While members of the Rwandan political élite, both in power and in exile, do not deny that the 1994 genocide occurred, there are significant differences in how they explain its cause, and the degree of context they consider necessary for a ‘true’ representation of the conflict to emerge. Despite this, recent journalistic works on Rwanda present a monovocal, ‘factual’ narrative about the genocide. While such narratives are clearly attractive to a wider readership, anthropology’s concern to give voice to competing representations as a cause, rather than subsidiary feature of conflict places researchers in a sensitive and difficult position. To what extent do anthropological treatments of conflict necessarily oppose such ‘factual’ narratives, or can they be regarded as complementary?
Read Linda Melvern's response, to this article (May 2002) here.
The following is a personal reflection on some of the difficulties and dilemmas involved in writing a thesis about Rwanda. While this article makes reference to many ideas current in anthropology, it is not intended either as a review of anthropological theory or as a précis of my thesis, although I refer to both indirectly.
The subject of my research is the post-genocide discourse of Rwandese in both Rwanda and Europe. I conducted interviews with former and current government officials, civil society leaders, and journalists. Additional interviews were carried out with non-Rwandese academics, journalists, and officials of various international agencies. The intention of my research is to evaluate discursive features such as the use of historical narratives and human rights terminology in the representation of Rwanda. As would be expected, the substantive content of the discourses of those in exile and those now in government differ at key points. It should be made clear, however, that none of those interviewed denied that there had been a genocide in Rwanda in 1994. Rather, disagreement concerns the historical processes that led to that event, and the degree of contextualisation required to adequately explain it. Certain key themes dominate the discourse, including the degree to which the genocide must be contextualised, the use of historical narratives to ‘explain’ the genocide, the image of an ‘international community’, and the question of the institutional guilt of the Rwandan ‘church’.
An example of one of these themes dealt with in my thesis is the manner in which Rwandese invoke historical narratives to ‘explain’ the cause of the 1994 genocide. The past is presented as inevitably leading to the present. These narratives demonstrate a desire to explain an event that, although now in the past, is held to hold the key to the present. The striking feature of these narratives is the implicit consensus about which moments in time are considered contentious. The narratives consequently cluster around particular points in time. The instrumental potential of these narratives relies on the assumption, entrenched in the Western historical genre, that an appeal to the past is inherently objective and authentic, and hence beyond dispute. Given the central part played by European, colonial academics in ‘explaining’ ethnicity, and the Western approach to how-the-past-creates-the-present, it comes as no surprise that Rwandese frame much of their explanation for 1994 by reference to this same history. Following the European tradition, Rwandese narratives fulfil the ‘criteria’ of facticity, and reproduce the claims of Western historiography. Rwandese look to the ‘other country’ of the past to demonstrate the ‘truth’ of the present.
In my thesis, I also consider the polyvocal representations of the genocide made possible by the ambiguity inherent in the universal template of the United Nations Convention on the Prevention and Punishment of Genocide. This ambiguity rests not on what genocide is - that much is made explicit - but on what further contextual details should be included to achieve an ‘authentic’ description of the genocide. This chapter in my thesis seeks to demonstrate how universal human rights law has become a discursive resource within which actors produce multiple representations of an event, while still maintaining that it was a case of genocide.
By presenting competing narratives, the ‘history’ chapter of my thesis does not provide a single diachronic history of Rwanda. The actual substance of the discourse is secondary to an exploration of how historical narratives are imbued with authenticity and objectivity. Likewise, the ‘genocide’ chapter does not try to explain why and how the genocide occurred, but explores the contest over where the parameters for requisite ‘context’ should be placed. Both chapters explore how multiple representations are articulated, and how they are formed and re-formed in response to competing discourses. As one would expect in a conflict situation, these competing discourses are not articulated in isolation, but are inherently dialectical.
Such an approach is in sharp contrast to ‘popular’ writings on the 1994 Rwandan genocide. A recent example is Philip Gourevitch’s We Wish to Inform You That Tomorrow We Will be Killed With Our Families (1998), winner of the Guardian First Book Award in 1999. By using interviews and personal testimony, Gourevitch provides an absorbing narrative of the genocide. Likewise, Linda Melvern’s book, A People Betrayed: The Role of the West in Rwanda’s Genocide (2000) takes the reader through events in chronological sequence, providing dates, names, and statistics, all the necessary hallmarks of a ‘factual’ narrative. Both books are powerful contributions to the corpus of work on Rwanda and should be valued as such. They both fulfil the dominant criteria for works on conflict by responding to the demand that we be taken, step-by-step, through its causes and effects. Both reflect the just-give-us-the-facts approach.
I use the term ‘popular’ with caution. In academic circles, the term ‘popular’ is often understood pejoratively. The term, however, also suggests accessibility. This accessibility is premised on the basis that if something exceptional has happened, it must have a causal explanation. My argument is not that either work is intrinsically wrong or misleading. Both provide an explanation of the Rwandan conflict. No one can deny that informing the ‘general public’ about genocide is a worthy endeavour. Rather, I pose the question (both to myself and other anthropologists) why a monovocal narrative, in which facts are presented as objective, should remain so appealing? Why is the just-give-us-the-facts approach so pervasive?
One can speculate about how, if I had been in a different position (a journalist like Gourevitch and Melvern, for example), I might have gone about writing about Rwanda. Is it merely my presence in an academic institution that requires I find new and ‘innovative’ (pedantic?) ways to represent Rwanda? Am I motivated to write in a way that will fulfil the criteria of an especially select readership; that is, my thesis examiners? In other words, would I have been better off writing a ‘factual’ narrative of Rwanda? Such an approach would have enabled me to side-step postmodern concerns, and avoid the deferential nods to Foucault. Above all, such an approach would have enabled me to stay well away from the ever present charge of revisionism, the charge that by going beyond conventional narrative and accepted facts I have some hidden agenda that questions the veracity of one million Rwandese dead. In one departmental seminar I was asked why I felt I had to make explicit my reasons for believing that the events in Rwanda in 1994 should be defined as genocide. My reason for doing so is that, unlike other areas of study (and in parallel with writings on the Holocaust), any movement away from the received representation of events may be treated with suspicion and denounced as revisionism. Any revision of the accepted framework, even if it maintains the fundamentals of that framework, can be regarded as denial-by-stealth. I feel it pays to be explicit.
Given these high stakes, what exactly is the value of my research? It is not a question of whether or not my research is superior to ‘factual’ narratives. There are no real criteria to make such a judgement, or judges who could make that decision. On the level playing field of partiality there are no winners. Rather, the question I should ask is, if the story has already been told as an eloquent and coherent narrative, what is left for me to contribute?
Whether we like it or not, anthropologists must accept that certain readerships require ‘facts’ to attribute meaning to particular events or occurrences. For them, ‘meaning’ is not hidden, but is evident in the ‘facts’ themselves. Arguments of representation and partiality do not carry the same weight. The question we are left with is what is the added value of a deconstructionist approach? Ethically, are we not duty-bound to give voice to the reality of events such as genocide in all their horrifying detail? By deconstructing discourse, am I not simply obscuring reality while claiming to demystify it?
Clearly, the key to appreciating the power inherent in post-genocide discourse is its partiality. I am prepared to admit, for instance, that my own attempts at deconstruction are necessarily partial. These familiar anthropological concerns regarding explicit reflexivity and the partiality of the author are noticeably absent, however, from the journalistic accounts mentioned above. The persuasive feature of the books by Gourevitch and Melvern is that they are presented as factual narratives, allowing witnesses to tell-the-story-in-their-own-words. As such, they are powerful testimony to lived experience. And yet we are aware that an ‘analytical’ position has been taken in the process of excluding or including certain facts and testimony. That authors claim to be disinterested conduits of impartial information, rather than arbitrators of knowledge, is a critique familiar within anthropology. Such an ‘in-house’ critique, however, does not solve the question of what my own research contributes. Does my own approach unintentionally mystify straightforward testimony?
Having worked for a conflict prevention/resolution NGO in Rwanda I am all too aware of how ‘facts’ are rarely incontestable. It is ‘facts’ that incite people to kill, and dissuade others from intervening. We can all present ‘facts’ as being what has gone before, or as indisputable. Conflict, however, is processual. At any given moment ‘facts’ are being cited, constructed, dismissed and reworked. Today’s fact is tomorrow’s fiction. To stop the clock at any moment is an arbitrary, albeit necessary, decision for any academic, writer, or journalist. We should remain aware, however, that facts are discursive resources, synchronically multiple and diachronically fluid. They are brought to the foreground and/or pushed out of sight. To make them concrete in our writing is to freeze-frame this process. The plasticity of ‘facts’ themselves, and the negotiation of what is and is not accepted as being ‘factual’, is not subsidiary to conflict but is central to it. If we treat ‘facts’ as anything other than contingent, we may be left not in a position of understanding conflict but of asking why, if everything was so clear at the time, there was any need for conflict in the first place. ‘Facts’ rely on hindsight, they are literally post facto. This can be seen in the context of history. We can describe a narrative as ‘history’ and consider it a fact. We can describe another narrative as ‘myth’ (in the general meaning of the term) and consider it fiction. But, this misses the point of instrumental historical narratives. Whether narratives are ‘fact’ or ‘fiction’ is immaterial. It is how people perceive and use them strategically at any given moment in time that is important.
In my own research endeavour, I am aware that two kinds of ‘facts’ are missing. First, in the drive to unpack and compare representations, I have omitted dates, names, and places. Second, I do not explicitly state my own position of what is ‘true’. One eminent academic who has written extensively on Rwanda suggested that while my history chapter includes historical narratives from those I interviewed, should I not also have introduced the chapter with a sketch of Rwandan history? He immediately saw the problem, of course. I cannot write a ‘history’ of Rwanda without pre-empting and/or dismissing aspects of my informants’ narratives. I cannot write a ‘true’ history of Rwanda. Another colleague questioned my dismantling of discourse on logical as well as anthropological grounds, suggesting that I was in danger of losing touch with reality in the process. Hemmed in by the desire to be ‘impartial’, to avoid accusations of revisionism, am I elevating discourse above any concrete reality or facticity? I do not think so. Any utterance is potentially both factual and fictional. It is in the act of interpretation, of trying out that utterance against our own biography, our own experiences and life circumstances, that we decide whether an utterance is true or not. None of this is new to anthropology.
I would like to return to the point mentioned above about how my research omits dates, events, names, and places unless referred to in interviews. If for a moment I imagine that my thesis was published, and a member of my imagined ‘popular’ audience entered a bookshop, why would they buy my book? My thesis is based on an unwritten preface; that is, that I have read and considered the ‘factual’ material as best as I can, and am now moving on to a more nuanced stage of interpretation. What right do I have, however, to take as my starting point the halfway point of a narrative? Should my book have a warning on the cover that states that for any of this to make sense the reader should first read four or five books containing an established chronology, narrative framework, requisite facts and figures?
By omitting this substantive material, am I not as guilty of assuming the unsaid or the unspoken as the ‘popular’ writers I mentioned above? Of course, I could be criticised for failing to contextualise my material correctly. Should each deconstructionist exercise be preceded by explicit reference to the ‘reality’ to which those concepts and notions purport to refer? To fully satisfy the twin calls for deconstruction and ‘facticity’ would require two different theses, however, with different epistemological foundations.
Perhaps all this angst is a reflection of my own position sitting at a computer trying to represent not only a country thousands of miles away, but also an atomised, amorphous discourse. In the process of writing we create a ‘virtual’ imagined landscape of our place or region of interest, a landscape around which we navigate as we try to form our arguments and arrange our material. Occasionally, we try to anchor that landscape in a vaguely remembered conversation, a single personality. By this, I do not mean the sanitised interview transcript or the name on the page, but the emotional reality of an encounter. And yet, writing insists that we create our own universe. The inadequacy of all of this is thrown into sharp relief on returning to the ‘field’ and finding it is not, and never will be, tamed by our writing. Is it not the case, however, that the actors implicated in our research are also navigating around a ‘virtual’ and imagined landscape of their own?
For those who write on Rwanda, any representation runs the risk of being considered partial or ‘misleading’. In the context of history, it is assumed that one can identify a writer’s position from the stance he or she adopts on key issues. For example, the debate about whether ethnicity was the creation of colonialism, or whether it existed prior to colonisation. Likewise, it is not just these alternative positions that are of importance, but the degree of context a writer considers appropriate. As context is potentially infinite, there can be no definitive contextual boundary. In such a minefield, desired impartiality is almost impossible to achieve. However creative (and careful) a writer attempts to be, there will always be room for accusations of bias towards one ‘side’ or the other. This has important implications for the question of ethics. As anthropologists we are rightly concerned with the ethical dimension of our work, particularly as it relates to the ‘intellectual property rights’ of those with whom we carry out research. Much is made of taking findings back to the field. In the context of my research, however, those who I interviewed were not giving me privileged access to a private and encoded world, my representations of which should rightfully be made available for their assessment and comment. Very few of the Rwandese who I interviewed had experienced the genocide at first hand. Their discourse was concerned less with personal testimonies about the genocide, than with ‘objective’ analysis of the causes of the genocide and the response of the ‘international community’.
Rather than victim testimony, the Rwandese I interviewed articulated a public, analytical discourse, one that they had already used on numerous occasions with journalists, international representatives, and other academics. More importantly, they were aware that I would talk to the ‘other side’. As the objective of the research was to investigate both the nature and the substance of this discourse, neither ‘side’ could be privileged. To put it bluntly, in a conflict situation (one marked by persistent discursive forms) the incorporation of these different discourses would automatically make my work suspect to either ‘side’. Neither ‘side’ wants to share the stage. I have consequently found myself in a position where the results of my research could not be usefully assessed and critiqued by those involved.
The fundamental characteristic of these various discourses is that they cannot exist in the same space. Their incorporation in my thesis is, in itself, contrary to their intended purpose. Given that from the outset the placing of these discourses side by side would be contrary to the expectations of those involved, how can I hope to satisfy informants that I have correctly presented their points of view?
The very act of giving voice to other discourses questions the veracity of their own point of view. In a situation in which dialogue between informant and researcher is made problematic from the outset, how does one apply an ethical framework to the question of representation?
I should still, of course, provide my informants with the opportunity to read and assess my findings, but I am aware that I will not receive feedback that would question what I have presented in a particularly constructive manner. The very act of placing these discourses side by side will always make my research problematic, however much I try to highlight the generic nature and power of discourse itself.
To return to the questions raised at the beginning of this paper. What is the integral value of a deconstructionist approach? If multiple perspectives are unacceptable to the actors involved, why should one expect a general readership to accept such a position? Ultimately, of course, there would be no conflict if there was only one diachronic process of cause and effect. Conflict, by nature, is about competing perspectives. While, as individuals, we make moral judgements about those involved in conflict, anthropologists cannot shy away from listening to and incorporating different perspectives. Of course, anthropologists must always avoid unnecessarily rationalising and relativising violence and conflict. And yet our contribution to understanding conflict places us in a difficult position. It puts us at odds with our informants, and it puts us at odds with a general readership who want to know why a conflict has occurred, rather than being asked to choose from a menu of disparate causes.
In conclusion, books such as those by Gourevitch and Melvern are important because they respond to, and are written for, a readership that demands the just-give-us-the-facts approach. In a world in which drawing attention to unsavoury events like genocide is difficult at the best of times, we should welcome books that find a responsive readership. Anthropology, however, has the unenviable task of providing alternative approaches to the understanding of conflict. Hopefully, such approaches will be recognised as being complementary to these other works, rather than being dismissed as obscurantist. Conflict is ultimately about disagreement. To properly understand conflict we must give voice to these disagreements and demonstrate how they are articulated. It is to be hoped that such an approach helps to highlight the processual nature of conflict itself. From such a perspective, disagreements about the nature and ‘truth’ of conflict have less to do with the sanitised, objective, and inevitable progression of ‘facts’, than they do with an informed engagement with the confused and confusing world of discursive strategies, partial ‘truths’, and conflicting subjectivities.
Read Linda Melvern's response, to this article (May 2002) here.
Nigel Eltringham is a research student in the Department of Anthropology and Sociology, School of Oriental and African Studies (SOAS), London. His doctoral research focuses on post-genocide discourses about Rwanda in the international arena. He has previously worked for a conflict resolution and development NGO in Rwanda and in rural development in Latin America. Aspects of his research have been published in R. Doom and J. Gorus (eds.), Politics of Identity and Economics of Conflict in the Great Lakes Region (2000), and D. Goyvaerts, Central Africa Revisited (2000).