Humanitarian Shapeshifting: Navigation, Brokerage and Access in Eastern DR Congo

ABSTRACT

This article examines the experience of Congolese humanitarians negotiating access with armed groups in eastern DRC. It describes how humanitarians become shapeshifters: they play different roles for different audiences as a tactic of social navigation in a context of uncertainty. Because humanitarians encounter risks related to their perceived identity, they play upon identity categories and personal histories, situating themselves tactically during encounters with different armed actors through creative modes of self-fashioning. Whilst there remains a focus on performing a distinct humanitarian identity, NGOs in practice draw upon the embedded local knowledge and skills of employees to work in conflict areas.

KEYWORDS:

• Humanitarianism
• conflict
• security
• DRC
• negotiation
• identity

Introduction

Dieudonné¹ is an experienced humanitarian who began his career over a decade ago as a driver for an international NGO in North Kivu, a province in the east of the Democratic Republic of the Congo (DRC), which has been at the epicentre of protracted violence in the Great Lakes for the last twenty-five years. Humanitarian agencies have been providing assistance in the region for decades. In order to operate, they need to negotiate access with a range of different political authorities and armed groups, who control different areas in a manner that shifts and evolves over time. Such negotiation is not an ‘event’ involving the head offices of international NGOs, but a continuous everyday process conducted by the humanitarians on the ground, the majority of whom are ‘local’ employees. I had come to talk to locally hired humanitarians, such as Dieudonné, to learn more about how this works in practice.

Dieudonné drove a white Land Cruiser – the vehicle that features in NGO fundraising material, transporting people and assistance to those ‘in need.’ But being a driver also included a set of social practices. Humanitarians like Dieudonné possess a situated savoir faire which helps them to read the political context, build contacts with a range of different armed authorities, and negotiate at roadblocks. Key to this work, Dieudonné explained, was ‘trying to play different roles to different people.’ When brokering humanitarian access, Dieudonné described the importance of ‘emphasising’ or ‘selecting’ different aspects of his identity and history (whilst tactically hiding others) depending on who he was talking to. In effect, Dieudonné described a form of creative identity work as central to humanitarian access: he tactically positioned himself in relation to different actors as a means of navigating a conflict environment.

This article describes this creative identity work as shapeshifting: a process of tactical self-representation which constitutes a form of social navigation in contexts of uncertainty (Utas 2005; Vigh 2008). Because people encounter risks or possess advantages depending on how they are perceived, they construct and reconstruct their identities in real time for different audiences in order to facilitate the delivery of aid, and to manage their own safety. Despite the humanitarian sector’s focus on constructing and performing an exceptional and distinct humanitarian identity as a legitimising tool to work in conflict settings (Abu-Sada 2012), on the ground, access also depends on who is representing an organisation, and the way that person is received and perceived in relation to the political histories of the environment in which they are working. Rather than being neutralised by an NGO’s logo, people’s personal identities, histories and networks bring advantages and risks for different employees. Tactical self-representation which draws on these personal identities therefore becomes an everyday tool for negotiating humanitarian access.

Below, I set these ideas in a theoretical background and explore them through the case of Dieudonné and that of another Congolese humanitarian, Benjamin. The article draws on 15 months of ethnographic fieldwork in eastern DRC between 2017 and 2021. During this time, I conducted nearly 250 interviews with humanitarians, political authorities and former rebels in Goma, Bukavu, and the Petit Nord territories of North Kivu. The analysis is based on these interviews and on informal conversations with Congolese humanitarians working to facilitate the delivery of aid – drivers, logisticians, or experts in ‘security management.’ I give prominence to the perspectives of these ‘local’ humanitarians who are fundamental to how NGOs work in conflict, but whose experiences are rarely foregrounded in institutional narratives or the academic literature on intervention (Peters 2020). In North Kivu, I negotiated my own access to what one interviewee described as ‘the Congolese space of aid’, which operates alongside the world of mobile ‘expatriates.’ I first came across this ‘space’ when I accidentally walked in on a meeting of a group of Congolese humanitarians who meet once a week to analyse security issues away from their ‘muzungu bosses’ (foreign, mobile, and often white ‘expat’ colleagues). Eventually, I earned the trust of some in this group and was invited to attend meetings as an observer. Many had worked in the sector for more than a decade, and over time, they shared their personal stories with me. Those mentioned in this article have been anonymised.

The article is structured as follows. First, I situate the concept of shapeshifting in the literature on social navigation, impression management and brokerage. The second section sets out the complex social terrain which Congolese humanitarians such as Dieudonné and Benjamin navigate. It describes the slippery notion of ‘autochthony’ which shapes, and continues to be shaped by, processes of violent conflict and illustrates how such notions of ‘autochthony’ play out in the humanitarian sector, bringing advantages and risks for different employees. The third section draws on two case histories to describe how Congolese employees navigate this complex social terrain by shapeshifting. I conclude that shapeshifting is an example of the creative processes of social navigation conducted by people in contexts of uncertainty, and part of the relational and interpretive labour of local humanitarians (Peters 2020) which often remains hidden behind the discourse on humanitarian principles.

Shapeshifting: Performance, navigation and brokerage

Shapeshifting refers to changing one’s shape or form into another. In order to facilitate the delivery of aid in a violent context, people tactically construct their identities for different audiences as a form of social navigation. I adopt the concept of shapeshifting to describe this process, after one humanitarian explained that he could switch roles because his ‘morphology was flexible.’ Shapeshifting incorporates several concepts – social performance (Goffman 1956), social navigation in war (Vigh 2008) and brokerage (Bierschenk, Chauveau, and de Sardan 2002).

Aid work, like all social practice, is performative: it involves impression management. Goffman (1956) describes how, when an individual comes into contact with other people, that individual tries to manage the impression that others might have of them by fixing the setting, their appearance and manner. By doing so, performances communicate information that confirms the identity of the actor in that situation. Most performances of social life are neither cynical nor sincere, but somewhere in the middle. Some people misrepresent or hide things; others maintain a distance as a means of projecting an idealised image to the audience. Crucially, we all segregate audiences: people put on different performances, or emphasise different parts of themselves, for different groups of people. Different settings require us to adjust: identities are constructed in tactical ways when interacting with different people (Goffman 1956). People working as humanitarians have overlapping social roles, especially when working in areas where they have their own histories. They perform different roles to different people (Rossi 2006; Heaton Shrestha 2006). Shapeshifting describes the process of shifting between these roles and constructing different identities during social interaction.

In a sense, everyone shapeshifts – for instance, few of us follow the same social script at work as when we are out with friends. Yet, in volatile environments, shapeshifting is not just a way to interact in everyday life but also a form of social navigation and tactical agency in a context of uncertainty. As Vigh (2008, 5) describes, crisis should be approached as context: a chronic condition, ‘a terrain of action and meaning’, rather than a singular event. This conceptualisation of crisis reveals how people navigate through volatile environments, where social positions ‘are reconfigured and reshaped in relation to stable instability and chronic crisis’ (Vigh 2008, 13). Here, people continually assess their social environment as well as how they ‘position themselves in relation to others’ (Vigh 2008, 20). In effect, tactical forms of self-representation and self-staging become a practice of social navigation: a form of tactical agency in uncertain and disempowering circumstances (Utas 2005, 403).

This observation is particularly relevant for a context like eastern DRC, where people live by the ethos of débrouillardisme - ‘fending for oneself’ or ‘muddling through.’ After decades of economic downturn, political volatility and protracted violence, people have developed improvisational methods for manoeuvring through an informal economy and complex social terrain (Trefon 2004). Practices of débrouillardisme rely on processes of reinvention and interpersonal relationships (Ayimpam 2014). In eastern DRC, moving between different social positions is itself a form of débrouillardisme: there exists a revolving door between civilian and combatant life, as well as rebel groups and the national army, as people navigate different positions and remain part of multiple social environments (Vlassenroot, Mudinga, and Musamba 2020; Eriksson Baaz and Verweijen 2013). People working as humanitarians in the area also switch social positions and maintain multiple ties. For many, joining the growing aid sector was itself a means of ‘muddling through’ and finding employment. But shapeshifting in everyday aid practice aims to accomplish particular ends: the delivery of assistance in a context of conflict. The penalty for getting this social positioning wrong can potentially be life-threatening both for humanitarians and the people who depend on the delivery of aid.

The concept of brokerage helps theorise how shapeshifting operates during humanitarian negotiations with armed groups. The defining feature of brokers is that they are well-connected and straddle and connect social worlds, negotiating competing pressures and expectations (James 2011). To do so, they construct tactical identity positions. The concept has been revived to explore the fragmented politics of postcolonial governance (Themnér and Utas 2016) and state power (James 2011), and extended to describe development workers situated in between aid funders and recipients in Africa (Bierschenk, Chauveau, and de Sardan 2002). These brokers become ‘identity entrepreneurs’ who learn to ‘change roles’ and ‘play on collective identities’ for their own advantage (Bierschenk, Chauveau, and de Sardan 2002, 28). I describe how humanitarians become creative brokers operating in the fragmented governance networks of contemporary conflict. These individuals are not brokers operating in ‘the middle’ but are members of aid organisations who are also embedded in local political histories and networks. They conduct creative processes of self-fashioning in order to broker access for their aid organisation and to manage their own security. Humanitarians are involved in these processes of impression management because access depends on the tactical deployment of different identities.

This sits in tension with the focus in the humanitarian sector on performing a distinct and exceptional humanitarian identity in order to work in conflict. After a series of security incidents in the early 2000s, humanitarians recognised a ‘perception gap’ between how they viewed themselves and how they were viewed by those with whom they interacted – this ‘gap’ was understood as harmful to security (Abu-Sada 2012). Impression management became key to security management (Givoni 2016). Despite the limited evidence that the principles of neutrality, impartiality and independence act as protective shields that guarantee ‘access’ (Magone, Neuman, and Weissman 2011), on the ground, they remain framed as important symbolic tools for working in conflict (Terry 2011) which need to be translated and then performed to onlookers through the everyday behaviour of employees, both inside and outside work hours (James 2020; Givoni 2016). In effect, humanitarians are supposed to present themselves as ‘blank slates’ with no history, no links, or allegiances – a form of ‘civilian plus’ which clearly distinguishes them from political and military actors (Sutton 2020). Codes of conduct define appropriate behaviour for employees as a matter of security (Beerli 2018). The T-shirt becomes a symbol of the individuals’ separateness from broader societal conflicts – whatever the identity of the person, their organisation and its principles are supposed to become their new ‘identity marker’ (Hilhorst and Schmiemann 2002). Yet, as this article shows, humanitarians do not simply rely on the performance of a detached humanitarian identity in order to work in conflict. Instead, the deeply personal and political identities of people in those NGO T-shirts matter: different political and social histories, networks and perceived identities affect who gains trust and wields power and influence, and who is at particular risk. Identities of different people working for NGOs become tools for access, because access depends on who is representing an organisation, and how that person is perceived in relation to the histories of the environment in which they are working.

This article thus contributes to the literature on the politics of everyday aid practice. The way that the humanitarian principles of neutrality, impartiality and independence are translated in practice is subjective (Hilhorst and Schmiemann 2002): decisions are guided by political culture and institutional interests as well as personal interpretations of principles in practice, as aid workers ‘fumble in the dark’ (Autesserre 2014). Humanitarian access relies on negotiation, interest seeking and compromise (Magone, Neuman, and Weissman 2011; Hilhorst and Jansen 2010). A complicated politics of difference operates in humanitarianism: procedures based on ideas of universal humanity often reproduce inequities between intervenors and the people they are supposed to help, as well as between humanitarians themselves (Redfield 2012; Fassin 2007; Benton 2016; Pailey 2019). For example, Read (2018) examines ‘difference’ as a gendered relation of power in aid, explaining how women feel distanced from the imaginary of an aid worker, and therefore try to perform their aid worker identity. ‘Aid worker’ identity is not a homogenous category that overwrites ‘all other kinds of difference’ (Read 2018, 305). Instead, gender intersects with race and nationality to produce forms of situated power, or disadvantage, in particular contexts.

Indeed, this article illustrates how forms of difference bring advantages or risks for different people working for the same humanitarian organisation who broker access in conflict. Pottier’s ‘roadblock ethnography’ (2006) explicitly explores these dynamics, describing the importance of Congolese humanitarians working for NGOs in negotiating with armed groups at roadblocks. In Ituri, he describes how ‘relief workers in conflict zones do not (and cannot) shed their ethnic identities, instead they accept that a perceived ethnic identity brings strategic advantages as well as disadvantages’ (2006, 252). Local staff had no choice but to ‘position themselves politically’, accepting that they had meaningful identities in the micropolitics of the zone (2006, 176). Perceived ‘ethnic’ identity impacted on staff safety and how they were perceived in relation to histories of violence, so NGOs had to adapt.

I build on Pottier’s work but focus on the ambiguity of identity: there is a degree of performativity involved. In her examination of peacebuilding organisations in DRC, for example, Martin de Almagro (2018) coined the concept of ‘hybrid club’ to describe clusters of local and foreign actors with complex identities who choose to perform sameness or difference to audiences depending on their particular goals in that context. As she illustrates, difference is ‘performative and relational’, it is created through encounters, and the same ‘clubs’ can present themselves differently in different contexts. Categories such as ‘local’ and ‘international’ need to be deconstructed, because ‘difference and commonness come into existence through the relational process of identity (re)production’ (Martin de Almagro 2018, 324). Given the advantages of being seen as an ‘outsider’ or a ‘local’ at different times, people construct their identities for different audiences whilst performing a humanitarian role. Some may distance themselves from the ‘local community’ to manage contradictory demands (Rossi 2006), whilst others purposefully construct a ‘local’ identity by hiding parts of their histories in order to qualify for ‘local’ contracts (Peters 2016). Humanitarian practice involves the strategic production of different identities (Heaton Shrestha 2006), as people ‘adjust strategically and instrumentally to their ascribed roles’ (Rossi 2006, 29). Below, I describe how people perform different forms of ‘localness’, depending on the armed group in front of them, in order to navigate a complex politics of belonging in eastern DRC. Humanitarians try to situate themselves tactically within politicised discourses of insiders and outsiders.

The politics of belonging in eastern DRC

Conflict in Kivu

This section describes the complex social terrain in which people shapeshift in eastern DRC. Conflict in the North and South Kivu provinces of eastern DRC became regionalised after the influx of Rwandan refugees following the genocide in 1994, and the Rwandan army launched attacks against refugee camps to eliminate former génocidaire elements, which later reorganised as the Forces Démocratiques de Libération du Rwanda (FDLR). In 1996, the Rwandan army and L’Alliance des Forces Démocratiques pour la Libération du Congo/Zaïre (AFDL), invaded Congo and overthrew President Mobutu. The ‘second war’ began in 1998, when the Rwandan-backed Rassemblement Congolais pour la Démocratie (RCD) invaded the Kivus. The conflict escalated to involve eight countries and more than twenty-five armed groups. Although the war ended in 2003, violence continued, with the rebellions of the Rwandan-backed Congrès National pour la Défense du Peuple (CNDP) between 2006 and 2009, and the Mouvement du 23-Mars (M23) between 2012 and 2013. Since 1996, at least 270,000 people have been going in and out of armed groups (Vogel and Stearns 2018). There are rarely clear frontlines, but rather shifting alliances and rivalries in an increasingly fragmented landscape of armed groups. Networks of state and non-state actors govern this region through complex negotiations (Vlassenroot and Raeymaekers 2008).

During this prolonged conflict, eastern DRC has become the base for a plethora of NGOs and the UN’s largest peacekeeping mission which aims to bring ‘stability’ to the region (Autesserre 2014). Aid organisations have transformed the economic, social and political landscape, dominating the political economy and becoming a key source of employment in the region (Büscher and Vlassenroot 2010). Humanitarian agencies and armed groups are two persistent features of the social landscape and have interacted for years. In order to work, NGOs must negotiate access with a range of political and armed authorities. NGOs rely on the networks and savoir faire of their Congolese employees in order to operate (James 2020). In this context, the ‘fixing’ sector is on the rise: the ability to network and broker has become a valued skillset in the job market. People have joined the humanitarian sector from a diverse range of political and social backgrounds - including former rebels themselves.²

‘Autochthony’ and armed mobilisation

Humanitarians work in a rebel kaleidoscope in the Kivu provinces, with over 100 armed groups which shift and fragment over time. The majority of these groups are local ‘Mai-Mai’ which draw on discourses of self-defence and ‘autochthony’. Mai-Mai groups mobilised in North Kivu in the 1980s, and then proliferated during the first and second Congo wars, and again during the CNDP and M23 wars. Mai-Mai groups draw on notions of belonging to legitimise their violent practices (Verweijen 2015): in particular, discourses of autochthony, which refers to the supposed natural belonging of autochthones to the ‘soil’, in contrast to foreigners (Hoffmann 2021; Jackson 2006). Mai-Mai groups are rooted in specific imagined ‘ethnic’ communities and claim to defend those ‘born from the soil’ from foreign Rwandan invaders who threaten their power, land, livelihoods, and survival (Hoffman and Verweijen 2019). A key technique of Mai-Mai power is ruling through ethnicity and custom – collaborating with customary chiefs and appealing to the language of autochthony to make sense of recent political and military developments in the region (Hoffman and Verweijen 2019).

Autochthony is a slippery concept – politically manipulated and itself a colonial import (Jackson 2006). The Kivu region is home to dozens of groups: some speak Kinyarwanda (the language in Rwanda), but this group is not homogenous. It is made up of ‘Hutu’ and ‘Tutsi’ who settled in Congo before the colonial period, immigrants who were settled in the Kivu as labour for the colonial administration, and refugees who fled periods of violence in Rwanda since the 1950s (Mathys 2017). Intertwined notions of territory, ethnic identity and customary authority both shape, and are shaped by, armed mobilisation (Verweijen and Vlassenroot 2015). The political salience of autochthony and ethnicity was forged during the colonial era. Through remapping the frontier zone of eastern Congo into distinct, ethnic homelands, colonialism transformed relationships between people and space (Hoffmann 2021). Communities were reorganised into monoethnic chiefdoms with specific ‘homelands’, defining who belongs where and who gets what (Mamdani 1996). In other words, territorialised ethnic identity became a means of accessing the state and resources. This had a profound impact on political order in the region, as well as local subjectivities (Hoffmann 2021; Hoffmann, Vlassenroot, and Mudinga 2020). In land-related struggles, autochthones claim to have ‘arrived first’, whereas populations lacking ‘homelands’ recognised during the colonial era are branded as ‘immigrants.’ At a territorial level, autochthony pits ‘autochthonous’ Nande, Hunde, Tembo and Nyanga against speakers of Kinyarwanda. At a provincial level, it refers to authentic provincial identity, which became increasingly politicised in processes of decentralisation. At a national level, autochthony has become entangled with debates about citizenship and electoral participation (Nguya-Ndila Malengana 2001).

In Masisi, the territory in North Kivu where Benjamin and Dieudonné worked, autochthony continues to shape contemporary political disputes for several reasons: because of the continued centrality of customary authority to govern rural areas, and because such discourses of self-defence against ‘Tutsi invaders’ flourished during a series of Rwandan-backed armed rebellions since 1996. The Belgian colonial regime created the Bahunde Sector or Chefferie in 1921 which rapidly expanded and included all of Masisi by 1935. Subsequently, the colonial administration oversaw mass Rwandan immigration to provide labour for mines and plantations, especially where the Hunde population was reluctant to work for settlers. In order to accommodate these immigrants, the Belgian administration purchased approximately 45,000 hectares of land from the Hunde Mwami for a small sum, and Hunde were obliged to leave (Stearns 2012, 13–20).

This sowed the seeds of conflict: Hunde authorities resented the loss of their land, and tensions led to the outbreak of violence during periods of democratisation and contests over authority and resources in the 1960s and 1990s (Willame 1997; Mararo 1997). In the discourse of so-called autochthones, the land sale and the mass immigration during the colonial era mark the beginning of Rwandan domination, which has continued with the invasion of Rwandan-backed armed groups (AFDL, RCD, CNDP and M23). In this article, Benjamin and Diuedonné describe their work during the CNDP and M23 conflicts. In 2006, Nkunda launched the CNDP armed movement drawing from ex-RCD officers with Rwandan backing and installed himself in Kitchanga, Masisi. The group claimed to defend the rights of the Tutsi community and combat the FDLR – the former génocidaires (Stearns 2012). In response, a dozen Mai-Mai mobilised in the area, in particular the Patriotes Résistants Congolais (PARECO), a collection of militias who saw themselves as defending ‘autochthones’ against ‘Tutsi domination.’ In 2007, PARECO began cooperating sometimes with the FDLR and FARDC against the CNDP, and in 2008, the Hunde faction of PARECO created the Alliance Patriotique pour un Congo Libre et Souverain (APCLS). APCLS brand themselves as ‘sons of the soil,’ once again defending autochthones from CNDP and then M23 foreign invaders (Stearns 2013).

Humanitarian ‘sons of the soil’

Humanitarians working in this region are not somehow immune from politicised discourses of belonging. At the outset, who gets to represent an NGO is a fraught issue. Humanitarian agencies are powerful actors in the political economy they bring resources and employment and become entwined in local power contests. Aid workers in North Kivu, for instance, described chronic tensions surrounding recruitment. One programme manager explained, ‘many security problems came from recruitment, accusations of partiality,’³ and a humanitarian from Masisi similarly described ‘balanced’ recruitment as essential to ‘security and access. If people saw our hiring as partial, things would degenerate very quickly.’⁴

The identities of Congolese employees were politicised using discourses of autochthony. Mai-Mai groups and their associated civilian networks, as well as local civil society groups and political authorities, contested the employment of ‘foreigners’ over ‘autochthones’ on several scales. At a territorial level, NGOs were accused of favouring one group over another. Sometimes, this was expressed in ‘ethnic’ terms. In Rutshuru, one NGO faced protests at the perceived favouritism of the Nande community in their recruitment, whilst in Masisi, there were tensions surrounding recruitment between Hunde and Hutu communities.⁵ At a provincial level, agencies faced criticism for the ‘importation’ of non-originaire staff from other provinces. On several occasions in Masisi, youth groups blocked NGOs from entering the territory in opposition to this ‘importation’ of workers,⁶ whilst NGOs in Rutshuru received anonymous letters objecting to the recruitment of non-originaires.⁷ In the end, the inter-community Baraza (a group of leaders from the province’s major ‘ethnic’ groups who resolve low-level conflicts) suggested that NGOs recruit ‘locals’ to ‘avoid insecurity.’⁸

Given the politicisation of personal identities, Congolese humanitarians described the need to decide who was going to represent an NGO during interactions with different armed actors, considering how that person may be perceived in relation to an armed group’s discourses, and the potential impact on the NGO’s access.⁹ Personal identities and histories brought advantages, and risks, depending on the situation. Perceived ethnic identity was one element: it mapped onto these discourses of autochthony, as well as histories of past violence and armed rebellion. Humanitarians from Masisi, for instance, stressed that they were sometimes associated with the armed group ‘from my community.’ During times of tension, this could place them at risk: ‘the problem is we are viewed as implicated in historical violence’, one aid worker explained.¹⁰ NGOs often travel in ‘balanced’ convoys (Pottier 2006). One driver in Masisi described this process. First, he placed a well-connected so-called ‘autochthonous’ employee in the convoy as a ‘representative’ – this employee had a form of ‘ethnic capital’, a political resource which marked their status and inclusion in that political space (Hoffmann, Vlassenroot, and Mudinga 2020, 127). This was then balanced with the inclusion of another employee – for instance, someone with the ethnic identity of an opposing armed actor. The driver explained that the humanitarian identity of ‘local’ employees was sometimes questioned by armed groups, but ‘when they see one of their brothers with us’ – someone who is known locally – ‘they think ok, it’s one of ours.’¹¹

Other humanitarian employees had useful personal links which marked their significance in a particular political space. In Masisi, one NGO employee told me that he was wearing ‘two hats’ or performing two roles simultaneously: customary authority and humanitarian. He was related to the Bahunde customary chief (Mwami) and argued that his overlapping position helped with access and ‘acceptance’: people trusted and confided in him. He had an extensive network among Mai-Mai groups who respect and collaborate with customary authority, as well as the army and state authorities.¹² Humanitarians like this logistician are not just technicians, nor are they simply well-connected brokers. Instead, they are humanitarian professionals who are also embedded in authority structures. This logistician possessed a symbolic form of identity-based resource, or ‘customary capital’, which enabled him to exercise authority through the notion of ‘custom’ (Hoffmann, Vlassenroot, and Mudinga 2020, 128), to the benefit of his NGO employer.

Whilst the identities and networks of some employees became tools for access, for others, their personal history, situation in society, or their perceived identity could place them in danger. For instance, in Pinga, a town on the border of Walikale and Masisi territories in 2013, power shifted between Nyanga- and Hunde-dominated armed movements, the Nduma Défense du Congo (NDC) and Alliance des Patriotes pour un Congo Libre et Souverain (APCLS). Locally hired humanitarians found themselves associated with one armed group or the other. When NDC took power, many Hunde humanitarians fled, and those who stayed were accused of being APCLS spies. When humanitarian convoys were travelling to APCLS areas outside of town, the APCLS refused to let them pass with a humanitarian employee from the Nyanga community in the car.¹³ Other aid workers described the sensitivity of being labelled ‘a Tutsi’ whilst working in North Kivu, given the autochthony discourse among Mai-Mai and their opposition to supposed ‘invaders’ from Rwanda. One Congolese logistician from North Kivu described the difficulty of working as a Tutsi: ‘armed groups don’t look at the T-shirt you are wearing. We have a problem with ethnicisme here, a gift from the Belgians. I don’t feel safe working in Mai-Mai areas.’¹⁴ Indeed, given the sensitivity of an association with Rwanda, many NGOs do not hire Rwandan nationals for positions in eastern DRC.¹⁵

Case histories: from sons of the soil to shapeshifting

The above stories illustrate how personal identities impact the security of humanitarian employees. However, such identities were not necessarily fixed or self-evident: people tried to play with the slipperiness of autochthony itself. Below, I examine two case histories – the stories of Dieudonné and Benjamin – to illustrate how people working for humanitarian organisations shapeshift in a specific political landscape. Whilst working in Masisi territory, Dieudonné and Benjamin tactically positioned themselves in relation to different armed actors, playing on constructions of autochthony, but also personal and military histories and networks. These personal histories also become a resource, or risk, in a way which complicates notions of autochthony.

Dieudonné

I was put in touch with Dieudonné by a former colleague of his, who told me that Dieudonné had been pivotal to their humanitarian organisation’s ability to operate in the CNDP conflict. Over brochettes and chips in Goma, Dieudonné told me about his decade-long experience of working as a humanitarian – first as a driver, then as a logistician. Dieudonné grew up in an army family and moved around the country. However, he spent his late adolescence in Goma. During the second war, Dieudonné had worked for the RCD rebel administration which controlled the Petit Nord territories. When the war ended in 2003, Dieudonné had hoped to pursue a political career. However, he found few openings in the new political scene. In 2007, he was in Goma after an unprofitable business venture and found a job as a journalière driver for an international NGO. He travelled to Masisi to begin his new career, which was at the epicentre of CNDP rebellion at the time. Dieudonné told me that his friends thought he was ‘mad’ to go and work in Kitchanga, but that he was curious. ‘Being a driver was always just a steppingstone,’ he explained, ‘it was a way into the growing aid sector, which seemed the best option at that time.’

In 2007, Dieudonné arrived in Kitchanga, a town controlled by the CNDP. Dieudonné was on a ‘local’ contract, but Masisi was quite foreign – he had never been. He remembered the journey from Goma to Kitchanga vividly, the rolling green hills and dramatic landscapes. Yet Dieudonné did have some attachment to the territory; his father was from Masisi and the community, Bahunde, that defined itself as ‘autochthonous.’ As a result, Dieudonné possessed forms of ‘ethnic capital’, a political resource in the area (Hoffmann, Vlassenroot, and Mudinga 2020). His name, for instance, placed him locally – a signifier of his roots, his ‘belonging’ according to autochthony discourses. At first though, his work as a driver was difficult and required a degree of improvisation. Dieudonné had to learn how to travel through an array of different roads or ‘axes’ which were unfamiliar to him, but that he was supposed to be able to navigate as a ‘local’ employee. His colleagues helped by drawing maps on napkins. Although based in a CNDP-controlled zone, Dieudonné also crossed frontlines and worked in areas controlled by a range of actors on ‘the other side’ who were fighting the CNDP: the FDLR, the Congolese army, and Mai-Mai rebellions such as PARECO and later, the APCLS.

As Dieudonné familiarised himself with the area, he came to realise that he had some useful personal contacts which enabled him to make sense of the context and to travel through different areas. ‘My background was useful for Kitchanga at the time,’ he told me,

I had the advantage of not growing up there, so I had distance from the local politics, arguments, everyday personal stuff that exposes you. But I had ties, contacts from my past which meant I could gather security information, talk to people, and a name which could help build trust if I needed it.

When he arrived in Kitchanga, Dieudonné kept bumping into people he knew, especially among the CNDP. He had studied with some members of the group and realised that some of his former RCD colleagues had subsequently joined the CNDP and were based nearby. Whilst interpreting for a foreign NGO colleague during a meeting with the leader of the CNDP Nkunda, for instance, Nkunda recognised Dieudonné from the RCD era. ‘I got in contact after that for the needs of the NGO, but obviously because of humanitarian principles and protocols we were not close whilst I worked in Kitchanga,’ Dieudonné explained.

Because of his background, Dieudonné explained that there were ‘two ways to play it’: either positioning himself as ‘a local boy’ – Hunde from Masisi – or a former RCD member with links to the CNDP. ‘I tried to do both,’ he summarised, ‘I was flexible.’ While brokering access across Masisi territory in areas controlled by warring groups, Dieudonné described ‘trying to play different roles to different people.’ When at roadblocks manned by Mai-Mai, Dieudonné often gave his full name. In other words, he constructed himself as an ‘autochthone,’ and did not disclose his history with the RCD rebellion. However, when interacting with the CNDP, Dieudonné described doing the exact opposite: he did not disclose his surname or his father’s attachment to the territory, but instead used his personal contacts in the CNDP to his advantage. Using existing contacts, he built an extensive network which enabled him to gather information, to check the security conditions on the road, and ultimately, to navigate travel through CNDP areas with greater ease. Dieudonné stressed the fact that ‘ethnicity is not always self-evident. Lots of people actually thought I was a Tutsi.’ He did not speak the dialect of his father’s community, Kihunde, and told me he could ‘pass as a Tutsi’ because he was identified as a ‘proche of the CNDP,’ given his personal contacts. Ultimately, Dieudonné endeavoured to use his ‘flexible’ position in Masisi tactically to navigate zones controlled by different warring groups.

If aid brokers are ‘identity entrepreneurs’ (Bierschenk, Chauveau, and de Sardan 2002, 28), then humanitarians try to be identity entrepreneurs for access: they cast and recast themselves for different audiences at different times, whilst maintaining multiple relationships. As Dieudonné’s story illustrates, identity markers were not always fixed, and ‘autochthony’ was slippery: humanitarians tried to situate themselves tactically during encounters with armed actors using the signifiers assigned to them by others. Dieudonné described trying to ‘do ethnicity’ differently with armed groups whilst performing a humanitarian role: tactically constructing his identity during interactions in different settings (West and Fenstermaker 1995). Perceived ethnic identity became part of a bricoleur ‘tool kit,’ along with personal experiences, networks and histories, which humanitarians used to ‘construct their identity’ and play with slippery concepts of autochthony in different instances (James 2011, 318).

Benjamin

In contrast to Dieudonné, Benjamin grew up in Masisi where he began working for international aid agencies during the M23 rebellion – his role was building a network among armed groups and collecting ‘security information’ about the evolving context. Benjamin was perfect for the role because he had grown up locally and had extensive personal connections. Benjamin gradually told me about his personal history working in a range of different social positions in North Kivu.

During the early 1990s, Benjamin worked for an international NGO in Masisi. During Congo’s ‘first war’, Benjamin joined the AFDL rebellion, but left just before the AFDL took Kinshasa: ‘after seeing what AFDL did, some of the violence, I left.’ Benjamin returned to work for several NGOs which were responding to the influx of Rwandan refugees, and then joined the church and became a pastor for several years. During the second war, Benjamin returned to Masisi to work in the mining sector, and then for the RCD rebel administration. After the end of the second war, Benjamin continued to work in mining, before becoming a leading figure in a local civil society organisation. The position required Benjamin to work closely with government authorities and CNDP officials. Over time, he became an informal intermediary between his contacts in armed groups and newly arrived NGOs in the area. In 2011, he decided to join the NGO sector again. Benjamin concluded that his job had always been brokering relationships with the same network of people, ‘but with different hats on.’ Benjamin was now wearing a humanitarian hat but explained that his former hats remained useful resources.

Despite this rich political history, Benjamin distanced himself from what he termed the ‘local context.’ He constructed a degree of distance, positioning himself as an outsider. In fact, Benjamin described himself not as a ‘local’ staff, but a ‘delocalisé’ – someone who had been brought in from another province in DRC to work. As Benjamin explained, although he was born and raised in Masisi, his father came from a different province in DRC. Employing local discourses of autochthony, Benjamin positioned himself as an outsider, someone impartial to local historical contests surrounding land and governance. Benjamin claimed that he could maintain a network with a range of armed and political groups because ‘I am seen as more neutral than other local staff, there is that distance, I am not ethnically aligned.’ In other words, Benjamin positioned himself as ‘local’ in terms of social connections, but ‘foreign’ from discourses of autochthony – both to armed actors, and to his humanitarian colleagues.

Benjamin had extensive personal contacts in a range of armed groups which meant he could gather information about conflict dynamics and negotiate with different sides. Benjamin had studied at secondary school with many key political figures: ‘they are my generation, we grew up together,’ he explained. Benjamin’s mother worked at a local hospital whilst he was growing up and was well known and respected locally. His mining career provided links to economic elites, whilst he knew several members of the M23 from the church. Meanwhile, he maintained contact with his former colleagues from the RCD and AFDL.

Yet, whilst interacting with a range of political and armed authorities, Benjamin highlighted his foreignness ‘by signalling my surname and family history.’ By doing so, he endeavoured to use his personal contacts in the region as a resource to build trust and access information, all whilst distancing himself from local power struggles and contests over resources. Benjamin’s local humanitarian colleagues described his position as exceptional: ‘he isn’t from here … well, he is not from one of the communities from here,’ one told me.¹⁶ His colleagues doubted that anyone else who grew up locally could negotiate access with all the armed groups like Benjamin: ‘For example, if you are from a community from here, you will be associated with the armed group that is linked to your community. How are you going to deal with the other side?’¹⁷

Benjamin described trying to play simultaneous roles. In 2011, for instance, he told me about ‘a gridlock between Hutu and Hunde communities’ in Masisi - a Hutu Nyatura armed group had blocked access to the road and ‘the area became divided in two, Hunde on one side, Banyarwanda on the other.’ Benjamin told me that he had a unique position: ‘I was from neither side, but I grew up with the key actors, so I tried to try talk to them. I rang up a local leader here who has influence over Nyatura armed groups and said “look, we know each other well, we grew up together, let’s talk.”’ Benjamin concluded that he was able to be both ‘impartial’ and well-connected, negotiating with both sides and the UN peacekeeping mission in order to remove the barrier.

Essentially, Benjamin endeavoured to perform both localness and distance (sometimes simultaneously) in Masisi: a local boy who was not a ‘son of the soil.’ Later in his career, Benjamin went to work in his father’s province of origin, where a conflict had recently begun: ‘My role was to go and make a new contact base, to negotiate access and analyse the political security context.’ There, Benjamin told me he played the same role – local, but not local – in a new context. He positioned himself as originaire using his name and language skills, but one who was distant from the recent conflict – he had minimal personal connections or interests in the province. Benjamin concluded, ‘I can go all around Congo and do this work, I speak four languages and eleven dialects, I can use them all in different ways, in different combinations.’

Risks

But what are the limits of this shapeshifting? It was a process fraught with tension – humanitarians involved in this delicate balancing act faced distrust. In Dieudonné’s words, ‘the ambiguity caused me problems.’ His loyalties were questioned, and over time, it became more difficult to tactically self-fashion in this way. Dieudonné described growing tensions with some of his CNDP contacts about his new position as a humanitarian. Some asked why he had joined the humanitarian sphere (‘working for foreigners’) or asked why Dieudonné did not rejoin the armed struggle. Others were concerned about how much he knew about them: rather than isolated in CNDP territory, Dieudonné also travelled to areas controlled by the government and opposing rebel groups. His contacts in the CNDP feared that he could denounce them to the ‘other side’ and that their families living in other areas would be exposed. Meanwhile, Mai-Mai combatants on the road also began to distrust Dieudonné’s loyalties and his self-representation as ‘autochthone’, especially given the fact he could not speak Kihunde, the local dialect and mother tongue of his father. In other words, there were limits to his ‘flexibility.’

Dieudonné also described the difficulty of presenting himself as an impartial humanitarian in the conflict given his politico-military history. Humanitarians with experience in Rwandan-backed rebellions endeavoured to keep their background hidden from the Mai-Mai militia who had mobilised against such rebellions. They feared being directly associated with the historical violence attributed to their former armed group. Dieudonné omitted his history with the RCD as much as possible and avoided discussing his background with his colleagues from Masisi who themselves had experiences of violence in the region and complicated social ties to Mai-Mai. But in 2013, it became more difficult to negotiate local identity politics when fighting broke out in Kitchanga between a branch of the Congolese army led by a Tutsi Colonel with links to the M23 rebellion, and the APCLS which had been controlling part of the town’s surrounding areas. In this context, animosities between Hunde (so-called ‘autochthones’) and Banyarwanda (labelled as ‘non-autochthonous’) became particularly fraught. ‘It became a Hunde versus Tutsi atmosphere, even in the team,’ Dieudonné explained; ‘my colleagues had links, some of them even joined the fighting. It became a Cold War and I tried to stay somewhere in the middle.’ However, with fighting in Kitchanga itself, it was no longer possible to separate audiences when performing different social roles. In the end, Dieudonné’s impression was that some of his colleagues saw his name, which suggested he was ‘one of them,’ but realised he was different: ‘I was clearly a city boy.’ Dieudonné was promoted and was even involved in recruitment decisions. Subsequently, he received anonymous death threats. One of these threats accused him of having favoured Banyarwanda in recruitment, a perception he thought was fed by his links to the CNDP and his perceived lack of commitment to autochthone discourses. It was no longer safe to work in Kitchanga, and he was posted elsewhere.

What about Benjamin? His self-representation as local, but not local, was a means of distancing himself from conflict dynamics, of shapeshifting between provinces, but also of tactically positioning himself within the aid bureaucracy. In effect, Benjamin presented himself to foreign NGOs as the best of both worlds: a well-connected broker, but someone supposedly further removed from local histories of violent conflict and able to position himself as ‘impartial’. Benjamin’s foreign colleagues also seemed to think his position was exceptional: Benjamin had all the right contacts but was ‘less at risk’ than other ‘locals’ because he was ‘ethnically neutral,’ they explained. In reality, of course, Benjamin’s personal, political and economic histories in Masisi meant his position and interests were more complex than assumptions about ‘ethnic alliances’ – a framing which is embedded in colonial discourses of Otherness (Nzongola-Ntalaja 1999) – but one that he nonetheless reproduced strategically.

When we met in 2020, Benjamin told me that he had left his post in Masisi. He was vague on the details but described the difficulty of representing himself as ‘just a humanitarian’ where he had grown up. He described harassment from his personal contacts: friends had made demands for information, for phone credit, had wanted to pressure him to set up systems of embezzlement, to influence hiring decisions, or to become more involved in the current political struggles. He told me that armed and political actors made demands on Benjamin that they would not, for example, make on his foreign employers: ‘because they know me personally, I was not just a representative of the NGO.’ Like other brokers, Benjamin had to mediate an ambiguous status – a man with many ‘hats’ – and balance the conflicting demands from his personal networks and his humanitarian employer. He described the difficulty of performing the role of a neutral and impartial humanitarian in an area where he had extensive personal networks – networks which were themselves, ironically, a resource for his humanitarian role (James 2020). Humanitarians like Benjamin found themselves in a difficult situation: how could they handle the tension between the demands of their employers to act in line with humanitarian values, and the demands of their local networks, when keeping both satisfied was central to their role and safety? Benjamin wanted to avoid making risky commitments to keep his local contacts happy, and in the end, felt safer and more comfortable working outside of Masisi.

Conclusion

Through an ethnographic focus on the personal experiences of people working in conflict settings, this article shows how humanitarians conduct modes of self-fashioning in order to negotiate access for the delivery of relief and manage their own safety. In effect, people working for humanitarian agencies learn to play different roles for different audiences. This reflects the broader social dynamics in regions of protracted ‘crisis’ (Vigh 2008), where being part of different social spheres has become a permanent condition of life (Vlassenroot, Mudinga, and Musamba 2020) and the ability to broker between overlapping authority structures has become important to everyday life (Vogel and Musamba 2017). Shapeshifting is a novel example of how shifting self-representation constitutes a form of tactical agency for survival in a context of violence (Utas 2005). By focusing on individuals, their histories and networks, it becomes clear that humanitarian agencies do not occupy a bounded enclave or ‘land’ separate from these broader social dynamics (Autesserre 2014), but form part of the messy web of relationships that determine how power is negotiated in the everyday.

Shapeshifting is also illustrative of the relational and interpretive labour of locally hired humanitarians during aid interventions (Peters 2020), which often remains hidden behind discourses on humanitarian principles. There appears to be a disconnect between these discourses of aid institutions at headquarters, and the situated savoir faire necessary to implement aid programmes in practice. Humanitarian institutions are represented by people who have their own identities, stories, and histories, which in turn shape how intervention unfolds on the ground. Access to work in conflict does not solely rely on the symbolic performance of principles or a distinct humanitarian identity as a public relations tool (Terry 2011), or shared interests negotiated with political leaders (Magone, Neuman, and Weissman 2011) - specific individuals on the ground, their histories, identities and relationships matter too. Although fieldworkers are supposed to position themselves as distinct and detached humanitarians when working in conflict (Abu-Sada 2012; Sutton 2020), most humanitarians are locally situated and must balance their own histories with this performance of a humanitarian role (Goffman 1956). In practice, humanitarian agencies rely on the improvised practices of local employees (such as shapeshifting) which draw from personal identity markers and histories. At an individual level, this means that employees like Dieudonné and Benjamin are simultaneously required to present themselves as embodiments of their humanitarian organisation, conducting themselves in ways that reflect their organisation’s image, all whilst shapeshifting with different personal and political identities, histories and networks in order to carry out their work. Both their career and security depend on their ability to balance these multiple self-representations and competing expectations. An examination of aid through the lens of the ‘everyday’ reveals the complexities of balancing these overlapping identities in practice, as well as the risks it involves.

Finally, these descriptions of shapeshifting reveal the complex politics of difference that humanitarians navigate in the everyday, beyond the local/international binary. Since the 1990s, there have been calls in humanitarian sector to hand power to ‘local’ actors, rectifying post-colonial power imbalances and domination (Roepstorff 2020). At the same time, there has been critique of the tendency of foreign NGOs to retreat into bunkers, transferring risks to ‘local’ staff who are already more exposed, with less institutional support (Collinson et al. 2013). The majority of aid workers working on the ground are, indeed, ‘local’. However, in this article, I highlight the complexity of the heterogenous identities and experiences within this ‘local’ category (Peters 2016). Humanitarians are positioned in society in complex ways – with political, military, professional and family histories which may influence their ability to work. It is also important to consider the ‘non-essential character of difference’ (Martin de Almagro 2018, 330): aid work involves a web of interactions in which individuals tactically produce difference (Rossi 2006), and the meanings of the same identity markers change depending on the geographical and temporal space. It is necessary to recognise the complexity of peoples’ positioning in volatile environments, the room for manoeuvre for some individuals, but also the considerable personal risks that many continue to face during the process. An examination of the everyday practices of social navigation conducted by ‘local’ aid workers, such as shapeshifting, can provide a better understanding of the risks different people face in specific contexts, without further essentialising identities.

Acknowledgements

This research was supported financially by the ESRC Doctoral Training Programme and the St John’s College Oxford 450th Anniversary Fund Scholarship. I thank the reviewers and editors for their detailed feedback which greatly improved this article. I am thankful to everyone who helped to facilitate my research, and, to those who took the time to talk to me. A special thank you; Dieudonné and Benjamin. Any errors remain my own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author bio

Myfanwy James is a Lecturer at the Department of International Development, University of Oxford. Her research examines issues of power, security and social identity in aid intervention during violent conflict, with a focus on the eastern Democratic Republic of the Congo.

Additional information

Funding

This work was supported by Economic and Social Research Council: [ES/J500112/1]; Paris: St. John's College, University of Oxford: 450th Anniversary Fund Scholarship.

Notes

1 All names are pseudonyms.

2 I interviewed 10 former rebels who have become humanitarians, 2017-2018.

3 Interview, Goma, February 2018.

4 Interview, Bukavu, November 2017.

5 Interviews, Goma, November 2017.

6 VOA. 2016. ‘Les jeunes du territoire de Masisi au Nord-Kivu bloquent l'accès aux humanitaires.’ 11 August. https://www.voaafrique.com/a/jeunes-territoire-de-masisi-nord-kivu-bloquent-acces-humanitaires/3460081.html

7 Interview, Goma, January 2018.

8 ‘Réunion avec société civile et Baraza provenaient de différents territoires.’ 2012. MSF archives, Paris. Interview, Goma, November 2020.

9 Interviews, Goma, October-December 2017.

10 Interviews in Masisi, August-September 2018.

11 Interview, Masisi, September 2018.

12 Interviews, Masisi, September 2018.

13 Interviews with humanitarians, Goma and Bukavu, 2017-2018.

14 Interview, Goma, February 2018.

15 Interviews, Goma, 2017-2018.

16 Interview, Masisi, September 2017.

17 Interview, Masisi, September 2017.