By Dr. Gregory H. Stanton[1] ©2002 Gregory H. Stanton

In 1994, 500,000 to one million Rwandan Tutsis along with thousands of moderate Hutus, were murdered in the clearest case of genocide since the Holocaust.  The world withdrew and watched.  To borrow a Biblical metaphor, we passed by on the other side. Samantha Power, in her searing article in The Atlantic Monthly, “Bystanders to Genocide: Why the United States Let the Rwandan Tragedy Happen,” and in her book, A Problem From Hell: America and the Age of Genocide[2], says “The story of U.S. policy during the genocide in Rwanda is not a story of willful complicity with evil.  U.S. officials did not sit around and conspire to allow genocide to happen.  But whatever their convictions about ‘never again,’ many of them did sit around, and they most certainly did allow genocide to happen.”[3]  Ms. Power concludes that her extensive research, including interviews with most of the U.S. policy makers who made the decisions, “reveals that the U.S. government knew enough about the genocide early on to save lives, but passed up countless opportunities to intervene.”[4]


Early Warnings

There were plenty of “early warnings” of the Rwandan genocide, but they were systematically ignored. The best book on the Rwandan genocide, Linda Melvern’s superb A People Betrayed: The Role of the West in Rwanda’s Genocide sets them forth in detail.  To list just a few, in the spring of 1992, the Belgian ambassador in Kigali, Johan Swinner warned his government that the Akazu, a secret group of Hutu Power advocates organized around the President’s wife, “is planning the extermination of the Tutsi of Rwanda to resolve once and for all, in their own way, the ethnic problem….” [5] In October 1992, Professor Filip Reyntjens organized a press conference in the Belgian Senate in which he described how Hutu Power death squads were operating and named their leaders, including Colonel Théoneste Bagasora, who later coordinated the genocide.[6]  In March 1993, four human rights groups led by Human Rights Watch and the International Federation of Human Rights issued a report on mass killings in Rwanda.  Although the word “genocide” was excised from the final report, the press release announcing it, written by Canadian law professor William Schabas, used the word genocide to describe the mass killings of Tutsis.[7]  The U.N. Special Rapporteur on Summary, Arbitrary, and Extrajudicial Executions, René Dégni-Ségui, conducted a mission to Rwanda in April 1993 and reported to the U.N. Human Rights Commission in August 1993 that the trial massacres of Tutsis, already begun by then, constituted genocide under the Genocide Convention.[8]

During the months prior to the Rwandan genocide, General Roméo Dallaire, commander of the U.N. Assistance Mission in Rwanda (UNAMIR), warned the U.N. Department of Peacekeeping Operations (DPKO) that Hutu extremists were planning a campaign to exterminate Tutsis.  In a now famous cable to New York on January 11, 1994, which DPKO authorized him to share with the U.S., French and Belgian Embassies, General Dallaire asked for authority to search for and seize the caches of machetes and other weapons that had been shipped into Rwanda for the Hutu militias, the Interahamwe.[9]  Iqbal Riza, deputy to then Undersecretary General for Peacekeeping Kofi Annan, in a letter signed by Annan, denied him permission to act, as exceeding UNAMIR’s mandate, and instructed him instead to take the information to the Rwandan government, many of whose members were planning the genocide. DPKO’s refusal to authorize action was confirmed on January 14 by Secretary General Boutros-Ghali himself.[10]

General Dallaire’s early warning of genocide was corroborated by the assassinations and further trial massacres of January to March 1994, which were also reported in cables to the U.S. State and Defense Departments.[11]  On January 21-22, UNAMIR seized a planeload of Belgian arms (shipped on a French plane) purchased by the Rwandan Armed Forces, which were then kept in joint UNAMIR/Rwandan government custody.[12]  At the request of DPKO, Dallaire provided confirmation of arms shipments and was finally authorized by the DPKO on February 3, 1994 to “assist the government of Rwanda” in recovering illegal arms.  In mid-February, the Rwandan Minister of Defense requested landing authorization for three planes carrying arms, but General Dallaire refused.  On February 27, General Dallaire repeated his request to DPKO for authorization to seize the caches of weapons the Interahamwe militias had hidden all over Rwanda. (General Dallaire had sent a Senegalese UNAMIR soldier to see some of the arms caches with his own eyes.) But U.N. authorities, including his direct superior, Canadian General Maurice Baril, again refused, referring privately to General Dallaire as a “cowboy.”

Belgium explicitly warned the U.N. Secretary General of impending genocide on February 25, 1994, but Belgium’s plea for a stronger U.N. peacekeeping force was rebuffed by members of the U.N. Security Council, particularly the U.S. and the United Kingdom.[13]


The Eight Stages of Genocide

In a paper I prepared at the State Department in 1996, which I am now expanding into a book, I suggested that there are Eight Stages of Genocide, and that each stage has distinctive warning signs.  There are also specific strategies at each stage to prevent and stop the genocidal process.[14]  I hope that a better understanding of the genocidal process will help policy-makers prevent future genocides.  Each of the eight stages was manifest in Rwanda.


1. Classification:  At this stage, social groups are classified into “us versus them.” Traditional Rwandan society was already classified into three groups, Tutsi, Hutu, and Twa.  A Tutsi royal clan, the Ganwa, ruled the country.  Although many African historians have pointed out that the groups did not fit the normal definition of ethnic groups, since they shared the same language, culture, and religion, there was nevertheless preferential endogamy, marriage within the group, a key characteristic of ethnic groups as well as of castes. In this strictly patrilineal society, a person took the group identity of his or her father.  Mixed marriages did not result in mixed children.  Des Forges[15] says these groups came to be seen as “castes,” by their German and Belgian colonial rulers, who ruled indirectly through the Tutsi elite.  Germans and Belgians developed the “Hamitic hypothesis” that Tutsis were the lost tribe of Ham and had migrated from Ethiopia.  The racist theories of the colonial era attributed superiority to Tutsis because of their aquiline noses and other “white” features. Tutsis were given preference in education, the church, the economy, and the government service.  Colonial rulers thus exacerbated the traditional classification divisions.  Ironically, the Hutu Power movement adopted these same theories, in order to portray Tutsis as foreign invaders who had dispossessed Hutus of rightful control over Rwanda.  The most notorious expression of the Hamitic hypothesis was in the famous speech by Léon Mugesera on November 22, 1992 when he said the Tutsis “belong in Ethiopia and we are going to find them a shortcut to get them there by throwing them into the Nyabarongo River [a source of the Nile.]”[16]


2. Symbolization: At this stage, the classifications are symbolized.  Groups are given names and other symbols (yellow stars, for example) and are required to wear them either by cultural tradition or laws.  In Rwanda, Belgium began to issue identity cards (ID’s) around 1926 and required them in the 1933 census.  The identity cards included each individual’s group identity, Tutsi, Hutu, or Twa.  They thus reified group identity for each person, and made changes from one group to another much more difficult.

Having studied the genocidal process and the history of genocidal massacres in Rwanda, I recognized the danger of the ethnic ID cards during my first stay in Rwanda in 1988, when I did a study of judicial administration for the Rwandan Ministry of Justice.  I had dinner with Joseph Kavaruganda, President of the Cour de Cassation (Supreme Court), and we agreed that the designation of ethnicity had to be removed from the ID cards.  I met with President Habyarimana several weeks later and urged him directly to issue new ID cards without the ethnic designation. “Someday they will be used for genocide,” I told him.  He remained impassive and non-committal.  Others also urged abolition of the ethnic ID’s, and that reform was included in the Arusha peace agreement signed in August 1993.  New ID cards were even printed.  But they were never issued.  Hutu Power advocates wanted the ethnic designation retained.  We now know why.  During the genocide, ID cards became facilitators of killing, because they permitted the killers to quickly determine who was Tutsi.  Those who refused to show their ID’s at Interahamwe roadblocks were presumed to be Tutsi unless they could quickly prove otherwise.  Nearly all Tutsis were immediately murdered.


3.  Dehumanization: This stage is where the death spiral of genocide begins.  The victim group is dehumanized.  It is called the names of animals or likened to a disease: vermin or rats, cancer or plague, or in Rwanda, “inyenzi” – cockroaches.  The reason this stage is necessary is that it gives ideological justification to the genocidists, who claim they are purifying the society.  It overcomes the normal human revulsion against murder.  If the other group is not human, then killing them is not murder.

In Rwanda, the dehumanization of Tutsis had already been a feature of genocidal massacres in 1959, 1962, and 1972.  In December 1990, the Hutu Power hate newspaper, Kangura, published the “Ten Commandments of the Hutu.”  They included the injunction, “The Bahutu should stop having mercy on the Batutsi.”  The Ten Commandments called for continuation of the Habyarimana government’s policy that the army be exclusively Hutu, and that officers be prohibited from marrying Tutsi women.  Cartoons and articles in Kangura referred to Tutsis as cockroaches and snakes, and regularly expounded the myth that they had invaded from Ethiopia.  Tutsis were “devils” who ate the vital organs of Hutus.  Twenty other extremist newspapers also published regular hate propaganda against Tutsis.[17]  Radio Télévision Libres des Milles Collines amplified the hate propaganda from 1993 onward, and brought it to every corner of Rwanda using repeater antennae provided by Radio Rwanda, the government network.  David Rawson, the U.S. Ambassador, said RTLMC’s euphemisms were subject to various interpretations and he defended its right to broadcast as “freedom of speech.” [18]  (This same misunderstanding of constitutional law was still prevalent in the State Department when I began work on Rwanda in July 1994.  The public affairs officer responsible for U.S. policy on Rwanda explained that this was why the U.S. opposed jamming RTLMC.  I explained, as a former law professor, that incitement to commit genocide is not “protected speech.”  Indeed if there were ever a case that met the “clear and present danger” test of U.S. First Amendment jurisprudence, this was it.)


4. Organization: All genocides are organized. At this stage, hate groups are organized, militias are trained and armed, and the armed forces are purged of members of the intended victim group as well as officers and others who might oppose genocide. Propaganda institutions, such as the hate newspapers and radio station, are also strengthened and funded.

After the RPF invasion in October 1990, the Rwandan Armed Forces (Forces Armées Rwandaises or FAR), the all-Hutu government army, expanded almost overnight from 5,000 to 28,000 men.[19]  It got considerable assistance in training and arms from the French government.  President Mitterand’s son, Jean-Christophe, headed the Africa office at the Elysée Palace, and was a close friend of President Habyarimana.  He was  reputed to own a plantation in Rwanda and to be personally involved in the arms trade.[20]  600 French paratroopers secretly took control of the counter-insurgency campaign.[21]   The Egyptian government, with the intervention of Foreign Minister Boutros Boutros-Ghali, sold $5.9 million in ammunition, rifles, mortar bombs, rockets, and rocket launchers to Rwanda on 28 October 1990.[22]  South African arms dealers were also a major source. Between 1990 and April 1994, Rwanda spent an estimated $112 million on arms, making it the third largest arms purchaser in Africa, after oil-rich Nigeria and Angola.[23]  The purchases were likely made with money diverted from loans by the World Bank.[24]

It was the organization of extremist militias, however, that marked the organizational turn toward genocide.  In 1992 the Interahamwe, the militia of the ruling MRND party, was organized.  It was soon followed by the Impuzamugambi, the militia of the CRD, an extreme Hutu Power party organized by the Akazu elite to make the President’s MRND seem moderate by comparison.  These militias were secretly trained in camps run by Rwandan army officers, armed with machetes, Kalashnikovs, and grenades from arms shipments to the government.


5.  Polarization: Moderates are targeted and assassinated.  Hate propaganda emphasizes the “us versus them” nature of the situation.  “If you are not with us, you are against us.”  There is no middle ground.  Moderates who attempt to negotiate peace are denounced as traitors.

Rwandan moderates had formed several opposition parties and had won seats in the National Assembly.  On 6 April 1992, Agathe Uwilingiyimana, a moderate Hutu, was named Minister of Education.  When she proposed ending the quota system that restricted Tutsi access to higher education, she was attacked in her home by twenty armed men.[25]  In November 1993, after she had been named Prime Minister in the government formed after the signing of the Arusha Accords, Radio Télèvision Libre Des Milles Collines publicly called for her assassination.  She was one of the first officials to be murdered during the genocide on April 7.  (Her ten Belgian UNAMIR guards were also slaughtered.)  Kangura and RTLMC called anyone who opposed Hutu Power an “accomplice” of the Tutsis and a secret ally of the R.P.F.

Joseph Kavaruganda, President of the Cour de Cassation (Supreme Court), another moderate Hutu, was also targeted by the extremists.  In January 1994, the head of the Interahamwe in Rugendo threatened Kavaruganda, and he complained to the President on January 15.  On February 21, thugs broke into the Supreme Court building and did considerable damage.  On March 19, 1994, Captain Pascal Simbiyangwa warned Justice Kavaruganda’s guards that the judge was a “cockroach” whose days were numbered and that the group who would kill him had already been chosen. On March 23, 1994, an Interahamwe, Enoch Kayonde told Justice Kavaruganda he could be killed at any time.  On the same day, Kavaruganda wrote a letter to President Habyarimana informing him of these death threats and asking for protection against the Presidential Guard.[26]  His pleas were to no avail.  Justice Joseph Kavaruganda, my personal friend, was murdered on the first day of the genocide.

It is significant that General Dallaire’s famous cable warning to the UN DPKO of the coming genocide was entitled, “Request for Protection of Informant.”  General Dallaire’s informant asked to be evacuated from Rwanda, possibly after temporary asylum in a foreign embassy.[27]  UN DPKO rejected the General’s plan.  Thereafter, the informant, who was personally opposed to the extermination plan, understandably stopped informing UNAMIR about it.  Physical protection of moderates is among the most important steps that can be taken to prevent genocide at this stage.  The UN refused to do even that, although it was clearly within UNAMIR’s mandate.


6. Preparation: During the preparation stage, plans are made for the genocide. Death lists are compiled.  Trial massacres are conducted, both as training for the genocidists, and to test whether there will be any response, such as arrests, international denunciations, or sanctions.  If the murderers get away with their crimes, if there is impunity, it is a green light to finish the genocide.

The trial massacres began in Rwanda soon after the Rwandan Patriotic Front invaded.  Hutus slaughtered 300 Tutsi civilians in Kabirira in October 1990.  In January 1991, 500 to 1000 Tutsi were murdered in Kinigi.  In March 1992, 300 Tutsi were massacred by Hutu militias in Bugesera.  No one was ever arrested for these crimes, and there were no demands from international diplomats for such arrests.  But the diplomatic community knew about the crimes.  Cables from the U.S. Embassy in February 1994 described the Interahamwe massacre of 70 Tutsis in Kigali between February 22 and 26.  On March 1, 1994, the Belgian ambassador reported that station RTLMC was broadcasting “inflammatory statements calling for hatred – indeed for extermination.”[28]


7. Extermination: At this stage, the killing legally defined as genocide begins.  Those who do it often think they are “purifying” their society, by “exterminating” those who are less than human and are a threat to them.  In Rwanda, the mass murder began within hours of the crash of President Habyaramana’s French plane on April 6, 1994.  He was shot down after conferring with regional leaders about implementation of the Arusha Accords, which he had signed in August 1993.  The Hutu Power elite saw the Accords as a direct threat to their power, because they called for sharing power with the Rwandan Patriotic Front.  To this day, it is unclear who shot down the President’s plane.  What is clear is that the Hutu Power genocidists were well prepared, and began the slaughter immediately.


8. Denial: During and after every genocide, the perpetrators deny they committed the crime.  They portray their murders as justified killing during war or repression of terrorism.  They dig up and dispose of the bodies and try to minimize the number of victims.  They try to blame the victims, often claiming that the victims’ own behavior brought on the killing. They portray the murders as spontaneous outbreaks in response to the victims’ depredations, or as the actions of rogue army commanders, rather than as intentional government policy.  They challenge the veracity of the eye-witnesses and assassinate the character of their accusers. The perpetrators claim to have been powerless to prevent the killings by others, and even have the audacity to claim they assisted their victims. All of these strategies of denial operated during and after the Rwandan genocide.  The presence of the Rwandan government representative at the very U.N. Security Council meetings that considered the situation provided an ideal forum for such denial.  Since the genocide, despite massive evidence against them, this denial by perpetrators has continued both at the International Criminal Tribunal for Rwanda and in Rwandan courts and prisons.


When did U.S. diplomats and policy makers know the mass murder was genocide?

Dr. Alan Kuperman, in his recently published book, The Limits of Humanitarian Intervention: Genocide in Rwanda, challenges “the common wisdom” that simple political will could have stopped the Rwandan genocide. He argues that well-meaning mediations and dilatory promises to back them up (e.g. the Arusha peace agreement on Rwanda) can actually increase the likelihood of genocide as they threaten the interests of ruling groups.[29]   He takes aim particularly at those who blame the U.S. for its inaction after the Rwandan genocide began.  The U.S. and United Kingdom played the leading role at the U.N. Security Council during the genocide.  Dr. Kuperman argues that neither U.S. nor U.K. policy makers recognized the killings as genocide for at least three weeks, and that even if they had acted immediately thereafter, it would have taken three more weeks to send in U.S. reinforcements.  He says that by that time, three-quarters of the killing was done.

Although the U.S. Defense Intelligence Agency (D.I.A.) recognized from radio intercepts as early as April 7 that centrally organized mass killing of Tutsis was underway, D.I.A. warnings went unheeded in the American government.  Some U.S. diplomats in Kigali began calling the killings genocide on the same, first day, and directly communicated their views to the State Department in Washington, DC.  The U.S. Embassy’s Deputy Chief of Mission Joyce Leader has told me personally that she began using the word genocide in her daily telephone calls to the State Department from the start.  It was clear to her that the Interahamwe and Presidential Guard were committing genocide. Dr. Kuperman questions whether Leader’s reports and the D.I.A. warnings were shared with top officials of the State and Defense Departments and the National Security Council. The answer is that although these reports were shared with top officials, including Assistant Secretaries and other policy makers, at their daily interagency secure teleconferences about the Rwandan catastrophe, other reports from the U.S. Ambassador to Rwanda and the C.I.A. contradicted them.[30]  Dr. Kuperman observes that although reports of the mass killing quickly reached mid-level officers in the U.S. State and Defense departments, the surfeit of information served to cloud rather than clarify the situation.


Refusal to invoke the G-Word

Why did policy makers at the State Department and National Security Council refuse to recognize that genocide was underway in Rwanda?  There are probably two reasons, both compelled by a prevenient group decision to avoid U.S. involvement.

First, the facts were resisted.  The U.S. government was forewarned of the impending genocide.  Communications were sent by cable, e-mail, and secure telephone from the U.S. embassy in Kigali informing the State Department about General Dallaire’s premonitions months before April 6.  But in 1993, President Clinton had ordered U.S. forces withdrawn from Somalia after General Aideed’s militia (possibly trained by Osama bin Laden’s Al Queda) killed eighteen Army Rangers.  Policy makers in Washington, D.C., especially Anthony Lake, Dick Clarke and Susan Rice at the National Security Council, George Ward at the State Department, and the Joint Chiefs of Staff at the Defense Department, distrusted U.N. peacekeeping missions and did not want the U.S. to get involved in another African “civil war,” another “quagmire.”[31]  In response to Somalia, President Clinton had just signed Presidential Decision Directive 25, which the same policy makers had drafted, limiting U.S. involvement in U.N. peacekeeping operations.  But it specifically allowed such intervention in cases of “genocide.”  They therefore resisted the “cognitive dissonance” of reports of impending genocide in Rwanda, which might have created at least a moral duty to intervene. The anti-interventionists dismissed General Dallaire’s reports as “unconfirmed,” meaning that U.S. embassy staff or intelligence personnel had not independently written about the arms caches and reported them through official cable channels.  They utilized cable reports from the American ambassador, David Rawson, in the early days of the genocide, to argue that this was just another episode of bi-lateral civil war, not a one-sided genocide.  Ambassador Rawson had grown up in Burundi with the Tutsi – Hutu conflict and he spoke Kirundi, the language of Burundi, which is closely related to Kinyarwanda, the language of Rwanda.  The Ambassador’s appraisal of the violence, however confused, therefore carried considerable credibility.  After the entire U.S. mission left for Burundi on April 10, with Ambassador Rawson in the last car, no further official channels existed to “confirm” reports from Kigali.  The first defense against action was denial of the facts.

The second reason for inaction was legal malpractice.  The State Department Bureau of African Affairs asked the State Department Legal Advisor’s office whether the massacres constituted genocide.  On April 26, Carl Pendorff issued an intelligence estimate calling the Rwandan massacres genocide.  At a crucial interagency meeting called by Deputy Assistant Secretary of State Prudence Bushnell, she asked, “Is this genocide?  And if it is, what are we going to do about it?”  Ms. Joan Donoghue of the Legal Advisors Office gave her opinion that the word genocide should be avoided, because she questioned whether the killings possessed the requisite “intent” and because use of the G-word, “genocide,” would obligate the U.S. to take action to stop it.[32]  Her oral opinion was soon followed by a written opinion from the Legal Advisor saying the same things.  Sadly, the lawyers were wrong on both points.  Intent can be proven by direct statements, but it is more often inferred from actions, like the systematic pattern of killing of Tutsis in Rwanda.  And unfortunately, the Genocide Convention imposes no legal requirement to take action to stop a genocide.  It only requires passage of national legislation to outlaw genocide,[33] and prosecution or extradition of suspected perpetrators.[34]  The Convention’s Article 8 states, “Contracting Parties may call upon the competent organs of the U.N.” to take action to suppress a genocide.  But that is not legally required.

For over two months, the Legal Advisors told the American government not to call the Rwandan killings genocide.  The State Department ordered the U.S. mission at the U.N. to vigorously oppose use of the term.  The U.K. rewrote a Presidential Statement proposed on April 29 by New Zealand’s Colin Keating, that month’s President of the Security Council, to avoid use of the word. On May 4, the U.N. Secretary General declared a “real genocide.”[35]

The U.S. continued to avoid the G-word until June.  In a now infamous press conference on June 10, State Department press spokesperson Christine Shelley, reading from talking points prepared by the Legal Advisors, declared that “acts of genocide have occurred in Rwanda.”  But when pressed by a reporter, she was unprepared to call it “genocide.”  This false distinction was finally buried the same day by Secretary of State Warren Christopher, himself a lawyer, who knew that Article 2 of the Genocide Convention defines genocide as acts of genocide.  An act of genocide is genocide, just as an act of rape is rape, or an act of murder, murder. The U.S. Secretary of State finally called it genocide on June 10, after most of the killing was over.

State Department lawyers and policy makers did not want to use the G-word because they wanted to avoid a duty to act.  So they chose another name for what was happening in Rwanda, one that would result in non-intervention: “civil war.”  Civil wars are two-sided (or multi-sided.)  The lesson the Clinton Administration learned from Somalia was, “Don’t get involved in African civil wars.”  Policy makers, including U.S. Ambassador David Rawson in Kigali, saw the killing as a continuation of the civil war that had plagued Rwanda since 1990, a war the Arusha Accords were supposed to settle. What they missed was the turn toward genocide of the Hutu Power movement.  Because they did not know much about genocide, they ignored the fact that most genocides have been committed during wars, including civil wars.  Robert Melson has shown in Revolution and Genocide that it is precisely during wars that pariah groups are most likely to become identified as threats, and therefore objects of genocide.[36]  Genocide and civil war are correlative, not mutually exclusive.  The second defense against action was legal definitionalism – denial that mass murder fit the legal definition of genocide.
The press and human rights groups also failed to name the crime until two weeks into the genocide. French newspapers were an exception. The first newspaper that called it genocide was Libération in an article by Jean-Philippe Ceppi on April 11, 1994. Libération had also been the first to use the word “genocide” in an early warning article about death squads in Rwanda in February 1993.  But the left-wing Libération is not given much weight by French foreign policy makers, and is not read by anyone in Washington. Le Monde followed with a story by Jean Hélène on April 12.  It, too, was ignored.   Human rights groups held back until Ken Roth, Executive Director of Human Rights Watch, wrote Colin Keating, President of the U.N. Security Council on April 19.  The Pope waited to call it genocide until April 27.

Besides the mis-reporting of the Rwandan killing as civil war, Dr. Kuperman notes that other factors contributed to inaction:

“Second, after a few days, violence was reported to be on the wane when in reality it was accelerating.  Third, most early death counts were gross underestimates, sometimes by a factor of ten….  Fourth, the initial focus was almost exclusively on Kigali, a relatively small city, and failed to note the broader scope of the violence.[37]


What the U.N. did and what it might have done

The U.N. did not wait to intervene in Rwanda until the beginning of the genocide.  Acting under Chapter VI of the U.N. Charter, the U.N. Department of Peacekeeping Operations had deployed 2,539 U.N. Assistance Mission in Rwanda (UNAMIR) troops to Rwanda by April 6, 1994.[38]  Dr. Kuperman claims they were too lightly armed to deter the Rwandan genocidists, who he says numbered 100,000, including the heavily armed Presidential Guard.  He agrees with General Dallaire that UNAMIR needed heavier weapons, full deployment of its 2548 authorized troops plus an equal number of reinforcements, all of them well-trained and well- supplied, with a clear mandate giving them authority to forcefully stop killing.  That could have been written into U.N. Security Council resolution 872 that created UNAMIR.  But the U.S. and U.K. had opposed a robust mandate with the 4,500 troops recommended by General Dallaire because it would have been too expensive. [39]

When the genocide began, policy makers in Washington and at the U.N. believed that UNAMIR forces lacked the strength to arrest the spread of the conflagration, and they refused to consider sending in their own troops.  In U.S. government parlance, that was a “non-starter.”  When that word is used, it really means, “We don’t want to think about it.”  It is the product of what social scientists have called “groupthink.”  Those who dissent are afraid to step forward to challenge the group assumptions.  State Department policy makers who attended a crucial meeting in the International Organization Affairs bureau on UNAMIR’s future have told me that after Assistant Secretary of State for African Affairs George Moose, National Security Council Peacekeeping Advisor Susan Rice, and International Organizations Deputy Assistant Secretary George Ward had all agreed that UNAMIR could not fulfill its mandate and should be withdrawn, they felt as subordinates that they could not object or contradict them.  They did not consider changing UNAMIR’s mandate because they assumed that troop-contributors had only committed to a peacekeeping operation, not an operation to stop genocide. No one suggested asking the troop-contributors if they would stay.  No one suggested sending in U.S. troops.  The U.N. Security Council’s earlier failure, because of U.S. and U.K. reluctance, to send a strong UNAMIR force created the self-fulfilling prophecy that nothing effective could be done.

In the U.N. Security Council, the U.S. took an active stance against keeping the UNAMIR troops in Rwanda. Ambassador Karl Inderfurth announced that position on April 15 in  “Informals”, closed meetings of the Security Council, with the representative of the genocidal Rwandan regime present.  Ambassador Inderfurth’s announcement of U.S. policy had fatal consequences.  The next day, the Rwandan Interim Government met, and knowing it could now act with impunity, decided to extend the genocide to Southern Rwanda.[40]

In the first week of the genocide, General Dallaire asked for a change in UNAMIR’s mandate that would authorize him to take action to stop as much killing as possible.  But instead, on April 21, the Security Council, led by the U.S. and the U.K., ordered reduction of UNAMIR to a token force of 270 troops.[41]   Over five hundred thousand Rwandan Tutsis were murdered while the U.N. “did a Pontius Pilate,” as General Dallaire told State Department officials in Fall 1994.[42]


Would UNAMIR intervention have saved lives?

Dr. Kuperman states, “Indeed, by my calculations, three-quarters of the Tutsi victims would have died even if the West had launched a maximum intervention immediately upon learning that a nationwide genocide was being attempted in Rwanda.”[43]   He concludes that although intervention during the Rwandan genocide would have been less effective than some think, saving 125,000 lives would have justified maximal intervention.  He notes that even the belated, minimal response proposed in May 1994 by the U.S., which would have unrealistically expected Tutsis to walk through militia infested areas to reach “safe zones” outside Rwanda might have saved 75,000 lives.[44]  (The cruel fate awaiting people who relied on weakly defended U.N. “safe areas” was demonstrated a year later in Srebenica, Bosnia.)

How many lives could have been saved?  We will never know.  But General Dallaire, the commander on the ground who knew the situation best, was and still is, convinced that a robust UNAMIR mandate plus reinforcements, demonstrating the international political will to stop further genocide, could have saved hundreds of thousands of lives.

Dr. Kuperman argues that reinforcements could not have arrived in time to save most victims’ lives.  But he only considers U.S. troops sent from the continental U.S.A. as reinforcements, a strangely self-defeating concept for a U.N. peacekeeping force.  Perhaps the most telling refutation of his view is the fact that over 1000 heavily armed French and Belgian troops flew into Kigali by April 10 to evacuate their own nationals.  If they had, instead, been used to reinforce UNAMIR, they might have had a powerful effect in deterring the spread of the genocide. An additional 500 Belgian reserves were available in Kenya, and 800 more French troops were stationed in central Africa.[45]  Two hundred and fifty U.S. Special Operations troops stood by in Burundi to assist, if necessary, with the evacuation of U.S. citizens. There were also tens of thousands of U.S. troops stationed in Europe, the Persian Gulf, the Indian Ocean, and other places much closer to Rwanda than the continental U.S.A.

Even without these reinforcements, according to General Dallaire, the UNAMIR troops could have used the weapons they had, which were superior to the machetes of the Interahamwe, to take down the roadblocks by force, and protect Tutsis who had gathered in defensible places.  The fact that the remaining 456 UNAMIR peacekeepers were able to save at least 25,000 lives by guarding people who had gathered in churches, stadiums, and hotels, leaves the question open whether the full 2,500 member force could not have saved many more lives had the U.N. Security Council immediately mandated it to do so.  In places protected by the 456 UNAMIR volunteers who stayed, most people survived. Even against the better-armed Presidential Guard, a robust response by UNAMIR might have deterred plans to extend the genocide. International outrage at attacks on U.N. peacekeepers might have also helped forge the political will necessary to obtain reinforcements.  Instead the U.N. Security Council, led by the U.S. and the U.K. decided to cut and run.  As General Dallaire later told State Department officials, “A peacekeeping force that is trying to stop genocide must expect to take casualties, or it is worthless.”

The major problem from the beginning of UNAMIR was that all but one of the Western powers were unwilling to send troops to intervene, or even to provide airlift and financing for an international force.  The result was that poorly trained troops from Bangladesh, lacking any equipment, were the largest contingent, followed by the Ghanaians, who arrived without a single vehicle.  The Belgian force numbered only 420,[46] and withdrew within days after the massacre of ten Belgian soldiers guarding the Prime Minister. The attack was consciously planned to drive out the Belgians.[47]  The Hutu Power militants had learned the lessons of Somalia, too.  If you kill them, they will leave.

Late in the genocide, France, which had supplied the Rwandan Armed Forces (FAR) with arms and advisors and may have helped train Interahamwe militias, launched Opération Turquoise.  After getting U.N. authorization[48], France sent in Senegalese and French troops on June 23.[49]  Opération Turquoise saved more than 10,000 lives in western Rwanda, but also permitted the leaders of the genocide to escape into Zaire.[50]

What finally stopped the genocide was the victory by the Rwandan Patriotic Front (RPF), which took Kigali on July 4 and declared a ceasefire on July 18. From July 14 to 16, a million refugees streamed into Zaire, the fastest migration of people in history. Refugee camps quickly fell under the control of the Hutu Interahamwe. The RPF committed its own atrocities, such as the massacre of at least 1000 Hutu holdouts at Kibeho.[51]  The camps were not emptied until the 1997 invasion of Zaire by Rwanda, Uganda, and Laurent Kabila. During their march to Kinshasa, Kabila’s troops and the Rwandan Patriotic Army committed more genocidal massacres against Hutu refugees in the Kivus, south of Kisangani, and at Mbandaka. The war that ensued in the Democratic Republic of the Congo has since cost over two million lives.[52]


Why Did We Pass By On The Other Side?

The major Western governments did know from the first days that mass killing was underway.  The U.S. Deputy Chief of Mission used the word genocide in her calls to the State Department from the beginning. Much of the communication was done by secure phone calls, since both Joyce Leader and Ambassador Rawson were cut off from access to the U.S. Embassy for long periods.  Classified documents confirm this very early recognition of mass killing.  The information did reach the top levels of government.

The real problem was genocide denial, first through denial of the facts, and then through denial that the mass murder was genocide.  State Department and Defense Department lawyers who were opposed to intervention, either because of their own views or to please their anti-interventionist superiors, denied that the mass murders constituted genocide.  That this denial was intentional can be seen from the fact that they continued to deny the genocide for two months, until long after it was obvious to nearly everyone else that one of the worst genocides of the twentieth century was underway.

General Dallaire and a panel of military experts assembled by the Carnegie Commission on Preventing Deadly Conflict concluded that it would not have taken weeks to put troops in place who could have significantly reduced the killings.  2539 UNAMIR troops were already in Rwanda. Dr. Kuperman dismisses this stubborn fact because they were only lightly armed.  But over 1000 heavily armed airborne troops from France, and Belgium were immediately available and did arrive by April 10.  Another 1550 Belgian, French, and U.S. troops were in nearby African countries, rendering Dr. Kuperman’s lengthy calculations of airlift capacity from the mainland U.S. irrelevant.  The Carnegie Commission panel concluded that prompt international denunciation of the genocide, accompanied by forceful military resistance by UNAMIR and Western troops could have saved many lives, as Colonel Scott Feil argues convincingly in his book for the Carnegie Commission, “Preventing Genocide: How The Early Use Of Force Might Have Succeeded In Rwanda.”[53]
Dr. Kuperman’s most telling point is that for intervention to be effective, it must come before genocide begins, not after it has begun.  Early warning must be coupled with early preventive action months before genocide. Dr. Kuperman comments:

“I have argued in previous writings that if UNAMIR [the United Nations Assistance Mission in Rwanda] had been reinforced several months before the outbreak of the violence, as Belgium’s government urged at the time, genocide might have been averted.  Such a reinforcement was justified by the flood of warning signs from Rwanda – available primarily to Belgium and France but often shared with the UN and United States – which indicated that extremist elements were seeking to provoke renewed civil war and massive civilian killing, if not necessarily genocide.”[54]

The problem is that early warnings of “mere” civil war and massive civilian killing seldom result in international intervention, whereas early warnings of genocide might.  But in Rwanda, except for a few who used the sacred “G-word,” none dared call it genocide – until it was too late. The Rwandan conflagration burned so fast, that reacting after it started would, for most of its victims, have been too late.  Within three weeks, at least 300,000 lives had been consumed.[55]



The Rwandan genocide could have been prevented.  The early warning signs were clear.  UNAMIR troops were already on the ground in Rwanda, though with inadequate training and material support.  But in a failure of political will, the U.S., the U.K., the U.N. Secretariat and the U.N. Security Council refused to act to prevent or stop the genocide. At least 500,000 people perished.

Two questions remain:

Why, with all the early warnings, did the U.S., U.K., France and the U.N. not do what needed to be done to prevent the Rwandan genocide?

Why, once the genocide began, did the U.N. Security Council order UNAMIR to withdraw, rather than sending reinforcements to stop the genocide?

Lack of political will is at the heart of the answers to both questions.  But to muster political will, governments must perceive and understand the crisis and have realistic options to resolve it.


1. The early warnings were ignored.  In Rwanda, the U.N. DPKO and the U.S., as well as other governments, refused to discern the signs of genocide.  Because they did not understand the genocidal process, they missed the early warning signs.  There was, at the time, no systematic understanding among policy makers of how genocide develops so that the warning signs could be noticed.  (That is why I am now writing my book, The Eight Stages of Genocide, to provide a powerful explanatory model of the genocidal process, along with specific policy recommendations for what can be done to prevent and stop the process at each stage.)  Misplaced hopes for the Arusha Accords led the U.S. Ambassador and diplomats to ignore the planning for genocide within the Rwandan government and the Hutu Power militias.


2. After the genocide started, policy makers resisted and misconstrued the facts. In the post-Somalia era, policy makers did not want to get involved in another African “quagmire,” so they minimized the facts.  The first cables from the U.S. Ambassador treated the killings as a bilateral continuation of the Rwandan civil war, rather than as a one-sided genocide.  The number of deaths in the early weeks was grossly under-estimated.  Closure of embassies and withdrawal of personnel and press prevented adequate reporting on the genocide, especially “confirmed” reporting in official cables from embassy staff.  Generally, policy makers require “confirmed” fact- finding before they will take action.


3. Lawyers who did not understand the law refused to call it genocide. Lawyers at the U.S. State and Defense departments and at the British Foreign Office had little training in the law of genocide. What knowledge they did have, they misapplied.  They created conceptual uncertainty among policy makers who relied upon them for advice.  Their power to block policy determinations, press guidance, and instruction cables saying that genocide was underway in Rwanda meant that for over two months, the U.S. and U.K. refused to call the Rwandan genocide by its proper legal name.  It also meant that the U.S. and U.K. refused to permit the U.N. Security Council to declare that genocide was being committed in Rwanda.  Words, especially legal words, have consequences.  During the Rwandan genocide, refusal to name it genocide meant that policy makers in denial could continue to obstruct action because they could argue there was no imperative to intervene.  The lawyers even misconstrued the law on the duty to intervene, arguing that the Genocide Convention creates a legal, rather than simply a moral duty to do so.


4. Groupthink ruled out effective options for intervention. When policy makers finally recognized the facts, they thought they had no acceptable options to prevent the genocide.  In the interagency policy meetings in the U.S. government, dispatch of U.S. troops was ruled out as a “non-starter,” and was never seriously considered.

UNAMIR was perceived as too weak and undersupplied to stop the rapidly spreading killing.  U.S. or British airlift for UNAMIR re-supply and reinforcement was also ruled out because of danger to American and British personnel and because of fear that such a step would lead the U.S. and U.K. into another African “quagmire.”

Policy makers had not considered options available when the genocide started They believed that UNAMIR’s Chapter 6 mandate would have to be changed to Chapter 7 to permit intervention without the permission of the Rwandan interim government. (General Dallaire has always rejected this contention because he asserts that UNAMIR’s Chapter 6 rules of engagement already authorized the use of force to protect civilian lives.)  They also ruled out asking UNAMIR troop contributors whether they would keep their troops in Rwanda under a changed mandate.  Unfortunately, the U.S. and U.K. did not consider it feasible to change UNAMIR’s mandate and place UNAMIR troops at risk, though there was little chance of a veto by any of the Permanent Five members of the Security Council.

Those who engaged in groupthink policy making also believed their responsibility would never be known.  Because policy memos and cables were classified, and because all personnel evaluations in the State and Defense departments are top-down, they might be right. This paper is one small crack in the wall of bureaucratic irresponsibility. Books by Linda Melvern, Alison des Forges, Samantha Power, and Michael Barnett have done much more.  Eventually, policy-makers need to know they will be held responsible for their decisions.


5. The U.S. was unwilling to financially or militarily support a reinforced UNAMIR. Despite a Ghanaian offer to keep its troops in Rwanda, along with offers from several other African states to reinforce UNAMIR, the U.S. was unwilling to make the financial commitment to support an expanded operation.  The U.S. and other military powers were unwilling to risk the lives of any of their own citizens.  Instead they sacrificed the lives of over 500,000 defenseless Rwandans.  This U.S. position led the ever-realistic U.K. to conclude that because U.S. approval in the Security Council would be necessary for such an intervention force, there was no possibility it would be approved.  Hence, an endangered, undersupplied UNAMIR force should be withdrawn.


6. Rwandan lives were not worth saving.  Although the U.S. and U.K. were willing to commit billions to save lives in Bosnia, where people are white, and the war was close to the interests of the European community, they were unwilling to do so in Rwanda, where people are black, and neither country has strategic or economic interests.  This racist double-standard was pointed out repeatedly by Nigeria’s Ambassador Gambari in the U.N. Security Council.  Our circle of moral concern excluded people of a different race in a continent far away. We ignored our common humanity.


Ultimately the failure to prevent the Rwandan genocide was a political failure.  Those with power failed to protect the powerless.  The world still lacks the international institutions and the political will to stop genocide.  To address this fundamental problem, as I suggested at a conference on genocide held in London in October 2000,[56]a global movement is needed in the twenty-first century like the anti-slavery movement of the nineteenth century.  To launch that movement is the purpose of Genocide Watch and The International Campaign to End Genocide.[57]


[1] President, Genocide Watch.  Fellow, The Woodrow Wilson International Center for Scholars. P.O. Box 809, Washington, D.C., 20044 USA. This paper was presented on 27 January 2002 in London, England at a conference, “Generations of Genocide,” sponsored by the Institute of Contemporary History and the Weiner Library.

[2] Power. Samantha, A Problem From Hell. America and the Age of Genocide, New York: Basic Books, 2002.– 108.

[3] Power, Samantha, “Bystanders to Genocide: Why the United States Let the Rwandan Tragedy Happen,” The Atlantic Monthly,September 2001, p. 86.

[4] Power (cf. note 3), p. 84.

[5] Belgian Senate, Commission d’enquête parlementaire concernant les évenéments du Rwanda, Rapport, 6 December 1997, p. 493.  Quoted in Melvern, Linda, A People Betrayed. The Role of the West in Rwanda’s Genocide, Zed Books, London, 2000, p.49.

[6] Prunier, Gérard, The Rwanda Crisis 1959 – 1994, History of a Genocide, London, Hurst and Company, 1995, p. 168.

[7] International Federation of Human Rights, Africa Watch (Human Rights Watch), Inter-African Union of Human Rights, and International Centre of Rights of the Person and of Democratic Development. Report of the International Commission of Investigation of Human Rights Violations in Rwanda since October 1, 1990 (January 7 – 21, 1993), New York: Human Rights Watch, 1993.

[8] Ndiaye, B.W., Report by Special Rapporteur on his mission to Rwanda from 8 to 17 April 1993, Commission on Human Rights, U.N. Economic and Social Council document E/CN.4/1994/7/Add.1, 11 August 1993, p. 23.

[9] Dallaire, Roméo, Cable to General Baril, UNDPKO, 11 January 1994, in Adelman, Howard and Suhrke, Astri (eds.), The Path of A Genocide. The Rwanda Crisis from Uganda to Zaire, London: Transaction Publishers, p. xxi.

[10] Des Forges, Alison, Leave None To Tell The Story, Genocide in Rwanda, Human Rights Watch & FIDH, New York, 1999, p. 154.

[11] Des Forges (cf. note 10), pp. 159 – 171.

[12] Des Forges (cf. note 10), pp. 156 – 157.

[13] Alan J. Kuperman, The Limits of Humanitarian Intervention. Genocide in Rwanda, The Brookings Institution Press, Washington, D.C., 2001:  p. 85.

[14] Stanton, Gregory, “The Eight Stages of Genocide,” @

[15] Des Forges (cf. note 10), p. 38.

[16] Melvern (cf. note 5), p. 47.

[17] Kakwenzire, Joan and Kamukama, Dixon, “The Development and Consolidation of Extremist Forces in Rwanda 1990 – 1994,” in: Adelman, Howard and Suhrke, Astri (eds.) (cf. note 9), p. 76.

[18] Melvern (cf. note 5), p. 71.

[19] Melvern (cf. note 5), p. 33.

[20] Kakwenzire, Joan and Kamukama, Dixon, in: Adelman, Howard and Suhrke, Astri (eds.) (cf. note 9) , p. 83.

[21] Melvern (cf. note 5), p.31.

[22] Melvern (cf. note 5), p. 31.

[23] Melvern (cf. note 5), p. 32, 67.

[24] Melvern (cf. note 5), p. 67-68.

[25] Melvern (cf. note 5), p. 46-47.

[26] Copies of this letter is in the Melvern archive at the University of  Wales, Aberystwyth and with the Gregory Stanton, who may be reached at

[27] Adelman, Howard and Suhrke, Astri (eds.), note 9.

[28] Caplan, Gerald, Rwanda. The Preventable Genocide, The Report of the International Panel of Eminent Personalities to Investigate the 1994 Genocide in Rwanda and the Surrounding Events, Organization of African Unity, 1998.

[29] Kuperman, Alan J.,“False Hope Abroad: Promises to Intervene Often Bring Bloodshed,” in Washington Post, June 14, 1998, Outlook Section.

[30] Joyce Leader lived next door to Prime Minister Agathe Uwilingiyimana, who was murdered on the first day in a nearby UNDP compound after failing in her desperate efforts to scale the wall to Leader’s compound.  Members of the Presidential Guard invaded Leader’s house searching for Prime Minister
Uwilingiyimana.  She was a moderate Hutu who opposed the Hutu Power conspirators. The Rwandan Army at Camp Kigali then murdered ten Belgian soldiers who had been dispatched to protect the Prime Minister.  The murders were an intentional terror tactic to convince Belgium to withdraw its troops, a
tactic that succeeded brilliantly.

[31] All of the American policy makers who made the decisions during the Rwandan genocide, including the decision to order withdrawal UNAMIR troops, were later promoted.  Dick Clarke became chief of counter-terrorism operations at the National Security Council.  Susan Rice became Assistant Secretary of State for African Affairs.  George Ward became Ambassador to Namibia and then, after being recalled from that job, left the State Department and became Director of Training at the U.S. Institute of Peace.  The military personnel moved up the chain of command.

[32] The State Department lawyers who gave this erroneous legal advice still work for the State Department with promotions and higher salaries.

[33] Genocide Convention, Article 5.

[34] Genocide Convention, Articles 5 – 7.

[35] Kuperman (cf. note 13), p. 31.

[36] Melson, Robert, Revolution and Genocide, University of Chicago Press, 1992, p. 273.

[37] Kuperman (cf. note 13), p.24.

[38] Melvern (cf. note 5), p.106.

[39] Melvern (cf. note 5), p. 85.

[40] Melvern (cf. note 5), p. 163.

[41] U.N. Security Council Resolution S/Res/912 (1994), 21 April 1994.  See also Des Forges (cf. note 10), p. 631.

[42] Dr. Kuperman disputes the commonly accepted death toll of 800,000, claiming that Rwanda’s pre-genocide Tutsi population was only 650,000, 8.27 percent of the population, based on the 1991 census.  Kuperman (cf. note 13), p. 20. However, the most recent count of the dead, carried out locality by locality, yields a figure of over one million dead from 1990 through 1994, with 800,000 dead during the genocide. Associated Press, “More Than One Million Rwandans Killed in 1990’s,” NY Times Online News Report,  14 February, 2002.

[43] Kuperman, (cf. note 13), p. viii.

[44] Kuperman (cf. note 13), p. 77.

[45] Melvern (cf. note 5), p. 147.  Kuperman’s figures are similar.

[46] Kuperman (cf. note 13), p. 40.

[47] Dallaire, Roméo, Cable to General Baril, UNDPKO, 11 January 1994, in Adelman, Howard and Suhrke, Astri (eds.) (cf. note 9), p. xxi.

[48] U.N. Security Council Resolution 929 (S/PV.3392) 22 June 1994.

[49] Melvern (cf. note 5), p. 213.

[50] Prunier, Gérard, “Operation Turquoise: A Humanitarian Escape,” in Adelman, Howard and Suhrke, Astri (eds.) (cf. note 9),  p. 303.

[51] Kuperman (cf. note 13), p.20.

[52] International Crisis Group, Disarmament in the Congo,, Nairobi/Brussels, 14 December 2001, p. 2.

[53] Feil, Colonel Scott,“Preventing Genocide: How the Early Use of Force Might Have Succeeded in Rwanda,” Report to the Carnegie Commission on Preventing Deadly Conflict, New York, April 1998.

[54] Kuperman (cf. note 13): p. 97; See also Alan J. Kuperman, “The Rwanda Failure,” in Washington Post, December 29, 1998, op-ed.

[55] Kuperman (cf. note 13), p. 122.

[56] Stanton, Gregory, “How We Can Prevent Genocide,” @,  forthcoming in the proceedings of the Conference in Honor Raphael Lemkin, London, October 2000, Leo Kuper Foundation.

[57] See


Comments are closed.