Genocide and the canon of historical tragedy

Stepping from 14th Street in Washington, DC into the U.S. Holocaust Memorial Museum, the visitor to the main exhibition is immediately placed in the shoes of the soldiers of the U.S. Army as they liberated the concentration camp at Dachau in 1944. The personal recollections of General Eisenhower as he visited the camp provide the opening tableau that allows the visitor to make the imaginative leap from U.S. citizen—until this moment unable to comprehend such horrors—to personal witness to genocide.

This moment of liberation has taken on meaning well beyond its immediate historical context. In popular, scholarly, and activist imaginations, international military intervention is not simply historical fact for the Holocaust, but it has taken on prescriptive value in relation to all cases of genocide. International military intervention is how genocides ought to end.

In beginning with such prescriptive assumptions, analysis of genocides past conventionally fall into a pattern of simplified representation of the complexities and uncertainties that existed at the time, in favor of a certain knowledge that had international military intervention occurred (or occurred earlier, or with more conviction), that genocide would not have resulted. Events are portrayed as inexorable tragedy, in which everything unfolds upon a familiar path. The early-warning signs of ethnic discrimination and dehumanization are catalogued (and years later, in other countries, monitored). Victims are stripped of all agency as perpetrators mark them for death. At each stage in the unfolding story we see how the final outcome could have been prevented, had the villains flinched or the heroes been more decisive or lucky. The chorus, moralizing tragic events for an audience of spectators, decries the absence of a deus ex machina: international military-humanitarian intervention. Throughout, we know what happens in Act Five: the horror that will unfold in the denouement.

Prevention of genocide is typically constructed using the same assumptions of predictability, but with an idealized ending. This time—a theoretical future, this time—the international (i.e. western) troops do arrive, the people are saved, and the villain is chased from the scene. This is the activists’ clamor for Darfur. While underplaying the role of the international system in generating the forces that create genocide, this narrative posits salvation in the actions of the international community.

A true appreciation of tragedy means looking at the events without flinching, bearing witness to the real choices and challenges facing all the actors, perpetrators, victims, bystanders, and those with mixed roles or who changed sides. This perspective alerts us that there are other possible endings, not circumscribed by the dominant narrative of genocide and its ending. The killers may get tired of their bloody work and call a halt, deciding “we have killed enough.” Or like thieves the killers may fall out among themselves. Alternatively, the victims themselves may resist and survive, or unexpected events may intrude and alter the course of events.

To those involved at the time, matters are typically confused and complex. After the fact, it is difficult for us to make the imaginative leap to a state of uncertainty that the narrative might not have unfolded as it did. Even in a case such as Darfur since 2003, where mass killing has occurred under the world’s scrutiny, the most resonant narratives are those that stress the inevitability that Khartoum’s alleged genocidal project will be completed unless the world acts. Now, some three years later, it is time to admit that the urgency to help does not reduce the necessity to fully consider what actions might sustain improvements in the situation.

To understand genocide and how to prevent it, we must escape three traps. We must not be seduced by teleologies that allow for only two endings: a completed genocide and a foreign military intervention. While it is important to acknowledge responsibility for violence, we must recognize the agency of all the actors, the complexities of their motives, and the difficulty of applying simple and unconditional labels such as “perpetrator” and “victim.” We must also not fall prey to the desire to advocate doing something which might be satisfying to demand when lives are at risk, but which is incapable of actually improving the situation for any duration. This allows us to take a perspective that studies the genesis, escalation and de-escalation of genocidal violence without making any assumptions about the ultimate outcome, and without obscuring political complexities by using simplified labels.

A focus on de-escalating violence

Genocide scholarship, for years the concern of only a handful of academicians, has since the 1990s grown considerably in depth and range.1For example, Steven Jensen (ed.), Genocide: Cases, Comparisons and Contemporary Debates, Copenhagen, Danish Center for Holocaust and Genocide Studies, 2003. See also: A. Jones, Genocide, War Crimes and the West: History and Complicity, London: Zed Books, 2004; Alexander L. Hinton, Genocide: An Anthropological Reader, Macmillan, 2002; Samuel Totten (ed.) Teaching about Genocide: Issues, Approaches and Resources, Greenwich CT, IAP, 2004.[footnote] But it has largely neglected an empirical study of that crucial part of the story: how genocide ends. A closely-related question is how mass killing, short of genocide, is halted before reaching a level that qualifies it as genocide. Commonly, the ending is rapidly covered in a few concluding paragraphs of a study focusing on the origins and unfolding of the crime. This neglect of how genocides end is a remarkable gap in the literature. This forum is a foray into the field.

Genocide and mass killing of civilians typically happen during war, but our concern here is not with how wars are ended—we are narrowing our attention to how mass killing of civilians comes to an end. Genocide scholars have already broadened the corpus of cases considered within the remit of “genocide studies” to include many more than the consensually-accepted instances of the Armenian genocide, the Nazi’s Final Solution and Rwanda in 1994, plus the cases of Cambodia and Bosnia, to consider a range of cases such as Indonesia in 1965-6, Bangladesh, Biafra, and Stalin’s purges, deportations and the starvation of the Ukraine. In the context of ending mass killing, or stepping back from the brink of genocide, we can broaden our remit still further, embracing the fact we can legitimately include cases that stopped short of a point readily identifiable as genocide.

Let us take a preliminary, cursory overview of episodes of mass killing during the 20th century, some of which have been treated as instances of genocide, and others which fell short, in each case asking the question: how was this brought to an end? The list below is incomplete and the summaries are exceptionally brief—intended more as a challenge to the experts in the field than a considered assessment. Six of the cases (Stalin, Biafra, Bangladesh, Guatemala, Uganda and the Nuba Mountains of Sudan) were discussed in a seminar at the U.S. Holocaust Memorial Museum in January 2004, and the summaries draw on the presentations at that meeting.

The Herero genocide, Namibia, 1904. The main reason why the German colonialists’ campaign of starvation and killing ended was that the generals in charge considered they had completed their task. There was also an outcry at home in Germany, leading to the decision to intern the Herero survivors rather than exterminate them entirely.

The Armenian genocide. The killings, deportation and starvation of the Armenians came to an end when the Ottoman Turkish authorities considered their aims had been accomplished. Although there was exposure and condemnation by some of the Allies, this seems to have been too modest and too late to have had a significant impact.

The Soviet Union under Stalin
. Genocidal episodes during Stalin’s rule included the starvation of the Ukraine in 1932-3, deportations and terror. In each case, Stalin used extreme violence instrumentally, as a means to a specific political end. In command of an exceptionally powerful and centralized state, Stalin was able to end the violence at will, while retaining the capability for restarting it at will. These episodes were definitively halted with the death of Stalin. Dissent within the leadership was impossible. There was neither organized domestic resistance nor effective international opposition.

The Nazi Final Solution
. For Hitler, violence was more than a means to an end: it was an end in itself. The Nazis’ violence escalated during the entire period of the Third Reich, without lull, constrained only by the loss of territory in the last two years of World War II. Domestic opposition played no significant role. Military defeat by the Allied armies brought the killing definitively to an end.

Mao’s China. Genocidal actions in revolutionary China include the repression of Tibet, Inner Mongolia, the Uigurs and other minorities, and also waves of terror, notably the Cultural Revolution. These were started and stopped by Mao himself, who retained the capacity to re-launch similar extreme actions at any time.

Indonesia: the massacre of the Communists in 1965-6. The power struggle between President Sukarno and General Suharto involved widespread violence, notably the systematic murder of suspected communists. This was brought to an end when the communists were eliminated as a political force and the power struggle was decisively resolved in favor of Suharto.

The Biafran war 1967-70. Genocidal violence during the Nigerian civil war was a product of the Federal Government’s determined attempt to stop the secession of Biafra. The ruling coalition supported the war, but only some of its members—not including the head of state, General Gowon—supported genocidal actions. When the Federal forces won the war, the violence rapidly de-escalated. Although there was an international outcry against the massacres and the blockade of Biafra, this contributed little to the government’s decisions.

Bangladesh 1970. Attempting to thwart the independence of Bangladesh, Pakistani army generals sought to eliminate the Bangladeshi leadership. However, there was limited support for the massacres among Pakistan’s rulers, and the army was reluctant to pursue the policy of mass killing for long. The pace of killings slowed, and the conflict became a more regular confrontation between the army and the Bengali guerrillas. The killings finally came to an end with intervention of the Indian army. While international public opinion was sympathetic to the Bangladeshi cause, Pakistan remained a valued ally of the west (especially given its pivotal role in the U.S.-China rapprochement) and there was no significant international pressure to halt the killings.

Burundi 1972
. The Tutsi-dominated government of President Micombero undertook the elimination of Hutus who were suspected of rebel sympathies and much of the Hutu educated class as well. The killings lasted three months, after which it appears that the government and army concluded that they had accomplished their task. There was minimal international protest.

The Red Terror in Ethiopia. In the context of an urban insurgency, the military junta headed by Colonel Mengistu Haile Mariam unleashed a wave of terror aimed at the regime’s political opponents. Successive waves of killings targeted new categories of suspects, until Mengistu felt his grip on power was consolidated. International condemnation had no appreciable impact; domestic opposition was silenced or driven to armed rebellion in the hills—where the government attacked them and the civilian populations they lived among with extreme violence. The counter-insurgency campaigns in Eritrea and Tigray in the 1980s, using population control, forced relocation and famine as weapons of war, came close to genocide in both intent and effect.

Cambodia’s Year Zero. The mass killings perpetrated by the Khmer Rouge were, like the political murders of other modern totalitarian regimes, mounted in waves. The government and party machinery could order an end to mass killing just as it could order their (re-)start. There was no domestic opposition to speak of. A decisive end to the genocide was brought about by the Vietnamese invasion and military defeat of the Khmer Rouge. Protected by a peculiar twist of Cold War politics, international moral condemnation of the killings meant little.

The Guatemalan counter-insurgency
. Culminating in the mid-1980s, the counter-insurgency was genocidal in its targeting of ethnic Mayans in the countryside. The killings slowed when the government had achieved its aim of establishing firm military control of the countryside. The guerrillas’ resistance was ineffective: they were unable to protect the people. Constitutional safeguards meant nothing. Abuses were reported by human rights activists and their foreign colleagues, but to little effect. Although Congressional sanctions prevented the U.S. from directly supporting the Guatemalan military activities, the Reagan Administration was firmly behind the Guatemalan government, and did nothing that might have discouraged its vicious counterinsurgency.

The “Luwero Triangle” massacres in Uganda 1983-4. The government of Milton Obote perpetrated genocidal massacres in the course of a vicious counter-insurgency of a type that has become all-too-common in Africa. The extremity of violence was a product of the weakness of the state and its inability to consolidate power, and the massacres were ended when the ruling coalition fell apart. Armed rebellion by the National Resistance Army was the spark for the massacres. The NRA provided limited physical protection but eventually did succeed in bringing down the government and ending the killings. The NRA did not, however, end the cycle of violence in Uganda—instead displacing it to other areas and other groups. The Luwero massacres caused hardly a blip on the radar screen of international concern.

Saddam Hussein’s Anfal against the Kurds. Having miscalculated in attacking Iran, Saddam Hussein used ever-more extreme measures both on the battlefront and in suppressing the Kurdish rebellion. The Anfal was a uniquely comprehensive and brutal counter-insurgency that consolidated Iraqi control over a rebellious region. It was halted when the military objectives were attained, and the regime retained the capability and readiness to enact similar measures again, notably against the Marsh Arabs. Domestic dissent was impossible and the Kurdish guerrillas provided little protection. International condemnation was muted by Iraq’s then-favored status in the west vis-à-vis Iran. After 1991, the combination of sanctions and a no-fly zone enforced by U.S. and British warplanes curbed the prospect of new Iraqi government military offensives against both Kurds and Marsh Arabs, a possibility that was definitively removed by the invasion and overthrow of Saddam Hussein in 2003. However, the legacy of intolerance and readiness to address political problems with extreme violence remains. Iraq is not free from the threat of genocidal massacre.

Extreme clan-targeted violence in Somalia (including the destruction of the northern Somali cities in 1988 and the brutalization and deliberate starvation of Bay Region in 1991-2). The military regime of President Mohamed Siad Barre massacred civilians in its counter-insurgency, a means of warfare that continued after his government was overthrown and the country was riven by factional warfare. Episodes of extreme violence in Somalia have ended when political realignments changed the logic of war, local resistance became too strong, or peace agreements were negotiated. The U.S.-UN military intervention in 1992 briefly interrupted the cycle of violence before becoming party to that violence itself. The potential for continuing extreme violence continues.

Mass killings and relocations in the civil war in Sudan, especially in the South and the Nuba Mountains. The Sudanese civil war from 1983 to the present provides multiple examples of mass killings of civilians that arguably rise to the level of genocide, including the militia massacres in Bahr el Ghazal in 1985-8, the assault on the Nuba Mountains in 1992, the killings in Juba in 1992, the clearances of the oilfields in 1997-2000, and several episodes of internecine conflict between Southern Sudanese factions. Of these, the jihad in the Nuba Mountains was the most sustained and most genocidal, insofar as there was a clear plan for the comprehensive relocation of the Nuba people and the destruction of their identity in pursuit of ideological goals. The Nuba jihad was ended because the ruling elite disagreed on strategy (those who wanted a more limited counter-insurgency won out) and because of resistance by the Nuba people themselves. International pressure played a very minor role. While the genocidal assault was ended, the war continued with ongoing killings and abuses, and the government retained the capacity for other extremely violent campaigns. The pattern of internal disagreement and local resistance has usually brought transitory reductions in violence and ultimately brought a negotiated end to the war in the South and the Nuba Mountains. International engagement in support of the Sudanese peace process was instrumental in achieving that negotiated settlement, which now looks fragile, in the shadow of the ongoing war and killing in Darfur.

Rwanda 1994. The genocide of the Rwandese Tutsis orchestrated by the Hutu Power regime was ended when the rebel Rwandese Patriotic Front scored a decisive military victory. Domestic opposition was eliminated and the regime showed few signs of internal dissension (possibly because its stay in power was so brief). A UN force saved thousands of people but did not stop the killing. The French Operation Turquoise did save some survivors in western Rwanda, but it also protected the genocidaires and allowed them to flee to Zaire in safety.  While the prospects of a new genocide against the Tutsis in Rwanda is remote—at least as long as the current government remains in power—the genocide and its aftermath unleashed violence across the Great Lakes region, especially the Democratic Republic of Congo, that arguably includes episodes of genocidal killing.

Bosnia-Hercegovina. Despite intensive international scrutiny and engagement on Bosnia throughout the conflict—manifest in a large humanitarian aid effort, international peacekeepers, the creation of an international criminal tribunal, and diplomatic efforts—civilians were conspicuously not protected. It was not until the maps were “cleaned up” following the fall of the Bosnian government outposts (“safe havens”) in eastern Bosnia that a negotiated solution was possible. The fact that large-scale massacres, including the killing of some 8,000 men and boys at Srebrenica, attended this development prompted increased NATO support for the Bosnian government. A counter-offensive by the Croat forces in alliance with the finally well-armed Bosnian government, coupled with disagreement between the Bosnian Serbs and their patron in Belgrade, Slobodan Milosevic, pushed the Serbian leadership to make key concessions at the negotiating table. These concessions were also possible because the core gains of ethnic cleansing were protected in the final agreement. The negotiated settlement at Dayton formalized the end to the violence.

Kosovo. In the subsequent case of Kosovo, sustained NATO bombing ultimately caused the Milosevic regime to capitulate, but there are questions as to whether the beginning of the NATO air campaign accelerated the very atrocities it was supposed to prevent.

General themes

These nineteen instances, cursorily presented, allow us to draw out some general themes and unanswered questions.

When do killers stop killing? A repeated unanswered question in these instances is, why did the perpetrators decided to stop the killing? We have little insight into the point at which the commanders decided “we have killed enough,” and almost as little into the internal debates among ruling elites, which may have led to a decision to de-escalate the violence. In short, we know remarkably little about the precise calculations that have helped bring these episodes to an end.

What is an “end” to the killing? In some cases—for example on Stalin’s death, Hitler’s defeat, the Nigerian victory in Biafra, or Bangladeshi independence—we can in retrospect say that the killing was ended for good, although a legacy of bitterness remains. More commonly, what is the ending of mass killing is in fact the ending of an episode in mass killing, with the perpetrators (usually the state) retaining the capacity to restart at any point. Stalin while alive, Mao, Suharto in Indonesia, Mengistu in Ethiopia, the Milosevic regime after Dayton, and the Sudan government are examples. In some instances, one pattern of killing is ended, but a cycle of killing has been set in motion that witnesses comparable levels of massacre, directed against different groups. Uganda, Rwanda and Iraq are examples. A related issue is the existence of peaks and lulls in the violence. Many cases demonstrate an escalation of violence from an already high level to a genocidal peak. The pattern is different in each case: in the USSR, purges and relocations were very discrete events, whose beginnings and endings can be dated to exact dates and decrees. In Sudan, Uganda and Guatemala, a steady background of counter-insurgency violence was the basis for a rapid explosion into genocidal killing.

Ideologies and their roles. In all cases, the motives for genocidal violence were mixed, including control of territory and people and the pursuit of an ideological and/or political goal. Several cases were Communist totalitarian regimes, which used violence in an instrumental manner, in pursuit of power and social transformation, a noted contrast with the “essentialist” violence of Nazi Germany. Leninist-Stalinist organization proved capable of the most extraordinary feats of social engineering, including mass murder. Ethnic essentialism in Rwanda rivaled the industrial powers and bureaucratic states in its efficiency of killing, combining both an instrumentalist use of violence (to create an ethnically homogenous state) and an essentialist use, in which violence was to transform the identity of the perpetrators. Islamist jihad as exhibited in Sudan possesses an essentialist element, but in this case it is as much associated with the mujahid dying as with killing. The weak socio-political theory possessed by Islamism proved (fortunately) a hindrance to successful prosecution of genocide in the Nuba Mountains. State-building and state preservation motivated violence in Indonesia, Biafra and Bangladesh, with political leaders clothing their strategies in nationalist garb. Other cases of genocidal killing were driven by motives in which ideology was less prominent, notably cases in which mass killing was part of a counter-insurgency strategy by a state under threat.

The army, war and counter-insurgency. In every case looked at, an army or militarized security force was the main instrument of mass killing and relocation. An army, as a centralized, hierarchical and professionally violent institution, is the obvious instrument for a state set on genocide. But professional armies are ambiguous instruments. While soldiers are schooled in obedience and the great majority will carry out their orders, they may also demonstrate disgust and weariness after some time. Soldiers are not necessarily professional killers, especially outside combat, and the perpetrators of genocide have regularly rediscovered that they may require special forces from outside the regular army in order to instigate mass killing.

Many of the genocidal killings were perpetrated in the context of counter-insurgency campaigns that involved draconian controls on the civilian population, suspected of sympathizing with rebels. The Herero genocide and the Iraqi Anfal are paradigms of counter-insurgency taken to genocidal extreme, and the Ethiopian military campaigns in Eritrea and Tigray come close (at least). The Armenian genocide had at least as its pretext the fear of rebellion. The combination of an ethnic dimension to the conflict and a weak state, which relies on franchising paramilitaries to conduct much of its killing, creates perfect conditions for genocidal violence. Thus in Uganda, Guatemala, Somalia, Sudan and (arguably) Indonesia, ethnically-targeted violence is instigated as a mechanism for combating insurgency, and is escalated to the point of genocidal massacre. These cases have the special difficulty that once these forces have been unleashed, they are difficult to rein back in.

In secessionist wars the logic can be different. During the Biafran war of independence from Nigeria, the Federal Government’s violence was aimed precisely at keeping the Biafrans within the state, though relatively powerless. The Pakistani violence in Bangladesh had a similar logic. Bosnia and Kosovo combine counter-secessionism and ethnic agendas in a particularly virulent mix. In all these cases, extreme violence was instrumental in pursuit of a state agenda.

What all these cases have in common is that the state concerned suspended ethical rules for the duration. Servants of the state were permitted or encouraged to act with impunity. The logic can be encapsulated in the command, “do what is necessary to ensure total submission, and don’t report back on the details.” In Sudan, the war areas have been described as “ethics-free zones.”[footnote]Alex de Waal, “Starving out the South,” in Martin Daly and Awad Alsikainga, Civil War in the Sudan, London, British Academic Press, 1994.

Famine as a tool. A related question is the use of starvation and destruction of livelihoods as a weapon of both war and genocide. Man-made famine was an instrument for the elimination of groups in Namibia in 1904 and the Ukraine in 1932, but the provision in the 1948 Genocide Convention that specifies “deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part”2Article 2(c). had not been used in a formal determination of genocide or a prosecution for genocide until the case of Darfur was investigated by the U.S. State Department and the Coalition for International Justice in 2004.3Samuel Totten and Eric Markusen (eds.) Genocide in Darfur: Investigating the Atrocities in the Sudan, New York, Routledge, 2006. Given that counter-insurgencies typically involve the forcible displacement and the destruction of their livelihoods, this represents a potentially very significant broadening of the scope of episodes identified as “genocide” by scholars.4The dominant genocide scholars in the U.S. have tended to emphasize closeness to the Holocaust paradigm as a definitional prerequisite for genocide rather than conformity with the letter of the 1948 Genocide Convention. The Darfur genocide determination dramatically reverses these emphases. If this broadening becomes consensually adopted, then this has implications for what it means to end a genocide that has been conducted primarily through these methods. Is humanitarian assistance and socio-economic redevelopment sufficient to represent the end of such a genocide?

Civil society and constitutional rule. When we consider the factors that give rise to a society prone to mass killing and the recovery of a society that has been riven by conflict and genocidal massacre, the role of civil society is pivotal. But during the episodes of genocidal killing themselves, civil society is an irrelevant force. This should not surprise us: wars are almost always accompanied by states of emergency in which civil rights are repressed; those committing genocidal massacre commonly justify it through facing an external enemy (often an effective means of creating social unity); and for the intended victims of a genocide, civil society-type responses are no longer useful.

In most cases, constitutionalism had broken down or was a sham, masking dictatorial rule. But there are intriguing instances in which constitutional safeguards were in place throughout the period of genocidal violence, but signally failed to prevent it. These cases include Indonesia, Biafra and Guatemala, where governments were able to overcome constitutional restraint and disregard dissent. A constitution is only as good as those prepared to enforce it.

There is a potential exception to this rule: if local civil society can in some way generate an international response, using its contacts with foreign groups to launch a moral boomerang that hopefully comes back to strike the perpetrators of genocide.5For a wider analysis of these domestic-international linkages in human rights activism, see: Thomas Risse and Kathryn Sikkink, ‘The socialization of international human rights norms into domestic practices: introduction,’ in Thomas Risse, Stephen Ropp and Kathryn Sikkink (eds.) The Power of Human Rights: International Norms and Domestic Change, Cambridge, Cambridge University Press, 1999. In some of our cases (Biafra, Guatemala, Sudan, and notably, Bosnia) civil society groups managed to feed information to international human rights groups, and that information was sometimes used to good effect, though only very rarely did it spark international action sufficient to stop genocide. The case of Darfur may yet be the exception that proves the rule.

Negotiations. In many cases, final political settlements have been negotiated when there is a military fait accompli. In just a few cases—notably Bosnia and the Nuba Mountains and Southern Sudan—negotiations have helped to end mass killing of civilians. These are both recent cases, reflecting the fact that until the end of the Cold War, the major powers allowed civil wars to continue until there was a military victory by one side or the other. International readiness to bring negotiated solutions to civil wars is a recent phenomenon. In parallel there has been increased concern with genocide and with accountability for human rights violations. There is a rather obvious tension between these two trends. In both the Bosnian and Sudanese instances, although humanitarian concern and the mass killing of civilians was undoubtedly a motivation for the international initiative to end the wars, the negotiations were framed primarily as peace talks to end a civil war rather than the ending of a genocide. Trying to frame negotiations as “ending genocide” has the obvious drawback that the label itself criminalizes one party.

International pressure and intervention

In only three cases in the compendium did foreign military intervention decisively bring the killings to an end. Those cases are the Allied victory over the Third Reich, Indian intervention in Bangladesh and the Vietnamese invasion of Cambodia. All were undertaken more for political self-interest than humanitarian concern. The NATO interventions in Bosnia and Kosovo had more debatable impacts on ending the violence. Only in two—Bosnia and Sudan’s North-South peace agreement—were international actors significantly involved in mediating an end to violence.

The rarity of international pressure or intervention having a significant impact is closely associated with the fact that most cases under consideration occurred during the Cold War, when one superpower or another protected genocidal regimes. The end of the Cold War opened up the possibility of international force being used to this end. In practice, this has yet to happen, but international scrutiny of human rights has at least meant that extreme atrocities rarely unfold in the dark, as was common in earlier decades.

The role of international troops is complex in part because of the ambiguity of the term “humanitarian,” which may refer either to humanitarian law in general (including the prohibition on genocide) or specifically to the provision of relief assistance to stricken populations. For most of the 1990s, it was much easier for western governments to agree on the narrower form of “humanitarian intervention,” and thus troops were dispatched to Iraqi Kurdistan, Somalia, Bosnia and Rwanda with a limited mandate. The presence of UN troops signally failed to stop killings in Bosnia and Rwanda. The centerpiece of Darfur advocacy has been that this failure should not be repeated. In Bosnia, humanitarian activities even served to deter armed intervention, for fear that the UN troops would be endangered and humanitarian operations halted should international forces attack the Serb forces.

International military intervention has become the principal focus for anti-genocide activism. This reflects the shadow of the Holocaust as much as the failures of the subsequent sixty years. The paradigmatic instance of stopping genocidal killing by morally uncomplicated means is still the Allied liberation of Nazi concentration camps. The 1990s NATO bombings of the Serbs, the only other instance that comes close, suffer from too many question marks to meet the standard of an ethically-clear anti-genocide intervention. The Bosnian intervention was too late and the Kosovo air campaign had many complicated side-effects.

The relevance of international military intervention is over-determined in writings on genocide prevention, existing throughout the body of literature as the final tool in the bag or as the culmination of a national or international policy. Almost without fail military intervention is presented as a politically neutral engagement on behalf on innocently suffering civilians. Thus the Brahimi Report advocates for an early warning system tied to international capacity to militarily intervene and the report “The Responsibility to Protect” by the International Commission on Intervention and State Sovereignty (December 2001)6http://www.dfait-maeci.gc.ca/iciss-ciise/pdf/Commission-Report.pdf argues there is a responsibility to intervene (with military intervention as the last resort) on behalf of threatened civilians. Similarly this is the clear direction of work by prominent genocide scholars, such as Samantha Power.7Samantha Power, ‘A Problem from Hell’: America and the Age of Genocide, New York: Basic Books, 2002. Given the preliminary findings of this survey, current emphasis on this tool should be reconsidered.

Preliminary conclusions

This section addresses four questions: how violence was de-escalated, why the international community played so small a role, why the cases do not appear to match the consensus of the genocide literature, and what lessons we can draw for the future.

What de-escalated the violence? In some of the cases, de-escalation occurred when those perpetrating the violence achieved their goals. Clearly, this is a discouraging conclusion, especially when those goals include genocide. But Stalin, however cruel, did not intend to eliminate entire nationalities, even when he had labeled them as “enemy.” The Nigerian and Guatemalan governments were concerned with military control. Eliminationist racism is rare, most obviously present in the Herero killings, the Final Solution and the Rwandese genocide. This is disturbing insofar as goals that fall short of eliminationist racism can also lead to very high levels of violence including genocidal massacre.

A second reason for de-escalation was successful resistance. This in turn depended on the extent to which the targeted group could call upon an armed force able to mount effective military campaigns. Several of these (the Sudanese Nuba, the Eritreans and Tigrayans, several Somali groups, the Ugandan National Resistance Army and the Rwandese Patriotic Front) resisted without external support, while the Bangladeshis had the assistance and ultimately the direct intervention of India.

The third reason was elite dissension. In several cases there is a clear lack of consensus among elites, commonly united on the need for an effective war effort, but not on methods that involve genocidal massacre. The military leadership itself can serve as a restraining force: those instructed to carry out the killing may be more reluctant than their civilian superiors.

Invasion—military intervention primarily for self-interested reasons—was a fourth means whereby genocide was ended. Invading forces brought an end to genocide in Bangladesh and Cambodia as well as ending the Nazis’ Final Solution. The impurity of both motive and method of the invaders should not detract from the reality that these were effective in bringing genocides to an end. Other instances of intervention with a humanitarian rationale, such as the Tanzanian invasion of Uganda to overthrow Idi Amin in 1979, the French Operation Turquoise in Rwanda in 1994, and the U.S.-led overthrow of Saddam Hussein in 2003, may have achieved short-term political and humanitarian objectives, but did not decisively end the violence, and indeed contributed to new cycles of killing.

Why was the international community so marginal? The commonest explanation for international indifference to suffering is lack of knowledge. This was definitely the case for Stalin’s and Mao’s campaigns, but the communications revolution and the growth of human rights advocacy means that secret genocides are more and more difficult. The case of the Nuba Mountains in the early 1990s may be the last case in which a threatened population was almost completely cut off from the rest of the world (though we should be alert to the dangers of North Korea). But, as studies of the Holocaust, Bosnia and Rwanda have repeatedly shown, ignorance was never the principal reason for failure to act. This is glaringly so in the case of Darfur.

The most important reason for international non-engagement is mixed political motives of both the perpetrators and foreign governments. Perpetrators may have goals that stop short of genocide, and some of their other policies may command the backing of powerful international friends. Their internal opponents (likely to be those targeted in the genocide) may be groups that foreign governments are reluctant to support, such as left-wing insurgents. A perpetrator government’s internal debates and external public relations may obscure the genocidal effects of their actions. These factors complicate how the violence is perceived and responded to by outside forces. And, invariably, foreign governments have their own competing interests in play, some of them domestic (keeping out of overseas adventures) and some of them foreign (supporting a friendly regime). The importance of sustaining Sudan’s North-South peace process played a large role in lowering international interest in Darfur at the time of the peak of violence in 2003-04. Stopping mass violence has rarely been a high priority for those foreign governments that have the capacity to intervene. Arguably, one of the greatest successes of the international human rights movement has been to insist that this changes, and indeed, over the years, western governments have become more sensitive on this issue.

But greater sensitivity to human rights and humanitarian concerns has its complications too. Potential genocide may not be the sole or overriding concern of an ambassador with a conscience. In Bosnia, Sudan and Uganda, diplomats and humanitarian professionals alike fell into a “humanitarian trap”: they compromised with abusive authorities in order to maintain a humanitarian presence in the country. Essentially, they traded ongoing humanitarian programs, which were providing assistance to thousands of people, against speaking out against mass killing. It is interesting to note, however, that in Biafra and Bangladesh, humanitarian assistance (by NGOs) was combined with political solidarity for the victims, a tradition that has reasserted itself with the advocacy over Darfur.

Another version of this is the “best of all worlds” trap. This is the failure to advocate a realistic response, because it is not a perfect response. This was significant in Rwanda. When the genocide was unleashed, the international community’s search for simultaneous solutions to the massacres, the war, the humanitarian crisis, the protection of UN personnel, and the restoration of a broad-based government committed to democracy, stood in the way of finding a response to the most severe problem, the genocide. In the event, the Rwandese Patriotic Front halted the genocide, and could have done so earlier had it received international support. But the international community’s justifiable concern that the RPF would not solve problems of democracy, human rights, humanitarian access and armed conflict, meant that it was not supported as an instrument to end the genocide. A similar analysis would fit for Bosnia, where both the government and Serb forces were treated as equal “warring parties” by international negotiators, despite the proclaimed commitment of the Serbian forces to ethnic cleansing. If genocide is indeed an absolute evil, and it is justifiable to suspend other considerations in order to stop it, then support for those groups that are actively resisting genocide, however unattractive they may be in other ways, must be a serious option.

What are the lessons for the future? The final question concerns drawing lessons for policy. This forum is deliberately not an exercise in identifying lists of doables for western governments or the UN. On the contrary, it is an exercise in provoking reflection on how relevant are today’s typical lists of policy recommendations. The final question continues in this spirit with an observation: genocidaires learn. Not only are advocates and policymakers concerned with preventing and punishing genocide carefully seeking how to do their task better, but aspiring criminals-against-humanity are doing so as well.

The prime example of would-be genocidaires learning their trade and sharpening their tools is the three years preceding the April 1994 Rwanda genocide. Over this period, methods of mass killing were tested, alongside methods of silencing and confusing the international community (for example by killing peacekeepers to ensure the withdrawal of the UN force, and sending bishops abroad to plead for the regime).[9. African Rights, Rwanda: Death, Despair and Defiance, London, 1994, chapter 2.]  On that instance, the genocidaires learned faster than their adversaries.

Undoubtedly, it is getting more difficult for aspirant genocidal states. Totalitarian rule is almost gone (North Korea being an obvious exception), and so one particular configuration of political circumstance is much rarer than in past decades. The sophistication of the media and civil society means that it is harder to control information, both for one’s own citizens and the world. Diplomats and journalists are much more attuned to the familiar signs of potential genocide. In the case of war-time genocides, humanitarians are more attuned to the trap of silence in exchange for access—not least because journalists are more ready to criticize them for it. The debate about Darfur has conspicuously avoided this danger.

But there are reasons for caution. First, governments have strategies to neutralize these pressures. These may include professional public relations and protection provided by a major power, which in turn may be associated with strategic interests such as possession of oil reserves or cooperation in counter-terrorist activities or the suppression of the narcotic trade. Second, events can move quickly. A genocide can be launched so quickly as to take the international community off guard (as was the case in Rwanda, if we overlook the warning signs). Or a country can slide from a favorite to a pariah very quickly, as happened with Iraq in 1990 and Zimbabwe in the last few years. Counter-insurgency and counter-secessionist genocides remain an ever-present danger whenever there is a serious military challenge to a state. The basic ingredients for genocide can be found in many different countries: too often, it is just the motive and political organization that is lacking.

References:

1
For example, Steven Jensen (ed.), Genocide: Cases, Comparisons and Contemporary Debates, Copenhagen, Danish Center for Holocaust and Genocide Studies, 2003. See also: A. Jones, Genocide, War Crimes and the West: History and Complicity, London: Zed Books, 2004; Alexander L. Hinton, Genocide: An Anthropological Reader, Macmillan, 2002; Samuel Totten (ed.) Teaching about Genocide: Issues, Approaches and Resources, Greenwich CT, IAP, 2004.[footnote] But it has largely neglected an empirical study of that crucial part of the story: how genocide ends. A closely-related question is how mass killing, short of genocide, is halted before reaching a level that qualifies it as genocide. Commonly, the ending is rapidly covered in a few concluding paragraphs of a study focusing on the origins and unfolding of the crime. This neglect of how genocides end is a remarkable gap in the literature. This forum is a foray into the field.

Genocide and mass killing of civilians typically happen during war, but our concern here is not with how wars are ended—we are narrowing our attention to how mass killing of civilians comes to an end. Genocide scholars have already broadened the corpus of cases considered within the remit of “genocide studies” to include many more than the consensually-accepted instances of the Armenian genocide, the Nazi’s Final Solution and Rwanda in 1994, plus the cases of Cambodia and Bosnia, to consider a range of cases such as Indonesia in 1965-6, Bangladesh, Biafra, and Stalin’s purges, deportations and the starvation of the Ukraine. In the context of ending mass killing, or stepping back from the brink of genocide, we can broaden our remit still further, embracing the fact we can legitimately include cases that stopped short of a point readily identifiable as genocide.

Let us take a preliminary, cursory overview of episodes of mass killing during the 20th century, some of which have been treated as instances of genocide, and others which fell short, in each case asking the question: how was this brought to an end? The list below is incomplete and the summaries are exceptionally brief—intended more as a challenge to the experts in the field than a considered assessment. Six of the cases (Stalin, Biafra, Bangladesh, Guatemala, Uganda and the Nuba Mountains of Sudan) were discussed in a seminar at the U.S. Holocaust Memorial Museum in January 2004, and the summaries draw on the presentations at that meeting.

The Herero genocide, Namibia, 1904. The main reason why the German colonialists’ campaign of starvation and killing ended was that the generals in charge considered they had completed their task. There was also an outcry at home in Germany, leading to the decision to intern the Herero survivors rather than exterminate them entirely.

The Armenian genocide. The killings, deportation and starvation of the Armenians came to an end when the Ottoman Turkish authorities considered their aims had been accomplished. Although there was exposure and condemnation by some of the Allies, this seems to have been too modest and too late to have had a significant impact.

The Soviet Union under Stalin
. Genocidal episodes during Stalin’s rule included the starvation of the Ukraine in 1932-3, deportations and terror. In each case, Stalin used extreme violence instrumentally, as a means to a specific political end. In command of an exceptionally powerful and centralized state, Stalin was able to end the violence at will, while retaining the capability for restarting it at will. These episodes were definitively halted with the death of Stalin. Dissent within the leadership was impossible. There was neither organized domestic resistance nor effective international opposition.

The Nazi Final Solution
. For Hitler, violence was more than a means to an end: it was an end in itself. The Nazis’ violence escalated during the entire period of the Third Reich, without lull, constrained only by the loss of territory in the last two years of World War II. Domestic opposition played no significant role. Military defeat by the Allied armies brought the killing definitively to an end.

Mao’s China. Genocidal actions in revolutionary China include the repression of Tibet, Inner Mongolia, the Uigurs and other minorities, and also waves of terror, notably the Cultural Revolution. These were started and stopped by Mao himself, who retained the capacity to re-launch similar extreme actions at any time.

Indonesia: the massacre of the Communists in 1965-6. The power struggle between President Sukarno and General Suharto involved widespread violence, notably the systematic murder of suspected communists. This was brought to an end when the communists were eliminated as a political force and the power struggle was decisively resolved in favor of Suharto.

The Biafran war 1967-70. Genocidal violence during the Nigerian civil war was a product of the Federal Government’s determined attempt to stop the secession of Biafra. The ruling coalition supported the war, but only some of its members—not including the head of state, General Gowon—supported genocidal actions. When the Federal forces won the war, the violence rapidly de-escalated. Although there was an international outcry against the massacres and the blockade of Biafra, this contributed little to the government’s decisions.

Bangladesh 1970. Attempting to thwart the independence of Bangladesh, Pakistani army generals sought to eliminate the Bangladeshi leadership. However, there was limited support for the massacres among Pakistan’s rulers, and the army was reluctant to pursue the policy of mass killing for long. The pace of killings slowed, and the conflict became a more regular confrontation between the army and the Bengali guerrillas. The killings finally came to an end with intervention of the Indian army. While international public opinion was sympathetic to the Bangladeshi cause, Pakistan remained a valued ally of the west (especially given its pivotal role in the U.S.-China rapprochement) and there was no significant international pressure to halt the killings.

Burundi 1972
. The Tutsi-dominated government of President Micombero undertook the elimination of Hutus who were suspected of rebel sympathies and much of the Hutu educated class as well. The killings lasted three months, after which it appears that the government and army concluded that they had accomplished their task. There was minimal international protest.

The Red Terror in Ethiopia. In the context of an urban insurgency, the military junta headed by Colonel Mengistu Haile Mariam unleashed a wave of terror aimed at the regime’s political opponents. Successive waves of killings targeted new categories of suspects, until Mengistu felt his grip on power was consolidated. International condemnation had no appreciable impact; domestic opposition was silenced or driven to armed rebellion in the hills—where the government attacked them and the civilian populations they lived among with extreme violence. The counter-insurgency campaigns in Eritrea and Tigray in the 1980s, using population control, forced relocation and famine as weapons of war, came close to genocide in both intent and effect.

Cambodia’s Year Zero. The mass killings perpetrated by the Khmer Rouge were, like the political murders of other modern totalitarian regimes, mounted in waves. The government and party machinery could order an end to mass killing just as it could order their (re-)start. There was no domestic opposition to speak of. A decisive end to the genocide was brought about by the Vietnamese invasion and military defeat of the Khmer Rouge. Protected by a peculiar twist of Cold War politics, international moral condemnation of the killings meant little.

The Guatemalan counter-insurgency
. Culminating in the mid-1980s, the counter-insurgency was genocidal in its targeting of ethnic Mayans in the countryside. The killings slowed when the government had achieved its aim of establishing firm military control of the countryside. The guerrillas’ resistance was ineffective: they were unable to protect the people. Constitutional safeguards meant nothing. Abuses were reported by human rights activists and their foreign colleagues, but to little effect. Although Congressional sanctions prevented the U.S. from directly supporting the Guatemalan military activities, the Reagan Administration was firmly behind the Guatemalan government, and did nothing that might have discouraged its vicious counterinsurgency.

The “Luwero Triangle” massacres in Uganda 1983-4. The government of Milton Obote perpetrated genocidal massacres in the course of a vicious counter-insurgency of a type that has become all-too-common in Africa. The extremity of violence was a product of the weakness of the state and its inability to consolidate power, and the massacres were ended when the ruling coalition fell apart. Armed rebellion by the National Resistance Army was the spark for the massacres. The NRA provided limited physical protection but eventually did succeed in bringing down the government and ending the killings. The NRA did not, however, end the cycle of violence in Uganda—instead displacing it to other areas and other groups. The Luwero massacres caused hardly a blip on the radar screen of international concern.

Saddam Hussein’s Anfal against the Kurds. Having miscalculated in attacking Iran, Saddam Hussein used ever-more extreme measures both on the battlefront and in suppressing the Kurdish rebellion. The Anfal was a uniquely comprehensive and brutal counter-insurgency that consolidated Iraqi control over a rebellious region. It was halted when the military objectives were attained, and the regime retained the capability and readiness to enact similar measures again, notably against the Marsh Arabs. Domestic dissent was impossible and the Kurdish guerrillas provided little protection. International condemnation was muted by Iraq’s then-favored status in the west vis-à-vis Iran. After 1991, the combination of sanctions and a no-fly zone enforced by U.S. and British warplanes curbed the prospect of new Iraqi government military offensives against both Kurds and Marsh Arabs, a possibility that was definitively removed by the invasion and overthrow of Saddam Hussein in 2003. However, the legacy of intolerance and readiness to address political problems with extreme violence remains. Iraq is not free from the threat of genocidal massacre.

Extreme clan-targeted violence in Somalia (including the destruction of the northern Somali cities in 1988 and the brutalization and deliberate starvation of Bay Region in 1991-2). The military regime of President Mohamed Siad Barre massacred civilians in its counter-insurgency, a means of warfare that continued after his government was overthrown and the country was riven by factional warfare. Episodes of extreme violence in Somalia have ended when political realignments changed the logic of war, local resistance became too strong, or peace agreements were negotiated. The U.S.-UN military intervention in 1992 briefly interrupted the cycle of violence before becoming party to that violence itself. The potential for continuing extreme violence continues.

Mass killings and relocations in the civil war in Sudan, especially in the South and the Nuba Mountains. The Sudanese civil war from 1983 to the present provides multiple examples of mass killings of civilians that arguably rise to the level of genocide, including the militia massacres in Bahr el Ghazal in 1985-8, the assault on the Nuba Mountains in 1992, the killings in Juba in 1992, the clearances of the oilfields in 1997-2000, and several episodes of internecine conflict between Southern Sudanese factions. Of these, the jihad in the Nuba Mountains was the most sustained and most genocidal, insofar as there was a clear plan for the comprehensive relocation of the Nuba people and the destruction of their identity in pursuit of ideological goals. The Nuba jihad was ended because the ruling elite disagreed on strategy (those who wanted a more limited counter-insurgency won out) and because of resistance by the Nuba people themselves. International pressure played a very minor role. While the genocidal assault was ended, the war continued with ongoing killings and abuses, and the government retained the capacity for other extremely violent campaigns. The pattern of internal disagreement and local resistance has usually brought transitory reductions in violence and ultimately brought a negotiated end to the war in the South and the Nuba Mountains. International engagement in support of the Sudanese peace process was instrumental in achieving that negotiated settlement, which now looks fragile, in the shadow of the ongoing war and killing in Darfur.

Rwanda 1994. The genocide of the Rwandese Tutsis orchestrated by the Hutu Power regime was ended when the rebel Rwandese Patriotic Front scored a decisive military victory. Domestic opposition was eliminated and the regime showed few signs of internal dissension (possibly because its stay in power was so brief). A UN force saved thousands of people but did not stop the killing. The French Operation Turquoise did save some survivors in western Rwanda, but it also protected the genocidaires and allowed them to flee to Zaire in safety.  While the prospects of a new genocide against the Tutsis in Rwanda is remote—at least as long as the current government remains in power—the genocide and its aftermath unleashed violence across the Great Lakes region, especially the Democratic Republic of Congo, that arguably includes episodes of genocidal killing.

Bosnia-Hercegovina. Despite intensive international scrutiny and engagement on Bosnia throughout the conflict—manifest in a large humanitarian aid effort, international peacekeepers, the creation of an international criminal tribunal, and diplomatic efforts—civilians were conspicuously not protected. It was not until the maps were “cleaned up” following the fall of the Bosnian government outposts (“safe havens”) in eastern Bosnia that a negotiated solution was possible. The fact that large-scale massacres, including the killing of some 8,000 men and boys at Srebrenica, attended this development prompted increased NATO support for the Bosnian government. A counter-offensive by the Croat forces in alliance with the finally well-armed Bosnian government, coupled with disagreement between the Bosnian Serbs and their patron in Belgrade, Slobodan Milosevic, pushed the Serbian leadership to make key concessions at the negotiating table. These concessions were also possible because the core gains of ethnic cleansing were protected in the final agreement. The negotiated settlement at Dayton formalized the end to the violence.

Kosovo. In the subsequent case of Kosovo, sustained NATO bombing ultimately caused the Milosevic regime to capitulate, but there are questions as to whether the beginning of the NATO air campaign accelerated the very atrocities it was supposed to prevent.

General themes

These nineteen instances, cursorily presented, allow us to draw out some general themes and unanswered questions.

When do killers stop killing? A repeated unanswered question in these instances is, why did the perpetrators decided to stop the killing? We have little insight into the point at which the commanders decided “we have killed enough,” and almost as little into the internal debates among ruling elites, which may have led to a decision to de-escalate the violence. In short, we know remarkably little about the precise calculations that have helped bring these episodes to an end.

What is an “end” to the killing? In some cases—for example on Stalin’s death, Hitler’s defeat, the Nigerian victory in Biafra, or Bangladeshi independence—we can in retrospect say that the killing was ended for good, although a legacy of bitterness remains. More commonly, what is the ending of mass killing is in fact the ending of an episode in mass killing, with the perpetrators (usually the state) retaining the capacity to restart at any point. Stalin while alive, Mao, Suharto in Indonesia, Mengistu in Ethiopia, the Milosevic regime after Dayton, and the Sudan government are examples. In some instances, one pattern of killing is ended, but a cycle of killing has been set in motion that witnesses comparable levels of massacre, directed against different groups. Uganda, Rwanda and Iraq are examples. A related issue is the existence of peaks and lulls in the violence. Many cases demonstrate an escalation of violence from an already high level to a genocidal peak. The pattern is different in each case: in the USSR, purges and relocations were very discrete events, whose beginnings and endings can be dated to exact dates and decrees. In Sudan, Uganda and Guatemala, a steady background of counter-insurgency violence was the basis for a rapid explosion into genocidal killing.

Ideologies and their roles. In all cases, the motives for genocidal violence were mixed, including control of territory and people and the pursuit of an ideological and/or political goal. Several cases were Communist totalitarian regimes, which used violence in an instrumental manner, in pursuit of power and social transformation, a noted contrast with the “essentialist” violence of Nazi Germany. Leninist-Stalinist organization proved capable of the most extraordinary feats of social engineering, including mass murder. Ethnic essentialism in Rwanda rivaled the industrial powers and bureaucratic states in its efficiency of killing, combining both an instrumentalist use of violence (to create an ethnically homogenous state) and an essentialist use, in which violence was to transform the identity of the perpetrators. Islamist jihad as exhibited in Sudan possesses an essentialist element, but in this case it is as much associated with the mujahid dying as with killing. The weak socio-political theory possessed by Islamism proved (fortunately) a hindrance to successful prosecution of genocide in the Nuba Mountains. State-building and state preservation motivated violence in Indonesia, Biafra and Bangladesh, with political leaders clothing their strategies in nationalist garb. Other cases of genocidal killing were driven by motives in which ideology was less prominent, notably cases in which mass killing was part of a counter-insurgency strategy by a state under threat.

The army, war and counter-insurgency. In every case looked at, an army or militarized security force was the main instrument of mass killing and relocation. An army, as a centralized, hierarchical and professionally violent institution, is the obvious instrument for a state set on genocide. But professional armies are ambiguous instruments. While soldiers are schooled in obedience and the great majority will carry out their orders, they may also demonstrate disgust and weariness after some time. Soldiers are not necessarily professional killers, especially outside combat, and the perpetrators of genocide have regularly rediscovered that they may require special forces from outside the regular army in order to instigate mass killing.

Many of the genocidal killings were perpetrated in the context of counter-insurgency campaigns that involved draconian controls on the civilian population, suspected of sympathizing with rebels. The Herero genocide and the Iraqi Anfal are paradigms of counter-insurgency taken to genocidal extreme, and the Ethiopian military campaigns in Eritrea and Tigray come close (at least). The Armenian genocide had at least as its pretext the fear of rebellion. The combination of an ethnic dimension to the conflict and a weak state, which relies on franchising paramilitaries to conduct much of its killing, creates perfect conditions for genocidal violence. Thus in Uganda, Guatemala, Somalia, Sudan and (arguably) Indonesia, ethnically-targeted violence is instigated as a mechanism for combating insurgency, and is escalated to the point of genocidal massacre. These cases have the special difficulty that once these forces have been unleashed, they are difficult to rein back in.

In secessionist wars the logic can be different. During the Biafran war of independence from Nigeria, the Federal Government’s violence was aimed precisely at keeping the Biafrans within the state, though relatively powerless. The Pakistani violence in Bangladesh had a similar logic. Bosnia and Kosovo combine counter-secessionism and ethnic agendas in a particularly virulent mix. In all these cases, extreme violence was instrumental in pursuit of a state agenda.

What all these cases have in common is that the state concerned suspended ethical rules for the duration. Servants of the state were permitted or encouraged to act with impunity. The logic can be encapsulated in the command, “do what is necessary to ensure total submission, and don’t report back on the details.” In Sudan, the war areas have been described as “ethics-free zones.”[footnote]Alex de Waal, “Starving out the South,” in Martin Daly and Awad Alsikainga, Civil War in the Sudan, London, British Academic Press, 1994.

2
Article 2(c).
3
Samuel Totten and Eric Markusen (eds.) Genocide in Darfur: Investigating the Atrocities in the Sudan, New York, Routledge, 2006.
4
The dominant genocide scholars in the U.S. have tended to emphasize closeness to the Holocaust paradigm as a definitional prerequisite for genocide rather than conformity with the letter of the 1948 Genocide Convention. The Darfur genocide determination dramatically reverses these emphases.
5
For a wider analysis of these domestic-international linkages in human rights activism, see: Thomas Risse and Kathryn Sikkink, ‘The socialization of international human rights norms into domestic practices: introduction,’ in Thomas Risse, Stephen Ropp and Kathryn Sikkink (eds.) The Power of Human Rights: International Norms and Domestic Change, Cambridge, Cambridge University Press, 1999.
7
Samantha Power, ‘A Problem from Hell’: America and the Age of Genocide, New York: Basic Books, 2002.