Even since Weber and Durkheim made it a major intellectual topic, violence has been a research object with significant emotional, partisan, and ideological power (Arendt 1972; Howell and Willis 1989; Scheper-Hughes and Bourgois 2004; Jones and Rodgers 2019). Although it increasingly raises questions that concern the human and social sciences in general, the topic simultaneously requires an interrogation of methods and ethics concerning access to fieldwork, the nature of collected sources and the conditions in which they were collected, as well as self-reflexivity on the part of the researcher. Furthermore, the definition of violence as a social fact is relative and subject to debate (Heitmeyer and Hagan 2003). It is not enough to rely on an axiological neutrality which, when dealing with violent practices, risks ratifying the erasure of persecutors’ actions and responsibilities (Naepels 2006). Research on violence carries psychological and moral implications, which is why it requires rigour to ensure its legitimacy among the social sciences and humanities and to protect it from any form of voyeurism. One of the most important questions that arises in the analysis of such phenomena, however, is the compatibility between empirical requirements and the physical and psychological integrity of the researcher. This special issue aims at contributing to the important, if not critical, reflection on violence in the humanities and social sciences, on the African continent in particular. For researchers, the concern is indeed not to abandon the research and analyses of violent settings to humanitarian actions and media expertise. Although the term “field” originated from the military vocabulary, meaning a place of tension and conflict (Pulman 1988), we must also consider the risk involving the study of violence in Africa as a challenge or a fascination with exoticised danger, especially when fieldwork in a violent setting is systemically highlighted in spite of physical and psychological risks for the researchers, but also for their assistants and their interlocutors.
Theoretical, methodological, and ethical questions surrounding the study of violence have engendered multiple, sometimes contradictory positions among scholars of African studies. This is reflective of a major problem within the field, which is that research efforts on violent topics, and/or within violent contexts, are inherently complex and risky. During fieldwork, the nature of intersubjective relationships producing or (co)producing materials is potentially altered by the threats, imminent danger, or ongoing insecurity encountered by researchers. Although the field of study on violence is known for its large diversity of methodological approaches, it often seems limited to two opposing types of work: on the one hand we have ex situ analyses—disconnected from the field and reliant on existing literature and second-hand data; on the other hand, research arguing the need for direct contact with violence itself. These two diametrically opposed positions, which form the basis of a substantial existing body of literature on violence, are nevertheless both unsatisfactory. In order to move beyond this opposition, this introduction contributes to the approach promoting a combination of methods as well as formulating alternative working methods.
Although there are “as many ways to be present in the fieldwork, to be visible and engaged in the social relationships, as there are fieldworks,” as underlined by Michel Naepels (2012, 86, our translation), here we want to reflect on the different ways to study violent phenomena—especially when conducting fieldwork—that do not lead the researcher, their assistants and/or informants to dangerous and out-of-control situations due to their immersion. This is a way for us to continue the discussion initiated by others on different ways to conduct fieldwork in violent contexts and/or on violent objects, by circumventing the issue rather than tackling it directly (Ayimpam and Bouju 2015). We examine the tangible strategies for understanding violent phenomena whilst minimising the physical and psychological risk-taking, and reflect on the conditions and procedures for reflexivity, in order to show that a plurality of methodologies is not only possible but desirable.
The approaches introduced in this special issue, based on the production and collection of empirical data, avoid the deadlock often associated with too much or too little ethnographic immersion. They demonstrate the value of studying violent settings in sub-Saharan Africa by focusing on original data—gathered or produced by the researchers themselves—which is both the starting point and bedrock of the analysis. Each contribution illustrates the importance of methodological detour as a way to avoid endangering researchers, assistants and participants, whilst paving the way for a better understanding of violent phenomena. The articles from this special issue illustrate various proposals which outline the opportunities afforded by field experience and knowledge in the collection of data and access to sources.
Among the multiple forms that violence can take, this issue is interested in armed conflicts and destructive or deadly social phenomena that can physically harm those involved, whether voluntarily or not. These types of violence can be defined as “extreme” in the sense that they translate into devastating, observable and quantifiable consequences. Although this typification does not reflect the reality of complex and multifaceted processes, it enables us to differentiate this type of physical violence from symbolic, psychological, and institutional violence. Whether they deal with the role of the Catholic Church in the genocide of the Tutsis in Rwanda, xenophobia in South African townships, Boko Haram in Nigeria, the civil war in Mozambique, or quantitative data on violence in Nigeria—the authors have not collected or produced their data directly where violence occurred, nor have they endangered their interlocutors. Only the paper on Boko Haram relies on fieldwork carried out in North-Eastern Nigeria while the region was affected by the conflict. However, the author, who met with actors and victims of violence, carefully chose the places and times of his study in order to avoid any violence.
Although work on data collected outside of any violent context is certainly easier for historians, who represent the majority of the contributors in this issue, papers by sociologists and political scientists demonstrate the value of a material-based approach in the study of contemporary violent phenomena, while granting a central role to the in-depth knowledge of the field that allows their production.
This issue thus gathers contributions which showcase the data collected by researchers engaged in the study of past or ongoing violent phenomena. By choosing to expose the processes of data and materials collection or production in the field—archives, official documents, leaflets, press articles, accounts—the aim is to better understand the conditions in which these materials were produced, but also the constraints they impose and the strategies researchers employ to bypass them. In this regard, the subject of “violence” allows for the acute emergence of scientific questions around ethics and critical detachment. The aim of this introduction is to put into perspective what constitutes, for us, a dominant—although not exclusive—approach to the study of violent phenomena on the African continent (1). To this end, we will discuss the current uses of reflexivity in research relating to violent contexts and phenomena (2), as well as the limitations on its application, by examining questions of reflexivity within a systemic framework (3). This special issue therefore adopts a particular methodological and ethical position, backed by the papers: if the study of violence must rely on in-depth knowledge of the phenomena being studied and their context, as well as on proximity to the field in order to access data and primary sources, it does not necessarily require the researcher to experience violence (4).
The field of area studies—which the journal Sources, and thus this special issue, subscribe to—is undergirded by a framework that necessitates an interrogation of the way in which the topic of violence has been addressed thus far in the analysis of social and political phenomena on the African continent. The current calls for the decolonisation of knowledge are particularly prominent within African studies, although they struggle to translate into practice. Yet, they invite us to examine and unsettle some routinized frameworks used to understand the social world, that are reproduced in academia.
Firstly, research from the humanities and social sciences on Africa is rooted in a political and intellectual history of violence, and a fascination with violence. Achille Mbembe talks of a “caricatural dramatisation” of the African continent, against the historical backdrop of a morbid attraction to violence, and brings to light the current ramifications of this colonial heritage (Mbembe 2000). We must take into consideration the weight of the “colonial library” (Mudimbe 1988) in the construction of the representation of a continent inherently violent, chaotic and/or “barbaric,” a construction which largely echoes the justification for conquest and colonial domination, which is then reproduced in contemporary representations. Regardless of the tangible administrative system in the colonised territories, concepts such as “civilisation” and “pacification”—contrary to the supposed violence and brutality of the colonised—were inherent to the colonial project and have deeply affected European imaginaries about the “development” of Africa (Cooper 2005). Furthermore, African societies have often been analysed using violence as a frame of reference for contemporary socio-political dynamics, assuming that the “colonial situation” (Balandier 1951) emerging from European domination (evangelisation, implementation of Westphalian States, development of global capitalism, rural depopulation and urbanisation, etc.) has produced a latent state of violence in the colonised societies due to its destabilising effect.
This propensity for violence within African studies must also be analysed in light of more recent changes in international research. While Marxist, development-focused, or structuralist approaches dominated the decades post-independence—the economic, political and social transitions in the 1980s and 1990s lead to the emergence of new approaches and topics, such as “politics from below” (Bayart, Mbembe and Toulabor 1989), neopatrimonialism (Médard 1991; Chabal & Daloz 1999) or the “institutional turn.”1 The impact of dynamics considered more or less exogenous to African societies (the end of the cold war, neoliberal reforms, and democratic transitions) has often been interpreted as the cause of the disruption of fragile political structures in a context of globalisation. This led to the emergence of new research objects, caught within the same prism of analysis: violence. Civil wars and intra-state conflicts, “failed States,” contemporary religious dynamics, land and resources-related issues, the use of occult practices or even elections-related violence have thus become rising topics in African studies. African societies are indeed faced with deep political, economic and social changes, but the growing interest in such research topics is also a consequence of the functioning of academia in the global North. Research priorities there are increasingly influenced by “project-based” research, which is subject to more and more competitive funding and scientific priorities defined by development agencies (Aust 2014). This context partly diminishes the capacity for researchers to define their own research interests. It simultaneously tends to favour topics prioritised by public decision-makers and private interests, in which violence plays a prominent role. In the case of violent phenomena, the current oversaturation of studies on terrorism, Islamism, trafficking, crime, etc.—is partly a consequence of this relative loss of autonomy of the knowledge production field. These external influences are also implicit in the development of Security Studies and the growing number of Peace & Conflict Studies research centres, which rely on an often-uniform approach to violence and a dual prism: violent situations and peace-building processes. In France, although this trend seems to be less prevalent in the recruitment of researchers in public institutions, which rarely focus on “violent objects,” it largely influences the public and private funding of research projects.2
These dynamics obviously do not represent entire fields, but they are partly responsible for how frameworks of understanding articulated around violence have been over-used in contemporary African studies. They have furthermore been reinforced by the weakening of public universities on the African continent, undermined by neoliberal restructurings (Provini, Mayrargue and Chitou 2020). Projects related to the decolonisation of knowledge led by African Studies Institutes in Africa (such as in Ibadan and in Legon) as well as critical and Marxist schools (such as in Zaria and in Dar es Salaam in particular) which prevailed in the 1960s and 1970s, have gradually been replaced by academic fields affected by financial insecurity, a dependence on consulting activities and brain drain both to countries in the global North and to private institutions. These evolutions have favoured the epistemic dominance of Western researchers and Western institutions on the production of knowledge. This special issue perpetuates these inequalities since authors from the continent remain in the minority. Despite recurring calls for the transformation of academic journals, and notably for a greater diversity of authors and contributions, it is clear that this special issue opens with an admission of failure of which the reasons are as systemic as the responsibility is individual. The journal Sources as well as the research institutions that created it are already attempting to actively participate in the still too timid process of scientific inclusiveness and the co-production of knowledge on Africa. The challenge remains, however, to carry on these attempts in order to offer new models of scientific collaboration. Even if the call for a reflexivity effort of this introduction is less an outcome than a process, it cannot exempt itself from critique towards our own practices. Efforts remain clearly inadequate to produce results up to the standards of an on-going decolonisation process in an academic field built on renewed partnerships.
Since the 1960s, ethnographic studies have become riskier as anthropologists became interested in new forms of conflict emerging out of the cold war, both from the side of governmental forces and from rebels. This was in opposition with dominant forms of analysis, often disembodied and informed by second-hand data (Sluka 1990). As research on violent phenomena increased, questions on the conditions of knowledge production were more closely considered, especially within anthropology, which is the discipline from the human and social sciences most preoccupied with fieldwork as a method as well as the practice of reflexivity and its limits. Since the 1990s, anthropology has chosen to tackle these rarely-mentioned yet widespread questions. In her report entitled “Surviving Fieldwork,” written for the American Anthropological Association (AAA) in 1990, Nancy Howell notes that 42% of surveyed anthropologists declare that they have been victims of criminal violence on the field (theft, physical harm, rape, killing), 9% were arrested, 22% experienced violent political events (revolution, war, riots), 15% were accused of espionage and 12% were faced with intense hostility during their fieldwork.
Following these findings, researchers tried to attend to these ethical and epistemological questions concerning fieldwork in violent contexts, and started publishing a robust collection of works that are still debated today. These works have discussed how to best pursue ethnographic fieldwork in violent contexts, shedding light on three main issues: 1) the control or use of ethnographic research by some forces active in the conflicts, such as the US army in the counter-insurgency wars in the 1960s and 1970s3 (Price 2016), the Spanish State in its fight against ETA (Faligot 1983) or the apartheid regime in Namibia (Lee and Hurlich 1982); 2) the growing number of research papers on violent topics detached from the reality on the ground (Nordstrom 1997); 3) the influence of positivist methods taught in anthropology departments in English-speaking countries (Kovats-Bernart 2002). By thinking about alternative approaches to work on objects and in contexts considered violent, these works have offered thought-provoking discussions both in terms of ethics and methodology. Moreover, they emphasised the value of the human and social sciences, and the ethnographic method in particular, in the production of first-hand original data with a deeper understanding of the phenomena in question.
Despite its diversity, this literature was mainly written by researchers who themselves have worked in violent contexts. They thus presupposed that carrying out fieldwork in these violent contexts is crucial to an on-depth understanding of the social, economic and political dynamics at play. For example, in their seminal work, Fieldwork under Fire, Nordstrom and Robber (1995) compare accounts and analyses by anthropologists who worked on violent objects and in violent contexts, from the Palestinian Intifada to civil wars in Guatemala and Somalia, as well as the Tiananmen Square crackdown. Confrontation with violence, which is described by the authors as a “fieldwork crisis,” constitute for them a privileged way of accessing, and therefore understanding, the experience of the respondents, their ways of understanding the violence they faced, and the strategies they rely on to use it or protect themselves from it. The authors argued that if the writing process causes an attenuation and a transformation of the account, physical and emotional proximity is the only adequate method for researchers to report on the everyday nature of violence or on the dynamics at stake during violent events. In other words, “the ontics of violence—the lived experience of violence—and the epistemology of violence—the ways of knowing and reflecting about violence—are not separate. Experience and interpretation are inseparable for perpetrators, victims, and ethnographers alike […] sociopolitical violence can be approached in many ways. At some level, however, to be able to discuss violence, one must go to where violence occurs, research it as it takes place.” (Nordstrom and Robben 1995, 14).
Alongside promoting the immersion of researchers in violent settings for heuristic reasons, this literature also defends the idea that ethnography is relatively more useful to produce data as close to the field as possible, as opposed to overextended research produced away from realities on the ground. For example, Sluka writes about his PhD fieldwork in Northern Ireland in the early 1980s: “It is true that there are very real dangers involved in doing research there. I was aware of these dangers [...], but believed that anthropologists had to do research where it had immediate social relevance in order to justify our existence.” (Sluka 1990, 116). As this quote illustrates, the achievement of such fieldwork is often justified by the “vital” function played by ethnography in the production of data and analyses, especially to fight against propaganda and/or misinformation (Malejacq and Mukhopadhyay 2016).
These two arguments—heuristic interest and social usefulness—fundamentally rest on the idea that fieldwork is possible “even in the most dangerous contexts” (Sluka 1990, 124), especially since journalists and aid workers, for example, also visit such places. Thus, since “ethnography is an inherently rewarding but at the same time ‘risky’ research methodology” (Rodgers, Kruijt and Koonings 2019), the discussion is not about the legitimacy of conducting fieldwork in a violent context but rather about the conditions and ethics of it. In other words, violence can only create “hazard of fieldwork” (Howell 1988) whose effects can be controlled with adequate methodological precautions in order to minimise the risks and to guarantee the safety of researchers, of their data, and of respondents (Sriram et al. 2009), especially during a time when digital technologies increase risks as much as they offer solutions to avoid them (Grimm et al. 2020). Consequently, many publications—such as the one by Greenhouse, Mertz and Warren (2002)—primarily focus on practical and methodological aspects of conducting fieldwork in violent contexts, in order to enable researchers to protect themselves and their data, and to guarantee the safety of their informers. Other works focus on more specific aspects of such methodological considerations, including the role and the protection of research assistants (Hoffman and Tarawalley 2014), the use of participatory methods (Wheeler 2009), and fieldwork on specific topics such as police (Beek and Göpfert 2013), armed conflicts (Woods 2006), or authoritarian regimes (Glasius et al. 2018). Contrary to the rigid methodological recommendations and codes of ethics that, mostly in the United States, set strict regulations on researcher-informant relationships,4 this literature instead proposes a flexible approach to ethnographic fieldwork. The researchers are given the ultimate responsibility to evaluate both the risks incurred and the scientific value of their approach. For example, Kovats-Bernart defends a pragmatic view of fieldwork and “flexible, blended, inclusive and pliable” methodological strategies (2002, 210). Dennis Rodgers (2007), using participant observations of street gangs in Nicaragua, suggests the idea of a “situational” ethics which, he argues, is the best way to approach fieldwork in cases where researchers are forced to act or react inappropriately in terms of ethics, for example when placed in a situation where they must take part in violent acts or in unlawful activities (see also Verhallen 2016).
The literature in the social sciences and humanities discussing the conditions for data collection is generally anchored in the “reflexive turn” of the 1970s and 1980s. Within the sub-fields for the study of violence, it has successfully contributed to challenging positivist research paradigms, reinforced the legitimacy of the social sciences in proposing balanced explanations based on first-hand data, and questioned the unethical ways to conduct fieldwork. However, one of its blind spots has been to limit reflexivity to a process of self-questioning, without always situating researchers within their institutional and professional realm marked by competitiveness and power struggles. This shortcoming is all the more significant as debates over reflexivity in violent contexts have led to the idea that ethical considerations were processual, relative and context-specific, against the standardised protocols of ethics committees.
The issue here is not to call into question each researcher’s ability—depending on their personality, training, and ambition—to assess the danger they face, especially when dealing with the pressures of conducting fieldwork. It is nonetheless necessary to replace these “choices,” first and foremost the decision to work on a violent object, within a more systemic reflection on the political economy of knowledge. Indeed, even when the many researchers criticising methodological and ethical “one-size-fits-all” solutions stand against their depoliticising effects, they often struggle to turn the critical lens around and question their own practices in light of issues related to the structure of the academic field. The reflexive approach they stand for remains relatively pragmatic, individual, and self-justifying, without always questioning the way research on violent objects and/or violent contexts takes place within a general economy of knowledge production influenced by power relations. The “participant objectivisation” model proposed by Pierre Bourdieu can fill this gap and replace the choices and practices of researchers regarding their position in the academic field. Bourdieu indeed argues that “what needs to be objectified is not the anthropologist conducting anthropological research on a foreign world, but the social world which made the anthropologist, and the conscious or unconscious anthropology which is used in their anthropologic practice […] especially their position within the microcosm of anthropologists” (Bourdieu 2003, 44-45).
This approach enables us to reposition research on violence within an academic field that values “the heroic—and mythical—figure of the researcher defiant in the face of danger” (Boumaza and Campana 2007). The time when “field” researchers, outside the discipline of anthropology, had to legitimise their practice and their ethnographic stance compared to other methods, has mostly come to an end. Contemporary human and social sciences, and area studies in particular such as African studies, are largely characterised by a growing reliance on the ethnographic method, without always linking it to its epistemological heritage developed in anthropology and essential for analysis (see Marcus and Fischer 1999 ; Clifford and Marcus 2009 , among others). The undertaking of “immersive” fieldwork tends to be more and more valued and encouraged in various disciplines, such as history or political science, and to be considered as the definitive method for data collection, without its implications always being discussed. The sacredness of fieldwork is particularly felt in African studies, where authors have noted the over-valorisation of monographs and the sidelining of theory (Mamdani 2004). Yet, the promotion of the ethnographic method in the study of violent phenomenon, when incomplete or inappropriate, has fostered what Lake and Parkinson call “out-dangering” (2017), which is a fascination for “’hot topics’ because of the recognition, and even potential celebrity that could come with such research” (Grimm et al. 2020, 3). In a recent volume on security during fieldwork, the authors argue that the pressure to publish, combined with career ambitions in a highly competitive context, often push researchers to conduct fieldwork under extremely dangerous circumstances (ibid).
However, immersion does not necessarily imply that the researchers studying violence take risks. Many works on conflicts, for example, have been carried out without being dangerous. Nonetheless, it requires some particular and continuous precautions. Political scientist Marielle Debos insists that for her research in Tchad, a “constant reassessment of risks—for both the researcher and their informants—for each trip, meeting, or interview” was very important, as well as a distancing from time of armed conflict in order to prioritise the “in-between-war” periods (Debos 2016, our translation). Long-term, regular fieldwork can also allow for researchers to analyse violent phenomena outside of critical moments, building on their knowledge of the context and the actors in order to identify safe places and stretches of time.5 If these approaches can be very heuristic, they nonetheless require great autonomy and scientific freedom which is rarely the case for non-tenured researchers and/or those in training.
As required by the practice of “participant objectivisation,” ethical considerations and methodological debates among researchers working on violence must be set within a collective framework and at a systemic level. Yet, little has been written, to the best of our knowledge, on the specific position of young researchers working in violent contexts, beyond the fact that a lack of experience can create additional risks. This is all the more harmful given that fieldworks on violent phenomena are often carried out by researchers-in-training. Indeed, these “junior” researchers, especially PhD students, have time to conduct long-term fieldwork, unlike tenured researchers who are often busy with teaching obligations and/or administrative duties. Funding by private and/or non-academic institutions also multiplied research grants and contracts tied with specific topics often determined with the strategic, security and political restrictions already mentioned. In a context of scarce resources, peer-pressure and high level of competition from the PhD level—and sometimes even Masters—carrying out a “dangerous” fieldwork is often perceived as a competitive advantage, valued by institutions (Knott 2019; Browne 2020). Moreover, although the situation of research assistants, who enjoy privileged access to the field, has been debated for close to a decade, structural inequality can still be found in the implementation of methodological protocols, choice of theoretical frameworks, and access to publication in international journals, as well as in risk-taking during fieldwork, especially in violent contexts. In the case of African studies, resorting to assistants perpetuates these unequal academic patterns and also plays a role in the reproduction of a colonial model of knowledge production (Jenkins 2018; Moss and Hajj 2020; Nyenyezi et al. 2020).
Additionally, despite its centrality, the question of the responsibility of research institutions in the way researchers are supported, trained, and led to question their practices on the field is often mentioned without further elaboration.6 The situation, of course, varies according to the universities, research centres, and disciplines, but there is generally limited consideration given to the physical, psychological, and emotional risks involved, leaving researchers only partially prepared for fieldwork in high-risk environments. Usually, the institutions respond to these issues by forbidding any travel to what is referred to as “at-risk” areas, which contributes to another bias already widely denounced in English-speaking academia by researchers concerned about the growing evaluation of research activities solely against security criteria (Peter and Strazzari 2017). This approach is mainly based on a risk assessment disconnected from realities on the ground, which relies on questionable geographical generalisations and driven less by the protection of researchers than that of the institution itself. Moreover, this approach absolutely does not respond to the requirements, such as training and group discussions, that are necessary for participant objectification during fieldwork in violent contexts.
All these elements, analysed through the framework of “participant objectivisation,” help put into perspective the statement made by certain researchers—that one can have “consciously taken the decision to engage in risky research” (Rodgers, Kruijt and Koonings 2019). The fact that some researchers working on violent topics defend the idea of “high risk, high gain” (ibid.) should rather lead to a questioning of the consequences that such valorisation of risk-taking brings to bear on the practice of conducting fieldwork. There are indeed numerous risks that researchers can face, ranging from threats to physical harm, but also psychological trauma. The consequences of perilous fieldwork are sometimes not only drastic but also irreversible. The murder of Giulio Regeni, PhD candidate at the University of Cambridge, by Egyptian security forces in early 2016 has profoundly affected the British academia and the field of Middle-Eastern studies. At the same time, few changes have been implemented by institutions in terms of training researchers—particularly junior scholars—before carrying out fieldwork, especially in contexts involving significant risks or challenges (Grimm et al. 2020). Although researchers working on such topics do not generally have to put themselves and their colleagues in danger, the selection of research topics and fields is never solely an individual process and always goes through institutional and discipline-specific validation steps (Naepels 2012). It is necessary, then, to frame these fieldwork practices within a general economy of knowledge production built on competitiveness, work insecurity for researchers (especially early-career ones), and the valorisation of “bankable” topics.
The objective here is not to delegitimise the use of the ethnographic method for the study of violent phenomenon, but to highlight the blind spots within the contemporary praxis of reflexivity by situating them at the systemic level. These blind spots enable us to reflect upon the practices valued in the field of African studies, addressed in this special issue, as well as the role played by institutions in the preparation of researchers to conduct fieldwork. While such a debate extends well beyond the scope of this introduction, the aim is to highlight, through the contributions gathered here, alternative field strategies and to allow for new approaches as they relate to violent objects. Indeed, these diverse contributions offer various practices of fieldwork aiming to produce, gather, but also make sense of original materials and sources which constitute, beyond the personal experience of the researchers, their material for analysis.
In our opinion, such ethical and methodological precautions are critically necessary for the analysis of violent phenomena. Not only do they facilitate the conduct of the research, but they can also enhance the “proximity” between the researcher and his/her field with the aim to produce and contextualise original data, thus accessing their full potential. Articles in this special issue all demonstrate, in various ways, that a deeper understanding and a finer interpretation of violent phenomena is made possible through an empirical approach. The authors have indeed all proven their capacity to move away from “armchair anthropology” and step “off the veranda” (Geertz 1990) in order to conduct fieldwork where their research objects emerge from original materials, some of them being reproduced for the first time.
The two articles written by historians in this special issue particularly underline the benefits of such an approach. The authors have taken advantage of their extensive field experience—which, for one of them, spanned across several decades—by replacing the sources they have produced into their respective environment. Even though only excerpts are presented here, both authors avoid the pitfall of isolating materials from their environment or studying only a fraction of them while ignoring the rest, thus limiting the risk of simplification or overinterpretation. Indeed, materials are analysed as a whole as the integrity of the source is essential, especially in the study of violent conflicts where the manipulation of sources is always a risk, and where there is a thin line between information and propaganda. Thus, Rémi Korman’s work on the controversies surrounding the role of the Catholic Church during the genocide against the Tutsi in Rwanda, is based on a large collection of press archives. However, the author makes use of his research experience in Rwanda to meticulously analyse this corpus. Social proximity with witnesses and survivors of this dramatic episode allows him to rely on his own sensitivity in reading the debates in the press between the Catholic clergy and the Rwandan State, largely concerned with finalising the “de-escalation” of conflict within the society initiated in the early 2000s.
In his article on the Renamo’s notebooks, Michel Cahen provides a variety of transcribed radio messages—which required decoding and prior analysis of the whole materials—and analyses a source that had remained invisible until now because it was previously inaccessible. The author explains that the uncovering of these documents necessitated hard work and relentless round trips to the field. These notebooks represent an extremely valuable testimony of daily life within a guerrilla movement, but take on their true meaning within a framework afforded by a thorough understanding of the Mozambican Civil War (1977—1992) of which the author is an expert. These notebooks are some of the few traces left by the Renamo itself, and they allow for a political analysis of its military strategy, throwing light on the specific relation it had with the people who lived on its territory, as well as the type of political and moral economy it had put in place.
The two subsequent articles illustrate how the authors, a political scientist and a sociologist, chose to tackle violent objects in difficult or complex situations. In the first one, Ini Dele-Adedeji investigates the Jihadist movement Boko Haram in North-Eastern Nigeria, and in the second one, Léo Fortaillier analyses xenophobia in South African townships. Both expose situations where difficult access to the field, which necessitated negotiations or resulted in hostility (Boumaza and Campana 2007), drove them to operate a methodological detour in order to construct their research object. It seems that the authors have managed to distinguish between “difficult” fieldwork—where the researcher faces problems accessing interlocutors or sources and has to develop alternative approaches—and “dangerous” fieldwork—where the physical security and mental integrity of the researcher cannot be ensured. Both contributions underscore the importance of ethnographic fieldwork in the production of data in the social sciences and humanities, when it is carried out with the intent to reduce risks. They also emphasise that such a method cannot be successfully implemented without prior knowledge and an “adequate social proximity” (Naepels 2012, our translation) to the studied spaces and contexts. The articles also testify to the fact that ethnographic immersion reaches its full potential when it is combined with other methods of data collection—such as semi-structured interviews, archival research, etc.—thus allowing for the production of various written, visual and audio materials for in-depth analysis. In other words, immersion is not enough, especially in potentially dangerous and violent contexts.
In that respect, Ini Dele-Adedeji worked in North-Eastern Nigeria, the region where Boko Haram emerged, and which was deeply affected by the group’s attacks as well as by the military crackdown since the beginning of the conflict in 2009. Having met victims and insurgents, the author stands out from the multiple analyses by war, guerrilla movement, and conflict “experts” who are often disconnected from the field. This immersive experience allowed the author to collect and analyse rare materials. Such materials include a leaflet—reproduced and translated here—distributed by Boko Haram to residents of Kano, informing them of a coming attack targeting law-enforcement officials. Written documents coming from the group are very rare and collecting such material shows proximity to the social environment. As the author argues, choosing an approach that avoids danger despite real risks, was particularly appropriate in order to establish a trusting relationship with the interlocutors, considering that simply owning such a leaflet can lead to an arrest. The author also managed to collect testimonies from detained former Boko Haram members involved in a rehabilitation project funded by the European Union. The interviews conducted in prison, as well as the observation of the Nigerian prison institution—which very few researchers have managed to carry out—afforded him a unique perspective on an already over-invested topic.
In order to explore xenophobic violence and its social impact in South Africa, Léo Fortaillier volunteered with an organisation fighting xenophobia and chose to study the career of a social worker dealing with young people in the townships. Starting his analysis with the working document used by his main interlocutor to host awareness and prevention seminars, the author’s approach is based on the material he produced thanks to an immersive study combining observations, informal discussions and semi-structured interviews. Focusing on the history and the usage of this tool enables us to understand its users’ intentions, to grasp its underpinnings and effects through the workshop participants’ reactions, on-the-spot discussions, and the ordinary categories used to describe the social world, which tend to discriminate against migrants and foreigners.
Finally, the article by Ismaël Maazaz, Vitus Ukoji, Victor Eze and Abiola Ayodoku introduces the quantitative database on lethal violence in Nigeria called “Nigeria Watch,” and the research projects that have been developed from it. The paper exemplifies a particular analytical method of violence using numbers and statistics. The authors discuss the collective way of reflecting on the conditions of the production of such data, on its potential uses, and more generally on the valuable contribution of quantitative data for qualitative research.
The approaches employed by the authors in this special issue show that if the production of materials on violence does require a sufficient proximity, research can nevertheless be conducted without direct and immediate exposure to violence, especially when conducted ex situ or ex tempo. Likewise, the scientific value of certain materials obtained prior to an event is sometimes only revealed later on, thus documenting a phenomenon that had not previously been conceived or defined as a research object a priori. However, these approaches are not in opposition to one another, but rather supplement each other, offering different options—with physical distance providing a sense of safety on the ground.
Methodological choices undertaken by the authors are the result of their scientific and personal backgrounds. We can nonetheless hope that the protection of researchers and of their informants during fieldwork will be the subject of further reflection and discussion, pushing research institutions to approach this issue in a more systemic way. Other approaches are also conceivable, such as the research group method, which generates discussion and facilitates an exchange of viewpoints. This method can also provide protection for researchers, in some situations during their fieldwork, by making them more visible and increasing the speed of production and collection of material in order to shorten the time spent in contact with danger. By mobilising concurrently different approaches, research groups can also allow for a cross-fertilisation of disciplinary backgrounds, their heritage, and preferred methodologies in order to increase the number of perspectives and to complement the ethnographic type of fieldwork. At the French Institute for Research in Africa based in Ibadan (IFRA-Nigeria) for example, where both authors of this introduction have worked, most projects are undertaken in groups, international and interdisciplinary, and where issues about working conditions and security are crucial. The Nigerian context, marked by strong travelling restrictions for foreign researchers and real security risks for many research projects on the ground, requires a strong sense of protection and vigilance. Security issues must remain a continuous concern and must be effective, given that the consequences of taking risks are not the same for researchers, depending on whether they belong to an African or Western institution, whether they are men or women, or whether they are familiar or not with the social environment in which the study takes place. The success of the research group method fuels this debate on the conditions of production and analysis of materials. This is a perfect example of what Marie Rodet, Aïssatou Mbodj-Pouye, Mamadou Sène Cissé and Mariam Coulibaly discuss in their discussion of the destruction of the Kayes archives in the “Digital archive workshop” section. The authors, researchers and archivists as well as collaborators, share their experience with projects for the protection of archival sources and reflect together on their relationship with research materials in such a volatile context such as today’s Mali. Witnesses to the deterioration of the security situation, they show how it has directly impacted their work, but also how it reveals specific challenges for the preservation of archival heritage.