañeda 2009;Mac Ginty and Firchow 2016;Zambakari 2017). One significant externality of a peacebuilding culture aligned to technocratic, managerial, and diplomacyoriented approaches is that it does not sufficiently take into account local actors and stakeholders. This defies what the academic community has widely acknowledged: that it is most often local dynamics -not national or regional ones -that have the greatest impact on the achievement and sustainability of peace (Autesserre 2014;Campbell, Chandler, Sabaratnam 2011;Leonardsson and Rudd 2015). Experience from UN peacekeeping missions, however, have repeatedly demonstrated that these operations are embedded in an organisational culture that equates peace to macro-level institution building, elections, and rule of law (i.e., liberal peace), which are largely disjointed from local realities. Similarly, UNHRC has discursively considered repatriation as an emblem of returned peace in refugee-producing countries despite the complexity and uncertainty that these movements produce in post-conflict environments (Feller 2009).
The above section has drawn specific links between return migration and ground-level efforts for conflict termination and durable peace. Most notably, it shed light on some structural flaws that suggest disjointedness between peacebuilding priorities and the need for ground preparations before largescale return movements; a disjointedness that, if left unaddressed, can, in fact, threaten fragile peace in post-conflict environments. This analysis has been done largely within the purview of refugees having already returned to their country of origin. The next section moves backward in this chronology, prior to their return and while in exile. It builds a case for their engagement and agency in peacebuilding from afar.

Recognising Voice and Refugee Attachment to the Homeland
In UNHCR's most recent Global Trends report, 15.9 million refugees, representing 78 per cent of the total global population, were in a state of 'protracted displacement,' defined as a situation 'in which 25,000 or more refugees from the same nationality have been in exile for five consecutive years or more in a given host country' (UNHCR 2019: 22). While this population is increasingly urbanised, many still remain in both formal and informal camps (Humanitarian Policy Group 2015). In the latter case, and particularly in parts of sub-Saharan Africa, these camps have been likened to widespread 'warehousing of refugees' (Smith 2004) in what Malkki (2002) calls 'vast zones of asylum.' Some have qualified these living conditions as 'liminal' (Mortland 1987;Purdeková 2011;Siganporia and Karioris 2016), described in anthropology as the period of limbo and transition between social positions in life (Turner 1967) and where refugees and asylum seekers are perceived to experience a ' conditional emplacement … and the ambiguity characteristic of that position' (Svensson and Eastmond 2013: 163).
Ethnographic work revealing the political and productive lives of refugees, however, has challenged claims of liminality on the grounds that it dehumanises, decontextualises, and dissociates the biological from the biographical (Brun and Fábos 2015;Brun 2016;Hammond 1999;Ramadan 2012). Explorations in refugee conceptions of home, homemaking, and homeland have been instructive in articulating that while refugees are effectively ' out of place' from their point of origin, they continue to exhibit strong feelings of attachment to them. In Black's (2002) study of Bosnian refugees' conception of 'home,' he draws the conclusion that 'the more distant "home" is in time and space, or the more unlikely or impractical a return "home" might be, the stronger the group's identification with, and yearning for, such a return becomes' (Black 2002: 126). Similarly, Ghanem (2003: 27) states that ' even if the association of "home" with "home country" is cast into doubt during pre-flight events, the irony of exile is that the geographical distance from one's country of origin often brings the forced migrant emotionally closer to it, sometimes even closer than before his/her flight.' This is also an area where anthropology has been particularly apt in engaging with migration studies, where scholars such as Chatty (2014: 81) have characterised the 'home' and 'homeland' as ' one of the most powerful unifying symbols for the dispossessed.' This perspective was instrumental to developing her methodology in Displacement and Dispossession in the Modern Middle East (2010), using ethnographic research to call on refugees' memories of ' childhood and youth…of forced migration or those of their parents, their recollections of places where they sought refuge, the institutions and networks in their new places as well as their perceptions and aspiration regarding home and homeland' (Chatty 2010: 3). In Malkki's (1992) ethnographic study of Burundian refugees in Tanzania, she found that by 'trying to understand the circumstances of particular groups of refugees illuminates the complexity of the ways in which people construct, remember, and lay claim to particular places as "homelands" or "nations"' (Malkki 1992: 25).
The refugee camp as a 'landscape of exclusion' (Sibley 1992), however, reifies the invisibilisation of refugees. Too often, and this applies to the virtual absence of consideration for refugees in peacebuilding, forced migrants become ' excludable bodies' from key social, cultural, and political processes; what Agamben (1998) would perhaps consider as 'bare life.' The observations above, then, are important in dispelling sedentarist assumptions claiming life in exile as apolitical and without merit of inclusion in decision making (Allan 2014;Betts and Jones 2016;Chatty 2014;Horst 2013;Parent 2019;Zetter 1991). Refugees effectively continue to remember and feel a sense of attachment and rootedness to their homeland, therefore participating in the formulation of what Anderson (1983) calls an 'imagined community.' Voluntary repatriation stands as a case in point in establishing continued commitment to refugees' place of origin. When undertaken, return migration is typically representative of refugees' agency, choice, and empowerment.
Future outlooks on peacebuilding programmes, then, should explore the inclusion of exiled persons in the planning, managing, and brokering of homeland peace. As former Assistant High Commissioner for Protection at UNHRC Erika Feller (2009: 86) notes: Peacebuilding will not be effective if it ignores the fact that refugees and displaced people often find themselves returning home alongside ex-combatants, demobilized child soldiers, women associated with fighters, and others who have become disaffected from their communities.
In recent years, the international community has expressed a commitment to building greater refugee agency within intervention programmes. For instance, in its most recent landmark document related to forced displacement, the New York Declaration on Refugees and Migrants, UN member states convening at the 2016 UN Summit for Refugees and Migrants asserted the vital importance of refugee agency in both peacebuilding and development initiatives. Similarly, one core principle of the Solutions Strategy for Afghan Refugees in the Middle East is refugee participation in decisions linked to displacement responses and solutions (UNHCR 2018). These steps forward demarcate an important change of tide, where the international community has recognised what social scientists have long claimed: that those refugees and IDPs who stand to benefit from these efforts must play an integral role in how these projects and programmes are developed. A commitment to this is long overdue, and while it should be welcomed, the international community will need to do much more than sign declarations to put this into practice. The question, then, is how to collect and consolidate this information. A further question is how to develop a transmittable output that can effectively communicate refugees' needs, claims, and perspectives prior to their return to post-conflict environments. The remainder of this paper will describe a distinct methodology that can be deployed to the field in order to gather these informational needs in situ (i.e., in communities of displaced persons) and prior to large-scale return migration. A general definition and outline of this approach -' ethnographic mapping' -will be provided in the next section. This will be followed by a more targeted discussion and model on how to use ethnographic mapping in displacement settings.

Defining 'Ethnographic Mapping'
Ethnographic mapping is formulated here as mapping practices that are decided upon and are performed through ethnography, and whereby its outputs -items, features, and contents oriented in space and meaning -are enriched by descriptive ethnographic writing. Here, the map goes beyond its primary function as a tool of representation, taking on a formulative role that enriches the process of ethnographic field research. This approach seeks to expand the typically assumed roles of the ethnographer, whom Emerson, Fretz, and Shaw (2011: 1)  Efforts to socialise ethnographic knowledge making (Brown and Dobrin 2004), however, have proposed that research participants be more involved in the drafting and formulating of descriptions made about them. In some ways, these new developments extend the poststructuralist preoccupation with representation, where scholars critical of midtwentieth-century anthropology contended that the structural frames it had developed were laden with power (Foucault [1976(Foucault [ ] 1990(Foucault [ , [1975(Foucault [ ] 1995, historically inaccurate aesthetics (Said 1978;Minh-ha 1991), and prioritised some frames over others, leading to defects in reason and descriptions of reality (Derrida [1967(Derrida [ ] 1998. In addition to its interest in further engaging research participants in a participatory approach that generates information on meaning 'from the ground up,' ethnographic mapping seeks to add a spatial component that complements the interpretivist aims of 'thick description' (Geertz 1973).
Building from Kingslover et al.'s (2017: 307) description of ' cultural mapping' as ' a broad technique that may be used in two general ways by anthropologists: to keep track of the social scientist's own findings… or as a method for understanding how those being interviewed are themselves mapping whatever is being studied,' ethnographic mapping is interested in the latter function. In considering how these maps take form, there is an element of choice on the type of map selected for these mapping activities. With this comes a responsibility to present research participants with multiple mapping possibilities, some involving a certain level of training (Caquard et al. 2019), and all of which assume different degrees of involvement (or interference) from the ethnographer. These cartographic approaches can be classified under two categories: (1) professionalised/traditional cartography and (2) counter-mapping (Wood 2010).
For the former, a participatory approach involves using general reference and thematic maps 2 and adding cartographic elements based on participants' spatial descriptions (for an example, see Brennan-Horley et al. 2010). As described by Crampton (2010), widespread democratisation of spatial data and tools has facilitated this type of work, opening new methodological approaches in qualitative and public participation geographic information systems (QGIS and PPGIS, respectively) (Brown and Kyttä 2014;Sieber 2006). The development of these professionalised participatory maps has been successful in the areas of indigenous, rural, and community development, environmental and natural resource planning and management, urban and regional planning, and in developing mapping tools and technology (for a review, see Brown and Kyttä 2018). While the outputs from this approach are likely to be appealing if relayed to technocratic and management-focused professional circles -areas of work that may have greatest familiarity and comfort in interpreting traditional grid-system maps -they do so at the expense of rich ethnographic information that may be collected through the mapmaking process. As recognised by Caquard et al. (2019: 1): the elusive geographies of memories don't easily overlap with rigid Euclidean structure of the conventional map…to map memories would inevitably require that memories be distorted in a way that fits a rigid cartographic structure, or to distort this structure in a way that would accommodate memories.
Similar to the internal tension vis-à-vis representation within anthropology, critical geographers have made laudable critiques of professional cartographers' claim to scientific, authentic, and 'true' representation of geographic space (Harley and Woodward 1987;Wood 2010;Wood and Fels 2008;Woodward and Lewis 1998). Through problematising these representational limitations, ' counter-mapping' has emerged as a new approach over the last twenty years or so. Counter-mapping is generally described as mapmaking practices that provide alternative representations of space with a distinct interest in formulating maps through a methodology that is inclusive of research participants (Hodgson and Schroeder 2002;Peluso 1995). In developing these counterrepresentations, maps produced within this scholarly area have also helped articulate counter-narratives about people's relationships, histories, and cultural affinities to spatial elements. Where there is increased scholarly interest in decolonising methodologies (Cruz 2008 (Perkins 2009: 127). In yet another development in counter-mapping, hand drawn and computer-assisted mental maps (also termed ' cognitive maps') have been used to help decipher people's attachment to and perspectives on places in the present or through the recollection of memories (Gieseking 2013).
While the diversity of possible mapping approaches can certainly be overwhelming, this equally demonstrates their ability to represent a diversity of social, cultural, and political settings. The use of topographic maps may be more appropriate when the key priority of a community, say dispossessed through land grabbing, wishes to spatialise ancestral land tenure. Counter-maps may be preferred in areas where local people have never had the opportunity to participate in formal mapping activities, perhaps as a result of historical colonialism that denied locals participation in mapmaking exercises. Research participants may choose to spatialise memories through mental maps or imagine their futures in a utopian iteration of such. Yet again, there need not be a divide between professionalised and counter-maps, as several scholars have devised sophisticated methods to merge technological/quantitative/'representational' spatial methodologies with more descriptive/qualitative/'non-representational' ones (Boschmann and Cubbon 2014;Caquard et al. 2015Caquard et al. , 2019Kwan and Ding 2008;Merschdorf and Blaschke 2018). Ethnographic mapping, thus, is purposefully flexible in this regard, as the spatial representation component of its craft is tightly bound to the process of ethnography and the understanding and negotiation of community preferences that transpire through it. This constitutes an important core principle of ethnographic mapping: that the visual be informed by the community through open participation and narrative ethnographic analysis. Another guiding principle includes a specific consideration for the analytical interdependency between mapping and writing, which are both tools that assist the ethnographer to understand how meaning unfolds over space and is constitutive of relationships, symbols, institutions, and the like. As noted by Perkins (2009: 128), 'by observing and participating in the performances around mapping we can explore its relations to identity, how different spaces are co-constructed, and the ways in which people behave when carrying out mapping tasks.' This ' ethnographic package' of visuals (maps) and narratives (descriptive writing) requires an earnest commitment by anthropologists to engage with mapping, geographers with ethnography, or practitioners and other scholars to appreciate basic principles of cultural geography. Here, users of ethnographic mapping should understand that while absolute representation is impossible, interdisciplinary tooling may lead to richer descriptions of people, places, and spaces. This is based on an appreciation of the vibrant synergies between geographical and anthropological theory that emerged at the end of the twentieth century, one that saw an increased appreciation for pluralistic views on place, multivocality, and the shifting nature of the relationship between geographic space and culture in research contexts increasingly impacted by dislocation, displacement, and deterritorialisation (Appadurai 1988;Feld and Basso 1998;Gupta and Fergusson 1992;Rodman 1992). As contended by Low (2014: xxii), mapping and spatial techniques augment classic ethnographic methods of participant observation and in-depth interviewing, making up the 'toolbox for deciphering the role of culture in the production and construction of space and place.' Ethnographic mapping stands as an interdisciplinary tool that can provide multiple routes to understand social realities. For participants, it offers multiple ways to communicate their perceptions. For ethnographers and other practitioners managing its implementation, the set of tools ethnographic mapping assembles leads to the output of rich multidimensional datasets. In the next section, we focus on this article's interest in developing a method that can collect and synthesise peace and conflict-relevant information from refugees in exile. A general model for the use of ethnographic mapping in refugee contexts will be provided.

A Generalised Model of Use in Refugee Contexts
We now turn to the elaboration of a proposed generalised model for implementing the use of ethnographic mapping in refugee contexts. As visualised in Figure 1, the process of ethnographic mapping can be seen as a two-pronged approach where spatial/mapping exercises are operating in tandem with narrative/descriptive composition (i.e., ethnographic writing).
As described in the previous section, mapping strategies should be based on a participatory model of decision making and community preferences, interests, and perspectives. It is thus advisable to undertake substantial participant observation prior to the use of mapping exercises. Participant observation is understood here as ' a longterm intimate engagement with a group of people that were once strangers to us in order to know and experience the world through their perspectives and actions in as holistic a way as possible' (emphasis in original text) (Shah 2017: 51). Within the context of forced migration research, this should also include 'the collection of detailed oral histories and life stories; by telling their own stories to the researcher, refugees themselves and the ethnographer her or himself are able to develop a narrative analysis' (Loescher 2014: 322). As described by Chapin and Threlkeld (2001), participant observation is also a key groundwork procedure in assembling a community of research participants that is broadly representative. Here, it is proposed that participants be divided into two broad groups: ' cultural brokers' who stand as representatives of different cultural groups within the refugee setting (Smith 2009), and ' community leaders' who crosscut these cultural boundaries.
Once these groups have been formulated, training may be necessary. In potentially volatile contexts, or in settings where internal tensions exist within the refugee community, conflict-sensitivity training of participants may be considered and desirable. Furthermore, as this may be participants' first experience with mapmaking, exposure and training on a variety of spatial representation methods are important in order for participants to make an informed choice on which mapping method(s) they want to use (Caquard et al. 2019;Cochrane, Corbett, and Keller 2014). During these training exercises, refugees should be encouraged to articulate what type of information they would like to map (for instance, memories about their place of origin, the journey of migration, prospects of return, etc.). These discussions and negotiations should be recorded through detailed ethnographic writing.
Once recruiting and training is completed, the first round of mapping exercises takes place. It is suggested that the cultural brokers group take the lead on this, where the visualisations they produce then undergo a form of validation through comments and observations by community leaders who act as intermediaries and negotiate between potential disagreements across cultural groups. These comments, shown to be greatly informative in past work involving mapping with exiled persons (Caquard et al. 2019), may be in the form of writing, drawings, or new maps. Between both activities, the facilitator (researcher, ethnographer, or practitioner) may need to make small interventions in order to synthesise the maps produced and simplify them for ease of use. Comments from the community leader group are relayed to the cultural brokers group and subsequent rounds of mapping are performed until participants are satisfied with the results, changing hands between cultural brokers and community leaders. Throughout the entire process, detailed ethnographic notes are taken on both the procedural development of the maps and the comments and negotiations occurring about the maps. These notes are particularly instructive in identifying themes and topics that cause disagreement and/or collaboration between participants.
Following the data collection stage, the facilitator develops a synthesis document that must consider both spatial representations and narratives in tandem. This document, or ' ethnographic package,' can be the result of inductive analysis where categories, themes, and conclusions flow directly from the data. However, the analysis can also be done deductively if ethnographic mapping is deployed to the field with the aim of answering specific questions or to gather predetermined variables of interest. Notably, the deductive approach may be useful in the application of ethnographic mapping to inform needs-based peacebuilding programming from afar. In this case, ground-level practitioners can benefit from the rich data collected on specific issues such as land tenure, property claims, natural resource, conflict perception, ethnicity and demography, contested territories, and the like.

Risks and Ethical Considerations
While the two previous sections define ethnographic mapping and provide a guide for its implementation, best practice, and possible context-based adaptations, this method remains largely untested. As is common in good scholarship, ethnographic mapping -like any newly proposed methodology -should be explored through the lens of its risks and ethical considerations. This section should serve as a guide for researchers and practitioners who wish to use ethnographic mapping in forced migration settings. Equally important, institutional review boards or ethics review boards can use the following discussion when evaluating projects proposing ethnographic mapping as a central data collection and/or theorybuilding method.
The risks and ethical considerations of ethnographic mapping meet at the intersection of research with forced migrants, participatory methodologies, and mapping. In the case of research involving asylum seekers, refugees, stateless persons, and internally displaced persons, Clark-Kazak (2017) includes equity, right to self-determination, competence, and partnership as key guiding principles for research with people in situations of forced displacement, further aligning these principles to the practices of voluntary and informed consent, confidentiality and privacy, and minimising harm while maximising the benefits of research. This unique context of research with especially vulnerable persons, however, adds to the complexity of these considerations. While vulnerability is often of greatest concern while working in the field, Mackenzie, McDowell, and Pittaway (2007) ask that forced migration researchers see beyond the lens of vulnerability in order to explore how their work risks entrenching existing hierarchies that threaten participants' autonomy. This includes hierarchies that go beyond the community, and those that create power dynamics between participants and researchers, undermining the principle of free, voluntary, and informed consent (Hyndman 2000). Long-term ethnography where the researcher incrementally becomes closer to the community can certainly help shore up these concerns, but this is largely dependent on the ethnographer's field abilities and their commitment to consistently revise the ethical implications of their research. Participatory methods can also be instrumental in subverting power dynamics on the field, but no single method is the road to the El Dorado of total inclusion. Willingly or not, participatory methods such as those suggested for ethnographic mapping will nevertheless give voice to some while denying or obscuring those of others, and thus how they are used should be critically re-evaluated and adapted throughout the duration of field work.
Similar to our best efforts to develop inclusive methodologies, claims that ethnographic mapping produces accurate -or true -sociocultural representations should be examined with scrutiny. As explored in earlier parts of this article through the example of professional cartography, absolute representation is not possible. It is important that the material and symbolic value of maps produced through ethnographic mapping be seen in these terms. In their critical review of disaster-response participatory crowd mapping, Bittner, Michel, and Turk (2016) articulate that despite widespread rhetoric on the inclusivity of crowd mapping, this practice is usually organised and mostly used by privileged persons least affected by crises and disasters. They ask critical geographers to 'look at the crowd,' its composition and power structures, and evaluate the extent to which participatory methods of this kind are intrinsically participatory or -perhaps -entrench certain access barriers to vocality and participation. At the convergence of mapping technologies and the 'representational trap' are the potential risks of producing ethnographic maps that may express new forms of territorialisation or reify power dynamics. As Fox et al. (2016) observe, maps are tools of technology and 'technologies are complex systems promoting and institutionalizing relational patterns aimed at realizing particular ends' (206), and so ' even if the community can control the maps, it is important to understand the multiple interests and actors found within communities; and the political and economic relationships between communities and other social actors' (210). Thus, while ethnographic mapping embraces the notion that research participants be part of the data collection process, rather than simply act as sources of information, outputs generated through this process should not be assumed to be politically or culturally inert. This is where cycling between cultural brokers and community leaders (and perhaps others), as described in the previous section, is of great importance to the process of ethnographic mapping and the documenting of power dynamics between participants.
Despite the participatory penchant of ethnographic mapping, the way it takes form on the field and how its outputs are interpreted undeniably involve the researcher, ethnographer, or practitioner. This is well understood by some fields of knowledge and practice, and hence why forced migration anthropologists commit to long-term fieldwork, as there is wide agreement that with extended contact comes familiarity and trust with the community (Jacobsen and Landau 2003). Hermann (2001) warns that the researcher is not a tabula rasa devoid of biases or proclivities -that researchers in conflict settings are at the centre of the struggle between positivist/objective and hermeneutist/subjective epistemologies, travelling between the role of 'insider' and ' outsider,' the process of personal reflexivity and participants' personal histories, and how these align to this scale of involvement/investment. In accounting for this, ethnographic mapping embraces the principle of poly/multivocality (Clifford and Marcus 1986;Rodman 1992), but as Emerson, Fretz, and Shaw (2011) point out, how these voices are translated into meanings and understandings is filtered through mediating effects (Mills 1990) that align to one's positionality. Where misrepresentation stands as a risk to any form of ethnographic and qualitative research, there must be conscious recognition of biases and subjectivities. Research design should thus include measures, protocols, and opportunities for the community of participants to dispel, enrich, or correct the researcher's interpretations and perceptions they deem erroneous.
Reciprocity should also make up a key tenet of best practice to conducting ethical ethnographic mapping research. In the case of forced displacement research settings, recent literature has focused on the harmful, exploitative, and potentially retraumatising impact of 'parachute researchers' who swiftly enter research sites, extract data, and then speak/write authoritatively despite having established few meaningful relationships with those being spoken/written of or about (Migration Leadership Team 2019; Pittaway, Bartolomei, and Hugman 2010;Talhouk et al. 2019). While Smyth (2001: 5) characterises research that treats people as objects as ' ethically questionable, ' Mackenzie, McDowell, and Pittaway (2007: 301) suggest that mere 'minimisation of harm' as an ethical standard is insufficient; researchers should instead 'recognize an obligation to design and conduct research projects that aim to bring about reciprocal benefits for refugee participants and/or communities.' They continue, arguing that 'if researchers are in a position to assist refugees to advocate on their own behalf or on behalf of others…then it is morally incumbent on them to do so [but that] such involvement must be at the request of, or willingly agreed by, the refugees concerned" (ibid.). In this case, however, the researcher is faced with what Jacobsen and Landau (2003: 186) call the ' dual imperative': 'to satisfy the demands of academic peers and to ensure that the knowledge and understanding work generates are used to protect refugees and influence institutions like governments and the UN.' This brings us closer to understanding the distinction between ' ethical research' and 'research ethics,' where the latter is typically of greatest concern and follows certain sets of universal criteria such as voluntary and informed consent and the principle of ' do no harm,' whereas the former is more so a judgment of the intrinsic value of a research project in advancing normative objectives.
While other risks and ethical considerations certainly remain, the above discussion provides some signposts to the prospective user of ethnographic mapping and ethics committees that review such research projects. Free, voluntary, and informed consent and issues of confidentiality and privacy were not addressed here, largely because they are well-entrenched ethical norms. Established guidelines should be strictly followed, as is common practice in all forms of research involving human participants.

Conclusion
Literature on peace and conflict has often characterised return migration as a cause of new grievances and an important factor in destabilising fragile peace in post-conflict environments. Other literature in the area of refugee studies has asserted that while in exile, refugees maintain a strong attachment to their homeland. In an attempt to draw out practical intersections between both of these observations, and recognising that return migration remains the preferred durable solution for states and refugees alike, this article contends that more comprehensive information-gathering strategies in exiled communities can help the peacebuilding community understand the needs and claims of refugees prior to their return. As has been argued, data generated through these sorts of consultative exercises could prove critical in devising comprehensive peacebuilding approaches that address the 'return migration as conflict generating' problem head on. Scholars recognising this need have advocated for greater integration between localised peacebuilding initiatives and refugees at the margins. Milner (2015) has proposed three areas of research and policy that could further bridge the relationship between refugees in exile and homeland peacebuilding: (1) integration of displaced persons in reconciliation programmes; (2) greater interaction with and investment in refugees from the peacebuilding community; and (3) more cross-pollination between peacebuilding literature and refugee studies to develop rigorous understanding of their intersection and develop more robust intervention models. While these normative proposals can be reached through a diversity of programmes for research and practice, this article has developed ethnographic mapping as an approach that can contribute to the fulfilment of these priorities.
Through the process of mapmaking and descriptive notetaking during these exercises, ethnographic mapping can generate rich spatio-narrative data at the nexus of peace and conflict. Its outputs can be instructive to peacebuilders on the ground, whereby formulations of space and power based on memories, hopes, desires, and claims can help inform responses and pre-empt potential peace-crumbling conflicts in the event of voluntary repatriation. As an approach anchored in anthropology and geography, ethnographic mapping is founded on concepts of great interest to the peacebuilding community, such as: 'space' as geographic orientation within and between sites of social activity (Harvey 2005;Jiménez 2003;Massey 1993); 'place' as meaning, emotion, experience, and attachment to space (Agnew 1987;Cresswell 2004Cresswell , 2006Cresswell , 2009Entrikin 1991Entrikin , 1997Low 2009Low , 2014Low , 2017Packwood 2001); 'territory' as the distribution and negotiation of space through power (Appadurai 1996;Duarte 2017;Munn 1996;Vaccaro, Dawson and Zanotti 2014); and 'landscape' as the ever-changing character of interactions between people and their natural environment (Cosgrove 1998;Hirsch 1995;Howard 2011;Sauer 1925). 3 In the analysis of data generated through ethnographic mapping, these conceptual tools have a great interpretive capacity to describe refugees' perceptions and experience of power dynamics, how geographic space is negotiated, and potential pathways towards the reconciliation of conflicting identities. It is not without limitations, however, as this approach requires its implementor to have a certain degree of expertise in the practice of ethnography. An equally important limitation of ethnographic mapping is the time it takes to fully implement: the time period from arrival to the research site to the output of useful data for peace practitioners can potentially span several weeks, if not months. As a result, ethnographic mapping as a 'slow approach' does not seek to compete with fast, rapid-response, tools that focus on short-term outcomes. This is not to say, however, that qualitative data resulting from ethnographic mapping could not foreseeably enrich quantitative information gathered by 'fast approach' tools such as IOM's Displacement Tracking Matrix.
In this article, ethnographic mapping has been largely conceptualised as a datagathering tool -a methodology to better understand displaced persons' needs and claims, and how these relate to conflict landscapes in transitions towards peace. This is largely a functional response to the current state of peacebuilding where refugee perspectives are seldom considered in decision making. Ethnographic mapping can help bridge this gap. In the long term, however, ethnographic mapping could foreseeably facilitate refugees' direct participation in peacebuilding programmes, stimulating their engagement, instilling feelings of agency, and giving them opportunities to have ownership over conflict termination and durable peace.

Notes
1 For further literature on the importance of housing, land, and property restitution in post-conflict settings involving refugees, see Leckie (2007Leckie ( , 2009). 2 Described by Eckert and Joerg (1908) as the two primary classes of professional maps. 3 Some of the authors listed discuss these concepts in relation to each other.