CONVENTIONAL WISDOM AND RWANDA'S GENOCIDE: An Opinion.©
Certain thoughts about the Rwandan genocide of 1994 are commonly accepted. For the most part, these ideas have been used to explain the causes of the genocide and, by implication, propose solutions to Rwanda's continuing problems. The status of these ideas within the media and policy circles, however, is problematic. Despite the tentative nature of the propositions when they were first asserted only two or three years ago, they have come to represent what may be called the "conventional wisdom" about Rwanda. Rarely are the assumptions behind such ideas challenged.
The generation of such "conventional wisdom" is not unusual; every social situation requires explanation that becomes part of common accepted knowledge. The ultimate measure of such common knowledge is its utility in predicting the likely actions of participants in the situations described. In the Rwandan situation, however, this common knowledge has not always been a good guide for such predictions. Only rarely have particular policy prescriptions led to the desired outcomes. In particular, Western-generated humanitarian policy, focused on democratic political institutions, respect for human rights, principles of voluntary refugee repatriation, and open markets, has been repeatedly frustrated.
In large part, I think that such systematic misreading of the Rwandan refugee situation is due to the very nature of how information is gathered in emergency situations. Necessarily, the emergency workers on Rwanda's borders in 1994 quickly developed common knowledge about the crisis, the actors involved, and the solutions to the situation. This knowledge helped to rationalize their own interventions and guide their efforts. Experientially based perceptions obtained in the days of the four month-long genocide thus provided the foundation for distinct views about Rwanda's problems (1).
Such experientially based knowledge is not inherently inaccurate or bad. The problem is that such knowledge, when unanalyzed, presents a fragmented, superficial, or incomplete picture of the emergency situation itself. This is particularly true when the individuals writing situation reports ("sitreps" in the sub-cultural argot) about the overall political situation are not part of the societies being analyzed. For the emergency specialist, this is almost always the case. Unfortunately, what happens in emergencies like the Rwanda genocide is that such views are uncritically passed on by a headquarters where, because of the urgency of the situation, sitreps are translated unanalyzed into emotional donor appeals and ReliefWeb documents. In turn, fleeting impressionistic views or opinions become the conventional wisdom shaping definitions of problems, accumulation of knowledge, and interpretation of "facts", and ultimately policy prescriptions (2).
Complete faith in conventional wisdom is risky, however. To understand any social situation it is necessary to move beyond the limitations of experientially based conventional wisdom (3). Of course, doing this in an emergency is a source of discomfort, as it involves abandoning the few certainties already legitimated both in a remote headquarters and the field, and acknowledging the risky contingent nature of emergency management. Nonetheless, both analysis of and prescriptions for social problems stand to benefit from exploring and unearthing the potential oversimplifications located within conventional wisdom.
Below are seven items of "conventional wisdom" from the Rwanda refugee crisis which, in my view, misrepresent a complex reality and have served as poor guides for policy interventions. In particular, these assumptions are inconsistent with broader understandings of social behavior. The issues will be familiar to anyone who has followed the development of the Rwandan crisis as it was discussed on policy levels, on ReliefWeb, and in the popular press. And while the examples have to do with the Rwandan relief operation, my own bit of conventional wisdom is to point out that the problems illustrated here are probably inherent in the very nature of emergency relief programs, and are not unique to the Rwanda program. Certainly policy errors based on such conventional wisdom were a major focus in William Shawcross' book, Quality of Mercy, about the Cambodian crisis of 1979-83. My suspicion is that field personnel and policy makers associated with the ex-Yugoslavian, Somali, Liberian, and other emergency operations will recognize the pattern of quick conclusions based on fleeting experience, with no subsequent sociological analysis.
THE RWANDA CRISIS AND CONVENTIONAL WISDOM: Seven Assumptions
I. Assumption: The key to the resolution
of Rwanda's political problems is to be found in war crimes trials since
justice is a necessary and reasonable prerequisite for reconciliation.
Related to this is the assertion that war crimes trials are necessary
in order to break the cycle of Hutu-Tutsi violence. This is apparent because
the absence of any punitive response to the 1960's genocide in Rwanda
was a precursor to the 1994 genocide (4).
As relevant as the post-World War Two trials may be, there are also instances where rebuilding following mass murder and genocide-like crimes have occurred without trials. The Ugandan Ambassador to the United Nations in 1996 pointed out to the African Studies Association that Uganda, Zimbabwe (6), and post-Mau Mau Kenya are examples of countries where war crimes were committed and went unpunished, but nevertheless the "cycle of retribution" stopped. The Ambassador also made the point that South Africa is dealing with extremely sensitive reconciliation issues without appeals for mass arrests, tribunals, or support for a justice system which cannot possibly try all of the accused in a fair or just manner. Indeed, the legitimation of mass arrests in Rwanda on the basis of the genocide can easily be seen as an attempt by an authoritarian minority government to maintain control through terror or arbitrary arrest. Certainly, it is seen by the Hutu masses as having this effect (7).
Violence is not cyclical, except perhaps in the short-run, and often only in the context of politically-inspired legitimation processes. A quick look at the 20th century verifies this; Germany, France, and the United Kingdom are now allies, despite two world wars. No cycle of violence emerged following the dismantling of Soviet gulags. More recently, peace is emerging in places like Lebanon, El Salvador, Vietnam, and Cambodia. Indeed, there are examples of alliances and fights between ethnic groups, nations, and tribes which have formed and re-formed in different situations. There is nothing inherently "cyclical" about these processes, however. Neither is there a cycle of violence in Rwanda.
2) Assumption: The genocide was the
consequence of poorly funded development programs: had money been invested
in the region in the past, the crisis could have been avoided. Likewise,
had money invested in the camps around Rwanda instead been spent in Rwanda,
Rwanda itself would be in better shape.
Thankfully, modern genocide is still not a common enough event to draw any generalizable assumptions about causation; the only other definitive cases are Nazi Germany and Ottoman Turkey in Armenia (10). Neither of these countries suffered from the same type of poverty as there was in Rwanda. Reaching such a conclusion would even be difficult if cases of government-sanctioned mass murder (as opposed to genocide) were included in the mix, e.g. Pol Pot's Cambodia, Stalin's gulags, East Timor, and the 19th century American West. While the body counts may have been more or less than Rwanda, the proximate political, social, and economic situations were different. A similar consequence (mass murder or genocide) does not necessarily imply the same cause (11).
3) Assumption: The
Interhamwe (or armed elements) and refugees must be separated so that
the refugees can voluntarily return to Rwanda (12).
No one person is either Interhamwe or not.
In sociological terms, "Interhamwe" is not a discrete category.
Rather there is an alienated Hutu population sympathetic to the young
men, whatever they may call themselves, who mobilize and protect them
in crisis situations. This population is sympathetic to these young men
even though many are aware that some have committed
genocide, and at times used brutal methods to control the Hutu population
Such status ambiguity also is why it is
rarely possible to separate "legitimate refugees" from military
in virtually every refugee situation. Refugee situations are by definition
focused by both politics and crisis: running for one's life is by definition
a crisis, and the very definition of "refugee" implies political
alienation from one's home country. It is little wonder that the many
gray areas surrounding claims to legitimate refugee status occupy a good
portion of the UNHCR's legal and diplomatic staff around the world. Thus,
from a sociological viewpoint, sympathy with paramilitary movements in
refugee situations, while undesirable, are normal.
Again, who is and is not a murderer is a legal interpretation. Genocide is an organized crime by a government against an ethnic group. But governments are not tried or put in prison; individuals are. One can speculate about how many individual "murderers" there are -- most guesses are based on the death toll and the means of execution (gangs of machete wielders). Estimates based on such logistics range from some tens of thousands to three million "murderers." These guesses while often logically sound, do not represent specific guilty persons in the legal sense (15).
Thus, while the 100-120,000 people (mostly Hutu men) in jail represent a reasonable count given the scope of the genocide, without trial to legitimate their individual guilt as being greater or different from those who were not arrested, the prisoners do not represent any sort of individual responsibility, but rather a collective one. Given the lack of clarity about who the collective is (is it all Hutu, people who fled to Zaire, civil servants of the former government, etc.) these 100-120,000 represent simply the power of arrest by the government, not a tool for identifying and punishing the guilty.
The problem of the basic legitimacy of the current RPF-led government compounds the issue further. The international community is in general agreement that the authoritarian RPF is better than anarchy and that the genocidal MRND has no role in Rwanda's future. This, however, is different from deeper issues of national legitimacy and nation-building. Tragically, the current government seems caught in a vicious cycle. It is perceived by the Hutu masses as an occupying force maintaining power through the use of arrest and intimidation. The jails, filled with people who are the sons, brothers, cousins, nephews, or fathers of most Rwandan Hutu, are a persistent reminder of this power. But, from the government's perspective, without the arrests and consequent intimidation, the Hutu masses may revolt against the minority government. Indeed the inability of the government to control the killing of genocide survivors seems to indicate that this result is occurring.
From the perspective of many Hutu, the arrests are only the most current and internationally obvious tool of Rwanda's authoritarian state. Many refugees point out that large numbers of Hutu died in the war, and continue to die without recourse to any system of justice. The lack of a visible response to the massacre by the military at Kibeho (4,000+ Hutu dead in April 1994), and the steady flow of execution victims in the Kagera River in 1994-5 are just two of the foci which cause Hutu to doubt the sincerity of the government (16). Alain Destexhe has made the legal point that killing by the RPF is fundamentally different than genocide (he describes these killings by the legalistic term "exactions"). However, this legalistic distinction in international law is irrelevant to the 100-120,000 people in Rwanda's jails awaiting trial by weak Rwandan national courts which are part of the same political apparatus fighting Hutu militia groups in the country-side. I suspect that this legal distinction is also not clear to the Rwandan farmers, be they Hutu or Tutsi. As such, it does little to assist, and may exacerbate, the basic legitimacy problems of the RPF government.
5) Assumption: Repatriation of refugees
to countries of origin is the only viable political solution.
Other refugee crises in the region (including the Rwandan crisis of the 1960's, Burundian of 1970s, and to a lesser extent the Mozambican crises of the 1980s) were resolved through combined programs of voluntary resettlement, along with local and third country resettlement. The Indo-chinese refugee situations of the 1980s were solved through a combination of second and third country resettlement, voluntary repatriation, and in the end some forced repatriation. Premature repatriation to Afghanistan resulted in the emptying and refilling of camps in Pakistan. Post World War Two refugee policies in Europe also reflected a combination of second country resettlement, third country resettlement, voluntary repatriation, and forced repatriation.
these precedents, it is a good assumption that when the dust of Central
Africa's wars has settled, Rwandans and Burundians will have been scattered
across eastern and central Africa
with the assistance of the international community. It is only a question
of how much it will cost, and how many will die in the process.
Disarmament would have involved enforcing a perimeter around the refugee concentrations, isolating all males, and then doing a systematic hut to hut search. Weapons would undoubtedly have been found. Given that the primary weapon, the machete, is also an agricultural tool, most weapons would necessarily be returned. Assuming that this was feasible, it would have done nothing towards enhancing the legitimacy of the RPF government in Kigali among the refugee populations. In fact it would have done just the opposite by intensifying refugee resentment toward the government and international community.
7) Assumption. There are no more Hutu and Tutsi in Rwanda, only Rwandans.
This is the official policy of the current
Rwandan government. It is of course an appealing policy to pursue. However,
stating that this is so does not necessarily make it so. Institutionalized
discrimination based on any number of status markers (race, ethnic group,
language skills, accent, economic status, etc.) are persistent all over
the world, despite laws to the contrary. Indeed, virtually every country
in the world does have a minority which study after study shows is discriminated
CAN ANALYTICAL DEPTH BE BROUGHT TO POLICY ANALYSIS IN COMPLEX EMERGENCIES?
I think that part of the problem
with the "conventional wisdom" described above is rooted in
the nature of emergency management. Three factors of emergency management
make the type of over-generalizations described here more likely in emergencies
than in other endeavors.
The second problem is the emotional appeal which donors necessarily make to meet the high set-up costs in emergencies. In an emergency, crisis management in the initial phase leads to mistakes which in turn lead to a crisis in a second phase. Oftentimes these errors are magnified when the press focuses on quick and easy solutions. These solutions are seemingly obvious during the intense early days, but more problematic when the complexities of the situation begin to be revealed. In such situations, because little analysis has been done, the general conclusion is that more of the same should be tried. For example, the faith in military peacekeepers as a means of separating refugees and Interhamwe probably emerged in this fashion. The first mistake was violating international conventions by permitting refugee camps on borders. This policy emerged because it is well-known that politicized refugee groups often use border proximity to mount cross-border raids. In Goma, this "mistake" was made for good reason: the size and health status of the refugees required it in mid-1994. Nevertheless, despite the initial justification, rebel forces mounted cross-border raids into the camps from Zaire, destabilizing relations between Rwanda and Zaire. The obvious "big picture" solution to this problem would have been to move the camps away from the border once the health situation stabilized, and press for this solution in diplomatic channels. However, by that time, the idea that the problem was Interhamwe, not camp location had become dogma, and the call for "separation" of refugees and Interhamwe had become more insistent (20). The Western press beat this drum particularly loudly.
The third issue is the very complexity of response required to deal with caseloads involving two million people and six countries, as was the case in the Great Lakes crisis. The combination seems to mean that a premium is placed on large volumes of quick information which is rarely analyzed before attention is refocused by the next series of incidents. The quick information is useful; as my citations show, I am an avid reader of ReliefWeb. However, it is too easy to use this over-abundance of information, much of which is focused on hearsay from the many interested parties, and thus avoid identifying the longer-term trends of which any particular day's events may or may not be a part (21).
Misconceptions and poor social analysis have characterized many of the interpretations of the Great Lakes crisis. More precisely, "conventional wisdom" seems to be a hazard of the manner in which complex emergencies are handled. In the case of Rwanda, this was due to a commendable hope that a quick solution could be found to what is one of the most horrifying events in modern history. Were it that easy, answers would have been found long ago. Continuing to insist on the viability of such conventional thought is a poor substitute for well-informed commitment to long-term social and political analysis.
(1) In a normal bureaucratic setting, such conventional wisdom is often subject to checks by auditors, analysts, fixed contractual relationships, procurement procedures, and other forms of administrative oversight. In the inherently quick changing nature of the Rwanda emergency, there were few such administrative checks. This is both because there were perceived to be few precedents for the situation, and also because emergencies themselves are fast-changing. As a result, claims that analogy and comparison were not relevant became easily validated. In turn, conventional wisdom rooted in fleeting impressions fixed by an unusual time and place became the basis for decision-making far longer than it would normally. Particularly for expatriates working in the emergency, the "boiler room" society of high staff turnover, public relations officers, press attention, high adrenaline, and reliance on verbal transmission of culture (e.g. verbal handover reports) meant that experientially based conventional wisdom was fixed without reference to the broader social context. This is why it is so easy to speak of an "agency culture" for the UNHCR, CARE, Red Cross, and other emergency-focused agencies. See also Hugo Slim (1995) The Continuing Metamorphosis of the Humanitarian Practitioner; Some New Colors for an Endangered Species. Disasters. June 1995.
(2) See William Shawcross 1984. The Quality of Mercy: Cambodia, Holocaust, and Modern Conscience. New York: Simon and Schuster, p. 13.
3) Ruane, Janet M. and Karen A. Cerullo 1997. Second Thoughts: Seeing Conventional Wisdom through the Sociological Eye. Thousand Oaks: Pine Forge Press.
(4) See IRIN Special Feature 1/97, 19 February, 1997, http://18.104.22.168/POLICY/Pb050.htm. This is also a theme in the Medecins sans Frontieres publication Populations in Danger, 1995. Also, The International Response to Conflict and Genocide: Lessons from the Rwanda Experience, v. 2, Executive Summary, March 1996, http://22.214.171.124/Policy/Pb025.htm. US News 28/7/97 p. 17 indicates that this is a strongly held view of US Secretary of State Madeline Albright.
(5) Destexhe, Alain 1995. Rwanda and Genocide. New York: New York University Press. p. 64.
6) Ian Smith who led the often brutal fight for white minority rule in Zimbabwe continues to farm in Zimbabwe. He continues to write unrepentant books about the subject. See The Economist review of The Great Betrayal by Ian Smith, April 19, 1997.
(7) Amnesty International has implied that this is what is happening in Rwanda (Amnesty International, April 8, 1997, Grave Doubts about Fairness of the First Trials ReliefWeb http://www.reliefweb.int. Gerard Prunier "Rwanda: The Social, Political, and Economic Situation in June 1997, (Writenet UK), http://www.unhcr.ch/refworld/country/writenet/wrirwa07.htm. Also Medecins sans Frontieres, op cit., Human Rights Watch Africa, Newsletter, March 1997.
(8) Foreign Direct Investment, Trade, and
Aid: An Alternative to Migration, United Nations Conference International
Organization on Trade and Development for Migration, Journal of Humanitarian
(10) See Alain Destexhe op cit.
(11) See Mark Cooney 1997, From Warre to Tyranny, Lethal Conflict and the State, American Sociological Review v.62, n.2, p.333 for a comment on the exceptionalism of the Rwandan genocide relative to other examples of state-sponsored mass murder.
(12) Sadako Ogata, United Nations High Commissioner for Refugees. See for example. Statement to 1997 UNHCR Executive Committee Meeting, October 14, 1997. http://wwwnotes.reliefweb.int:81
(13) See Reyntjens, Filip 1995. Subjects of Concern: Rwanda, October 1994, Issue 13(2):39 and Gerard Prunier "Rwanda: Update to end of November 1994, Writenet (UK), http://www.unhcr.ch/refworld/country/writenet/wrirwa02.htm
(14) See Philip Gourevitch, "The Return," New Yorker January 20, 1997. Also Zarembo, Judgment Day in Rwanda, 92, 312 genocide suspects await trial, Harper's Magazine, April 1997, pp. 68-80.
(15) Alain Destexhe op cit. is careful in making this distinction.
(16) The Lutheran World Federation/Tanganyika Christian Refugee Service reported removing 917 bodies from the Kagera River in Tanzania between June 1994 and May 1995. The majority of the victims were removed in the early months, and were most likely killed in the genocide. However, 346 bodies were removed between November 1994 and April 1995. In the last month of the project (1995), 7 bodies were removed. Many of the victims removed during this period had their hands tied behind their back, and were killed by a gunshot wound to the head. See Feature Report and Final Report, Kagera River Body Removal Project, Tanganyika Christian Refugee Service (Dar Es Salaam and Geneva), June 1995, by Tony Waters. A memorial at the mass grave in Tanzania was dedicated by LWF/TCRS in May 1996.
17) Medecins sans Frontieres op cit.
(18) See Brain, James 1973. Tutsi and Ha, a Study in Integration, Journal of Asian and African Studies 8(39), Waters, Tony (1995) The Social Construction of Tutsi in Modern East Africa, Journal of Modern African Studies 33(2):343-48, Catherine Newbury, The Cohesion of Oppression, New York: Columbia University Press, Rene Lemarchand (1994) Burundi: Ethnocide as Discourse Cambridge: Cambridge University Press. For an extended discussion of why ethnic identity is persistent in some situations, and not others, see Tony Waters (1995) Toward a Theory of Ethnic Enclave Formation: the Case of Ethnic Germans in North American and Russia. International Migration Review. 29(2):515-544.
(19) See Philip Gourevitch, "Letter from Rwanda" The New Yorker, January 20, 1997, p. 49.
(20) Successive water crises in Benaco camp, Ngara Tanzanian 1994-6 were also of this nature. There was a rushed water drilling program in the first two months of the camp. This led to mistakes, and second crisis which led to essentially the same mistake, i.e. water drilling without careful assessment of geological conditions. This led to further crises, which meant that a river-based pumping station was funded before approvals were received from local authorities. This led to moving to another site which then needed a new road. In the end, the trucking station was not used when the planned for water truck fleet was not funded for vehicles or diesel, even though $1,000,000 had by then been spent on various quick solutions. The refugees themselves apparently developed their own means to deal with the water shortages, and gastro-enteric disease rates remained low as a consequence. I am sure that NGO and UN Staff who work in other sectors of the emergency in Ngara will recognize this pattern.
(21) Miskell and Norton Journal of Humanitarian Affairs (http://www jha.sps.cam.ac.uk/a/a014.htm reposted on 4 July, 1997) have commented on the nature of this problem with respect to emergency contingency planning.