During those first weeks of genocide, Western reporting was marred by four lapses. First, it mistook genocide for civil war. The country had been wracked by a low-level civil war from 1990 to 1993 between the government, controlled by the Hutu majority, and a rebel force comprised mainly of Tutsi. Though a minority, the Tutsi had ruled until the late 1950s when the Hutu took power and forced many Tutsi to flee as refugees. Conflict continued through the next three decades, so upon the outbreak of genocide on April 6, 1994, Western correspondents reported the initial burst of violence in the capital Kigali as resumption of a bloody civil war.

On April 11, an editorial in London's Times pondered international calls for a cease-fire and asked rhetorically, "Which parties would be asked to cease fire against whom?"

An April 12 report in Belgium's De Standaard on government violence in Kigali added that "it is absolutely certain that a large number of acts of terror were committed" as well "in the area controlled by the rebels." Early reports also indicated that the Tutsi rebels were winning the civil war and had rejected government offers of a nationwide cease-fire, which contradicted any notion of Tutsi as victims. By April 13, Paris Radio France International reported that "the fall of Kigali seems imminent." On April 14, Le Monde and The London Times reported that it was now the Hutu who feared vengeance from Tutsi rebels who had gained the upper hand in Kigali.

Violence was then reported on the wane when in fact it was mounting. On April 11, just four days after the fighting had started, The New York Times reported that fighting in Rwanda had "diminished in intensity." Three days later, Le Monde said that "a strange calm reigns in downtown" Kigali. On April 15, it reported this calm spreading to the capital's suburbs, allowing "humanitarian organizations to cautiously resume their activities." Only on April 18 did Brussels La Une Radio Network question this consensus by explaining that the decline in reports of violence was because "most foreigners have left, including journalists."

The exodus of reporters was so thorough that it virtually halted Western press coverage. European newspapers that had been providing daily coverage of the violence in Kigali stopped cold on April 18, for four days in France's Le Monde and seven in Britain's Guardian. Ironically, this was when the slaughter reached its peak.

Early published death counts were gross underestimates, sometimes by a factor of ten. On April 10, three days into the killing, The New York Times quoted estimates of 8,000 or "tens of thousands" dead in Kigali. However, during the second week media estimates did not rise at all. The estimates did not rise to levels that commonly would be considered "genocidal" for a country of 8 million people with 650,000 Tutsi. On April 16, the Guardian still reported only an "estimated 20,000 deaths." Two days later, The New York Times repeated this same statistic, underestimating the actual carnage at that point by about ten-fold. Not until a few days later did the scope of killing rapidly emerge.

Fourth, for nearly two weeks, Western news organizations focused almost exclusively on Kigali, a city that contained only 4 percent of Rwanda's population, and did not report the far broader tragedy unfolding around them. The few reports of violence in the countryside seemed to indicate renewal of mutual communal strife or civil war, rather than genocide. On April 11, Paris Europe No. 1 Radio reported that "Hutus are hunting down Tutsis throughout the country," but then added, "and the other way round." Brussels La Une Radio Network reported that killing and looting in Rwanda's southwest was targeted against the "opposition," rather than an ethnic group. Likewise, on April 12 The Washington Post wrote, "sketchy reports said fighting has spread to Rwanda's countryside," but in a context suggesting combat between government troops and armed rebels. The first report of a large-scale massacre outside the capital came on April 16.

American newspapers failed to convey the nationwide scope of the violence until April 22 when The New York Times belatedly reported that fighting bands had reduced "much of the country to chaos." Still, many foreign observers could not conceive that genocide was under way. On April 23, The Washington Post speculated that the dearth of Tutsi refugees fleeing Rwanda was because "most of the borders have been sealed." Only on April 25 was the riddle solved when The New York Times reported that violence had "widened into what appears to be a methodical killing of Tutsi across the countryside," and that the missing refugees "either have been killed or are trying to hide."

At least three factors help to account for these reporting lapses. First, the evacuation of foreign nationals left few reporters in the country after the first few days or in the capital after the first week. Second, the situation was legitimately confusing. Tutsi rebels were winning the civil war and retaliating against suspected civilian Hutu extremists at the same time that the civilian Tutsi population was being systematically exterminated. Third, even experts were slow to appreciate what was happening. The commander of Belgian peacekeepers stated on April 15 to Paris Radio France International that "the fighting has ... all but stopped." No human rights group even suggested the possibility of genocide until April 19.

In the wake of Rwanda's tragedy, the media harshly criticized the United Nations and its Western members for not immediately recognizing the killing campaign and reacting to prevent it. Such criticism is only partially valid. American and other Western officials dragged their feet after the genocide was reported, avoiding use of the word genocide for weeks afterward for fear of being compelled to intervene. But the media must share the blame for failing to provide prompt notice of the genocide. In obscure parts of the world, where Western governments do not invest significant intelligence assets, the news business is relied upon to serve as a surrogate early-warning system. In Rwanda, it did not fulfill this role.

Partially in reaction to this reporting failure in Rwanda, Western media have suffered from exactly the opposite problem ever since. They now exaggerate the extent of civilian atrocities in ethnic conflict. Around the world, rebels and human rights groups learned the lesson from Rwanda that they must declare "genocide" to have any hope of Western intervention. Because the press does not want to get caught napping again, it duly reports such claims even though it cannot confirm them. Thus, Western readers were told for months that genocide was raging in Kosovo, but forensic investigators have been able to find just 2,000 corpses to date, some of whom may have been armed rebels.

Likewise, Western media reported that genocide was occurring in East Timor after its vote for independence, but subsequently only about 200 bodies were found. This is not to say that a few hundred or thousand deaths are unimportant. But they do not comprise genocide by any other reasonable definition. The UN defines genocide as "acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group as such." The definition has been broadened in practice to include destruction of political groups.

Perhaps the main reason that Western correspondents have had difficulty reporting ethnic violence accurately is that at least one of the sides doesn't want them to, and reporters cannot confirm many allegations without risking their lives by visiting combat zones. There is no moral requirement for journalists to make such a personal sacrifice. But so long as reporters do not confirm the facts on the ground, they must try to do everything else possible to piece together the real story for readers - in full awareness that combatants, governments and private agencies are all trying to manipulate them.

Rwanda's Hutu government wanted reporters to think that violence was civil war rather than genocide. In a similarly manipulative way, the Kosovo Liberation Army wanted reporters to think that Yugoslav government violence prior to NATO's bombing was genocide or ethnic cleansing rather than counterinsurgency. In both cases, Western reporters were fooled. They should take a lesson from this as they continue their vital task of informing Western policymakers and publics about violent conflicts around the world.

Alan J. Kuperman is a MacArthur transnational security fellow at MIT's Center for International Studies and a fellow of the Institute for the Study of World Politics. His most recent publication is in the Jan./Feb. 2000 issue of Foreign Affairs. The article, "Rwanda in Retrospect," is an excerpt from his forthcoming book on the limits of humanitarian intervention.