Hot wars demand attention before cold wars or simmering, potential conflicts. But even as most of the energy of envoys and emissaries concentrates on the Balkans, it may be opportune to look at a crisis less violent at the moment but equally dangerous in the long run. The Armenian-Azerbaijani war over Karabakh, now in cease-fire but without solution, has gone on for more than a decade. The problems of extracting enormous oil and gas resources in the Caspian region and piping fuels through the South Caucasus and possibly Turkey to Western markets has encouraged new diplomatic efforts and visits to the region by prominent veterans of international negotiations like Zbigniew Brzezinski and Henry Kissinger. Less visibly, diplomats and peace activists have worked to untie the knot that has kept armies poised and hundreds of thousands of refugees languishing in camps. Negotiators have concluded that without resolving the Karabakh issue the region's security and economic development may be permanently threatened. Our conviction is that there is a way out; indeed, a solution is at hand. But thinking a way out of the Karabakh conflict requires rethinking some conventional notions of nationalism and ethnic conflict. Separation of antagonistic peoples may be necessary, at least in the short run, but a lasting solution also requires the building of links between those peoples who, after all, will be living next to one another in future centuries.
It is commonly held, and reported by journalists, that the hatred between Azerbaijan is and Armenians is of ancient, even tribal, origin and that, precisely because of its antiquity and persistence, the conflict defies easy solution. Little historical evidence supports this view. At the same time, a few historians have argued that the hostilities are actually about 70 years old, a resurgence of a suppressed struggle that flared up periodically, most violently in the last period of Russian state collapse, 1918-19. We argue that long-term antagonisms (and cooperation) between the Christian and Muslim peoples of the South Caucasus, stemming from the distant past, have only weak links to the contemporary conflict. Indeed, one might say that the origins of the current conflict are "shrouded in the mists of the twentieth century." Profoundly shaped by the 70-year experience of Soviet rule and the larger global context of twentieth-century nationalism, the war between Armenians and Azerbaijanis can only be resolved when local perspectives and interests that derive from the experiences of Soviet rule and nation formation are put in bold relief.
Three contemporary contexts frame the outbreak and persistence of the Karabakh war: the processes in which modern nations have been made, the specific form of nation-making that took place within the Soviet Union, and the dynamics of the Soviet collapse. First, Armenia and Azerbaijan live in a world in which nation-states, in order to be legitimate, are required to represent a cultural community of people who believe that their shared characteristics entitle them to sovereignty in their historic "homeland." The modern discourse of the nation confers upon national communities the right to political control over the specific territories that they inhabit, as well as those contested (like Karabakh with its overwhelming majority of ethnic Armenians, and Nakhichevan with its overwhelming majority of ethnic Muslims). This discourse is based in a narrative of the nation's antiquity and its people's (nearly) continuous presence in a historic "homeland." Even as it proposes rights and justice for oppressed nationalities, in fact the national discourse creates new problems of making political and cultural boundaries commensurate as Woodrow Wilson learned at the end of World War I. This is especially true in the Caucasus, where much of its history has been one of migration, intermingling of different religious and linguistic groups, not to mention overlapping polities and contested sovereignties from ancient to modern times. Yet nationalists persistently draw harder and clearer boundaries between their own people and those living closest to them (who share much of each other's culture), to obscure distinctions within their own nation and to exaggerate differences with their neighbors. For example, Patrick Donabedian, a French diplomat in Erevan, quotes the Greek geographer Strabo, who attests that by the second century BC the entire population of Greater Armenia ( including today's Karabakh) spoke Armenian, implying that today's Armenians are the direct descendants of those speakers.1 On the other side, A. Abbasov and A. Memedov of the Azerbaijan Academy of Sciences write that the early settlers were Caucasian Albanian tribes, precursors of today's Azeris, and that the Armenians, unlike most of the other minorities, do not have a long history in Azerbaijan, even in Karabakh.2 They too select cultural specks in the past in order to write an exclusive, continuous national history.
But analysts err when they reproduce the cultural geographies of nationalists and attempt to separate surgically the histories and claims of one people from another. Those with short historical horizons miss some important features of ethno-religious communities of the past that distinguish them from nations in modern times. In earlier centuries, the differences between ethnic and religious communities were less sharp. Rather than separate and discontinuous, ethnic groups shared many cultural features of their neighbors; the edges of their differences were blurred; and it required hard work by scholars and activists, journalists and teachers, statesmen and warriors through many centuries to sharpen differences between groups and homogenize distinctions within groups. If the boundary between us and them were not maintained, modem nationalists worry, our ethnic group would disappear (as many did), assimilating into other nearby ethnicities. This is, indeed, what happened to the ancient Caucasian Albanians, many of whom adopted Armenian Christianity, eventually identifying with Armenians, while others adopted Islam and eventually merged with Azerbaijan is. To avoid disappearance, which some refer to as "white genocide," representatives of ethnic groups ethnic entrepreneurs police ethnic boundaries, define acceptable cultural features, and sanction cross-overs.
For Armenians and Azerbaijanis, two kinds of "policing" go on. One raises the barriers to mixing, primarily by drawing sharp lines between "us" and the "other." Indeed, much of Armenian identity is wrapped up in what they have suffered at the hands of the Turks, and since the Azerbaijanis are "Turks" (Azeri is a Turkic language), hostility felt toward one people is transferred to another. To an Armenian nationalist standing guard at the-psychic boundaries of the nation, intermarriage with an Azerbaijani, which was something which happened occasionally in relatively cosmopolitan Baku in Soviet times, would be anathema, a betrayal of what it means to be an Armenian. Since Armenians and Azerbaijan is today are not particularly religious, given the long Soviet experience, the objection to intermarriage is not a religious proscription but based on a firm commitment to sharpen ethnic boundaries.
This leads to the second kind of policing: the enforcement within each group of what behavior is permissible and impermissible and the rules set down for which Armenians or Azerbaijan is are the true Armenians or the real Azerbaijan is. "Ethnic identity implies a series of constraints on the kinds of roles an individual is allowed to play," writes Norwegian ethnologist Frederik Barth, "and the partners he may choose for different kinds of transactions ns."3 Thus, Armenians who deny the fundamental historical role of Karabakh in Armenian national history are not coded as dissident; rather they are held to be traitors to the Armenian nation. Similarly, Azeris police themselves so that speaking openly about Azerbaijani "murders" of Armenians in Sumgait (February 1989) or in Baku (January 1990) would be heavily sanctioned. Authentic Azerbaijan is must blame the Russians, Gorbachev and the Communists, or the Armenians themselves.
Armenians and Azerbaijan is, like other ethnic nations with relatively exclusionary ideas of what constitutes the nation, have come out of broader, more inclusive communities like Christendom or Islam or "the Soviet people" or "Caucasian civilization" toward narrower, more exclusive communities characterized by ethnic nationalism. The international community recognizes the right of polities based on ethno-nations to sovereign statehood, thereby giving extraordinary political power to groups that manage to be recognized as nations. Yet the international community does not recognize that every ethnic group must be granted its own state or even that self-determination requires independent statehood. The claims to Karabakh, whether by demographically dominant Armenians or by the Azerbaijanis, in whose republic the region lies, only make sense in a political universe in which culture and history are given the opportunity to claim territory and statehood.
The second context in which the Karabakh conflict arose was the peculiar legacy of Soviet nationality policies. For much of the Soviet period the conventional wisdom of most Western writers on Soviet treatment of non-Russians was that Russification and repression weakened the nationalities of the USSR and made them pliable victims of totalitarian manipulation. What was largely missed in this bleak picture were the ways in which Soviet policies actually consolidated non-Russians in territorialized political units and fostered national consciousnesses. Generally overlooked by Sovietologists until the explosion of nationalism in the late 1980s under Gorbachev, this process of Soviet nation-making had several long-term effects on the post-Soviet states. First, ethnicity was matched to territory, generally imperfectly, but nevertheless a strong sense developed that each nationality ought to have its own territory, even its own polity. The well-respected U.S. State Department geographer Lee Schwartz could not, in fact, devise by computer program a set of republican boundaries that would coincide with nationality more accurately than did Stalin's henchmen, despite the significant residual minority populations that Stalin's scheme left in virtually all republics.4 Second, within those units the titular nationality (the one with its name on the unit) had certain advantages and privileges, and in actual Soviet practice (in contrast to stated goals of equal treatment) ethnic "minorities" were subordinated to the dominant nationality of the republic or region. Diaspora populations were encouraged to migrate back to their purported "homeland." Armenians steadily left Georgia and Azerbaijan for Armenia, and Azerbaijanis either migrated or were deported (most notably under Stalin in the late 1940s and again in the late 1980s, just after the outbreak of the Karabakh conflict) to Azerbaijan. Over time, Armenia, Georgia and Azerbaijan became more ethnically homogeneous, and even the "Russians" (who were actually a congeries of Russians, Ukrainians, Belorusans and Jews) gradually left the Caucasus for their "home" republic. In each republic, titular Communist elites held power, granted favors to their compatriots, systematically aggrandized power and privilege within their ranks, and limited the writ of the central Soviet state within the bounds of the republic. 5
Though nationalist expression was restricted by official policy, national loyalty and national consciousness were engendered by emphasis in educational and cultural programs on the achievements of the nation. The great irony of Soviet nationality policy was that a program that was intended to eradicate nationalism, eventually meld all the ethnicities into a single "Soviet people," and reduce the political salience of nationality, in fact embedded ethnicity into politics, granting advantages to some and disabilities to others. The durable legacy of the Soviet experience for those emerging from the grip of Soviet power was that it became almost impossible to imagine politics that was not infected by ethnicity.
The third context that led to the Karabakh conflict was the rapid collapse of the Soviet state, the resultant weakening of all state authority, and the general fragility of the post-Soviet nations. All over the territory of the former Soviet Union the first three years after the breakup were marked by interrepublic warfare (the Karabakh conflict), interethnic struggles (the Transdniestrian, Abkhaz-Georgian and Georgian-Osetian conflicts), civil war(Tajikistan and Georgia), massive refugee problems (Georgians expelled from Abkhazia, Azerbaijanis from in and around Karabakh, Armenians from Azerbaijan and other parts of the former Soviet Union, Russians from Central Asia and elsewhere), not to mention the collapse of local economies and the progressive weakening of state apparatuses. For some of the cognoscenti far from the scene, all these developments were euphemistically described as "transitions to democracy" and the building of market economies. For many closer to the pain, rather than a time of the founding of democratic institutions, the first three years appeared to be a period of destruction rather than construction, of social and political breakdown and coups d’état (sans etat), and the establishment of kleptocracies that transferred massive amounts of public property into the hands of the former nomenklatura, who sold themselves the wealth of the Soviet state at fire-sale prices.
Yet the whole story is more complicated still. By any definition this was (and continues to be) a period of revolutionary transformation. It was simultaneously a tale of the dismantling of a leviathan state and the replacement of old forms of state and economic power with the partial construction of fragile new state authorities. A chronic pathology of the rapid and chaotic collapse of the USSR was the general weakness of state authority in the post-Soviet states. The creation of modern democratic institutions based on a rule of law, a market system with protected property, enforcement of contracts, and a minimum of social order requires a competent, effective state apparatus. One might have a state without democracy or the rule of law, but the latter two are largely contingent on the first. The dilemma for post-Soviet republics has been that the old state was seen by nationalist or democratic counter elites as the major impediment to the reconstruction of the social order, and in their revolutionary fervor they accelerated the dissolution of state authority initiated by Gorbachev. The formerly ruling Communist parties, hardly conventional political parties, had been the sinews of the system, encompassing all the administrative and economic structures.
Their dissolution or removal from power both deprived the state of its disciplinary infrastructure (the loss of the "verticality of power") and at the same time left many of the old nomenklatura, with their specialized knowledge and personal affiliations, in places of influence but now without any superordinate authority. Extra-state forces - entrenched old elites, parvenu criminal mafias and more legitimate entrepreneurs - have essentially filled the space left by the retreat of the state and the dissolution of the party.
The effectiveness of the Soviet regime in blocking the emergence of alternative elites has left former Communists, stripped of their ideological baggage, among the most effective political players in most republics. In Georgia and Azerbaijan, former party bosses Edward Shevardnadze and Heidar Aliev rule their respective republics. In Armenia, anti-Communist nationalists remain in power, but the present head of state, Robert Kocharian, faces a challenge from the former Communist party chief, Karen Demirjian, who not only came in second to him in the last election but in a recent poll was overwhelmingly selected as the most popular politician in the republic. Rather than "transition to democracy," the political shift in the South Caucasus fits Jon Elster's characterization of the entire post-Communist transformation as “rebuilding the ship at sea."6 This phrase, in diagnosing that the core problem to be solved is order under chaotic conditions, suggests why democratic procedures appear to be so ineffective. Even initially well-intentioned and dedicated democrats, like independent Armenia's first president, Levon Ter Petrosian, were tempted by whatever state resources were at their disposal to enhance authority to its limits. State-building-the creation of authoritative (hopefully, not authoritarian) and legitimate states whose laws will be obeyed, taxes paid, internal security protected is an essential item on the agenda of all the post-Soviet states.
Just because independent states exist in the South Caucasus does not mean that they govern in the name of coherent, conscious nations. In the usual narrative of the fall of the Soviet Union, well-formed nations emerged from decades, if not centuries, of oppression to take the opportunity offered by Gorbachev to assert their natural, long-denied aspirations for independence and sovereignty. But most analysts of the Soviet collapse argue that the disintegration of the Soviet system was the result, at least initially, not of resurgent nationalism but of the weakness of, indeed abdication of power by, the central Soviet state. What resulted from the Soviet collapse was not the birth of fifteen fully formed nation-states but fledgling states whose only claim to legitimacy was that they were "owned" by titular nations.7 Laws on citizenship favoring the dominant nations became instruments to police the boundaries of who might be included within the national body. Georgia imploded in civil and ethnic war; Azerbaijan was fractured by the struggle with the Karabakh Armenians, faced some resistance from Lezgins in the north, and used its army to suppress a hastily formed "Talysh-Mughan Republic" in the south. In Ukraine and Russia compact populations of non-Ukrainian and non-Russian peoples put enormous pressure upon state authority. The Russian Federation faced complex threats from regions seeking ever higher levels of autonomy (e.g., Tatarstan) and fought two murderous wars against the breakaway region of Chechnya. Ukraine avoided war with non-titulars but has hardly resolved the issue of how to incorporate non-Ukrainians into a Ukrainian state. The legacy of Soviet rule was a complexly mixed multinational subcontinent with millions of people living outside what now had become their "homelands" and new minorities in republics face to face with new dominant majorities and without any appeal to an imperial center. It was within this context of weak states and fledgling nations, ethnically mixed populations, and Soviet traditions of ethnically based politics that Armenians and Azerbaijanis turned a political dispute into the prolonged and bloody confrontation over Karabakh.
THE CONTOURS OF THE KARABAKH CONFLICT
The anomalies that led to the outbreak of violence in and around Karabakh began with the application of Leninist nationality policy in this region. Though the population of Mountainous Karabakh (Nagorno-Karabakh) was overwhelmingly Armenian in the twentieth century (75-80 percent), for strategic and economic reasons Soviet authorities placed it within the wealthier Soviet Republic of Azerbaijan rather than in Soviet Armenia. Autonomous Karabakh was separated from Armenia proper by a six-mile swath of land - the Lachin corridor that was primarily settled by Muslim Kurds. With Lachin as part of the Republic of Azerbaijan, Armenia had no contiguous border with Karabakh.8 Though Armenians had nominal political control over the regional Karabakh soviet, in the pseudo-federal structure of the Soviet Union autonomy meant little, and Karabakh remained subservient to Baku and its Azerbaijani Communist party. Armenians in Karabakh and Armenia proper protested periodically against this infringement of the national principle, as well as what they perceived to be restrictions on the cultural and economic development of the Karabakh Armenians by their Azerbaijani overlords.
Azerbaijanis, on the other hand, saw Karabakh as part of their historic homeland, the cradle of poets and composers, and the victim of aggressive Armenian nationalism.9 When in the late-I 980s nationalist stirrings were felt throughout much of the Soviet Union, titular groups within union republics envisioned a possible sovereignty and a concomitant "ownership" of their republics. In this utopian vision, national minorities like Azerbaijanis in Armenia or Armenians in Azerbaijan were a thorn pricking the balloon of national fulfillment. Migration back to "home" republics accelerated. But with the opening of greater political expression under Gorbachev, the Karabakh Armenians called for the merger of their autonomous district with the Armenian republic. On February 13, 1988, street demonstrations began in Stepanakert, the capital of Karabakh, and six days later they were joined by mass marches in Erevan. In an unprecedented action, the Soviet of People's Deputies in Karabakh, up to this time a typical rubber-stamp Soviet-style legislature, voted 110-17 to request from Moscow the transfer of Karabakh to Armenia. A new era of nationalist politics had opened in the USSR that within three years would challenge the authority of a moribund superpower.
The years 1988-90 were crucial and complex. First in Sumgait, an Azerbaijani city near the capital, and later in Baku itself, ugly riots broke out with Armenians singled out for beatings, even murder. In the drab industrial town of Sumgait, Armenians were set upon by neighbors, hacked to death before the eyes of family members, several set afire. For Armenians the pogroms of Sumgait and Baku were bloody proof that Armenians could never live under Azerbaijani rule and feel safe. Armenian accounts refer to these events as evidence of Azerbaijani ethnic hatred, of the genocidal tendency among "Turks" that Armenians experienced in the Ottoman Empire in 1915 and which now Azerbaijani "Turks" were reviving.10 For Armenians genocide is a palpable threat, and their historical experience suggests that no outside power will come to their aid against Turkish extermination. They have developed a mentality, not unlike many Israelis, of a besieged and vulnerable nation whose only salvation lies in its own efforts to defend itself from overwhelming Muslim neighbors. In Armenia one frequently is told that only Karabakh and its army stand between the Armenians and another genocide. As horrific as the killings in Azerbaijan were, it should be noted that the initial tragic events in Sumgait and Baku were affairs of a few days rather than a methodical, prolonged genocide of local Armenians. Ethnic violence did not spread from city to city, village to village. There was no overall Azerbaijani plan to rid Azerbaijan of Armenians, certainly not to murder them systematically. Even today some Armenians manage to live in Baku without overt threat or ethnic slurs. Whatever the role of Azerbaijani officials and that remains murky—it is clear that the key actors in the pogroms, particularly those in Baku in 1990, were Azerbaijani refugees forced out of Armenia. Yet the riots and killings fatally colored the mutual understandings of these two nationalities, making each see itself as victim and the other as oppressor.
The Armenian view of their desperate situation is well known in the West, while the Azerbaijani vision of victimhood is far less appreciated. Azerbaijani claims to innocence are coupled with Armenian guilt in popular narratives. In direct contrast to the Armenian view, Azerbaijan is see Armenians as the aggressors, the first to start conflict, and a fantastic story has emerged that an Armenian led the rampaging mob in Sumgait. One lawyer claimed that one of the murderers in Baku had been recently let out of a Soviet prison and was known to be half-Azerbaijani and half-Armenian.
Some credulous Baku residents believe that the Soviet army, which was on the brink of occupying Baku in 1990, needed an event to justify their reestablishment of central Soviet rule over the republic. Many Azerbaijanis, particularly those who suffered directly from the Armenian military advances, are convinced that the Armenians had a plan to exterminate Azerbaijanis throughout Karabakh and in the corridor that lies between Armenia and the autonomous district. In extensive interviews carried out in August 1998 with refugees and "internally dispersed persons" (IDPs) now living in Tartar, Barda, Sumgait and Baku, a very clear story emerged.11 Armenian militias along with civilian compatriots systematically cleansed the corridor separating Armenia from Karabakh in a cold-blooded campaign. Armed bands relied on local Armenians to identify Azerbaijani villages and homes and then recruited these people to burn down the homes of their neighbors. One IDP recounted that his one-time Armenian neighbor told him, "We don't kill you because we want your land. We kill you because you are Muslim." Such narratives of betrayal are mixed with reports of inhuman atrocities, and several informants described Armenians as "animals." Finally, Azerbaijanis see Armenians as particularly privileged in their close ties to the Russian and European worlds, while they as Muslims suffer discrimination and condescension from the West and North.
Within Azerbaijan the refugee and IDP camps are the seedbeds for narratives of "return" and "revenge." From 1988 to 1993, an estimated 20,000 Azerbaijanis were killed, all but a few hundred in the fighting; 233,700 refugees were created along with 551,000 IDPs.12 The bulk of these refugees and IDPs were from Azerbaijani territory outside the formal borders of Karabakh itself. While many IDPs claim that they left because their government urged them to do so while the Azerbaijani army attempted to resist the Armenian incursion, a significant percentage told how they attempted to hold on to their properties until mortar shells hit their houses. Others who had left earlier heard later that their homes had been burnt to their foundations.13 Jobless, without hope, unintegrated in Azerbaijani society, the refugees construct and reconstruct their horrible past.
In both republics there exist deep hatreds that are usually imagined as ancient and primordial, permanent and ineradicable. Yet at the same time there are counter-memories of past times of peace and stability when Armenians and Azerbaijan is lived together without conflict. In interviews carried out by David Laitin and others in Azerbaijan in the summer of 1998, long-time town residents of Baku and Sumgait insisted again and again that they very much wanted the Armenians to remain, that they felt they were a well-educated community, much more like them than the rural refugees and IDPs who took the homes of Armenians. These townspeople stated that they were unable to discover who had instigated the riots but were sure that they were outsiders. Not only do some cosmopolitan Azerbaijan urbanites remember the contributions of Armenians to their republic's culture and well-being, but Armenians in Armenia remember how the rural Azerbaijan is where the purveyors of the best fruits and vegetables in the collective-farm markets.
War and murder have steeped some images in bloody hues, and the recovery of older patterns of coexistence is difficult to imagine. Nourished by resentments and material deprivation, the seeds of largescale war that could easily last for generations and draw in powerful states like Iran, Turkey and Russia continue to be planted, almost hourly, in the South Caucasus. But the collective suffering of Armenians and Azerbaijanis has not only hardened the divisions between these peoples but made it clear after seven years of war and tens of thousands killed that a political rather than a military solution is desperately needed. Despite these narratives and bitter memories, opportunities for solving the Karabakh problem are available. While no solution will adequately compensate the families of victims, those who remain embittered must move toward a compromise solution that will safeguard the future of their children and their children's children from wars fueled by oil revenues and fought by refugees in the name of national dignity.
POLITICS IN THE CAUCASIAN MODE
Armenia and Azerbaijan came to independence in quite different ways: Armenia through an anti-Communist nationalist movement that successfully replaced the Communist party in power through democratic elections, Azerbaijan with its Communist leadership intact and prepared to support the anti-democratic coup of conservative Communists in August 1991. Armenia enjoyed a reputation internationally as a brave, resistant anti-Communist supporter of Western-style reforms, while Azerbaijan stumbled from coup to coup until state power reverted to the former Communist party boss, Heidar Aliev. Though the move toward democracy in both countries was extremely uneven, the overall trend in the South Caucasus by the mid-1990s (including Georgia, where civil and ethnic war tore the country apart in 1991-92) was toward stabilization of existing state structures, a relative degree of public order, and a slower pace of reform and democratization. In Armenia the war, the Azerbaijani blockade, the failure to repair the damage suffered in the December 1988 earthquake, and the growing apathy and despair that encouraged migration to the West eroded the earlier popularity of the Ter Petrosian government (1990-98). The unity of the original band of nationalists who had led the Karabakh movement in Erevan splintered within the first year when key members broke with the government and formed opposition parties.
Banditry and armed militias in the streets of the cities, along with the growth of independent centers of economic power, threatened the almost non-existent state apparatus. A series of victories in the Karabakh war, beginning in May 1992 with the capture of Shushi and Lachin, and the expansion and stabilization of the front with a cease-fire in the spring of 1994, gave the Armenian government a short breathing space needed to bring civil order to its towns, Jay the basis for a restoration of the economy, and win over foreign friends and aid.
Ter Petrosian navigated a narrow course between open support of the Karabakh republic's policies, supported by leading parties in the parliamentary opposition, and the requirements of Russia and the United States to restrict its direct involvement in Azerbaijan. At first his government attempted to reverse the traditional Armenian reliance on Russia and to ameliorate relations with Turkey. The Armenian diaspora reacted angrily when it perceived that the Armenian government was playing down the genocide issue in order to "appease" the Turks. But the effort to improve relations floundered as Turkey drew closer to Azerbaijan. Armenia remained close to Russia, careful not to criticize its more aggressive policies, particularly in Chechnya. Erevan requested in 1992 that Russian troops remain in Armenia and eventually agreed to the formal establishment of a Russian military base in the republic.14 Alone of the Caucasian republics, Armenia joined the Commonwealth of Independent States (CIS) at its very initiation and never left it. To a greater extent than the other former Soviet republics, Armenia depended heavily on imports from Russia and the rest of the CIS. After an initial catastrophic collapse of the economy, which was tightly tied into Soviet military production, Armenia began a program of rapid privatization, particularly of agricultural lands. The World Bank and IMF committed to major grants to Armenia in late 1994 and early 1995, and Armenia remained the largest per capita recipient of American aid among post-Soviet states. Living and health conditions remain at a very low level, but the worst seems to be over if the conflict in Karabakh could be peacefully settled.
Armenia's brief stability was assisted by a number of factors. First, the absence of any serious minority problem has prevented the kinds of interethnic conflicts that plague Georgia and Azerbaijan. Second, a powerful, binding national identity to which much of the population subscribes has given the government, which is the heir to the nationalist movement that displaced the Communists, an abiding legitimacy. Third, victory in the Karabakh war also aided the government for a time, though chronic economic problems and the burdens of the unsettled conflict ate away at Ter Petrosian's political base. Until late 1994 the government's major opposition came from a diaspora party now established in the republic, the Armenian Revolutionary Federation or Dashnaktsutiun, which took a harder line on Karabakh and the genocide issue. At the end of 1994, Ter Petrosian expelled or arrested its leaders, charged them with terrorism and closed its newspapers. This breach of democratic practice continued through the parliamentary elections of June 1995, which were characterized by international observers as "free but not fair," and the presidential election in September 1996, when serious interference with the votes led to open street demonstrations and the deployment of the military against the protesters.
From that time until his resignation in February 1998, Ter Petrosian was dependent on the backing of the "power ministries" of defense, national security and interior, as well as his prime minister, the battle-hardened and highly respected former president of Karabakh, Robert Kocharian. When Kocharian succeeded his former patron, he adopted politics with a decidedly deeper nationalist tone and greater sensitivity to the views of the politically active Armenian diaspora and the Dashnaktsutiun. With his principal supporters in the military and among Karabakh hard-liners, Kocharian had very little room to maneuver on the issue of war and peace. His tenure also depended on the support of the power ministries. But for the "men with guns" in Armenia, as well as the leaders of Karabakh, the status
quo was acceptable for a few years. Their role as defenders of the nation remained intact; their political positions were nearly unassailable, not to mention the economic advantages that some gained from the absence of a settlement. But the fragile stability was shattered on October 27, 1999, when armed men burst into the Parliament and murdered the prime minister and the speaker of the legislature.
Azerbaijan's stability has long been contingent on its performance in the Karabakh war, which has not gone well for Baku since 1992. But other factors have also contributed to the fragility of the state. The nationalist movement in Azerbaijan at the end of the Soviet period achieved neither the mass following that nationalists in Armenia and Georgia enjoyed nor the cohesion and solidarity of the leadership that the Armenian national movement possessed. The Communists therefore had less incentive to give up power, and the old Communist elite survived the coup against Gorbachev and led the country into independence. At first Azerbaijani forces were able to push back the Armenians, empty Armenian villages outside of Karabakh, and use Russian and Chechen mercenaries effectively against enemy soldiers and civilians. But after the Armenians took towns outside Karabakh and hundreds of Azerbaijan is died in a massacre at Khojali, the Communist government of Ayaz Mutalibov fell (March 1992). When Shushi (Shusha), the mountain-top traditional capital of Karabakh, was captured by the Armenians, Mutalibov attempted to return to power, but fighting in the streets resulted in a victory of the Azerbaijani Popular Front in May.
The nationalist government of Abulfaz Elchibey reoriented Azerbaijan away from Russia, left the CIS and turned toward Turkey but proved equally inept in pursuing the war. After an Armenian victory at Kelbajar, which completed the effort by the Karabakh Armenians to form a broad land bridge to the Armenian republic, forces led by Suret Huseinov overthrew the Elchibey government in June 1993 and brought the former Communist chief Heidar Aliev back to power in Baku. The war dragged on, less as a confrontation between two armies and more as a struggle between one army or the other against civilians. When armies met, one side often broke and ran Armenians in the first years, Azerbaijanis from 1992 on. When the Azerbaijanis tried a winter offensive in December 1993, thousands of their troops, abandoned by their officers, froze to death or were picked off by the Karabakh forces in the mountain passes. Five months later the two sides signed a Russian-brokered armistice.
Though his troops were unable to secure victory, the skillful Aliev and his political allies managed to survive.15 Aliev's background is impressive indeed. Born in Armenia and raised in Nakhichevan, he is said to have captured Stalin's attention as a member of a counter-intelligence group on the Ukrainian front in World War II by intimidating Soviet deserters. He worked his way up to become first secretary of the Azerbaijan Communist party in 1969 and eventually as a member of the Politburo in Moscow, until his removal by Gorbachev in 1987. Well-connected to the elites entrenched in the economy and localities, Aliev's government managed to consolidate its power, at least in the environs of Baku, even without a legitimizing myth as the anti-Soviet movement. Despite the continued losses suffered in the Karabakh war, Aliev gained a significant degree of popular support as a competent and experienced politician who best represented a promise of peace and stability in the future. Aliev beat back an attempted coup (October 1994) by his prime minister, Huseinov, disarmed militias loyal to the Popular Front, which had retained significant support in Nakhichevan, and managed to escape a number of attempted assassinations and coups. He completed the negotiations on the oil concessions to Western companies in September 1994 (the so-called "deal of the century") and agreed to Russian and Iranian shares. Reassessing the "turn toward Turkey" under Elchibey, Aliev distanced his government somewhat from Ankara and drew closer to Moscow and Iran. Though he agreed to have Azerbaijan rejoin the CIS, Aliev respected the deep anti-Russian sentiments in Azerbaijan by steadfastly opposing the stationing of Russian troops in the republic.
Riding the surface of a fragmented and dispirited society, without the kind of unifying nationalist discourse and anticommunist legitimacy enjoyed by the Armenian government for a brief but crucial period, Aliev remains dependent on the economic and local political elites that actually run Azerbaijan. Re-elected twice to the presidency (1995, 1998), in elections that did not meet international standards for fairness and produced much dissension within the country, the Azerbaijani president seeks to hand his position and power over to his son. His health impaired by a weak heart, Aliev recognizes that both economic prosperity and political stability in Azerbaijan depend on the ending of the Karabakh war on conditions that do not humiliate his countrymen.
RESOLVING THE CONFLICT
Even before massive fighting had broken out in and around Karabakh, the stage for conflict had been set by completely incompatible ideas on the future status of that enclave. Armenia formally "annexed" Karabakh; Azerbaijan officially abolished its autonomy; and Karabakh declared itself an independent republic, even though no other state (not even Armenia) recognized it. For the better part of the next decade, efforts at settling the conflict were carried on by three sets of actors: the Armenians and Azerbaijanis themselves; Russia, usually acting unilaterally; and the "international community," in this case the CSCE (Conference for Security and Cooperation in Europe), later the OSCE (Organization for Security and Cooperation in Europe). So far, each effort has failed, though several have come close to agreement. A close review of the negotiations convinces us that the breakdown of these attempts stems not from intractable, irresolvable differences nor even from fear on either side that the security of their populations would be threatened by a post-settlement regime - but rather from contingent political factors. At first, the situation on the ground was in such flux that the side having the military advantage was unwilling to make concessions. Then, when the military situation stabilized, the international community was divided and sent mixed messages to the combatants, making it difficult to structure a peace plan. Finally, once a cease-fire was put in place and the international community unified around common principles, its proposed solutions failed to balance the goals of territorial integrity and self-determination in a way that all parties could accept. Yet it is clear that there has been movement and flexibility on both sides and that the political conditions of the bargaining, rather than the structural conditions of the conflict, have undermined resolution.
The first post-Soviet peace effort in the region was the Yeltsin-Nazarbaev plan put forth in the months after the demise of the Soviet Union. The two presidents proposed a cease-fire, the holding of elections within two months; and the creation of a constitutional government in Mountainous Karabakh. Refugees were to be returned, and peace would be guaranteed by an international military force, perhaps U.N. troops. The Russian Kazakh efforts, however, died on the vine. Neither of the belligerents in the war had yet tested itself in battle, and each was convinced that it could improve its position with arms. After the January 1992 entry of the former USSR republics into the CSCE, that organization became the principal mediator between Armenia, Azerbaijan and, eventually, Karabakh.16 The Khojaly massacre in early 1992 stimulated CSCE interest in the conflict, and John J. Maresca, the American ambassador to the CSCE, led a delegation to Karabakh. At the time, Armenians were steadily driving the informal Azerbaijani militias out of Karabakh. Mutalibov, the Communist party chief still in power, not only did not have his own army, but decided not to build one and to rely on the Russians to win the war for Azerbaijan. The CSCE soon agreed to organize a peace conference on Karabakh in Minsk, and, as a result of high-level American intervention, an informal "Minsk Group" was formed.
Patient but frustrating negotiations followed. The Minsk Group managed to bring the parties together despite Azerbaijani resistance to having Karabakh represented officially. Maresca arranged private meetings of American, Russian and Turkish representatives. Then the Italian chair of the Minsk Group was invited (3 + 1 ). Later still, Armenia and Azerbaijan (5 + 1) joined. Finally, the full Minsk Group met together. The war began to go better for the Armenians after the offensive at Khojali (March 1992) and their capture of Shushi, historically the most important town of the region, and the Lachin corridor (May 1992). At that moment Ter Petrosian was in Tehran, where the Iranians were attempting to mediate a cease-fire, but the success of the Karabakh troops ended the Iranian efforts. Armenia became emboldened, initiating clashes on its border with Nakhichevan that led Turkey to threaten intervention. A moment of acute danger of a widening war passed only after NATO cautioned against escalation.
The Minsk Group held its first "emergency preliminary" meeting in Rome in June 1992, without representation from Karabakh. Regrettably, the Minsk Group and the CSCE did not state clearly from the beginning that any final resolution of the Karabakh conflict would have to recognize and reconcile two fundamental principles the territorial integrity of recognized states and national self-determination of peoples. These principles were later invoked more forcefully in other conflicts such as Abkhazia and Kosovo.
This failure may have given the parties in the Karabakh conflict a false signal that unilateral shifts in borders might be acceptable to the international community. Its first co-chairs, Finland and Russia, worked diligently to bring the parties together, but events in the Caucasus moved too quickly for diplomacy. Mutalibov fell to Elchibey, who drew Azerbaijan closer to Turkey. In June 1992, Azerbaijani forces, along with Russians and Chechens, attacked northern Karabakh and captured the town of Mardakert, renaming it Agdere.
Russia, which in its first months after full independence had retreated from direct involvement in the South Caucasus, now indicated its renewed interest in the region. But it acted on its own, unpredictably, and without coordination with the CSCE effort. At first, the Russian government tilted toward Baku (until October 1992); then it turned more favorably toward the Armenians. The new Russian government of Victor Chernomyrdin began to take greater initiative in the Karabakh conflict early in 1993, organizing its own meetings with the principals, proposing its own cease-fires, and acting independently of the CSCE efforts. Yeltsin made it clear in a notable address on February 28 that Russia had a "vital interest" in the territory of the former USSR. "The time has come," he told a gathering of Russia's business and civic elite, "for the appropriate international organizations to grant Russia special powers as the guarantor of peace and stability on the territory of the former Union."17 As far as the Karabakh conflict was concerned, this new "Yeltsin Doctrine" represented a shift from a multinational approach to a unilaterally Russian one. In contrast to the international proposals advocating an internationally controlled "monitoring force" that was not authorized to use force, Russia began to push for a Russian-controlled CIS "separation force" able to use its weapons to maintain a cease-fire. Russians, it was suggested, would make up the bulk of this force, though under CSCE control.
Divided between president and parliament and with high officials enriching themselves from the chaos of economic and political collapse, the Russian state was unable to act as one. In September 1992, Russia's defense minister, Pavl Grachev, met his Azerbaijani and Armenian counterparts in Sochi and proposed his own plan for settlement of the war. With the Russian state reduced to competing fiefdoms and rival personalities, Grachev deliberately excluded Vladimir Kazimirov, the foreign ministry's point man on Karabakh, from the meeting. For his part, Kazimirov soon learned to keep in touch with both defense and the presidential apparatus, and for the next two years promoted a unilateral plan formulated by Russia. The Armenian offensive toward Kelbajar (April I 993), which lay between Armenia and Karabakh in the north, stymied further consideration of Moscow's plan. Tens of thousands of Azerbaijanis became refugees, many of them fleeing to Baku or Sumgait, others finding shelter in camps set up by the Iranians, as the Armenian forces swept them out of areas near the Armenian border. The tide of international opinion that had generally favored the Armenians until the killings at Khojali soured even more rapidly after Kelbajar. Yet the Armenians went from one battlefield success to another. In July they besieged and captured Agdam, a city of J 50,000 just outside the borders of Karabakh, facing condemnation by U.N. Security Council resolution 853 (July 29, 1993). In August, the Armenians turned south to clear the area separating Karabakh from Iran Jibrail (August 19), Fizuli (August 23) and Zangelan (October-November).
Azerbaijan was caught between the most powerful army in the Caucasus, that of the Karabakh Armenians, and a wavering Russia with its own internal political conflicts and a deep desire to re-enter the USSR's former southern sphere of influence. Back in power, Aliev quickly recognized Russia's importance and joined the CIS late in 1993. Azerbaijan, which increasingly saw Russia as an ally of Armenia, opposed a large Russian contingent in the international force. Grachev tried to bully Azerbaijan into going along with Georgia and Armenia and allowing Russian troops to be based there. Aliev not only rebuffed Grachev's demands but rallied the Turks to support an international peacekeeping force instead of a Russian one. The off-again, on-again relations between Turkey and Azerbaijan improved. Azerbaijan is, however, were bitter that Turkish support was so constrained. The Turks maintained the traditional Kemalist restraint against foreign adventures (Cyprus and Iraqi Kurdistan around Mosul excepted) and were careful about antagonizing Russia, with whom Turkish capitalists were doing billions of dollars in business.18
The United States, which also favored an international rather than a Russian peacekeeping force, turned greater attention to the Karabakh question in 1993. That summer, officials in the American government carried on a discussion of the role of peacekeeping in the Caucasus. Ambassador Maresca, who was attempting to put together a CSCE proposal, saw the Russian action as a rogue operation and a serious impediment to a multilateral solution. He offered his Russian counterpart, Kazimirov, seven conditions for supervising a cease-fire by an international force that included Russians. There was no real response from the Russians. "Their bad faith became increasingly obvious," Maresca later reported. "It was clear that it was their deliberate intention not to cooperate, thus to ensure that their own proposal would be understood by the parties to be the only game in town, and ultimately to supplant the international negotiating process."19 For his part, Kazimirov was disadvantaged by the divisions in the Russian government and by not being regularly informed by the defense ministry of its policies. Many Russians were deeply suspicious of increased Turkish influence in Azerbaijan, the North Caucasus, particularly Chechnya, and Central Asia and were upset at what they perceived to be American backing of Turkey as a participant in the Karabakh solution. On September 22, 1993, Foreign Minister Kozyrev reasserted Russia's paramount role in an article in Nezavisimaia Gazeta: "Whether we like it or not, there is no alternative to a Russian Federation peacekeeping contingent in this conflict...; immediately after a settlement mechanism is set in motion, this contingent should be given the status of a U.N. force and reinforced with U.N. units from neutral European CSCE countries. Here too, we ourselves, as well as the United Nations, must do our historic duty. It would be irresponsible to evade this."20
In the fall of 1993, Russia was temporarily sidetracked by the violent conflict between Yeltsin and the oppositional parliament, which he overthrew in a bloody confrontation. Negotiations over Karabakh were carried on by the new Swedish co-chair of the Minsk Group, Jan Eliasson, who decided to limit meetings of the Minsk Group and engage in shuttle diplomacy in the region. In Maresca's eyes this was "downgrading the U.S. role in the process, even though the United States was the only voice the Russians took seriously."21
The international actors were unable to act in concert. The U.N. Security Council reaffirmed the sovereignty and territorial integrity of Azerbaijan and called for maintaining the latest cease-fire (October 14, 1993), but a week later the Armenians launched attacks on Zangelan and Goradiz. Russia haltingly but effectively took command. What Europe could not accomplish, the Russians managed to achieve. Once a new constitution and parliament was established in Moscow, the Russians brokered a new cease-fire (February 1994) that was formally signed on May 12, 1994. The combatants were exhausted the Azerbaijanis demoralized, the Armenians over-extended. The armistice was initially to continue for three months; it has held for five years. Russia and the CSCE, however, continued to pull in opposite directions. The CSCE decided to upgrade its force from a "monitoring" to an international" peacekeeping" force, which is larger, armed and more active in controlling the cease-fire.
The Russians, particularly General Grachev and the Ministry of Defense, continued to push for a Russian or CIS force. Grachev gruffly informed a negotiating session on Karabakh, "Whatever I propose, that's what we are going to agree on."22 Russia wanted its soldiers back in Azerbaijan deployed in a strong forward position against Iran and Turkey. Armenia and Karabakh were pleased with the Russian proposal, but Azerbaijan was set against it. Armenians were most concerned that there be no Turkish participation in the international force.
American involvement intensified once the Clinton administration attempted to revive the CSCE efforts at mediation in the summer of 1994. The president told Ter Petrosian during his visit to the United States in August that he would follow the Karabakh issue personally. Ambassador Maresca presented a proposal for the settlement of the conflict. He suggested that Karabakh be recognized as the Republic of Nagorno-Karabakh within the sovereign Republic of Azerbaijan; that Armenia and Azerbaijan sign a treaty on mutual transit rights across each other's territory; that refugees be allowed to return to their homes; that all of Armenia and Azerbaijan, including Karabakh, be a free trade zone; and that the settlement be guaranteed by CSCE and the U.N. Security Council. The United States would not be involved on the ground but was willing to take the lead in building road connections. A conference of international donor organizations would raise funds for economic reconstruction.23
Regrettably Russia and the CSCE continued to pull against each other in the negotiations over Karabakh. The most active efforts were carried on by the Russians, who held a series of meetings of representatives of the defense ministries of Armenia, Azerbaijan and Karabakh. Eliasson shuttled between Stepanakert, Erevan and Baku but with no tangible results. The Minsk Group meeting in Vienna criticized the Russian efforts as unilateral and excluding the CSCE, which led to the Russians boycotting the remainder of the meeting. To many observers only the Russians, who had come back with renewed energy into the Caucasus, seemed to be achieving any results. Even the United States backed the Russian proposal for its troops to be deployed in the region of Karabakh, though contingent on direct supervision and control by the United Nations and the CSCE. More than ever before, Russia's aim to become the most influential power in the Caucasus and hold back advances by others had a determining effect on the contours of a resolution of the conflict. For their part, the Turks agreed that the international peacekeeping force would not involve the Turkish military, and they suggested Erzerum in eastern Turkey as a logistical staging area for the force. By this time, Karabakh, which had once been a local conflict in the southern Soviet Union, had now become a token in an international game of power politics, the stakes of which involved millions of barrels of Caspian oil.
The last months of 1994 were a period of particular turmoil both in Russia and Caucasia. One more unsuccessful coup in Baku was foiled by Aliev's forces. Yeltsin sent Russian troops into Chechnya. After the murder of the former mayor of Erevan, Ter Petrosian arrested Dashnak leaders and closed newspapers. In both Azerbaijan and Armenia state power remained fragile, and the popular mood was sullen and discouraged. The war in Karabakh had brought little material benefit to Armenia, and the cease-fire had simply frozen the problems of refugees and Armenian occupation in Azerbaijan. In both republics the political leadership faced oppositions that were unwilling to accept a compromise like that proposed by Maresca. After their victories and sacrifices, Armenians were resistant to the idea of leaving Karabakh within Azerbaijan, even as a fully self-governing republic. Azerbaijan is, smarting from defeats and humiliations and feeling victims of Armenian aggression, stood firmly on their right to territorial integrity of their republic.
Finally, in December 1994, about the time the Russians launched their war on Chechnya, they agreed to participate in a multinational peacekeeping force to be deployed in Karabakh after a political agreement was reached. Key questions, like the chain of command, were yet to be worked out. Russians wanted to command the force and have most of the troops, while the other CSCE members preferred to have less than 50 percent Russian troops and CSCE command over the force. But at least CSCE and the Russians were backing a similar arrangement.
For the next four years, negotiations wore on, often in Moscow or in European capitals, sometimes through shuttle diplomacy between Baku, Stepanakert and Erevan. Delicate, complex talks were held by Joseph Preset, the American special negotiator for Nagorno-Karabakh and CIS regional conflicts, and in a back channel, encouraged by the Americans, by Armenian presidential advisor Girair Libaridian and his Azerbaijani counterpart, Vafa Guluzade (from late 1995). Though at several moments success seemed at hand, in the end the question of the final status of Karabakh and questions of what to do with Lachin and Shushi made it impossible to come to a resolution of the conflict. In October 1995, Baku hinted publicly that it was prepared to consider international control of the Lachin corridor. Some suggested that the Dayton Accords over Bosnia or (still later) the Russian-Chechen agreement to postpone discussion of the final status of Chechnya serve as models for Karabakh. Finally, about the time that France, the United States and the Russian Federation became the co-chairs of the Minsk Group, a number of sources suggested the outlines of a draft peace plan: Karabakh would remain formally within Azerbaijan but with its own police, military and security forces. Its full autonomy would not include the right to establish diplomatic relations but would be guaranteed by international peacekeepers deployed in the Lachin corridor and formal security guarantees from NATO and the United States. The Armenian attitude toward autonomy was expressed in an anecdote that circulated at the time. It seems that Armenians from Karabakh were taken to the Aaland Islands, autonomously run by their Swedish inhabitants but within the political boundaries of Finland. When asked if such autonomy would work for them, the Armenians surprised their hosts by answering positively. "Are you sure that you would accept such autonomy?" the surprised hosts repeated. "Oh, yes," said the Armenians, "but only within Finland!"
Time had seldom been on the side of the Armenians in the Karabakh conflict, but in the four-and-a-half years after the May 1994 cease-fire it shifted even more to favor Azerbaijan. Armenians had won the war in Karabakh. But almost from the moment of victory their impeccable moral position had begun to blur in the eyes of the Western powers. Armenians occupied about 15 percent of Azerbaijan and had rendered hundreds of thousands of Azerbaijanis refugees in their own country. Corruption and economic polarization marked their move to a market economy. Their government outlawed a major oppositional party and manipulated elections. Within Armenia and Karabakh public opinion grew more suspicious of the West and drew closer to its traditional reliance on Russia. In a poll conducted by the Armenian Academy of Sciences early in 1996, Russia was overwhelmingly preferred as the peacekeeping force and chief mediator over the OSCE. Nearly three-quarters of those polled favored recognition of Karabakh as an independent republic, the other quarter calling for unification with Armenia. 24 Armenia's turning inward was matched by its growing international isolation, which was confirmed at the Lisbon summit of the OSCE in December 1996, when every state but Armenia accepted a resolution that supported the territorial integrity of Azerbaijan. Though a blow to the Armenians, Lisbon represented the belated assertion by OSCE of the fundamental conditions to which resolution of ethno-territorial disputes would be required to adhere.
At the same time, for economic reasons, Azerbaijan looked better and better to the West. A state with a population twice the size of Armenia's, Azerbaijan was blessed by accidents of nature. One of the great undeveloped oil reserves in the world lay under the Caspian Sea, and Azerbaijan stood to become a Caucasian Kuwait. The American government, which had hobbled its oilmen in Iran and Iraq, was reluctant to do it to them once again in the Caucasus. Baku now found new friends in Washington and Houston, among them former secretaries of state and national security advisors. Influential Americans, impatient with the failure to solve the Karabakh conflict, now appeared to be more tolerant of strong leaders who keep their small states stable and friendly to the West. Armenia had lost its democratic patina just as Azerbaijan was no longer required to have one.
At the end of May 1997 the co-chairs of the Minsk Group France, Russia and the United States put forth a proposal for further negotiation. The issues of cessation of hostilities and final status of Karabakh were separated, and the chairs stipulated that agreement on either issue could be implemented without waiting for agreement on the other. Agreement I, concerning the cessation of hostilities, was based on a paper that the Minsk Group had been trying unsuccessfully to have all sides accept. Agreement II, the difficult matter of status for Karabakh, had an unusual pedigree, originating in a paper of Libaridian that had made its way to Russian foreign minister Evgenii Primakov and then was circulated by Kazimirov. Ter Petrosian liked the agreement on status, confident that he could improve upon it in negotiations. But President Arkadii Gukasian of Karabakh rejected the document brusquely, likening the cochairs of the Minsk Group to the "troikas" of judges that had condemned innocent people to death in Stalin's purges. Baku cautiously encouraged further work on the proposal but remained dubious about the status plan.
The summer and early fall of 1997 was the moment when two of the three sides came closest to a settlement of the Karabakh conflict. Ter Petrosian suggested a "step-by-step" approach. The question of Lachin would be moved into Agreement II and negotiations would commence only on Agreement I. Agreement II would be negotiated only after Agreement I had been secured. When the Russian negotiator in Baku presented Ter Petrosian's proposal to the Azerbaijanis, he added the idea of delaying discussion of
The advantages of accepting Agreement I seemed compelling to the mediators. The Azerbaijanis would get back the occupied lands outside of Karabakh, which would then be demilitarized. Karabakh would continue to exist in its present form until the agreement on final status was reached, but very significantly it would have gained an internationally recognized interim status (in contrast, for example, to the Serbs in Krajina), which it would not have without the agreement. Lachin and Shusha would remain as they were, and Azerbaijani refugees would not be returned to those regions until the agreement on status was reached. No international troops, however, would be deployed until the agreement on status, but Karabakh would gain additional security by the demilitarization of the formerly occupied lands and a pullback of heavy Azerbaijani weaponry. All embargoes and blockades would end, including those by Turkey (though not a signatory to the agreement, Turkey agreed as well to end its embargo). Finally, the agreement stipulated that Armenia, Azerbaijan and Karabakh would continue negotiations on final status for Karabakh which would not be implemented until all sides accepted a formula. Karabakh Armenians, therefore, had a veto on any status agreement.25
Despite the evident advantages of the agreement, the Karabakh government rejected the proposal, probably because it was based on an understanding that Karabakh would remain in some form within the boundaries of Azerbaijan. It may also be the case that some officials in Stepanakert calculated that by withdrawing their troops from occupied Azerbaijan, Karabakh lost its direct access to Iran and the profitable trade that flowed from the south. Moreover, closing that route would make Karabakh even more dependent on Erevan and President Ter Petrosian, who had become increasingly suspect in the eyes of Karabakh hard-liners. Fearful that by giving up the occupied lands it would lose its leverage over the status question (despite the OSCE's granting of a veto over status to Karabakh), the Karabakh government came out once again for a "package deal" calling for resolution of both the status issue and the question of withdrawals simultaneously. To many observers it looked as if Karabakh was holding out for full de jure independence, something to which Azerbaijan and the international community would not agree. But it should be noted that Karabakh was not against a bargained agreement in principle. Nor did it reject this proposal because of fears of a genocide. Rather, Stepanakert rejected it because the "self-determination" element of the plan was not sufficiently favorable to Karabakh.
A gauntlet had been thrown down to the Armenians and Azerbaijan is, for the Minsk Group's proposal had the united backing of the international community, most important, the United States and Russia. Each side tentatively moved toward picking it up. In August, President Aliev used the occasion of his visit to Washington to make a dramatic speech at Georgetown University. Western observers were stunned to hear him tell the audience that Lachin and Shusha would not be returned to Azerbaijani control in the foreseeable future. In September, Ter Petrosian gave one of his rare press conferences. Carefully he laid out five options open to Armenia. The first was to maintain the status quo: Armenian occupation of parts of Azerbaijan, hundreds of thousands of Azerbaijani refugees living in camps, Azerbaijani blockade of Armenia, and continued pressure from the international community. This he rejected, for it would lead to the loss of "all we have gained during these years." He went on:
It happened in Bosnia. The Serbs lost everything. I don't think that the maintenance of the status quo is a real option. We may persist for a year or two, but the international community will become exasperated and lose all its patience.
He also rejected the option of either recognizing Karabakh as a completely independent state or annexing it to Armenia. Such moves would in effect be an ultimatum both to Azerbaijan and to the world. He also dismissed the option of renewing the war, with Karabakh somehow forcing Azerbaijan to give up Karabakh.
I do not think that Karabakh is capable of forcing Azerbaijan to its knees, because it will have to seize Baku. But if it tries to seize one more region now, let alone Baku, the world will not tolerate it.
Therefore, only two practical options remained. The first was the "package" approach to peace: a one-time settlement that would involve simultaneously the return of the occupied territories except for the Lachin corridor, the deployment of peacekeeping forces along the Karabakh-Azerbaijan borders, the lifting of the blockade, the return of refugees to their homes, and the establishment of buffer zones along the borders. What remained to be decided was the most difficult question of all: the future status of Karabakh. Ter Petrosian claimed that Armenia was ready to sign on to the package approach, but when Karabakh and Azerbaijan both rejected this approach in writing, the Armenian government felt compelled to drop it. The only option left, Ter Petrosian concluded, was a step-by-step approach to peace. The status issue would be postponed until the first steps had been taken. Ter Petrosian revealed that Azerbaijan had agreed to this approach and promised that Armenia would shortly respond.26
Ter Petrosian's openness and his suggestion that Armenia might be favorable to a gradualist solution, while Karabakh was not, began the steady slide into the political crisis that culminated in early February 1998 with his resignation. His cool realism about the political and economic trends in the South Caucasus mapped a future in which resource-poor Armenia would face a wealthy and powerful Azerbaijan. He reasoned that victory in the war presented Erevan with an opportunity, not to stand fast but to cut a deal now that would guarantee both Karabakh 's security and Armenia's political and economic well-being. The step-by-step approach beginning with the withdrawal of the Armenian forces from the occupied Azerbaijani territories outside of Karabakh would build confidence on both sides that could lead to a long-term solution.
The president's conviction, so difficult for Armenians to swallow, that neither the independence of Karabakh nor annexation of the region to Armenia was a possible outcome of the negotiating process, was truly extraordinary for a politician who needed to maintain a base of support within the elites of both Erevan and Stepanakert. He essentially told Armenians that they had to recognize that Karabakh would be formally (de jure) part of Azerbaijan, while Azerbaijan is had to recognize that the Armenians of Karabakh would be de facto fully self-governing. The details of this strange hybrid arrangement had yet to be worked out. Ter Petrosian 's move toward a position of compromise, which was acceptable to a large degree also to Azerbaijan, precipitated an open break with the government of Karabakh, which opposed the outlines of the OSCE agreement that they interpreted as suggesting "vertical subordination" of Karabakh to the Baku government. Within Armenia an intense intragovernmental struggle went on at the end of 1997 and the beginning of the new year. After the Karabakh authorities rejected Ter Petrosian's suggestions, many of the president's principal allies within the government and parliament of Armenia abandoned him. Ter Petrosian decided to resign rather than risk a confrontation with the powerful ministers of defense and internal security. A new government was formed, headed by the former president of Karabakh (at that time prime minister of Armenia), Kocharian, who a few weeks later was elected president of Armenia in his own right. Almost immediately the Armenian government took a harder line toward the Karabakh issue.
While Ter Petrosian had linked Armenia's future stability and economic development to settlement of the Karabakh conflict, Kocharian believed that Armenia could develop politically and economically without giving anything up in Karabakh. His government turned its energies to fighting corruption in Armenia, with which, he claimed, the former regime was riddled. Instead of improving relations with Turkey, Kocharian relied more on the Armenian diaspora, raising again the issue of the Genocide. His government rejected the Minsk Group's "phased" plan, which called for the return of occupied Azerbaijani territories before an agreement on the status of Karabakh. Instead the Armenian side now favored the "package deal," which would resolve all the disputed issues in a single agreement. A principal object of the Armenians, most strenuously argued by the Karabakh leaders, was to avoid any kind of “subordination" of Karabakh to Baku. They were willing to wait while the international mediators rethought their position. On June 17, 1998, the new Armenian foreign minister, Vartan Oskanian, stated that Armenia was prepared to consider less than full independence of Karabakh from Azerbaijan but not merely autonomous status within Azerbaijan. Upset with what he saw as Azerbaijani intransigence, Oskanian added that Armenia might take its own action to break the deadlock, perhaps even unification with Karabakh. The international reaction was swift and negative. The U.S. State Department called his words "disturbing" and "unacceptable," and within days Armenia clarified that it did "not reject a solution to the Karabakh conflict" and that the foreign minister's words should not be "taken out of context and ...characterized in ways that are unfounded."
In an attempt to resolve the status issue, the Minsk Group came forth in November with a new formulation: a package deal in which Karabakh and Azerbaijan would form a "common state." This formulation was largely the brainchild of Primakov, then prime minister of Russia, and Oskanian. Precisely what a "common state" meant in terms of the relationship between Stepanakert and Baku (horizontal, equal relations or vertical relations of subordination of the former to the latter) was deliberately vague, but, as Russian ambassador Andrei Urnov noted, any agreement would have to guarantee security to the peoples of the region, provide an adequate degree of self-government or autonomy for Karabakh, and keep the Lachin corridor between Azerbaijan and Armenia open. "Common state" may have referred to something like the present status of Serbia and Montenegro within the shared framework of Yugoslavia. Two self-governing entities would coexist within a largely symbolic structure. The "package deal," with its unconventional status for Karabakh, was formally accepted by Armenia and Karabakh as a basis for negotiations on November 26, but Azerbaijan rejected it. Baku refused to accept a structure in which Karabakh might have all the attributes of a sovereign state, including the rights to foreign representation, or concede that Azerbaijan be reorganized as a confederation. Basically, the "common state" plan was a non-starter, and eventually the Minsk Group quietly abandoned it. Once again, a proposal failed not because of rigid positions or unremitting hostility but because the framing of the proposal did not reconcile the goals of territorial integrity and self-determination in a way acceptable to all parties. Discussions continued, and as summer 1999 turned into fall, the presidents of Armenia and Azerbaijan met occasionally, quietly, to talk about Karabakh. But, as we go to press, Armenia is reeling from the assassinations of two of the most powerful and popular politicians in the country: Prime Minister Vazgen Sargisian and Speaker of the Parliament Karen Demirchian. Kocharian remains the man in charge, reminded once again of the intimate link between internal stability and the festering conflict over Karabakh.
TOWARD A SOLUTION
As Armenians and Azerbaijan is must realize, victories do not necessarily yield peace. The present situation remains intolerable and unstable. It is intolerable because of the suffering on both sides. Guiltless Azerbaijani civilians live in squalor, thousands in windowless freight cars that reach 110°F in summer and below freezing in winter. They teach their children what belongs to them and who now has it. Many of the informants interviewed by Laitin and his colleagues described eight-room houses with large fruit-bearing gardens in their lost homesteads. Many felt that their loss of dignity living as wards of the international relief community was as distressing as their loss of property. In Erevan and other parts of Armenia people are disillusioned with the new political realities with which they live. The victories in war have brought no work, no revival of industry, no opening of the borders to trade. Those who are able to do so leave Armenia for Russia or Los Angeles. Some of those who stay worry that Armenia's future has been mortgaged to a recalcitrant Karabakh.
The situation is also unstable because the world is descending on the South Caucasus. Untapped oil wealth will radically change Azerbaijan and the relations between the republics. Georgia is already involved in the building of a pipeline from Baku to Supsa on the Black Sea, while Armenia is not to benefit at all from the new wealth about to be generated. The present cease-fire has saved thousands of lives that might have been lost, but it also is costing Armenia and Azerbaijan a future of potential stability, cooperation and shared prosperity. It should be clear that it is in both parties' interests to bring about a peace that cannot be undermined by a large population that infuses in its children a clear sense of historical betrayal.
Despite the explosiveness of the situation, the outlines of a sustainable peace are not difficult to draw. As should be clear from our discussion of failed peace plans, on substantive grounds the parties to this conflict are not intransigent or against compromise, despite their public rhetoric. It seems to us that what has been lacking is a formulation that can reconcile the two principles of territorial integrity and national self-determination without one undermining the other. What follows is our framework for a negotiated settlement that might well bring these principles together.
A 13-POINT FRAMEWORK
- Karabakh must remain de jure in Azerbaijan, in accordance with the principle of the territorial integrity of states and the inadmissibility of changing borders unilaterally through the use of force. The symbolic sovereignty of Azerbaijan over Karabakh would be represented by an Azerbaijani flag flying over the government house in Karabakh and the appointment of an Azerbaijani representative to Karabakh, who would have to be approved by the Karabakh government. The formal aspect of sovereignty entail is that Azerbaijan represent Karabakh in the United Nations and other international bodies.
- Karabakh citizens would have proportional representation in the Azerbaijan parliament in Baku. A majority of Karabakh representatives in the Azerbaijan parliament would have the power to suspend within Karabakh any proposed law that directly affects Karabakh.
- It should be appreciated that even with representation and veto rights, the victorious Armenians would consider this recognition of official Azerbaijani sovereignty over Karabakh as an enormous, almost unbearable concession, not quite akin to that of the Israelis turning over part of Palestine to the Palestinians, but large enough indeed. This recognition by the Armenians would be offset by the establishment of full self-government for the Republic of Karabakh within Azerbaijan. This would observe the principle of national self-determination and must be confirmed in a referendum within Karabakh. Karabakh self-government would entail the following:
- The Republic of Karabakh would have its own parliament based on proportional representation of the population. No decision of the Azerbaijan government or parliament would be legal in Karabakh without the consent of the Karabakh president and parliament. There would be, in other words, no vertical subordination of Karabakh to Azerbaijan.
- The Republic of Karabakh 's government would have full rights in regard to policing, education, local investment in infrastructure, and culture;
- The Republic of Karabakh's government would collect its own taxes based upon Azerbaijani tax codes. It would receive transfer payments from Baku so that its per capita revenue would be equal to that of Azerbaijan (less Azerbaijan's expenditures for military and foreign affairs);
- A police force would be established by the Karabakh authorities for use within Karabakh. It would be paid for by the Armenians as long as the international peacekeeping force is in place. With the removal of the international force, the police would be paid for by part of the funding that previously supported that force.
- Azerbaijani soldiers or police would be allowed on the territory of Republic of Karabakh only under conditions acceptable to the parliament of the Republic of Karabakh.
- Armenians and Azerbaijan is living in Karabakh would have the right to dual citizenship or full citizenship in either republic with the right to permanent residence in Karabakh.
The above five conditions would be specified in the constitutions of the Republic of Azerbaijan and the Republic of Karabakh. The Republic of Armenia would amend its constitution where necessary to conform to this agreement.
- If the Karabakh government agrees, the Armenian government would have the right to aid Karabakh in setting up taxing authorities, school systems and voting districts within the Republic of Karabakh. These authorities could provide supplemental benefits to the Armenian population resident in Karabakh, and these benefits would not decrease the tax transfers owed to Karabakh from Azerbaijan.
- A fund would be set up under OSCE auspices for payments of compensation to victims on both sides and for the international policing of the agreement. While interested parties would be expected to contribute to this fund, a major input into the fund would be from Azerbaijani borrowing against future oil revenues and would be deducted as part of Azerbaijani contracts with international oil firms. (These revenues will be greatly enhanced with a peace agreement, so the peace dividend would be used to compensate the refugees and IDPs and to protect people from future violations of the peace accords.)
- Armenia and Azerbaijan must agree on a treaty guaranteeing transit rights through each of its territories. The blockade would be ended. Free and unfettered passage would be guaranteed between Karabakh and Armenia and between Karabakh and the rest of Azerbaijan. The Lachin corridor would be demilitarized and for an unspecified time would be a "transit zone" pending its final disposition.
- Former political Armenian enclaves outside Karabakh in Azerbaijan, like the Shahumian district, would be lost to Armenians; Azerbaijani enclaves within Armenia would likewise be lost. Compensation would be paid to those people displaced from their homes in Armenia and Azerbaijan who are not able to return to them.
10. Three sets of populations would be permitted to return to their prewar homes if they so desire. There would be an immediate return of Azerbaijani IDPs to their homes outside of Karabakh. The returnees would be provided loans to rebuild their homes and assurances for their safety. Likewise, Armenians who had lived in Azerbaijan would have their properties restored in Azerbaijan or receive compensation for their losses. Armenians who formerly lived in Azerbaijan would be given export and commercial licenses to develop businesses in Azerbaijan under highly favorable terms. Azerbaijanis who formerly lived in Armenia would be allowed to return to their homes or receive compensation for their losses.
- There would be a program of gradual return of Azerbaijani refugees to Karabakh proper. Generous compensation for the loss of property for those who chose not to return would be negotiated. Shushi (Shusha) would become a multinational town, with both Armenians and Azerbaijanis living there in approximately equal numbers.
- An international peacekeeping force, whose makeup is acceptable to both sides and to the government of the Russian Federation, would be deployed to monitor and enforce the peace. The force would be paid for by the OSCE-administered fund for 20 years, unless both parties agree to its disbanding. (This last point would give the Republic of Azerbaijan a strong incentive to earn the trust of the Armenian population, as it will be expensive to pay for the peacekeeping force; and it would give an incentive as well for the Armenians to earn the trust of the Azerbaijan authorities, so that they could quickly replace the international gendarmerie on the borders of Karabakh with local forces.)
- Armenia and Azerbaijan would agree to annual inspections and reports from an international organization, such as the OSCE or the Council of Europe, on any human-rights violations in their republics proper or in Karabakh.
With such a solution, Azerbaijan would regain formal sovereignty over its whole territory, while Armenians would have self-government and international protection. The refugee problem would be solved, and the way would be open for the economic development of the whole South Caucasus. Armenians, Azerbaijan is and Georgians could take advantage of the oil boom that is on the horizon whose rents could be spent on education, public services and commerce, rather than on revenge. (Azerbaijan would be paying an enormous cost for unilaterally funding the international peacekeeping force and the resettlement efforts; but this would be far cheaper than funding two generations of war.)
In this new solution, though each nationality would have self-government and control over distinct territories, economic and political links would be maintained and, hopefully, flourish. Thus, a shared sovereignty is envisioned, one in which different powers accrue to different authorities. Some powers remain with Baku, such as foreign representation; others remain with Karabakh, indeed almost all the necessary powers of local self-government and development; and some powers, largely having to do with taxation and finance, are shared. Security will be the province of both sides, as well as the international community. Finally, our solution tries to avoid the choice between a "package deal" and a "step-by-step" approach, for it suggests that the basic principles that affect status be agreed upon up front. Only then will the tough negotiations on details and the withdrawal of troops from occupied Azerbaijan (the Karabakh Armenians' principal means of leverage) be carried on.
The most unusual aspect of our proposal requires the Azerbaijani government (through funds deducted from oil revenues) to finance the international peacekeeping force and the resettlement programs. In one sense this proposal equalizes the sacrifice: Armenians give up territorial control over an area they consider "theirs"; Azerbaijan gives up funds from Caspian oil that is "theirs." In another sense, the proposal creates incentives for both parties to seek an end to external policing and to avoid a situation like that in Cyprus, where neither party has such an incentive. The benefits that would accrue to Azerbaijanis and Armenians from building trust would be seen by both sides as a plus.
Our proposed solution takes into consideration the international implications that settlement of one ethnic war would have in similar conflicts elsewhere (e.g., Kosovo, Abkhazia, South Osetia, Chechnya, Transdniestria, etc.). We are neither proposing full separation of the two peoples and the changing of territorial boundaries nor the acceptance of the consequences of ethnic cleansing. Instead we envision self-government for all peoples but within a new structure of shared sovereignty. In this way, the two principles of international law territorial integrity and national self-determination may be reconciled even though neither may be satisfied to its maximal degree. Each side wins, but neither wins everything it wants.
Despite the disembodied images that fill the media and people's minds, the struggle over Karabakh was never about indelible primordial cultural differences or hatreds fostered by religion. Both Armenians and Azerbaijan is are highly secularized populations that experienced similar educations and experiences in the years of Soviet rule. Azerbaijanis are barely influenced by mullahs, and in Baku they are marked by extreme cosmopolitanism. In Baku-based interviews, Laitin chatted with too many draft dodgers to believe in some Azerbaijani crusade against Armenia. The people of Baku wish fervently that a solution would be found to "buy off' the more rabid refugees so that everyone could do business in peace. It would be sad, indeed, if the residents of Baku were drawn into a generation of war by a population hardened by the common life experience of trying to survive in refugee camps. Armenians are exhausted by war and sacrifice and desperately desire a "normal" life and the prosperity that their leap into independence, democracy and capitalism promised. For them, Karabakh is not only irredenta regained but the first line of defense against another threat of annihilation. But for many Armenians there is also a growing sense that Armenia has sacrificed its future for Karabakh and that the latter's interests have overtaken those of Armenians more broadly. Mutual suspicions and memories of atrocities separate the two sides, and key politicians in each of the three capitals actually benefit from the current stalemate, profiting from war or the absence of peace and solidifying their positions as indispensable patriots. If they were to lift their eyes only slightly to more distant horizons, they should become convinced that their peoples' futures demand a resolution of the conflict soon.
As diplomats and scholars contemplate the future maps of the Balkans and the Caucasus, it has become clear that the old architecture of fully sovereign states dominated by a single entitled nation no longer works in places where populations have been mixed and mobile for centuries. To avoid perpetual conflict, some form of shared sovereignty must be imagined, one that gives each people security and rights of self-government while forging links to a common future. It is not hard to see what failure to solve the Karabakh problem will bring: unstable states and collapsed economies that make continued ethnic conflict more likely. Weak states in the South Caucasus create a positive incentive for the worst elements in Russia to entertain neo-imperial ambitions. From the late eighteenth through the mid-nineteenth centuries and again in the years of revolution and civil war at the beginning of the twentieth century, Russian expansion beyond the Caucasus and into Central Asia has historically been contingent on the weakness of the polities in those regions. No matter what kind of government comes to power in Russia, war and social collapse in the South Caucasus and Central Asia could encourage Russians back into the role of colonial power.
We have tried to show the benefits of shared sovereignty for Armenia and Azerbaijan. But there is another payoff that should interest the Great Powers. Unlike the Balkans, where Russian and Western interests have often clashed, the South Caucasus is a place where their interests more closely coincide. After years of frustration, Russia and the United States have come to similar conclusions about a solution to the Karabakh conflict, and they have jointly formulated the OSCE proposals. As a claimant to great-power status, Russia needs to show that it can resolve deadly conflict in its own backyard, and Moscow can only share in the oil revenues or pipeline profits from Azerbaijan if that oil flows securely.
Though rogue elements in Russia may see opportunities in an unstable south, others more sober are likely to calculate that stable states consolidating their authority and regaining a greater degree of legitimacy and popular support could be more reliable allies. Given their traditional ties to Russia, stable South Caucasian states may be less likely to respond to temptations offered by Iran and Turkey. With its leverage in Armenia, Russia is key to any solution. At a time when Russian-American relations are strained by issues like NATO expansion, revision of the ABM Treaty and Balkan politics, Karabakh offers a unique opportunity for these sometime "partners" to act in concert. This is one that both can win.
If we had to predict the future of the South Caucasus, we would propose two scenarios. The more pessimistic one sees the failure to solve the Karabakh problem, related to the failure to create stable states and economies in the region, leading to continued ethnic conflict and an incentive for the worst elements in Russia to intervene more directly. On each side there are those who want the whole pie and a few who would risk more deaths to defeat or destabilize the other. Influential Armenians calculate that they can wait until Aliev's death and stand by while the state of Azerbaijan dissolves into fragments. Important Azerbaijanis expect that the income from oil will bind them closer to the United States, Georgia and perhaps even Russia. In time, oil and the strategic partnership with great powers will shift the balance of power decisively to Azerbaijan. With such expectations in Baku, Erevan, and Stepanakert there is little incentive to settle the conflict now.
The optimistic scenario is contingent on ending the Karabakh war. It sees the states of the south continuing to consolidate their authority and regaining a greater decree of legitimacy and popular support. This development would give their governments the political muscle to move energetically to settle other outstanding interethnic conflicts, bringing peace not only to Karabakh, but to Abkhazia and South Osetia, and then to put greater effort into economic reform and development. The solutions to the ethnic-territorial conflicts will be resolved on the basis of full national self-determination of the respective nationalities (Karabakh Armenians, the Georgians and Abkhaz of Abkhazia, and the Osetians) within the existing independent republics, which retain formal sovereignty but accept full local self-government for their non-titular peoples. No international borders will be changed without consent on both sides. International peacekeeping operations would be required in Karabakh and Abkhazia, at least for the foreseeable future. Stronger republics in the south would encourage a more moderate policy on the part of Russia, for Russia's major strategic interest in the region is a secure buffer against intrusions from Turkey and Iran, not the full burden of colonizing a complexly mixed and resistant population.
Because of the interrelations of the economic, ecological and ethnic problems in the South Caucasus, regional solutions are essential. These small states with their long history of interaction and dependency can only grow stronger with the end of these debilitating conflicts. Oil development in Azerbaijan can be maximized only after peace and stability are achieved; Armenia and Georgia can only develop with open borders, invigorated regional trade, and the reduction of military spending. The piping of oil and gas through Armenia and Georgia would only consolidate the economic and security interests of the three republics. The future need not reflect the recent past of unbearable ethnic horrors. Though Karabakh will be distinct from the rest of Azerbaijan constitutionally and demographically, the fact that Azerbaijanis and Armenians will be allowed to return to their homes and that Shushi will be binational presages a renewal of older patterns of interethnic, interreligious coexistence. It is this optimistic scenario that some variant of our peace proposal can assure.
1 Patrick Donabedian, "The History of Karabagh from Antiquity to the Twentieth Century," in Levon Chorbajian, Patrick Donabedian and Claude Mutafian, The Caucasian Knot: The History and Geo-Politics of Nagorno-Karabagh (London and New Jersey: Zed Books, 1994), p. 53.
2 A. Abbasov and A. Memedov, "Natsional'nye i ethnicheskie men'shinstva v Azerbaidzhane," Azerbaidzhantsy: Istoriko-ethnograjicheskii ocherk (Baku: Elm, 1998), p. 63.
3 Frederik Barth (ed.), Ethnic Groups and Boundaries (Boston: Little, Brown and Co., 1969), p. 17.
4 Lee Schwartz, "Regional Population Redistribution and National Homelands in the USSR," in Henry R. Huttenbach (ed.), Soviet Nationality Policies (London: Mansell, 1990), pp. 121-161.
5 This new paradigm emphasizing the nation-making effects of Soviet nationality policy rather than exclusively focusing on the nation-destroying aspects has been developed in the works of Ronald Grigor Suny, particularly, Looking Toward Ararat: Armenia in Modern History (Bloomington and Indianapolis: Indiana University Press, 1993) and The Revenge of the Past: Nationalism, Revolution, and the Collapse of the Soviet Union (Stanford: Stanford University Press, 1993); Yuri Slezkine, "The USSR as a Communal Apartment or How a Socialist State Promoted Ethnic Particularism," Slavic Review, LIII, 2 (Summer 1994), pp. 414-452; and Arctic Mirrors: Russia and the Small Peoples of the North (Ithaca and London: Cornell University Press, 1994); and Rogers Brubaker, "Nationhood and the National Question in the Soviet Union and its Successor States: An Institutionalist Account," in his Nationalism Reframed: Nationhood and the National Question in the New Europe (Cambridge: Cambridge University Press, 1996), pp. 23-54.
6 Jon Elster, Institutional Design in Post-Communist Societies: Rebuilding the Ship at Sea (Cambridge: Cambridge University Press, 1998).
7 On the ownership of the state by nationalizing elites, see Brubaker, Nationalism Reframed, p. 5.
8 The Mountainous Karabakh Autonomous Region (Nagomo-Karabakhskaia Avtonomnaia Oblast') was formed in 1923. The Armenian population fell gradually. By 1959 Armenians made up 84.4 percent, but twenty years later they were just under 76 percent. Armenians feared a steady decline, even elimination from Karabakh, and often referred to the other autonomy in Azerbaijan, Nakhichevan, where they had almost completely disappeared demographically. [Suny, Looking Toward Ararat, pp. 193-195] A Kurdish autonomous region was formed around Lachin in the early 1920s and lasted about a decade before being abolished by the Soviet authorities. In subsequent censuses Kurds were counted as "Azerbaijanis." Robert O. Krikorian, "Red Kurdistan and the Struggle for Nagorno-Karabakh," Journal of the Society for Armenian Studies, 6 (1992-1993), pp, 61-81.
9 Audrey L. Alstadt, The Azerbaijani Turks: Power and Identity Under Russian Rule (Stanford, CA: Hoover Institution Press, 1992), pp. 207-208.
10 Visiting Karabakh in the summer of 1997, Suny observed that the local Armenians almost invariably referred to Azerbaijanis as "Turks."
11 These interviews were carried out from August I through 17, 1998, in collaboration with Jennifer Grocer and Carl Hershiser from the University of Texas and were supported by a grant from the Harry F. Guggenheim Foundation.
12 Figures are from UNHCR, "Refugees and Others of Concern to UNHCR1997 Statistical Overview," available at http://www.unhcr.ch/refbib/refstrat/l 998/lookeuro.htm.
13 Flying over the Lachin corridor in the summer of 1997 it was possible to see clearly dozens of buildings gutted, without roofs.
14 Most of the 8,000-9,000 "Russian" soldiers in Armenia are actually Armenians in Russian uniforms, serving under Russian officers.
15 Many of the top leaders in Azerbaijan are either from Nakhichevan or from Azerbaijani families originally from Armenia known as "Eraz" (Erevan Armenians). Origins and kinship connections are key to the formation of the "clans" that run Azerbaijan (and Armenia as well). Other groups, like Karabakh Azerbaijanis or the Tat-Talysh group, make up rival clans; many of those excluded from the Eraz gathered in the Musavat party, led by Isa Gambar, a non-Eraz.
16 John J. Maresca, "The International Community's Efforts to Resolve the Conflict over Nagorno Karabakh: A Case Study in Lost Opportunities for Conflict Resolution," a report prepared for The Carnegie Commission on Preventing Deadly Conflict, April 1995, p. 5.
17 Izvestiia, March 4, 1993, p. 2; cited in The Current Digest of the Post-Soviet Press, XLV, 9 (March 31, 1993), p. 17.
18 Svante E. Cornell, "Turkey and the Conflict in Nagorno Karabakh: A Delicate Balance,'' Middle Eastern Studies, XXXIV, I (January 1998), pp. 51-72. In 1998 Laitin observed uniformed Turkish Special Forces walking through Baku, and it is widely known that Turkish officers have been helping train Azerbaijani soldiers.
19 Ibid., p. 17.
20 Nezavisimaia Gazeta, September 22, 1993, p. 1; condensed translation in The Current Digest of the Post-Soviet Press, XLV, 38 (October 20, I 993), pp. 29-30.
22 See Elizabeth Fuller, "The Karabakh Mediation Process: Grachev versus the CSCE?" RFEIRL Research Report, III, 23, June 10, 1994.
23 Special Report of the United States Institute of Peace, "War in the Caucasus: A Proposal for Settlement of the Conflict Over Nagorno-Karabakh," (Washington, DC, 1994). Maresca in general saw Russia as imperialist and desiring to reestablish its influence, if not control, over the former Soviet Union. [See his, "The End of the Cold War ls Also Over," in Gail Lapidus and Renee De Nevers (eds.), The End of the Cold War Is Also Over (Stanford, CA: Center for International Security and Arms Control, Stanford University, April 1995), pp. 3-24. In the same brochure, Daniel Sneider proposes a transitory period for Transcaucasia of Finlandization political sovereignty but limits on foreign policy options. ("Comment: The Caucasian Test Case," The End of the Cold War ls Also Over, p. 53).
24 TransCaucasus, A Chronology, V, 2 (February 1996), p. 12.
25 In these negotiations Azerbaijan gave up its earlier insistence that Karabakh Azerbaijanis have an equal negotiating status with Karabakh Armenians.
26 "'Transcript of the Press Conference of President Levon Der Bedrosian," The Armenian Reporter Int’l, October 4, 1997, p. 21.