Navigation – Plan du site
Partager et faire dire la violence au quotidien

Conventions of Silence

Emotions and Knowledge Production in War-Affected Research Environments
Astrid Jamar et Fairlie Chappuis


Les projets de recherche académiques et appliqués entrepris dans des contextes affectés par les conflits se déroulent principalement à travers des approches interdisciplinaires rattachées aux études des conflits et du développement. Les conventions épistémologiques, méthodologiques et professionnelles qui dominent ces disciplines ont pour habitude d’ignorer ou de mettre sous silence les émotions générées par les expériences de violence et de pauvreté. Sur base de nos expériences professionnelles, cet article développe un cadre analytique pour déconstruire ces conventions qui excluent les émotions et dimensions normatives inhérentes à tout projet de recherche dans des contextes affectés par les conflits. Nous avons identifié trois niveaux à travers lesquels les conventions dominantes dans la recherche sur la paix, les conflits et le développement mettent sous silence aussi bien les émotions du chercheur que de son objet de recherche. A travers une illustration de ces modes de mises sous silence et un dialogue avec la littérature anthropologique et postmoderne, nous démontrons les effets négatifs aux niveaux personnels, méthodologiques et épistémiques.

Haut de page

Texte intégral


1Empirically driven research on war-affected contexts requires researchers to engage with the hard realities of life in the so-called field, and with the people who suffer them daily. These interactions create emotional responses for both researchers and the people they engage with. But if this seems to state the obvious, it may be surprising that it goes largely unacknowledged within the social science research most concerned with war-affected contexts : the interdisciplinary fields of peace and conflict studies, and development studies. Professionally, methodologically and conceptually, the innate emotionality of the research experience is silenced in these studies in war-affected environments. How can this silence about the role of emotions be understood in the context of research in war-affected contexts ? What does this silence mean for academic and policy research on peace, conflict and development issues in war-affected settings ?

  • 1 Fairlie Chappuis holds a doctorate in Political Science from the Freie Universität Berlin and an in (...)

2Conventions on ethics and the effects of emotional experiences on researchers in higher education differ substantially by country and by discipline. In professional or applied policy research settings most influenced by Political Science, such contingencies are rare and when it comes to integrating, or simply acknowledging the impact of these experiences on researchers and their findings, there is a deafening silence. In extension of this neglect, no allowance is made for the effect of the research process on the emotionality of the people whose experiences are central to the whole enterprise. Our own professional experiences in various academic and policy research settings addressing issues related to security, poverty, transitional justice and forced migration have exposed us highly sensitive and intrinsically emotional exchanges between respondent and researcher. We personally have interdisciplinary backgrounds working across political sciences, international relations, development studies and anthropology with a focus on West and Central Africa. Applying our research interest working in both NGO- and think thank-settings, we both experienced the recurring silencing of emotions linked to highly traumatic human experiences1.

3Discussions of war, poverty and insecurity between the visiting, foreign researcher and the local person are usually shot through with insurmountable power asymmetries, and frequently charged with feelings of empathy, frustration, anger, helplessness and hopelessness on both sides of the interaction. This paper collects our combined reflections, emanating from our experiences of the emotional dimensions of our own fieldwork and the professional, methodological and normative dilemmas these experiences exposed us to. We develop an analytical framework for ordering these varying effects and illustrate its usefulness with examples from our own experiences.

4The framework describes the pervasive silencing of emotionality in these particular sub-fields of academic and policy-orientated research. The first section briefly situates the discussion vis-à-vis existing strands of literature relevant to emotionality in war-affected research contexts. The second presents the analytical framework that distills the conventions we have observed into six distinct modes of silencing that permeate the theory, practice and purpose of research. On the basis of this analysis, the third section presents our reflections on the methodological, ethical and epistemological implications of these silencing conventions.

Status Quo : Qualitative Research, the Researcher and Emotionality

5Covering several strands of literature on emotions in professional research contexts, this brief theoretical review points to certain gaps in methods and ethics related to qualitative research in war-affected settings. In academic disciplines, such as sociology, anthropology and psychology, reflection on the role of emotionality in research is given more importance (e.g. Hubbard et. al. 2001), but there is limited analysis of the impact, either on the research findings, or the researcher personally (Kleinman and Copp 1993 ; Dickson-Swift et. al. 2009 ; Davies and Spencer 2010 ; Millar 2014). Social science research on emotion in sensitive labour over the last two decades has sometimes considered conflict-affected contexts but most often focuses on social and health services delivered to vulnerable individuals within Western societies, e.g. seriously ill persons, abused women or vulnerable families (Dickson-Swift et al. 2009), and with rare exceptions looking at human rights activists, aid workers and development researchers (Rodgers 2010 ; Patni 2011 ; Humble 2012).

6Emotionality is defined as marginal or irrelevant across most scientific and professional settings. It is relegated to the personal sphere, framed as a logistical challenge to be managed alongside other concerns such as accommodation, transport and visa formalities (Millar 2014). The treatment of emotion in professional working environments is often “downgraded to notions of female intuition” and consequently “the caring and affective aspect … is denigrated to secondary status” (Hubbard et. al. 2001 : 122). While acknowledging the role of the researcher’s interpretations in social construction, emotional responses are still rarely taken into account specifically. Scientific approaches and emotions have long been perceived as contradictory concepts within academia (Lutz and White 1986).

7Emotions do figure prominently in discussions of methodology. The literature often emphasises the need for caution in handling the emotions of respondents (Goodhand 2000 ; Mackenzie et. al. 2007). The researcher’s emotions continue to be treated with suspicion (Kleinman and Copp 1993 : 2 ; Wood 2006). Emotions in international relations, politics and humanitarian aid are increasingly problematized (Bleiker and Hutchison 2008 ; Anderson 2014 ; Fassin 2010). Yet these examples still describe emotions as external data and remain silent about the knowledge production process – e.g. by silencing what emotions meant for the researchers or how they were dealt with.

8Only a few authors draw attention to the contribution of the emotional reaction to research processes and outcomes. A seminal example is the anthropological work of Veena Das (2007 : 8) on violence in India that aims to discuss “what happens to the subject and world when the memory of such events is folded into ongoing relationships”. However, given the dominant conventions governing academic and policy research, few authors could afford to take the personal and professional risk involved in divulging the actual role of emotional engagement in their research processes. In our fields of research, acknowledgment of emotions is perceived as a serious deviation from what are in effect idealized and unrealistic disciplinary benchmarks. Thus, because emotionality has been denigrated as “feminine intuition” (Kanuha 2000), being too open about emotional challenges in published research could endanger the credibility of the researcher.

9Relating specifically to war-affected contexts, research on emotion remains rare (Nordstrom and Robben 1995 ; Behar 1996 ; Wood 2006 ; Begley 2009 ; Thomson et. al. 2012). Emotion is mostly perceived as a challenge and a threat if it is not accounted for. According to Wood, emotional dynamics need to be mentioned because “inadequate attention to them may lead field researchers to make errors in judgement that may have significant consequences for their research subjects as well as themselves” (2006 : 384). Such perception fully acknowledges the exchange of emotions between participants and field-researchers, but as a negative component that might lead to errors (rather than valuable data).

10Overall, ethical challenges and the management of emotion are not proportionately addressed, given the sensitivity inherent to research in war-affected contexts. As a result of this status quo, dealing with the impacts of emotionality on either the researcher, the research participant or the research itself during the data-gathering and analysis processes is relegated to the personal sphere. Even soft-positivist methodological conventions for systematizing data collection and analysis in mainstream peace and conflict, and development studies, frame emotionality as subjective interference with the researcher’s objective access to social reality.

11In other branches of social science, such as anthropology or sociology, most current literature on research methods and the position of the researcher revolves around the notion of subjectivity. Much of this discussion reflects a relative consensus that a researcher’s history and sensitivity impact every step of the research process from the topic definition, to data-gathering, analysis and presentation (Warren 2006 : 214). While acknowledging the role of the researcher’s interpretations in social construction, emotionality has tended to fall on the interpretivist side of the epistemological divide.

12Thus, writing within dominant positivist methodologies continues to ignore emotionality. Within this tradition, peace and conflict studies, as well as development studies, continue to exclude emotion as a means of access to potentially valuable theoretical material. This is an approach that aspires to erase the expression of emotional dimensions of research, because of the normative implications, expressed as a potential for bias. The approach actually fails to completely exclude emotionality (and hence normative or methodological bias), however, because even the acknowledgement of emotion introduces the possibility that the research has been polluted by their influence. In contrast, we believe that emotionality can potentially explain processes of motivation, preference formation and norms underlying meaningful social action and research itself.

13Some see the silencing of emotionality as a desirable quality that burnishes the scientific status of social science research. Conversely, the experiences of other researchers working on sensitive topics and facing vulnerable people led them to question this axiom of scientific practice : leading to the emergence of so-called engaged research. Following Nancy Scheper-Hugues (1995), a group of researchers challenges the idea that researchers must observe morally odious realities with feigned ethical neutrality in order to preserve an artificial theoretical distance (Sanford and Angel-Ajani 2006 ; Kalir 2006 ; Warren 2006). On this basis, they consider it part of their obligation as subject-area experts to engage with the politics and policy of the real world.

14On the basis of our own experiences, we believe that emotional sensitivity to the harsh realities of lived human experience at the centre of our fields of study both creates and encourages normative and political engagement, and that research benefits from an explicit acknowledgement of these normative and political commitments. Indeed, the starting point does away with the idea of the researcher as a neutral instrument of the research process : it is as patently absurd to believe that the external researcher does not bring his own normative and political predilection into the research context, as it is to believe that normative and political predilections do not already exist within and shape individuals and communities under study. It is striking that both the study and practice of peace-building are only starting to acknowledge the complexity and influence of already existing social systems in contexts of external intervention in conflict-affected countries. At the same time, the idea of the researchers as not-neutral levers of change is barely broached in the literature beyond the platitudes of intervention as “political” (Paris 2002 ; Chandler 2006 ; Schellhaas and Seegers 2009 ; Bonacker et. al. 2010 ; Duffield 2010 ; Richmond 2011).

15Consequently, we propose an open acknowledgement that our emotionality shapes our normative commitments, which in turn become epistemic filters for our research ; and that the same applies to the social actors at the centre of our research field. If we begin from the (surprisingly disputed) premise that research in war-affected contexts is motivated fundamentally by the willingness to identify and alleviate the causes of suffering, it follows that the contributions of research are necessarily affected by our emotional response to the subject and our normative perspective. Although this concern appears to emanate naturally from the above-mentioned literature, both academic and policy-oriented research have developed unspoken conventions that serve to silence emotionality. The following section defines and illustrates three modes of silencing that affect emotionality in research.

Listening for Silences : an Analytical Framework for Describing Research in War-Affected Environments

16We categorize the relationship between emotionality and research, and the modes and impacts of silencing across different interconnected levels : personal, methodological and conceptual on the one hand ; researcher and researched on the other. Drawing on our experiences, we aim to contribute to the insufficient discussion of the emotionality of research in war-affected settings, by stressing the impact on the researcher, the “researched”, the research process and ultimately the outcome.

Emotionality : Conventions of Silencing

17Two important distinctions structure engagement with emotions within development and peace and conflict studies : firstly, the location of silencing, that is, whether emotionality is silenced at a personal, methodological or conceptual level ; and secondly, the emotional referent, that is, whether these conventions silence the emotionality of the researcher himself, or the subjects of the research – the “researched”. On the basis of these two distinctions, the impact of silencing conventions can be distilled into six distinct modes of silencing, as summarized in the table below.

Table 1 : Conventions of Silencing

18Due to the interactional dimensions within emotionality, the categories defined above cannot be considered absolute, but instead serve to reveal specific facets of the many and varied ways that emotionality is silenced within the theory and practice of these disciplines. The empirical examples that follow are structured around these typological categories and serve to illustrate instances of each category and how these levels are interconnected and mutually constitutive.

Personal modes of silencing

19The fact that conducting research on matters of peace and conflict can be a deeply traumatic experience for the researcher is barely acknowledged in the theory and practice of research in war-affected settings. This exclusion has the effect of relegating emotion to the personal sphere, beyond the practice of research as a profession : “handling the situation” is construed as a personal matter, something like a test of character, something that shows whether the researcher is “cut out for this” or not. As an extension of this cavalier disregard for the emotional wellbeing of the researchers active in the research process, the feelings evoked by the researcher or the research project among those whose lives are the subject of study are also relegated to the realm of the personal. Relegating the emotionality of the research subject to the private sphere effectively creates an “emotional impunity” in the practice of research, absolving the researcher of responsibility for any frustration, anger, sadness or trauma that may result from the interaction. The following describes these two effects in more detail and gives some examples.

Restricting emotionality to the private sphere as professionalism : A1

  • 2 It is also because of the strongly gendered norms surrounding expression of emotion that we have ch (...)

20When emotions are construed as a personal issue of the researcher, emotions are treated as a private and individual affair, properly kept separate from the professional, scientific, public spheres where research is conducted. Emotions are thus constructed as an inappropriate intrusion in research and policy, and by convention, the researcher inherits a professional responsibility to suppress the expression or experience of emotion. Emotions are thus silenced, and transgression of this unwritten rule of research professionalism is stigmatized as reflecting poor judgement, and a lack of personal control, or even maturity, on the part of the researcher. Callousness in the face of human suffering becomes a mark of professional maturity. There are also heavily gendered expectations around this type of silencing, as existing norms around emotional expression, hysterics and overreaction make it arguably more risky for a female researcher to be seen as emotional compared with a male colleague2. In private discussions with other colleagues, we have been warned, e.g. that talking openly of emotionality damaged the credibility of certain researchers in our fields in the past.

21An example that demonstrates this type of silencing comes from the experience of a qualitative research project focusing on the plight of refugees in the African Great Lakes region in 2012. During several months of intensive fieldwork the researcher collected personal stories of traumatic experiences, such as rape, murder, escape from death and violence, and family separation, while at the same time becoming increasingly aware of the grim future awaiting these people. The personal frustration of seeing these people, who had already suffered so much, wait helplessly in false hope for help that would never come, led the researcher to attempt to develop practical coping mechanisms, such as sensitizing informants to the limits of the international aid system, and taking up a stronger role in practitioner research. Yet none of these strategies removed the overwhelming sense of powerlessness and guilt. When the emotional burden finally began to affect the researcher’s health and working capacities, he spoke openly about it to the project manager, only to be admonished, indirectly threatened, and ultimately told that this behaviour was “everything but professional.”

The emotional impunity of the research process : B1

22Where the emotions of the research subjects are construed as personal issues, this allows research (and policy that may result) to categorize these responses as marginal or even irrelevant : this means that the emotional responses of subjects to research do not properly fall within the sphere of relevant concerns, and researchers and policy actors escape responsibility for their role in creating or reproducing certain feelings. This creates a kind of emotional impunity whereby it is acceptable and expected for researchers and policy advocates to ignore how their work makes people feel. In war-affected settings, e.g. this means that the damaging hope or fear that an interview subject may experience can be ignored completely. There is also little space for considering the potential challenges of re-traumatization in asking people to relive especially difficult experiences typical of war-affected contexts. Because the emotionality of the experience is silenced, ethical imperatives to “do no harm” are insufficiently expressed in research methodology and design.

23An instance typical of this emotional impunity encouraged by silencing of research emotionality comes from the experience of interviewing low-income respondents in the slums of large cities. It was common for people to spend hours waiting for the researcher to arrive, and then for their interview turn to come. In the course of dissecting the elements of their livelihoods it became clear to the researcher that this was a significant investment of time that literally cost the person lost livelihood : practically an entire day’s earnings in one example of a woman who managed to feed her five children on barely two euros per day. Yet the researcher had been advised that financial compensation for these people would be inappropriate, might distort answers or incentives, and would create damaging precedents. This left the researcher in the position of being responsible for having inadvertently caused someone and their children to go hungry for a day, despite having taken required precaution to avoid causing harm. Despite all the appropriate caveats and careful explanations of the nature of the research, interviews often ended with the painfully naive question, “And now what will happen ?” revealing the false hope these people felt in being heard by someone they wrongfully perceived as powerful and important.

24The above example may be dismissed by some in view of the particularly anodyne nature of the feelings at stake. Some might be tempted to believe that the objective of research justifies the inconvenience of some frustration and false hope on the part of the respondent. Putting aside for a moment the corrosive effects of disappointment in such a bleak context, as well as the fact that such logic voids the concept of respondent consent, other more obviously traumatic examples make clear the damaging nature of emotional impunity. Thus pain, fear and terror appeared when a researcher’s questions dredged up experiences of torture by corrupt security forces, or of imprisonment and rape in the context of domestic abuse. The obvious harm in asking a respondent to relive the unlivable for the purposes of the research was relegated entirely to the sphere of the respondent’s own personal emotional management. Many similar examples of emotionality being silenced within the research process could be drawn from our experiences. Silencing emotionality from the professional environment has at least two direct identifiable consequences : first, it places an enormous emotional burden on field researchers : and second, offers an escape from responsibility for emotions that may be engendered in respondents. The wider methodological, epistemological and ethical consequences of this convention are explored in the final section.

Methodological modes of silencing

25The second mode of silencing considers how methodological decisions and processes relate to emotionality, and later how silencing impacts output to promote the scientific image – sometimes a façade – of data collection and analysis.

Researchers’ emotions as threat to scientific validity : A2

26Construed as a methodological problem, dominant methodologies treat the researchers’ emotions as a threat to scientific validity, because of the risk that emotion may introduce bias, skewing data collection or analysis. Despite the fact that emotionality is inherent to research, acknowledgement of it within the scope of the research would affect the perception of the work produced. Research findings that include a discussion of the researcher’s emotionality would be considered, at best, irrelevant, and at worst, a signal of poor quality. Thus within conventional approaches, a sound methodology identifies potential bias to reduce the risk of unscientific, and hence emotional, interference in the production of credible data. The verification of translations, coding and analysis can all be construed as dedicated efforts to strip away the influence of emotionality through a filter of shared interpretation among researchers. In data collection, the field researcher must gather rational pieces of information : during data analysis, unscientific bits and pieces are to be left aside. This approach fails to acknowledge emotion as theory-relevant and valuable data.

27This problem was poignantly demonstrated in an instance where an interview that took place in Burundi in 2012 about changes in post-war security conditions and livelihoods led a young woman to describe her tragic experiences of internal displacement, separation from her family, and ultimately forced prostitution at the hands of a trusted relative. The interview became so difficult for the respondent to give, and also for the researcher to hear, that certain questions relevant to the research methodology were simply impossible to ask. This meant that the interview was of limited use from the point of view of the research methodology, even though the emotional pain that surfaced through this interaction revealed something profound about the post-war environment, and was thus an expression of some key social dynamics central to the research itself. Nevertheless, from a methodological perspective, it was more relevant to record changes in this woman’s estimated income levels as a result of these events, than the personal annihilation that changes in income signified for her.

Emotionality interferes with data collection : B2

28The emotions of research subjects are construed as methodological challenges for researchers because of the idea that the emotions of research subjects can interfere with the researcher’s supposedly direct access to objective facts or authoritative interpretations of a situation. The researcher must attempt to contain the influence of the emotional reactions of research participants on data collection or analysis. Sifting through reactions and impressions in search of the so-called facts of the matter is only ever an imperfect process based on the extent of the researcher’s own human sensibilities and inevitably flawed knowledge of the culture and context. As a result, the participant’s state of mind, the emotional relation to the research topic, and his understanding of the research project, are all factors that affect the research process and outcome.

29Drawing once again on the research project focused on refugees in central Africa, one member of the research team recounted that even if he insisted on the research objective prior to the interview, the informant continued to believe we were undertaking screening for refugee resettlement in Western countries : “She started to cry, I could not tell whether she tried to play with my feelings so I would help her, or [if it was] because it was very hard for her to express herself”. In another example, one informant stood up to unplug his tablet computer at the end of a two-hour interview, during which he had lamented a lack of resources so severe he could not feed his family properly or cover other basic needs. In this instance, the emotional reaction of the researcher – of feeling betrayed – encouraged him to neglect data from this interview. Retrospectively, the possession of the tablet could be a sign of exaggeration from the informant as much as a false indicator of wealth. In this research project, the research team discussed on several occasions whether some refugees exaggerated their stories in order to receive empathy and consequently future support. Such judgements created suspicion towards the validity of information gathered. As a policy-oriented project, the “too-emotional” data was just left aside and not integrated in data analysis, as it was simply considered irrelevant. At the methodological level, reflexivity on these interviews resulted in balancing the data gathered with emotional exchanges between the participant and the researcher, which ultimately created another level of bias.

30Mainstream methodological concerns in peace, conflict and development studies take little account of how social and disciplinary conventions silencing emotions can lead to the suppression of relevant data. The examples above suggest that attempts to avoid emotional bias do not solve the problem. On the contrary, scientific precautions to limit emotional interference may ultimately have unintended consequences on the data-collection and reflexivity process. Paradoxically, such measures could actually increase biases by overemphasizing efforts to create distance from emotion. Stripping away emotional reactions and content not only strips out important information about actors and social context, but creates a different kind of bias in the research by introducing a systematic discrimination against certain facets of the data. Moreover, because of the nature of the way emotional responses are likely to be triggered, this bias may disproportionately affect the most marginalized and powerless ; thus we are more likely to be concerned our research is tainted by a sense of concern and empathy when studying the conditions of the poor and marginalized, than the wealthy elite.

Conceptual modes of silencing

31In extension of the methodological analysis, normative commitments are one of the elements impacting all stages of research. Jointly construed on the basis of emotion and reflexivity, normative commitments order relations between the researcher, data, and thus the research output at an implicit, and usually unacknowledged, level. Silencing emotion pushes normative commitments into the background, but as emotion influences our values and our ethics, being silent on what we feel leaves blank the rationale framing certain types of arguments. This lack of transparency about why research is framed a certain way inhibits rational evaluation of the research process and its contribution. Façades of neutrality also mask the reasons for decisive choices in research design that can in turn skew research towards particular political outcomes.

Normative engagement as a threat to credibility : A3

32The emotions of the researcher are part of the normative motivations underlying a research agenda and the use of specific concepts or approaches, because people are committed to political agendas at least in part on the basis of emotion. Academic research in peace and conflict studies is largely premised on normative commitments to liberal democracy as a preferable form of human social organization, but this creates bias in the research against other forms of social order that might be based on different social practices, traditions or understandings, for example : gender norms or traditional forms of political legitimation. Partly because emotionality is stigmatized in the professional research field and in methodology, normative political commitments, which are also influenced by emotionality, must also be muted even as they contribute to the conceptualization of research itself. In advocacy-driven research settings, commitments to specific social, economic and political models can be declared from the outset and the research is therefore more transparent in declaring its normative orientation. Acknowledging the normative framework shaping an approach to research is not, however, sufficient to protect it from the influence of unacknowledged normative biases, as the following example shows.

33The experience of writing a report on research findings for an NGO in central Africa illustrates how the researcher is requested to adopt a particular normative framework in an advocacy-driven setting. Even though we all have normative engagements, this specific example highlights clearly the requirement to adapt the report to reflect the normative engagement of the authoring institution. Although the research adopted a robust methodology within these parameters, the underlying normative context of the project impacted the scope of the research and therefore on all other dimensions of the research process. Due to its mandate and negotiated role vis-à-vis national institutions, donors and other NGOs, the NGO was normatively rooted in a tradition of western legalism. The NGO had committed to observe and document infringements of the judicial process related to a law it had also helped to advocate for. As a result, the values of Western legalism were also integrated into the legal framework of the system observed. Initially the researcher thought to address the limits inherent to the policy model under scrutiny, and then expand the analysis to infringements of its implementation. Ultimately an executive decision was made to delete sections critically analyzing the model and focus instead on instances of abuse of power by certain partners. According to the writing researcher, these elements were necessary to explain the lack of fairness of the process. However, the NGO’s normative engagement towards Western legal ideas and its commitment to support national institutions in undertaking justice defined key elements of the report. Its normative commitment is openly acknowledged but the effects on the research produced are silenced : specifically, the fact that the report focused on a single type of critique, due to their normative commitments, was left unspoken for concerns of credibility and coherence.

Lived experience as subject for research : B3

34In contrast to other modes of silencing, the area of study in war-affected environments that has given the most attention to emotionality is the study of the experience of people affected by war, violence and peace. In this type of research, the emotionality of the people involved is made the central object of research itself, opening new possibilities to understanding the social context of peace and conflict. Using the emotionality of people who experience war, peace and poverty as an explanatory factor can explain wider social processes. This approach fits with the ontological premise that human interactions provide a foundation for causal explanation, which is also open to empirical verification (Rueschemeyer 2009 : 33). Frequently making use of ethnographic research methods, this type of research can improve understanding of how the micro-dynamics of human experience in specific contexts interact with the levels of analysis more common to other types of research in this area, e.g. the institution and the state. This approach allows for the integration of emotionality into soft-positivist research designs by looking at the links between internal states of emotion, cognition, preference formation and wider social contexts – options for analysis that were previously sidelined from analysis by a refusal to engage with emotionality.

35There is thus increasing interest in the emotions of international relations, and many examples of this type of research project within the peace, conflict and development literature (Peters and Richards 1998 ; Narayan et al. 2000 ; Wood 2003 ; Weinstein 2006 ; Uvin 2009 ; Blattman 2009). These developments are positive in the sense that this focus on internal emotional states corrects against the silencing of emotionality. However, this development is problematic insofar as the researcher’s claim to direct access to the internal states of research subjects allows an appropriation of feelings and experiences for the researcher’s own broader external purposes. The fact that research subjects may have little understanding and no control over the use of such ‘data’ poses conceptual, methodological and above all ethical challenges.

36Perception-based research that seeks to address peace and conflict at the so-called local level is emblematic of this type of approach and is becoming increasingly common in peace and development studies. An example of such research comes from a study where the effectiveness of externally sponsored security interventions were analyzed in terms of the perceptions of residents of two post-conflict cities. The research sought to track the effectiveness of efforts to improve the performance of police and military institutions by asking how people’s feelings of fear and safety changed over time. The research usefully linked institutional reform to changes in community perceptions, demonstrating the power of this conceptual adaptation to reveal new facets of peacebuilding practice. However, the process inevitably involved reducing the complex personal experiences of fear and safety to anonymous data points in a study that could ultimately rationalize interventions over which the people directly affected have no control.

Implications of silenced conventions

37Conventions of silencing have epistemic, ethical and methodological consequences, as this section discusses in turn.

Normative commitments in academic and policy research

38Conventional approaches to research in peace and conflict presuppose that the emotions of the researcher are irrelevant to theoretical research. Other disciplines, e.g. anthropology, do confront emotion in the study of peace, war and development (Wood 2003 ; Fassin 2010 ; Rodgers 2010 ; including ourselves), but these are still carefully bracketed so as to preserve the relevance of the findings and the credibility of the researcher. Thus the premise remains that the researcher maintains a degree of personal, emotional detachment from the field of study – even while immersed in participant observation. Such detachment aims to prevent research becoming co-opted by personal, political agendas. From this point of view, emotions are dangerous because they can become a back-door through which bias could enter the research process, invisibly shaping research design, process and findings, ultimately threatening scientific validity.

39The conventions of silencing we describe in this paper are meant to protect the research process from the dangerous influence of emotion. Yet these conventions routinely fail to exclude emotional bias and its influence. Instead, the unintended consequence of these conventions is to disguise the actual influence of emotion. This is problematic because what is implicit and invisible, and unacknowledged and unspoken, cannot be directly evaluated or challenged. If emotion shapes normative commitments, which become the epistemic filter through which research passes, no degree of self-awareness can control for this type of influence nor systematically account for it. Moreover, countering this influence depends on the personality of the researcher, which exposes the limits of reflexivity as a viable mitigating strategy. From this perspective, emotions are dangerous not because they provide a back-door for normative commitments into the research process, but because emotionality is already inside the house, so to speak, inevitably built into the unacknowledged premises of research, part of its very architecture.

40The call to make normative commitments explicit in research led to the development of engaged or activist research approaches (Scheper-Hughes 1995 ; Sanford and Angel-Ajani 2006 ; Warren 2006). Rejecting disciplinary conventions that require a façade of scientific neutrality based on the imperfect analogy between social and natural sciences, advocacy researcher entails normative transparency, which allows all research and policy positions to be openly interpreted through these declared pre-existing political and ethical commitments.

41Of course, for those openly engaged in advocacy research, a declared normative mandate can also undermine their perceived credibility, allowing their work to be criticized as unscientific and politically motivated. For this reason, many such organizations seek to disavow the normative motivations underpinning their work. Instead they cling to claims of scientific methodology as a way to bolster their credibility in a chosen area. Alternatively, and more gravely, an overarching normative mandate is sometimes used as a shield behind which weak research is hidden : in this case commitments to certain ideals predetermine the outcome of research to an extent that it becomes a mere accessory to the argument itself.

42In contrast, we argue for the acknowledgement that emotion shapes normative commitments, which in turn are epistemic filters for research, and that this fact must be integrated across the professional, methodological and theoretical dimensions of research. As we have illustrated, silencing emotionality not only strips out important information ; it also generates a different kind of bias in the research by systematically rejecting certain data, thereby providing a skewed picture of social reality. When life and research in war-affected contexts typically involves reminiscences of violence and deprivation in daily activities and relationships, portraying an environment free of emotion is an illusion. Feelings of mistrust, vengeance, empathy, fear, relief and hope among individuals shape actions by communities, political stakeholders, policy makers and those who study them in ways that are currently largely neglected.

43Ethnography and ethnography of aid are already embedded in post-modern epistemologies (Lewis and Mosse 2006 ; Hammersley and Atkinson 2007) that engage reflexively with the production of knowledge. However, while looking at rapport, power, agency and hierarchies, both within the research process (through interactions between researchers and participants) and the research context (through interactions among participant communities), engagement with emotion remains superficial. As a result, gaps remain in ethical guidelines and requirements for research in poor and war-affected regions.

Greater emotional transparency in research requires greater methodological transparency

44Supporting the open acknowledgement of our positionality by approaching social realities through the prism of power, the subjective nature of social relations and representation, for some requires the rejection of objectivity and the idea of truth (Foucault 1982 ; Sylvester 1994 ; Said 1995). However, the representation of social reality as an intersubjective experience does not require renouncing entirely the goal of explaining social reality (Rueschemeyer 2009 : 33). Yet in either case, if we accept the premise that emotionality frames research, yet remains a silent influence, then it makes sense to encourage transparency in both emotional responses and normative engagement. In other words, because emotion and consequent normative commitments are integral to research, we ought to display transparency in writing and reporting it.

45Research approaches inspired by postmodern epistemologies already claim that researchers cannot credibly protect their research from the influence of their own underlying emotions and normative commitments, and have promoted the idea that such commitments should be made plain in research (Malkki 1995 ; Nordstrom and Robben 1995 ; Behar 1996 ; Wood 2003 ; Sanford and Angel-Ajani 2006 ; Thomson et. al. 2012). Postmodern epistemologies have won few friends within peace, conflict and development studies (and heterodox research methodologies are called such for the same reasons). Sociological, ethnographic and anthropological research traditions tend to have more sophisticated approaches to reflexivity. Partly for these reasons, these disciplines have gained increasing attention in the study of peace and conflict in recent years, evident in the so-called “ethnographic turn in IR” and increasing focus on micro-dynamics of conflict (Vrasti 2008). Nevertheless, these approaches still remain insufficiently developed to effectively counter the silencing that can be caused by the accusation of emotional bias in the methodology.

46There is a greater consensus around the idea of the researcher as inevitably entangled in the research context than there is about the consequences of this acknowledgement. It is only by exposing these fundamental influences that we can discuss their relative impact on the research process, and also begin to understand what alternative forms of knowledge they might bring. Both academic and practice-driven research need to complement a transparent declaration of emotionality and normative commitment with more robust answers to the question, “Why believe this ?” For academic researchers, this is a matter of professionalism that is all too often restricted to graduate school classes and methodology sections in PhD dissertations, while for practice-driven research, methodology presentation serves a political purpose of plugging gaps in credibility.

47More open acknowledgement of emotion and normative commitment requires methodological explanation in order to better distinguish between assessments we can all largely agree on for sound methodological reasons, and analysis motivated by other types of engagement. In this way, emotional transparency clarifies and enhances social scientific methodology. Transparency can thus help resolve normative conflicts that often underlie apparent research dilemmas. More pragmatic approaches to research make clear, on the one hand, why we prioritize a particular goal over another, and on the other hand, the nature of the trade-off to be struck in achieving balance.

Research ethics and doing no harm

48Conventional research ethics call us to “do no harm” (Goodhand 2000 ; Mackenzie et. al. 2007), yet the silencing of emotionality harms the people on both sides of the research process. Given practical limitations for the protection of respondents, the conflicts of interest inherent to the assessment of harm, and the fact that professional conventions systematically marginalize emotional experience, how can researchers credibly claim to do no harm ? More attention to the emotionality of the research experience would be one step towards taking greater responsibility for the potential emotional damage that research may do, both to the researcher and to the researched.

49Research can affect the researcher personally. Dickson-Swift et. al.(2009 : 70–71) point to the “human costs” that emotional labour can inflict on researchers, including feeling “phoney”, emotions of guilt and self-blame, difficulties sleeping, anxiety, gastrointestinal upsets and depression. Researchers may develop personal strategies to distance themselves from the data that are not necessarily conducive to the wellbeing of the researcher. On the other hand, silencing respondent emotions allows the negative emotions that research can stir to be conveniently ignored. Silencing emotions does not address the challenge of emotionality in such research : it merely creates a harmful fiction of a formal sphere free of emotionality. Greater recognition of the dangers that research poses would allow for the expression of emotional experiences. In institutional settings, this would include an onus to provide support throughout the research process.

50Moreover, the healing potential of being heard, or the idea that storytelling in itself may have a cathartic effect is sometimes evoked, as well as the more political project of giving voice to vulnerable people to alleviate their suffering, empower them, and contribute to awareness-raising. Denouncing commodification of trauma narratives, several authors criticize these benefits and underline insufficient consideration given to people sharing their stories (Hamber and Wilson 2002 ; Ross 2003 ; French 2009). For instance, on the basis of an analysis of storytelling organized for Apartheid victims in South Africa, Colvin underlines that :

Traumatic storytelling has not brought them reparations, it has not eased their poverty, it has not forced perpetrators to confess or beneficiaries to admit their own liability. Only on occasion has it seemed to ease the psychological effects of trauma. More often than not, after the brief ‘intervention’ ... they are left to go home alone, with little follow-up support. (2008 : 174)

51Such observations can be applied to most qualitative research projects with vulnerable people. In theory, of course researchers should take every precaution to assure their research does not re-traumatize or disturb respondents in any way. In practice, few precautions to protect respondents are possible, and even less are feasible given the limitations of resources to which silencing contributes. Moreover, there is an inherent conflict of interest in making the researcher, whose career may depend on access to data, responsible for judging whether the access gained comes at too high a price to the respondent. The fact that the emotionality of the respondent is often silenced means that the researcher is likely to systematically underestimate the harm their research might do.


52This paper has brought to fore how policy-orientated and academic research silence emotions, and therewith, normative engagement. From our research experiences in war-affected countries, it has identified three modes of silencing convention that define how a “good professional” should handle emotions. These conventions aim to preserve scientific credibility, even though they differ between disciplines within academia and policy research settings. The consequences of these silencing conventions for research have been discussed in terms of ethics, epistemology and methods. First, silencing functions to hide the emotional damage research may do to both the researcher and the “researched”. Silencing emotionality helps to gives the illusion of doing no harm, when, in fact, means for protecting researchers and respondents remain woefully lacking. Second, current conventions on method and conceptualization operate as epistemic filters leaving out data, in a way that contributes to another type of bias. Finally on this basis, it becomes part of the methodological process to silence emotions.

53Since the increase of internal conflicts since the end of the cold war, peace and conflict studies are increasingly looking into effects of violence at the individual level, widening the initial scope of military, political or institutional analyses most relevant to interstate wars. This evolution has led the interdisciplinary peacebuilding field to get closer to suffering and vulnerable individuals, and consequently to difficult emotions. In parallel of this evolution, Colvin observes : “traumatic storytelling is an increasingly common activity in post-conflict, democratizing societies, an activity that produces an ever-expanding volume of narratives of traumatic suffering and recovery” (2008 : 174).

54These data can only result from a research process heavily charged with a range of human emotions. A number of researchers have underlined the ethical issues implicit in this commodification of trauma narratives with insufficient consideration given to the people sharing their stories. Ethnographic research about war-affected settings stresses important unresolved challenges to current research approaches. It is simply not sufficient to affirm the usefulness of voicing the suffering of vulnerable people by undertaking and publishing research. More importantly, the silencing of emotions in the reflexive process may in fact inadvertently create other types of power abuse by giving complete impunity to the researcher to determine tolerable degrees of emotional discomfort. Finally, Begley (2012) observes that emotional challenges encountered by the researcher only reflect the tip of the iceberg compared with what the research participants face in their daily lives in war-affected environments.

55For all these reasons, this paper calls for more transparency in emotionality and normative engagement through all stages of knowledge production within development, peace and conflict studies in war-affected environments. From a methodological perspective, there is no single solution fit to handle all the ethical and emotional challenges encountered during fieldwork : no book could had tell us how to handle them during field research. We suspect that identifying challenges, and maintaining openness about the strategies adopted and integrated into the research process are keys elements to acknowledging emotionality, rather than silencing it.

56Through this paper, we have assembled different pieces of the puzzle, hoping to create grounds for complex debates in wider development and peace conflict studies in relation to emotionality. Engaging with the interdisciplinary debate contained in the “affective turn” (Massumi 2002 ; Gregg and Seigworth 2010 ; Ahmed 2013) and scrutinizing the cognitive role of emotions (Nussbaum 2003) is crucial to adequately approaching the complexity of emotions. To grasp the role of emotions and their relevance for understanding conflict-affected environments, and to develop policy embracing such complexity, further research is essential to bridge these two insufficiently connected research areas. This reflection thus constitutes an invitation to further research and reflection on this topic. In particular, we feel it important to end by noting that although we have focused this paper on negative emotions, as these were the most challenging in our personal research experiences, joy, comprehension, empathy, hope, and the raw beauty of human resilience were also features of our research experience. These experiences continue to provide us inspiration for further engagement in policy and academic research on development, peace and conflict issues in post-war environments.

Haut de page


Sara AHMED, The Cultural Politics of Emotion, New York, Routledge, 2013.

Rosie ANDERSON, “Playing the Fool : Activists’ Performances of Emotion in Policy Making Spaces”, Emotion, Space and Society, 13, 2014, pp. 16-23.

Larissa BEGLEY, “The Other Side of Fieldwork : Experiences and Challenges of Conducting Research in the Border Area of Rwanda/eastern Congo”, Anthropology Matters Journal, 11 (2), 2009, pp. 1-11.

Larissa BEGLEY, “The RPF Control Everything ! Fear and Rumour 70 under Rwanda’s Genocide Ideology Legislation”, in Susan THOMSON, An ANSOMS, and Jude. MURISON, Emotional and Ethical Challenges for Field Research in Africa : The Story Behind the Findings, UK, Palgrave Macmillan, 2012, pp. 70-83.

Ruth BEHAR, The Vulnerable Observer : Anthropology That Breaks Your Heart, Boston, Beacon Press, 1996.

Christopher BLATTMAN, “From Violence to Voting : War and Political Participation in Uganda”, American Political Science Review, 103 (02), 2009, pp. 231-247.

Roland BLEIKER and Emma HUTCHISON, “Fear No More : Emotions and World Politics”, Review of International Studies, 34, 2008, pp. 115-135.

Bonacker THORSTEN et. al., Interventionskultur : Zur Soziologie von Interventionsgesellschaften, Wiesbaden, Springer, 2010.

David CHANDLER, Empire in Denial : The Politics of State-Building, London, Pluto Press, 2006.

Veena DAS, Life and Words : Violence and the Descent into the Ordinary, Berkeley, University of California Press, 2007.

James DAVIES and Dimitrina SPENCER, Emotions in the Field : The Psychology and Anthropology of Fieldwork Experience, Stanford, Stanford University Press, 2010.

Norman DENZIN, On Understanding Emotion, New Brunswick, Transaction Publishers, 1994.

Virginia DICKSON-SWIFT et. al., “Researching Sensitive Topics : Qualitative Research as Emotion Work”, Qualitative Research, 9 (1), 2009, pp. 61-79.

Mark DUFFIELD, “Risk-Management and the Fortified Aid Compound : Everyday Life in Post-Interventionary Society”, Journal of Intervention and Statebuilding, 4 (4), 2010, pp. 453-474.

Didier FASSIN, “Heart of Humaneness : The Moral Economy of Humanitarian Intervention”, in Didier FASSIN and Mariella PANDOLFI (ed.), Contemporary States of Emergency : The Politics of Military and Humanitarian Interventions, Brooklyn NY, Zone Books, 2010, pp. 269-293.

Michel FOUCAULT, “The Subject and Power,” Critical Inquiry, 8 (4), 1982, pp. 777-795.

Brigittine FRENCH, “Technologies of Telling : Discourse, Transparency, and Erasure in Guatemalan Truth Commission Testimony”, Journal of Human Rights, 8 (1), 2009, pp. 92-109.

Clifford GEERTZ, The Interpretation of Cultures : Selected Essays, New York, Basic Books, 1973.

Jonathan GOODHAND, “Research in Conflict Zones : Ethics and Accountability”, Forced Migration Review, 8 (4), 2000, pp. 12-16.

Melissa GREGG and Gregory SEIGWORTH, The Affect Theory Reader, Durham and London, Duke University Press, 2010.

Brandon HAMBER and Richard WILSON, “Symbolic Closure through Memory, Reparation and Revenge in Post-Conflict Societies,” Journal of Human Rights, 1 (1), 2002, pp. 35-53.

Martyn HAMMERSLEY and Paul ATKINSON, Ethnography : Principles in Practice, New York, Taylor & Francis, 2007.

Gill HUBBARD et. al., “Working with Emotion : Issues for the Researcher in Fieldwork and Teamwork”, International Journal of Social Research Methodology, 4 (2), 2001, pp. 119-137.

Darryl HUMBLE, “‘This Isn’t Getting Easier’ : Valuing Emotion in Development Research”, Emotion, Space and Society, 5 (2), 2012, pp. 78-85.

Barak KALIR, “The Field of Work and the Work of the Field : Conceptualising an Anthropological Research Engagement”, Social Anthropology, 14 (2), 2006, pp. 235-246.

Valli KANUHA, “‘Being’ Native versus ‘Going Native’ : Conducting Social Work Research as an Insider”, Social Work, 45 (5), 2000, pp. 439-447.

Sherryl KLEINMAN and Martha COP, Emotions and Fieldwork, Newbury Park, Sage Publications, 1993.

David LEWIS and David MOSSE, Development Brokers and Translators : The Ethnography of Aid and Agencies, Bloomfield, Kumarian Press, 2006.

Catherine LUTZ and Geoffrey WHITE, “The Anthropology of Emotions,” Annual Review of Anthropology, 15, 1986, pp. 405-436.

Catriona MACKENZIE et. al., “Beyond ‘Do No Harm’ : The Challenge of Constructing Ethical Relationships in Refugee Research”, Journal of Refugee Studies, 20 (2), 2007, pp. 299-319.

Liisa MALKKI, Purity and Exile : Violence, Memory, and National Cosmology among Hutu Refugees in Tanzania, Chicago and London, University of Chicago Press, 1995.

Brian MASSUMI, Parables for the Virtual : Movement, Affect, Sensation, Durham and London, Duke University Press, 2002.

Gearoid MILLAR, An Ethnographic Approach to Peacebuilding : Understanding Local Experiences in Transitional States, New York, Routledge, 2014.

Patel NARAYAN et. al., Voices of the Poor : Can Anyone Hear Us ? , New York, Oxford University Press, 2000.

Carolyn NORDSTROM and Antonius ROBEEN, Fieldwork Under Fire : Contemporary Studies of Violence and Survival, Berkeley, University of California Press, 1995.

Martha NUSSBAUM, Upheavals of Thought : The Intelligence of Emotions, Cambridge, Cambridge University Press, 2003.

Roland PARIS, “International Peacebuilding and the ‘mission Civilisatrice’”, Review of International Studies, 28 (04), 2002, pp. 637-656.

Rachana PATNI, Emotional Fools and Dangerous Robots : Postcolonial Engagements with Emotion Management, PhD Theses, London, Brunel University, 2011.

Krijn PETERS and Paul RICHARDS, “‘Why We Fight’ : Voices of Youth Combatants in Sierra Leone”, Africa, 68 (02), 1998, pp. 183-210.

Oliver RICHMOND, A Post-Liberal Peace, Oxon and New York, Routledge, 2011.

Kathleen RODGERS, “‘Anger Is Why We’re All Here’ : Mobilizing and Managing Emotions in a Professional Activist Organization”, Social Movement Studies, 9 (3), 2010, pp. 273-291.

Fiona ROSS, “On Having Voice and Being Heard Some after-Effects of Testifying Before the South African Truth and Reconciliation Commission”, Anthropological Theory, 3 (3), 2003, pp. 325-341.

Dietrich RUESCHEMEYER, Usable Theory : Analytic Tools for Social and Political Research, Princeton, Princeton University Press, 2009.

Edward SAID, Orientalism : Western Conceptions of the Orient, London, Routledge & Kegan Paul, 1995.

Victoria SANFORD and Asale ANGEL-AJANI, Engaged Observer : Anthropology, Advocacy, and Activism, New Brunswick and London, Rutgers University Press, 2006.

Constanze SCHELLHAAS and Annette SEEGERS, “Peacebuilding : Imperialism’s New Disguise ?”, African Security Studies, 18 (2), 2009, pp. 1-15.

Nancy SCHEPER-HUGHES, “The Primacy of the Ethical : Propositions for a Militant Anthropology,” Current Anthropology, 36 (3), 1995, pp. 409-440.

Christine SYLVESTER, Feminist Theory and International Relations in a Postmodern Era, Cambridge, Cambridge University Press, 1994.

Susan THOMSON et. al., Emotional and Ethical Challenges for Field Research in Africa : The Story Behind the Findings, UK, Palgrave Macmillan, 2012.

Peter UVIN, Life After Violence : A People’s Story of Burundi, London, Zed books, 2009.

Kay WARREN, “Perils and Promises of Engaged Anthropology : Historical Transitions and Ethnographic Dilemmas”, in Victoria SANFORD and Asale ANGEL-AJANI, Engaged Observer : Anthropology, Advocacy and Activism, New Brunswick and London, Rutgers University Press, 2006.

Jeremy WEINSTEIN, Inside Rebellion : The Politics of Insurgent Violence, Cambridge, Cambridge University Press, 2006.

Elisabeth WOOD, Insurgent Collective Action and Civil War in El Salvador, Cambridge, Cambridge University Press, 2003.

Elisabeth WOOD, “The Ethical Challenges of Field Research in Conflict Zones”, Qualitative Sociology, 29 (3), 2006, pp. 373-386.

Haut de page


1 Fairlie Chappuis holds a doctorate in Political Science from the Freie Universität Berlin and an interdisciplinary Masters in International Affairs from the Geneva Graduate Institute of International and Development Studies. Her work revolves around the human security implications of external security interventions in post-war contexts. As a researcher she has worked in both academic and applied research settings ; as a practitioner she has worked in fields of livelihood development and security reform. She is currently a Programme Manager with the Research Division at DCAF : a centre for rule of law, security and development.

Astrid Jamar holds a doctorate in Development Studies from the University of Sussex, a Masters in Peace Studies from the International Christian University (Japan) and a Masters in International Relations from the Free University of Brussels. Her work addresses the professionalisation and technocratisation of transitional justice, peacebuilding and international aid. She has field experience working with international NGOs and local institutions on transitional justice. Reflection upon these experiences in academic and policy-oriented research formed the basis of her doctoral research. She is currently a Researcher at the Political Settlements Research Programme, University of Edinburgh.

2 It is also because of the strongly gendered norms surrounding expression of emotion that we have chosen to use masculine pronouns in this paper.

Haut de page

Table des illustrations

Fichier image/png, 50k
Haut de page

Pour citer cet article

Référence électronique

Astrid Jamar et Fairlie Chappuis, « Conventions of Silence  », Parcours anthropologiques [En ligne], 11 | 2016, mis en ligne le 20 décembre 2016, consulté le 24 juillet 2017. URL : ; DOI : 10.4000/pa.513

Haut de page


Astrid Jamar

University of Edinburgh

Fairlie Chappuis

Free University of Berlin

Haut de page

Droits d’auteur

Parcours anthropologique

Haut de page
  • Logo Université Lumière Lyon 2