Foreign Policy


In Retrospect

A review of NATO’s war over Kosovo, Part I

By Noam Chomsky

The tumult having subsided, it should be possible to undertake a relatively dispassionate review and analysis of NATO’s war over Kosovo. One might have expected the theme to have dominated the year-end millennarianism, considering the exuberance the war elicited in Western intellectual circles and the tidal wave of self-adulation by respected voices, lauding the first war in history fought “in the name of principles and values,” the first bold step towards a “new era” in which the “enlightened states” will protect the human rights of all under the guiding hand of an “idealistic New World bent on ending inhumanity,” now freed from the shackles of archaic concepts of world order. But it received scant mention.

A rare exception was the Wall Street Journal, which devoted its lead story on December 31 to an in-depth analysis of what had taken place. The headline reads: “War in Kosovo Was Cruel, Bitter, Savage; Genocide It Wasn’t.” The conclusion contrasts rather sharply with wartime propaganda. A database search of references to “genocide” in Kosovo for the first week of bombing alone was interrupted when it reached its limit of 1,000 documents.

As NATO forces entered Kosovo, tremendous efforts were undertaken to discover evidence of war crimes, a “model of speed and efficiency” to ensure that no evidence would be lost or overlooked. The efforts “build on lessons learned from past mistakes.” They reflect “a growing international focus on holding war criminals accountable.” Furthermore, analysts add, “proving the scale of the crimes is also important to NATO politically, to show why 78 days of airstrikes against Serbian forces and infrastructure were necessary.”

The logic, widely accepted, is intriguing. Uncontroversially, the vast crimes took place after the bombing began: they were not a cause but a consequence. It requires considerable audacity, therefore, to take the crimes to provide retrospective justification for the actions that contributed to inciting them.

One “lesson learned,” and quickly applied, was the need to avoid a serious inquiry into crimes in East Timor. Here there was no “model of speed and efficiency.” Few forensic experts were sent despite the pleas of the UN peacekeeping mission, and those were delayed for four months, well after the rainy season would remove essential evidence. The mission itself was delayed even after the country had been virtually destroyed and most of its population expelled. The distinction is not hard to comprehend. In East Timor, the crimes were attributable directly to state terrorists who were supported by the West right through the final days of their atrocities. Accordingly, issues of deterrence and accountability can hardly be on the agenda. In Kosovo, in contrast, evidence of terrible crimes can be adduced to provide retrospective justification for the NATO war, on the interesting principle that has been established by the doctrinal system.

Despite the intensive efforts, the results of “the mass-grave obsession,” as the WSJ analysts call it, were disappointingly thin. Instead of “the huge killing fields some investigators were led to expect,..the pattern is of scattered killings,” a form of “ethnic cleansing light.” “Most killings and burnings [were] in areas where the separatist Kosovo Liberation Army [KLA-UCK] had been active” or could infiltrate, some human-rights researchers reported, an attempt “to clear out areas of KLA support, using selective terror, robberies and sporadic killings.” These conclusions gain some support from the detailed OSCE review released in December, which “suggests a kind of military rationale for the expulsions, which were concentrated in areas controlled by the insurgents and along likely invasion routes.”

The WSJ analysis concludes that “NATO stepped up its claims about Serb ‘killing fields’” when it “saw a fatigued press corps drifting toward the contrarian story: civilians killed by NATO’s bombs.” NATO spokesperson Jamie Shea presented “information” that can be traced to KLA-UCK sources. Many of the most lurid and prominently-published atrocity reports attributed to refugees and other sources were untrue, the WSJ concludes. Meanwhile NATO sought to deny its own atrocities, for example, by releasing a falsified videotape “shown at triple its real speed” to make it appear that “the killing of at least 14 civilians aboard a train on a bridge in Serbia last April” was unavoidable because “the train had been traveling too fast for the trajectory of the missiles to have been changed in time.”

The WSJ analysts nevertheless conclude that the “heinous” crimes, including the huge campaign of expulsion, “may well be enough to justify” the NATO bombing campaign, on the principle of retrospective justification.

The OSCE study is the third major source concerning Serb crimes. The first is the State Department’s case against Milosevic and his associates in May; the second, their formal indictment shortly after by the International Tribunal on War Crimes. The two documents are very similar, presumably because the “remarkably fast indictment” by the Tribunal was based on U.S.-U.K. “intelligence and other information long denied to [the Tribunal] by Western governments.” Few expect that such information would be released for a War Crimes Tribunal on East Timor, in the unlikely event that there is one. The State Department updated its case in December 1999, with what is intended to be the definitive justification for the bombing, adding whatever information could be obtained from refugees and investigations after the war.

In the two State Department reports and the Tribunal indictment, the detailed chronologies are restricted, almost entirely, to the period that followed the bombing campaign initiated on March 24. Thus, the final State Department report of December 1999 refers vaguely to “late March” or “after March,” apart from a single reference to refugee reports of an execution on March 23, the day of NATO’s official declaration that the air operations announced on March 22 would begin. The one significant exception is the January 15 Racak massacre of 45 people. But that cannot have been the motive for the bombing, for two sufficient reasons: first, the OSCE monitors and other international observers (including NATO) report this to be an isolated event, with nothing similar in the following months up to the bombing; we return to that record directly. And second, such atrocities are of little concern to the U.S. and its allies. Evidence for the latter conclusion is overwhelming, and it was confirmed once again shortly after the Racak massacre, when Indonesian forces and their paramilitary subordinates brutally murdered 50 or more people who had taken refuge from Indonesian terror in a church in the remote Timorese village of Liquica. Unlike Racak, this was only one of many massacres in East Timor at that time, with a toll well beyond anything attributed to Milosevic in Kosovo: 3-5000 killed from January 1999, credible church sources reported on August 6, about twice the number killed on all sides in Kosovo in the year prior to the bombing, according to NATO. Historian John Taylor estimates the toll at 5-6000 from January to the August 30 referendum.

The U.S. and its allies reacted to the East Timor massacres in the familiar way: by continuing to provide military and other aid to the killers and maintaining other military arrangements, including joint training exercises as late as August, while insisting that security in East Timor “is the responsibility of the Government of Indonesia, and we don’t want to take that responsibility away from them.”

In summary, the State Department and the Tribunal make no serious effort to justify the bombing campaign or the withdrawal of the OSCE monitors on March 20 in preparation for it.

The OSCE inquiry conforms closely to the indictments produced by the State Department and the Tribunal. It records “the pattern of the expulsions and the vast increase in lootings, killings, rape, kidnappings and pillage once the NATO air war began on March 24.” “The most visible change in the events was after NATO launched its first airstrikes” on March 24, the OSCE reports. “On one hand, the situation seemed to have slipped out of the control of any authorities, as lawlessness reigned in the form of killings and the looting of houses. On the other, the massive expulsion of thousands of residents from the city, which mostly took place in the last week of March and in early April, followed a certain pattern and was conceivably organized well in advance.”

The word “conceivably” is surely an understatement. Even without documentary evidence, one can scarcely doubt that Serbia had contingency plans for expulsion of the population, and would be likely to put them into effect under NATO bombardment, with the prospect of direct invasion. It is commonly argued that the bombing is justified by the contingency plans that were implemented in response to the bombing. Again, the logic is interesting. Adopting the same principle, terrorist attacks on U.S. targets would be justified if they elicited a nuclear attack, in accord with contingency plans—which exist—for first strike, even preemptive strike against nonnuclear states that have signed the nonproliferation treaty. An Iranian missile attack on Israel with a credible invasion threat would be justified if Israel responded by implementing its detailed contingency plans—which presumably exist—for expelling the Palestinian population.

The OSCE inquiry reports further that “Once the OSCE-KVM [monitors] left on 20 March 1999 and in particular after the start of the NATO bombing of the FRY on 24 March, Serbian police and/or VJ [army], often accompanied by paramilitaries, went from village to village and, in the towns, from area to area threatening and expelling the Kosovo Albanian population.” The departure of the monitors also precipitated an increase in KLA-UCK ambushes of Serbian police officers, “provoking a strong reaction” by police, an escalation from “the prewar atmosphere, when Serbian forces were facing off against the rebels, who were kidnapping Serbian civilians and ambushing police officers and soldiers.”

For understanding of NATO’s resort to war, the most important period is the months leading up to the decision. Of course, what NATO knew about that period is a matter of critical significance for any serious attempt to evaluate the decision to bomb Yugoslavia without Security Council authorization. Fortunately, that is the period for which we have the most detailed direct evidence: namely, from the reports of the KVM monitors and other international observers. Unfortunately, the OSCE inquiry passes over these months quickly, presenting little evidence and concentrating rather on the period after monitors were withdrawn. A selection of KVM reports is, however, available, along with others by NATO and independent international observers. These merit close scrutiny.

The relevant period begins in December, with the breakdown of the cease-fire that had permitted the return of many people displaced by the fighting. Throughout these months, the monitors report that “humanitarian agencies in general have unhindered access to all areas of Kosovo,” with occasional harassment from Serb security forces and KLA paramilitaries, so the information may be presumed to be fairly comprehensive.

The “most serious incidents” reported by the ICRC in December are clashes along the FRY-Albanian border, and “what appear to be the first deliberate attacks on public places in urban areas.” The UN Inter-Agency Update (December 24) identifies these as an attempt by armed Albanians to cross into Kosovo from Albania, leaving at least 36 armed men dead, and the killing of 6 Serbian teenagers by masked men spraying gunfire in a cafe in the largely Serbian city of Pec. The next incident is the abduction and murder of the deputy mayor of Kosovo Polie, attributed by NATO to the KLA-UCK. Then follows a report of “abductions attributed to the KLA.” The UN Secretary-General’s report (December 24) reviews the same evidence, citing the figure of 282 civilians and police abducted by the KLA as of December 7 (FRY figures). The general picture is that after the October cease-fire, “Kosovo Albanian paramilitary units have taken advantage of the lull in the fighting to re-establish their control over many villages in Kosovo, as well as over some areas near urban centres and highways,...leading to statements [by Serbian authorities] that if the [KVM] cannot control these units the Government would.”

The UN Inter-Agency Update on January 11 is similar. It reports fighting between Serb security forces and the KLA. In addition, in “the most serious incident since the declaration of the ceasefire in October 1998, the period under review has witnessed an increase in the number of murders (allegedly perpetrated by the KLA), which have prompted vigorous retaliatory action by government security forces.” “Random violence” killed 21 people in the preceding 11 days. Only one example is cited: a bomb outside “a cafe in Pristina, injuring three Serbian youths and triggering retaliatory attacks by Serbian civilians on Albanians,” the first such incident in the capital. The other major incidents cited are KLA capture of eight soldiers, the killing of a Serbian civilian, and the reported killing of three Serbian police. NATO’s review of the period is similar, with further details: VJ shelling of civilian and UCK facilities with “at least 15 Kosovo Albanians” killed, UCK killing of a Serb judge, police and civilians, etc.

Then comes the Racak massacre of January 15, after which the reports return pretty much to what preceded. The OSCE monthly Report of February 20 describes the situation as “volatile.” Serb-KLA “direct military engagement...dropped significantly,” but KLA attacks on police and “sporadic exchange of gunfire” continued, “including at times the use of heavy weapons by the VJ.” The “main feature of the last part of the reporting period has been an alarming increase in urban terrorism with a series of indiscriminate bombing or raking gunfire attacks against civilians in public places in towns throughout Kosovo”; these are “non-attributable,” either “criminally or politically motivated.” Then follows a review of police-KLA confrontations, KLA abduction of “five elderly Serb civilians,” and refusal of KLA and VJ to comply with Security Council resolutions. Five civilians were killed as “urban violence increased significantly,” including three killed by a bomb outside an Albanian grocery store. “More reports were received of the KLA ‘policing’ the Albanian community and administering punishments to those charged as collaborators with the Serbs,” also murder and abduction of alleged Albanian collaborators and Serb police. The “cycle of confrontation can be generally described” as KLA attacks on Serb police and civilians, “a disproportionate response by the FRY authorities,” and “renewed KLA activity elsewhere.”

In his monthly report, March 17, the UN Secretary-General reports that clashes between Serb security forces and the KLA “continued at a relatively lower level,” but civilians “are increasingly becoming the main target of violent acts,” including killings, executions, mistreatment, and abductions. The UNHCR “registered more than 65 violent deaths” of Albanian and Serb civilians (and several Roma) from January 20 to March 17. These are reported to be isolated killings by gunmen and grenade attacks on cafes and shops. Victims included alleged Albanian collaborators and “civilians known for open-mindedness and flexibility in community relations.” Abductions continued, the victims almost all Serbs, mostly civilians. The OSCE report of March 20 gave a similar picture, reporting “unprovoked attacks by the KLA against the police” and an increase in casualties among Serb security forces, along with “Military operations affecting the civilian population,” “Indiscriminate urban terrorist attacks targeting civilians,” “non-attributable murders,” mostly Albanians, and abduction of Albanian civilians, allegedly by a “centrally-controlled” KLA “security force.” Specific incidents are then reported.

The last NATO report (January 16–March 22) cites several dozen incidents, about half initiated by KLA-UCK, half by Serb security forces, in addition to half a dozen responses by Serb security forces and engagements with the KLA, including “Aggressive Serb attacks on villages suspected of harbouring UCK forces or command centres.” Casualties reported are mostly military, at the levels of the preceding months.

As a standard of comparison, one might consider the regular murderous and destructive U.S.-backed Israeli military operations in Lebanon when Israeli forces occupying southern Lebanon in violation of Security Council orders, or their local mercenaries, are attacked by the Lebanese resistance. Through the 1990s, as before, these have far exceeded anything attributed to the FRY security forces within what NATO insists is their territory.

Within Kosovo, no significant changes are reported from the breakdown of the cease-fire in December until the March 22 decision to bomb. Even apart from the (apparently isolated) Racak massacre, there can be no doubt that the FRY authorities and security forces were responsible for serious crimes. But the reported record also lends no credibility to the claim that these were the reason for the bombing; in the case of comparable or much worse atrocities during the same period, the U.S. and its allies either did not react, or—more significantly—maintained and even increased their support for the atrocities. Examples are all too easy to enumerate, East Timor in the same months, to mention only the most obvious one.

The vast expulsions from Kosovo began immediately after the March 24 bombing campaign. On March 27, the United Nations High Commissioner for Refugees (UNHCR) reported that 4,000 had fled Kosovo, and on April 1, the flow was high enough for UNHCR to begin to provide daily figures. Its Humanitarian Evacuation Programme began on April 5. From the last week of March to the end of the war in June, “forces of the FRY and Serbia forcibly expelled some 863,000 Kosovo Albanians from Kosovo,” the OSCE reports, and hundreds of thousands of others were internally displaced, while unknown numbers of Serbs, Gypsies, and others fled as well.

The U.S. and UK had been planning the bombing campaign for many months, and could hardly have failed to anticipate these consequences. In early March, Italian Prime Minister Massimo D’Alema warned Clinton of the huge refugee flow that would follow the bombing; Clinton’s National Security Adviser Sandy Berger responded that in that case “NATO will keep bombing,” with still more horrific results. U.S. intelligence also warned that there would be “a virtual explosion of refugees” and a campaign of ethnic cleansing, reiterating earlier predictions of European monitors.

As the bombing campaign began, U.S.-NATO Commanding General Wesley Clark informed the press that it was “entirely predictable” that Serb terror would intensify as a result. Shortly after, Clark explained again that “The military authorities fully anticipated the vicious approach that Milosevic would adopt, as well as the terrible efficiency with which he would carry it out.” Elaborating a few weeks later, he observed that the NATO operation planned by “the political leadership...was not designed as a means of blocking Serb ethnic cleansing. It was not designed as a means of waging war against the Serb and MUP [internal police] forces in Kosovo. Not in any way. There was never any intent to do that. That was not the idea.” General Clark stated further that plans for Operation Horseshoe “have never been shared with me,” referring to the alleged Serb plan to expel the population that was publicized by NATO after the shocking Serb reaction to the bombing had become evident.

The agency that bears primary responsibility for care of refugees is UNHCR. “At the war’s end, British Prime Minister Tony Blair privately took the agency to task for what he considered its problematic performance.” Evidently, the performance of UNHCR would have been less problematic had the agency not been defunded by the great powers. For this reason, the UNHCR had to cut staff by over 15 percent in 1998. In October, while the bombing plans were being formulated, the UNHCR announced that it would have to eliminate a fifth of its remaining staff by January 1999 because of the budgetary crisis created by the “enlightened states.”

In summary, the KVM monitors were removed and a bombing campaign initiated with the expectation, quickly fulfilled, that the consequence would be a sharp escalation of ethnic cleansing and other atrocities, after the organization responsible for care of refugees was defunded. Under the doctrine of retrospective justification, the heinous crimes that ensued are now held to be, perhaps, “enough to justify” the NATO bombing campaign.

The person who commits a crime bears the primary responsibility for it; those who incite him, anticipating the consequences, bear secondary responsibility, which only mounts if they act to increase the suffering of the victims. The only possible argument for action to incite the crimes is that they would have been even more severe had the action not been undertaken. That claim, one of the most remarkable in the history of support for state violence, requires substantial evidence. In the present case, one will seek evidence in vain—even recognition that it is required.

Suppose, nevertheless, that we take the argument seriously. It plainly loses force to the extent that the subsequent crimes are great. If no Kosovar Albanians had suffered as a result of the NATO bombing campaign, the decision to bomb might be justified on the grounds that crimes against them were deterred. The force of the argument diminishes as the scale of the crimes increases. It is, therefore, rather curious that supporters of the bombing seek to portray the worst possible picture of the crimes for which they share responsibility; the opposite should be the case. The odd stance presumably reflects the success in instilling the doctrine that the crimes incited by the NATO bombing provide retrospective justification for it.

This is by no means the only impressive feat of doctrinal management. Another is the debate over NATO’s alleged “double standards,” revealed by its “looking away” from other humanitarian crises, or “doing too little” to prevent them. Participants in the debate must be agreeing that NATO was guided by humanitarian principles in Kosovo— precisely the question at issue. That aside, the Clinton administration did not “look away” or “do too little” in the face of atrocities in East Timor, or Colombia, or many other places. Rather, along with its allies, it chose to escalate the atrocities, often vigorously and decisively. Perhaps the case of Turkey—within NATO and under European jurisdiction—is the most relevant in the present connection. Its ethnic cleansing operations and other crimes, enormous in scale, were carried out with a huge flow of military aid from the Clinton administration, increasing as atrocities mounted. They have also virtually disappeared from history. There was no mention of them at the 50th anniversary meeting of NATO in April 1999, held under the shadow of ethnic cleansing—a crime that cannot be tolerated, participants and commentators declaimed, near the borders of NATO; only within its borders, where the crimes are to be expedited. With rare exceptions, the press has kept to occasional apologetics, though the participation of Turkish forces in the Kosovo campaign was highly praised. More recent debate over the problems of “humanitarian intervention” evades the crucial U.S. role in the Turkish atrocities, or ignores the topic altogether.

It is a rare achievement for a propaganda system to have its doctrines adopted as the very presuppositions of debate. These are among the “lessons learned,” to be applied in future exercises cloaked in humanitarian intent.                                     Z

Afterword to the French translation of New Military Humanism (Common Courage, 1999; Page Deux Lausanne, 2000.)