Thomas de Waal belongs to a very special order of journalists, the small corps of Western reporters who have covered events in the Caucasus over the last ten or twelve years—in Georgia, Armenia, and Azerbaijan, in Dagestan and in Chechnya. Some of them lost their lives. Others who come to mind—Thomas Goltz, Suzanne Goldenberg, Carlotta Gall, Wendell Steavenson, and Vanora Bennett—have all written books about the place in English, but have in most cases gone on to other lands and careers.

But one experience marks them all. They are privileged to have lived in something like the world of Shakespeare. Characters from his plays infest the Caucasus: heroes of stainless nobility and courage, villains steeped in treachery and cruelty, clowns and conquerors, fools in love with their fantasies, creatures from the pit who hobble out of the darkness to the throne. No wonder that the Rustaveli Theater in Tbilisi, Georgia, has mounted some of the most powerful productions of the tragedies and histories.

Traveling in the region, you may find a brooding Armenian history professor whose ambitions make you think of Richard III. Some Special Forces colonel in Baku, capital of Azerbaijan, can resemble a new Macbeth, waiting impatiently for his cue. And mad King Lear, driven from his kingdom by scheming and ingratitude? Today, he reappears as a demented refugee from Nagorny Karabakh—the mainly Armenian Christian enclave in the midst of mainly Muslim Azerbaijan. Wandering in the mountains, he raves about days when he was mayor of a happy place, a city on a hill where Christians and Muslims, Armenians and Azeris, once lived together as friends.

The war over Nagorny Karabakh began—or, more accurately, the killing and expelling of Armenians and Azerbaijanis began—in 1988 and lasted until a cease-fire in 1994. Few people know where Nagorny Karabakh is, or why there was a war there. Other, vaster events obscured it. The collapse of Soviet communism, first in Central and Eastern Europe and then in the Soviet Union itself, the post-Yugoslav wars, and the first war against Iraq all took place in the same few years.

Nonetheless, this conflict mattered then, and matters still. It had a uniquely poisonous quality which carried infection far beyond the Caucasus itself. The Karabakh war contributed powerfully to the failure of Mikhail Gorbachev’s perestroika experiment and then to the breakup of the USSR in 1991. And the poison is still there. The conflict remains unsettled, and its horrifying legacy of misery and hatred still destabilizes the region. And the worst of it may be yet to come. In a period when great and medium powers—Russia, the United States, Turkey, and Iran—are competing for the oil wealth around the Caspian Sea, renewed fighting over Karabakh between Armenia and Azerbaijan could drag those powers into a disastrous confrontation.

“Nagorny” means “mountainous” in Russian, and “Karabakh” means roughly “black garden” in Turkish. Up to 1988, Nagorny Karabakh could be described as a hilly territory, with a largely Armenian population, assigned to the Soviet republic of Azerbaijan; it was an enclave separated on its western side from the Soviet republic of Armenia by a belt of Azerbaijani-settled territory. The Armenians are traditionally Christian and speak Armenian; the Azeris are traditionally Muslim and speak a language close to Turkish. Large Armenian minorities lived in Azerbaijan, especially in its capital Baku on the Caspian shore, while large Azeri minorities lived in Armenia. Even the population of Nagorny Karabakh was mixed. The town of Stepanakert was mainly Armenian; the old hilltop city of Shusha was mainly Azeri.

Was this, then, a happy multicultural community, before the wicked fairy of nationalism arrived to set neighbor against neighbor? It was nothing so simple. The Caucasus has always been a region of diversity, often harmonious but sometimes lethal. The Western journalists who worked there in the 1990s—Thomas de Waal among them—found these complexities summed up in an extraordinary novel which they passed from hand to hand. Ali and Nino by “Kurban Said” (apparently born Lev Nussimbaum in Baku) describes the love of a proud Azeri boy for a Georgian girl in the turmoils of the early-twentieth-century Caucasus. Their passion transcends their differences, and yet the novel is also full of ironic stereotypes of this or that ethnicity.*

Perhaps the truth is this: that the Armenians and Azeris of Karabakh never fully trusted one another and yet each could not imagine life without the other. They could often speak both languages. They sat on the same school bench. Their mothers talked intimately as they waited for water at the same pump. They could form close and loyal friendships. And yet, somewhere, there was always a sense of otherness, of alien things which the neighbors, as a group, might be doing or saying behind closed shutters or in their shrines.


A pessimistic law of history was at work here. Many such mixed communities coexisted for centuries, not just in the Caucasus but throughout Eurasia and North Africa. And yet they were, in reality, only held together by fear—the fear of what a brutal outside authority would do to them all if mutual tolerance broke down. When the external pressure was removed—whether it was the Caliphate, the Tsardom, the Ottoman or British Empire, or Soviet power—then the current of fear which enforced that mutual tolerance was switched off. In the condition of “freedom,” people began to look at one another in a new, warier way. In the condition of “democracy,” people were invited to think about what divided them rather than what united them. And ethnicity—rather than wealth, class, or social function—was the only dividing category which came to mind.

This sad dynamic was bound to start operating in Nagorny Karabakh as the Soviet Union began to disintegrate. And yet there was nothing inevitable about the war that followed. There had been violence between Armenians and Azerbaijanis in 1905, at the time of the first revolution against the Tsardom, and then again in the interval following the Bolshevik revolution, during which Armenia and Azerbaijan won a short-lived independence. Their freedom was ended by the Bolshevik conquest in 1920. Joseph Stalin became commissar for nationalities, and in drawing up the frontiers for the new Soviet republics, he at first assigned Nagorny Karabakh to Armenia. But then, apparently provoked by a vain rebellion in Armenia, he made the fatal decision to award it as an “autonomous region” to Azerbaijan. In this sense, Stalin—a Georgian who understood the tensions of the Caucasus all too well—was directly responsible for the tragedy that happened almost a century later.

But the long Soviet era that ensued, though maintained by police terror, was a time of peace in the Caucasus and of ambitious economic development. As de Waal writes, “for seventy years there was almost no instance of mass violence between Armenians and Azerbaijanis. They lived side by side, traded with each other and intermarried.”

Reason—the rational calculus of what can be gained or lost—is not a helpful instrument for understanding why wars begin. But this war was utterly unreasonable. Its consequences were piteous and unnecessary. Between 17,000 and 25,000 people died, in battle or massacres. Almost all the 350,000 Armenians who lived in Azerbaijan were expelled or fled. Something like 750,000 Azerbaijanis—the 200,000 who had lived in Armenia, and the 500,000 who had lived in Nagorny Karabakh or in the other areas of Azerbaijan captured and occupied by Armenian forces—became refugees.

Today, nine years after the cease-fire, most of the “displaced” remain helpless and excluded, living in archipelagos of hostels, shacks, or tented camps, or squatting in commandeered public buildings. I remember speaking to an Armenian woman, once a prosperous intellectual in Baku, now living with her family in a single room in the Armenian capital of Yerevan. “I have no hope at all,” she said evenly. “My life will never change.” She was luckier than another refugee woman I met in the Armenian mountains, facing winter in a cabin of loose stones with no light and no heat except a fire of moist pieces of turf. Two of her children had already died of hunger, and the third was sick. Such are the conditions on the side of those who had supposedly won the war. The misery of the Azerbaijani refugees is worse.

In the prelude to this conflict, there are echoes of other twentieth-century disasters. Modern Armenia was born after a holocaust, the slaughter of a million people by the Turks in 1915, and in the early Soviet period Armenia had a “Zionist,” pioneering feeling about it, supported—like Israel—by an enthusiastic world diaspora. There is another parallel. The wars during the breakup of Yugoslavia became inevitable when Catholic Croatia gained independence without any convincing guarantee for its large Serb-Orthodox minority concentrated in Slavonia. For Slavonia, read Nagorny Karabakh. For Serbia, not lacking its own national victimology, read Armenia. As the enormous and ramshackle Soviet state began to weaken, the Armenians—haunted by their national memory of genocide—grew obsessed with the fate of their ethnic brothers and sisters “trapped” by “Turks” within what might soon become an independent Azerbaijan.

For decades, de Waal tells us, Armenia had petitioned Moscow to detach Karabakh from Azerbaijan. Moscow took no notice. It was not until the late 1980s, as Gorbachev’s reforms demoralized the Soviet power structure, that popular demands for democratic change merged with nationalist agitation. In February 1988, the local “Soviet” of Nagorny Karabakh dared to demand union with Armenia. Vast demonstrations began in Yerevan, where a million people met in Theater Square; a festival of new freedom which turned into a roaring demand for union with Karabakh.


Then came a fatal event which sealed off the chances of any reasonable bargain. Azerbaijan was also seething with nationalist excitement, and in the last days of February the mob broke loose in the Azerbaijan town of Sumgait, near the capital city of Baku. Almost all the 14,000 Armenians in the town fled, and something like thirty were murdered (most Armenians believe the total was far higher, but de Waal’s sober account is credible). The Sumgait pogrom locked the Armenians into their inherited expectation of genocide. Now, they concluded, the future of the Karabakh Armenians was being written on the wall in blood. They began to organize for defense, in Armenia and in Karabakh, and that fall the first attacks between the two peoples began. In Karabakh, the Azeris were chased out of the town of Stepanakert, and the Armenians out of the town of Shusha. A few months later, Armenians launched bloody onslaughts against Azeri minority villages within Armenia itself.

Nothing better illustrates Mikhail Gorbachev’s limitations than his utter failure to grasp the nature of what was happening in the Caucasus. When the Sumgait pogrom began, he issued an order to “send the working class… people’s volunteers into the fight with the criminals. That, I can tell you, will stop any hooligans and extremists.” And when he was told about the ethnic expulsions on both sides spreading across Karabakh, he could only threaten: “We will expel them from the Party!” Gorbachev was living in a sunken capsule of illusion, a Soviet dream world in which Party authority was still absolute and nationalism could not exist. And yet there were moments when he could seem to be the only sane person in the asylum. On December 7, 1988, a great Armenian earthquake killed almost 25,000 human beings (probably more, as de Waal observes, than the entire six years of the Nagorny Karabakh war). Gorbachev at once flew to visit the wrecked cities, only to be greeted in the ruins by screams of “Karabakh!” Furious, he ordered the immediate arrest of the “Karabakh Committee” which was by then running Armenian politics. That was a blunder, but it is hard not to feel a spark of sympathy for his outrage.

The Karabakh Committee, the radical nationalist Azeri “Popular Front” in Azerbaijan, and the Armenian militant groups in Karabakh now drove the region toward war. There were more huge patriotic rallies in Azerbaijan in late 1989, culminating in the “Black January” pogroms in Baku in the new year: ninety Armenians died, and the rest fled on ships across the Caspian Sea. Too late, Gorbachev sent in the army. There were heavy civilian casualties, and the Azeris were confirmed in their suspicion that Moscow had decided to take the Armenian side in the conflict. Soon the Russians and Jews of Baku were fleeing in the wake of the Armenians. They were replaced by wave after wave of frantic, destitute Azeri refugees driven from their homes in Karabakh and the border regions.

By now, weapons were pouring into both countries as communal fighting between partisan bands spread. A clumsy campaign by Soviet armed forces against Armenian guerrillas ended when the failed Moscow putsch of August 1991 paralyzed the military command. From then on the Soviet troops did little more than sell their tanks, guns, and even their skills as mercenary soldiers to the highest local bidder. Azerbaijan declared independence at the end of that month, and Armenia chose independence in a referendum held a few weeks later. The war became an all-out conflict. In Karabakh, the “capital,” Stepanakert, was under artillery and rocket bombardment all through the spring of 1992, until the Armenian forces relieved the siege by storming the old city of Shusha in May. A corridor across Azerbaijani territory was conquered, opening a direct connection between Armenia and Karabakh for weapons and troop reinforcements, although Armenia never formally admitted sending men to fight beyond its old frontiers.

De Waal recounts in detail the advances, retreats, and failed truces of the war. The Armenians were better trained and more determined, although Armenia itself had to endure years of virtual blockade without electricity, railways, or gas, supported only by a trickle of trade across the border with Iran. The Azerbaijani forces were much larger than their enemy: some 100,000 by the end of the war, against about 35,000 Armenians. But they were weakened by recurrent political crises and attempted military coups at home, each of which made it necessary for troops to be brought back from the front line to patrol the streets of Baku. In spite of this, the Azeri army was able to launch a final offensive in December 1993, opening the last and bloodiest phase of the war, which nearly recaptured the territory they had lost outside Karabakh. A successful Armenian counteroffensive, at the cost of thousands of lives on both sides, brought the war to a close with the cease-fire of May 1994. The truce line, still almost impassable, leaves the Armenians in possession not only of Nagorny Karabakh, but of large tracts of other Azerbaijani provinces. It also leaves nearly a million refugees in possession of nothing except their own unhappiness.

This was a savage war between neighbors. By day, the soldiers burned villages, fought battles, or hunted and sometimes massacred civilians. By night, they shouted across the lines for news of families and friends, and traded vodka for bread. Shusha, a sort of Jerusalem which both sides regarded as the cradle of much of their cultures, remains half-ruined. The great bell of its Armenian church was later found in the Ukrainian city of Donetsk; the busts of its Azeri musicians and poets turned up in a Georgian scrapyard. The city of Aghdam used to have 50,000 inhabitants, mostly Azeri. But when de Waal went to see it, he discovered only broken walls covered with thistles and brambles: “a small Hiroshima…. Now it is completely empty.”

De Waal, a wise and patient reporter, traveled throughout the region and talked to hundreds of people in his effort to discover why this disaster happened. He rejects three easy explanations. First, this was not an inevitable climax to “ancient hatreds,” which scarcely existed. Neither was it triggered by the top-down politics of a few extremists (“the fire began below,” as he puts it, and “many ordinary people must take their responsibility for the bloodshed”). Finally, this was no war fought for socioeconomic motives. Everyone lost by it. Azerbaijan was ruined (the so-called “oil wealth” of recent years stays in a few wealthy hands in Baku, and has done little or nothing to help the refugees). Armenia, though technically the victor, landed itself in deepening isolation and poverty. Nagorny Karabakh itself remains an unvisited statelet, claiming an independence recognized by nobody in the outside world.

De Waal concludes:

The Nagorny Karabakh conflict makes sense only if we acknowledge that hundreds of thousands of Armenians and Azerbaijanis were driven to act by passionately held ideas about history, identity, and rights.

These ideas were driven on both sides by fanatical national egoism, which gave no place whatever to the thought that the other side might also have legitimate identity or rights. De Waal interviewed Igor Muradian, the man who more than any other was responsible for focusing Armenian passions on Nagorny Karabakh in the years leading up to 1988, and asked him whether he had consulted or even taken into consideration the 40,000 Azerbaijanis whose homes were in Karabakh. He replied: “I will tell you the truth. We weren’t interested in the fate of those people. Those people were the instruments of power, instruments of violence over us for many decades, many centuries even. We weren’t interested in their fate and we’re not interested now.”


As far as history goes, recent and ancient, the two nations are now locked into totally incompatible versions of the past. Nine and a half years after the end of hostilities, there still is no agreement over the political future of Nagorny Karabakh. The October presidential election in Azerbaijan, which ensured a dynastic succession from the ailing leader Heydar Aliev to his son Ilham Aliev, seemed to confirm the uneasy status quo. During peace talks in Paris and Key West in 2001, the two countries made progress toward a settlement, but no final agreement was reached. Some observers hope that Ilham Aliev, who was less antagonistic toward Armenia than opposition candidates during the campaign, may provide new impetus for the talks. But Ilham Aliev made belligerent statements about Armenia during a recent session of the UN General Assembly, and there is as yet little indication that the two sides can overcome their differences.

Armenians often attribute the war to an Azerbaijani plot to extinguish the Armenian people, ultimately with the help of the Turks. Azerbaijanis think that it was the planned first step in a program of Armenian imperial expansion, covertly supported by the Soviet Union and then Russia, which aimed to break up Azerbaijan and obliterate Islamic power in the whole Caucasus region. As almost always happens in such arguments, history and archaeology are debauched in order to provide evidence of priority—“we were there first, but they are colonizing interlopers.”

There is a popular belief in Azerbaijan, unsupported by fact, that the Armenians were brought from Iran and deliberately settled in the Caucasus by Russia in the nineteenth century. In a fascinating chapter, de Waal examines the way Azeri intellectuals spin the “Albanian” myth. The Caucasus Albanians (not to be confused with the small nation on the Adriatic coast) were an obscure people who appear to have lived on the Caspian shores and who vanished from history around the tenth century AD. But modern Azeri scholars now claim that Albania covered the entire area which is now Armenia and Karabakh, and that the thousands of medieval Armenian inscriptions all over the region are nineteenth-century forgeries created in a campaign to conceal the memory of “Albanian civilization.”

And yet this mutual exclusion is not the whole story. Again and again, de Waal found that pitiless group rejection could live side by side with personal affection and understanding:

…The lines of division run straight through the middle of people. Hateful impulses coexist with conciliatory feelings in the same person. Armenians and Azerbaijanis…are torn between aggression and conciliation, personal friendships, and the power of national myths.

A bellicose Azeri politician boomed at de Waal about the coming reconquest of Karabakh and then, as the meeting ended, asked him to send his best regards to an old friend in Yerevan who was chief of the Armenian general staff. The Armenian minister of defense argued that it had been correct to expel the Azeri minority from Armenia, but then spoke “with real affection” of his old friends in the other nation—with whom he used to speak Azeri. In Baku and Yerevan, refugees loaded de Waal with news and fond messages for beloved friends and neighbors they had been forced to leave behind. Many hundreds of Armenians owe their lives to Azerbaijanis who rescued or hid them and their families during the Baku pogroms of January 1990.

This war was a tragedy that need never have happened. It was not a crime committed by one side against the other, and the weakness of Armenia: Portraits of Survival and Hope is precisely that it understands the Azerbaijanis only as aggressors and the Armenians only as victims. Two Christian academics from California, with Armenian family roots, Donald E. Miller and Lorna Touryan Miller, carried out an elaborate and expensive oral history survey in Armenia, recording memories of the earthquake and the Karabakh war. What their interviews show is that the Armenians are a unique and marvelously resilient people who have endured generations of injustice and physical suffering. What they do not show is that the Armenians carried any more responsibility for the war than they did for the earthquake.

Donald and Lorna Miller are obviously sensitive and compassionate people, and the pain they feel for the terrible stories they have heard does them credit. They take care to highlight instances of Azerbaijani humanity to persecuted Armenians, especially during the Sumgait and Baku massacres. But they swallow whole the version that the Azeris launched an unprovoked campaign of ethnic cleansing and extermination against innocent neighbors. An unprepared reader of this book would conclude that all the atrocities in the six-year conflict were committed by Azerbaijanis. This is simply untrue, as de Waal makes clear. To take one instance, something over four hundred people—most of them Azerbaijani civilian families—died at Khojali in 1992 when a crowd of fugitives was “hit by a wall of gunfire from Armenian fighters.” Several senior Armenians talked to de Waal about Khojali, and expressed their unease about it. But the Millers’ long questionnaire invited Armenians to recall only their own sufferings. Their book, in the end, is martyrology rather than history.

In retrospect, the Nagorny Karabakh war formed part of a specific episode in the history of nationalism. The collapse of the Soviet empire produced a “springtime of nations,” not unlike those of 1848 or during the years between 1918 and 1920. Suppressed nationalities reclaimed independence and statehood, or asserted it for the first time. But at the same moment—and this often happened in those earlier upheavals—the removal of oppressive external authority transformed the nature of ethnic feelings in many old multicultural societies. Passive distaste for the neighbors suddenly became dynamic and exclusive. Previously tolerant people discovered that “we” cannot share our town or our land with “them.”

Why the sense of liberation can lead so easily into paranoid xenophobia is a question so far unanswered. But one result of this period (which is far from over) has been violent ethnic separation—the statesman’s euphemism is “exchange of populations”—as minorities in many parts of the world are driven from their homes. In the last ten years, this has happened from the Balkans to the Caucasus and on into the Middle East, Iraq, in Africa, and in southern Asia. In Europe and western Asia, the same process has generated a string of mono-ethnic statelets and polities, often bankrupt and chaotic and frequently unrecognized by their neighbors: they include Kosovo, the republics which form Bosnia-Herzegovina, Abkhazia, Nagorny Karabakh, Kurdistan in northern Iraq, and one day, perhaps, Chechnya.

Thomas de Waal begins his admirable, rigorous book with a plea. He begs his Armenian and Azerbaijani readers not to quote it selectively, to suit their own political agendas. That is too much to hope for. But perhaps they will also remember some of the surreal scenes he records here, which convey the futility of all that happened. One such scene compresses all the war’s pathos into a single image:

An Armenian friend described to me how he went to the ravaged city of Aghdam…and saw the Felliniesque sight of men filling a line of flat-top Iranian trucks to the brim with rose petals. The petals came from the thousands of rosebushes scrambling over the ruins of the deserted town, and the Iranians bought them to make jam.

This Issue

November 20, 2003