Autonomy as a Source of Conflict
Caucasian Conflicts in Theoretical Perspective

Svante E. Cornell *

[Tables]

Since the 1950s, ethnopolitical conflict has grown as a source of concern in the international arena. It culminated after the cold war with the eruption of conflict in the former Soviet Union and Yugoslavia. A number of conflicts also broke out between ethnically defined social groups in Africa and south Asia, in the postcommunist states of Eastern Europe and Eurasia, as well as in Western Europe. ¹ The reigning assumption that ethnic conflict was a vestige of the primitive past was revised and eventually abandoned, particularly in view of the spread of ethnic conflict to less developed regions. This led to increased media coverage and public awareness of ethnic issues; more importantly, academic research on ethnic conflict and its resolution mushroomed. ²

Ethnic mobilization among minority populations in multiethnic states has often led to demands for self-rule (territorial autonomy) or for outright secession. ³ Especially in defined geographical areas where minorities are compactly settled, the creation of a separate state is a feasible goal and territorial control becomes a chief issue of conflict. In situations in which ethnic groups live in overlapping settlement patterns, such demands are less feasible and are made more infrequently, making control over or influence in the central government the most [End Page 245] contentious issue. ⁴ Indeed, Fearon and Laitin find "regional concentration of minority group (a) powerful and robust factor. . . far more likely to see large-scale ethnic violence than urban or widely dispersed minorities." ⁵

Many theorists have found that solutions involving regional autonomy are effective in dealing with ethnic conflict. Ted Gurr, for example, has argued that "negotiated regional autonomy has proved to be an effective antidote for ethnopolitical wars of secession in Western and Third World States." ⁶ Likewise, Kjell-Åke Nordquist has observed that creating an autonomy--"a self-governing intra-state region--as a conflict-solving mechanism in an internal armed conflict is both a theoretical and, very often, a practical option for the parties in such conflicts." ⁷ Regional autonomy implies the introduction of ethnoterritoriality--territorial control linked to ethnicity. It occurs either when a region is explicitly created as a homeland for an ethnic group or when a minority group constitutes a large majority of the population of an autonomous state structure and perceives it as its own.

Central governments are nevertheless almost universally reluctant to accede to demands for autonomy for several reasons. First and foremost, they fear that granting territorial autonomy to a minority group would be merely the first step toward the eventual secession of the region. Second, granting autonomy to one region may be perceived as discrimination against other inhabitants or groups. ⁸ Third, autonomy increases the risk of intervention by a foreign state affiliated with the [End Page 246] specific minority population. ⁹ In spite of such reservations, however, an increasing number of ethnopolitical conflicts over territory have been settled by compromises involving regional autonomy, such as the provision of autonomy to the Basques of Spain in 1980, the Miskitos of Nicaragua in 1990, the Nagas in India in 1972, and the Afars in Ethiopia in 1977. The popularity of autonomy as a solution undoubtedly stems from its being one of the few conceivable compromise solutions in conflicts over the administrative control of a specific territory. Indeed, as will be discussed further, autonomy represents a compromise on the issue of state sovereignty itself.

Advocates of ethnofederalism argue that autonomy solutions are effective conflict-resolving mechanisms and that further federalization of multiethnic states along ethnic lines will help prevent ethnic conflict. In some of the literature, ethnofederalism has been characterized as what David Meyer terms a "cure-all prescription" for ethnic tensions. ¹⁰ There is, however, considerable reason to argue that the institution of territorial autonomy may be conducive not to interethnic peace and cooperation but rather may foster ethnic mobilization, increased secessionism, and even armed conflict. Whereas the merits of federalism were widely lauded in the literature from the 1960s to 1990, developments since then have generated doubt that ethnofederal solutions can effectively prevent ethnic conflict. Several researchers have noted--usually in passing--how federal structures may be counterproductive under certain circumstances. ¹¹ Yet no systematic inquiry has been made into how and why federal structures, designed to mitigate centrifugal forces, instead may end up strengthening them. This article outlines a rudimentary theoretical framework that may explain why ethnofederal constructs, specifically territorial autonomy, may cause rather than prevent conflict. After presenting the logical case against territorial autonomy, the specific case of the South Caucasus and in particular the post-1991 [End Page 247] developments in the Republic of Georgia will be analyzed. Georgia contains five compactly settled minorities, three of which were autonomous at independence; it presents an opportunity to compare developments among minorities with different status. Moreover, given Georgia's small size, the similarities in political development at the central level and its effect on minorities, and an analogous international context, the five cases are comparable.

Theoretical Aspects on Autonomy

A number of authors have attributed the collapse of three communist federations--Yugoslavia, Czechoslovakia, and the Soviet Union--to their ethnofederal character. ¹² In each case, the weakening of the central state structure and its eventual dissolution were intimately related to the centrifugal policies pursued by powerful national elites in the ethnically defined component republics. In Yugoslavia, Slovenian and Croatian demands for a national restructuring along confederal lines was the starting point of its demise; in the former Soviet Union, centrifugal tendencies in the Baltic republics, the Caucasus, and Russia itself contributed significantly to its dissolution. However, literature on federalism and ethnopolitical conflict has not explored this problem at a lower level of analysis--that of autonomous regions. Indeed, Yugoslavia and the Soviet Union were federal states whose component republics technically were on equal footing and had only the nonterritorial, nonethnic federal center above them. The federal center was legitimized by the civic, ideological identity of the state. (Although it has been argued that the Soviet Union was dominated by ethnic Russians and Yugoslavia by ethnic Serbs, the extent of this assertion has been credibly challenged. After all, Stalin was a Georgian, Nikita Khrushchev a Ukrainian, and Marshal Tito a Croat.) The case of autonomous regions is different. As one or several specific minority regions have been granted autonomy, segregating them from the rest of the country, the legitimacy of the central government increasingly rests on an ethnic and territorial basis--the nonautomonous areas, practically meaning the majority ethnic group in the country.

Autonomy in a political and legal context refers to the power of social institutions to "regulate their own affairs by enacting legal rules." In [End Page 248] international law, autonomy is taken to mean that "parts of the state's territory are authorized to govern themselves in certain matters by enacting laws and statutes, but without constituting a state of their own." ¹³ This refers to territorial autonomy, which gives an ethnic group self-rule--political authority over a certain territory--in order to govern its internal affairs to a determined extent. Cultural autonomy, by contrast, is a scheme whereby members of particular ethnic communities are endowed with specific rights and duties in relation to the government. This is also used in certain countries for religious communities--in Israel for Muslims and Christians and in India for Muslims. Members of particular groups may also be given special rights to preserve their culture and language, often through the institution of native language schools for minorities. Cultural autonomy is not territorially based; it may still, nevertheless, be either individually or group based and either voluntary or compulsory. ¹⁴ Both forms of autonomy must be grounded in the legal system of the particular country and as such form a part of that state's system of government. ¹⁵ In a broader sense, autonomy can be defined as "the granting of internal self-government to a region or group of persons, thus recognizing a partial independence from the influence of the national or central government," which can be determined "by the degree of actual as well as formal independence enjoyed by the autonomous entity in its political decision-making process." ¹⁶ Territorial autonomy is usually considered synonymous with "self-government" as stated in the UN Charter--free of references to sovereignty or independence--thereby avoiding automatic conflict with the territorial integrity of states. ¹⁷ Cultural autonomy does not carry the far-reaching consequences implicit in territorial autonomy: territory and ethnicity are not linked, there is no creation of statelike institutions. Therefore, unless otherwise specified, this article will use the term autonomy to refer to territorial autonomy.

Autonomy Regimes in the Literature: Advantages and Disadvantages

The advantages of autonomy regimes are relatively well known. Given the multitude of ethnic groups in the world, advocates of autonomy [End Page 249] argue that group rights need to be recognized below the state level in order to avoid the proliferation of hundreds of additional states. The traditional structure of the international system is already threatened by the relative reduction of the role of states in international affairs by the increasing importance of substate entities such as ethnic, national, or religious groups, as well as by suprastate entities such as regional and international organizations. Autonomy is basically the only possible compromise to balance the conflicting territorial interests of the group and the state; moreover, the flexibility inherent in the concept of autonomy, as it may be tailored to each particular situation, enhances the ability to reduce ethnic tensions. Autonomy, therefore, may work both to prevent and to resolve ethnic conflict.

Autonomy regimes, however, by protecting diversity, necessarily rely on the assumption that differences and their institutionalization enrich the world more than they endanger it. According to Steiner: "Autonomy regimes of ethnic minorities defend cultural survival rights in counteracting [the trend toward homogenization that has accompanied Western development]." ¹⁸ Indeed, the international system seems to be moving toward a system of norms that protects difference by pressuring states into creating autonomy regimes for minorities. But such norms raise obvious and serious issues: "The ideal in the human rights movement of preserving difference cannot so readily be bent to support the creation of autonomy regimes." ¹⁹ Whereas such regimes are based on the norm of equal protection, autonomy schemes imply forms of institutionalized separateness that violate the very norm of equal protection in that they "explicitly discriminate among groups on grounds of religion, language, race, or national origin . . . [and thereby] drive home the lesson that socioeconomic life and career turn on ethnic bonds." Furthermore, autonomy regimes not only preserve but also lock into place historical differences between groups; it is difficult to disagree with the claim that "a state composed of segregated autonomy regimes would resemble more a museum of social and cultural antiquities than any human rights ideal." ²⁰

Other authors have argued that autonomy, by involving the differential treatment of a certain group, may result in protests by other groups and thus lead to conflict rather than preventing it. ²¹ A unitary state, by contrast, through integration--but with mechanisms for the full respect [End Page 250] of individual human rights as opposed to collective rights--provides equal opportunities and identical rules for all citizens of the state, irrespective of color, ethnicity, or religion. The unitary state thereby counteracts the polarization around such issues that would arise from solutions such as autonomy or federalism. Autonomy may in fact isolate the minority and prevent its members from political or economic participation in the larger sphere of the state. Accordingly, it makes dialogue between groups within the society difficult, alienates component groups from one another, and leads to segregation. ²² Lyck's analysis of the Faroe Islands' autonomy in Denmark illustrates one of the negative effects of autonomy--that it led the state to feel less responsible for the development of the region. ²³ The general sentiment of the literature is that the advantages of autonomy nevertheless supersede its possible drawbacks. This can only occur, however, if the autonomy is designed, created, and maintained with necessary safeguards providing mechanisms to ensure the regulation of possible future conflicts and for eventual alterations of the autonomy's status. Autonomy is not automatically a recipe for success; to the contrary, it is a solution that brings a number of dangers and risks.

Autonomy and Sovereignty

Autonomous regions, by their very nature, are conducive to secessionism. The relationship between the central government of a state and an autonomous region resembles neither the horizontal relationship between sovereign states nor the vertical relationship between a state and its citizenry, regardless of whether they be organized politically along ethnic, religious, or ideological lines. When a central government grants autonomy to a given region, it acknowledges the devolution of a certain portion of its own sovereignty to the representatives of that region's population; the central government concedes that it no longer has unlimited jurisdiction over the territory--herein lies the essence of autonomy. At the same time, however, the central government emphasizes the subordination of the autonomous region to itself in that the existence of the latter in no way compromises its own territorial integrity. Hence, the relationship between the two units can be described as diagonal; an autonomous region can be conceived of as a state within a state, even if neither party officially acknowledges it. [End Page 251]

Autonomous regions are typically constructed in the manner of states--with executive, legislative, and judicial bodies, as well as statelike symbols such as flags and coats-of-arms. In fact, autonomies may share most attributes of a state but never by definition the primary attribute--complete sovereignty. An autonomous region, while often claiming sovereignty, is by definition a part of a sovereign state; the sovereignty of the autonomy is always partial, confined to certain determined spheres such as culture and economy. Yet, the institution of an autonomous region nonetheless implies that the state itself is no longer completely sovereign; it has agreed to share its sovereignty with the autonomous region--albeit on an unequal basis. One extreme example, the Republic of Bashkortostan, is defined as "a sovereign state within the Russian Federation." The definition may seem a contradiction in terms, but it exemplifies the variety of tailored solutions available for navigating contending claims over sovereignty. ²⁴

That there is no blueprint for the conduct of relations between an autonomous region and the central government contributes to the usefulness of autonomy as a mechanism of conflict resolution or prevention--it is flexible and adaptable to the specific grievances of a specific minority. Within the "society" of independent and sovereign states, relations between all members are based on certain generally accepted principles, such as the equality of states, noninterference, and the inviolability of borders. The increasing universality of principles of human rights, and of democratic government are examples of growing but not yet fully accepted rules governing state-citizen relations. The relationship between a central government and its autonomous region(s) shares elements of both of these relationships.

Autonomy and Secessionism

The institution of autonomous regions is conducive to secessionism because institutionalizing and promoting the separate identity of a titular group increases that group's cohesion and willingness to act, and establishing political institutions increases the capacity of that group to act. ²⁵ [End Page 252] Autonomy affects each of the following areas: borders, group identity, state institutions, leadership, mass media, and external support.

Borders

An integral characteristic of autonomous regions is that they have recognized and clearly delimited borders--which in the Soviet case appeared on most maps of the union. The importance of borders is emphasized by Benedict Anderson in his classic work Imagined Communities and Thongchai Winichakul in his recent work Siam Mapped. Winichakul refers to Southeast Asia as "a map anticipated spatial reality, not vice versa. In other words, a map was a model for, rather than a model of, what it purported to represent . . . it had become a real instrument to concretize projections on the earth's surface. A map was now necessary for the new administration to back up their claims. . . ." ²⁶ Anderson points to the special importance of what he terms the "map-as-logo," a map on which place names, rivers, mountains, and neighbors all disappear and only the borders of the territory in question remain; the map is hence "pure sign, no longer compass to the world." As he demonstrates, the map in this format can then be used for "transfer to posters, official seals, letterheads, magazine and textbook covers . . . instantly recognizable, everywhere visible, the logo-map penetrated deep into the popular imagination, forming a powerful emblem for the anticolonial nationalisms being born." ²⁷ The same process occurred in the former Soviet Union with respect to its component entities. The maps--with the borders and shapes of the individual republics, autonomous republics, and autonomous regions--were long in existence by the 1980s and had been so for as long as most inhabitants could remember. As in Anderson's example, these maps and borders antedated spatial realities since the internal borders of the Soviet Union carried little historical and practical importance. For autonomous minorities, however, the republic or region's shape, map, or borders had significant symbolic importance, being a given for most of its inhabitants. With the breakup of the Soviet Union, this symbol became an important rallying point and an important tool in the hands of political entrepreneurs. On a more practical note, the task of delimiting the borders of the imagined new state was already completed, an obvious advantage over the situation of nonautonomous minorities. [End Page 253]

Group Identity

Several researchers have recognized the importance of autonomous entities in sustaining, promoting, and enhancing group identity and cohesion. Gurr, for example, notes that "the capacity for collective action is relatively high [in the case of] groups that control an autonomous regional government." ²⁸ The primary instrument for the promotion of ethnic identity is the education system. According to Dmitry Gorenburg:

Instilling a strong sense of ethnic community in individuals requires them to be exposed early and frequently to information about their ethnic identity. In the context of Soviet nationalities policy, this exposure came primarily through the education system. By establishing separate systems of native language education for most of the minority ethnic groups that had their own ethno-territorial administrative units, the Soviet government in effect created an institution dedicated to instilling a common and separate identity among the students . . . the identity was further reinforced in the classroom, where titular students were taught the culture and history of their ancestors, who were portrayed as having a direct genetic link with the members of the modern ethnic group. ²⁹

State Institutions

Autonomous regions typically possess statelike institutions that can be crucial factors in promoting ethnic mobilization. Unlike nonautonomous minorities, minorities in autonomous regions typically have governments and parliaments that act as legitimate representatives of their ethnic constituencies and constitute legitimate decision-making bodies. Parliaments can pass language laws, refuse to accept legislation from the central government, and issue declarations of sovereignty and independence. A minority with autonomous status hence has institutions for challenging state authorities in general and its specific policies and actions in particular. A minority lacking such institutions, by contrast, would find mounting such a challenge more difficult. Popular movements, petitions, and demonstrations may in certain contexts be effective ways to influence state policy; however, even organizing such shows of dismay are considerably easier if autonomous structures already exist. Beyond increasing the sense of legitimacy of the actions taken by the minority, decision-making structures are crucial in any attempt to raise ethnopolitical demands from the level of quiet dissatisfaction to that of direct action. As Meyer notes, autonomy "institutes a stratification of authority, subordinating administrative personnel into a [End Page 254] defined hierarchy"; moreover, it "establishes standard operating procedures and positive sanctions for the execution of its bureaucratic roles, and negative sanctions for poor performance, however that is defined by the Autonomous Structure leadership." ³⁰ In other words, the existence of a nationalist leadership in the autonomous structure often compels the entire bureaucracy to follow suit and adopt a more nationalist profile.

Leadership

The very fact that autonomous regions have governments also means that they have leaders--essential to any process of mobilization. The position of the leadership of an autonomous region, having a relatively strong base on which to stand, is therefore institutionalized in a manner that the leadership of a regular popular national movement cannot be. As Meyer suggests, autonomy gives "a stamp of legitimacy to its executives and to the rule of the titular ethnic group," and "facilitates improved cohesion of various ethno-politically mobilizing nationalists by providing a single institution around which they can unite." ³¹ Institutionalization also formalizes rules for succession, helping ensure that a "national struggle" could withstand a change in leadership. The existence of autonomous structures, especially in regions where the titular ethnic group is the demographic majority, also increases the likelihood of politicians promoting ethnic mobilization to further their own ambitions. Since the institution of autonomy is the source of power for leading regional elites, the leadership has a vested interest in increasing their region's level of self-government; elite power is positively correlated with the level of autonomy. Consequently, the elite has a vested interest in maintaining high nationalist sentiment among the population, thereby ensuring pressure from below to sustain or enhance the level of autonomy.

Mass Media

Governmental authorities in autonomous regions also often control the mass media--including television, radio stations, and newspapers. Accordingly, these authorities not only can influence the attitudes of the population in the long term through the education system, but often can influence the population directly through news coverage and depiction of events in media--plainly speaking, propaganda--and speed up the process of ethnic mobilization. [End Page 255]

External Support

The international political (and perhaps legal) standing of an autonomous minority is superior to that of a nonautonomous minority because, as previously mentioned, autonomous minorities possess institutions and the granting of autonomy entails the recognition by the state of the devolution of its sovereignty. Therefore, external support is more likely to be forthcoming for an autonomous minority since there is an institution in place to which funds and other types of support can be channeled.

As enumerated, there are a multitude of factors by which autonomy could hinder attempts to create interethnic harmony and peace. Even though the practical implications of this proposition, if proven correct, are large, they would not necessarily imply that autonomy is an unworkable construct to be avoided at all cost. In particular, it is necessary to draw the important distinction between postconflict situations and those in which armed conflict has not occurred. Where there is armed conflict, especially if it is over territory, it may be unavoidable to take territorial claims into account. Where a minority group has de facto control of a portion of territory, it is often unrealistic to assume it will surrender claims to any of its authority over that territory. In such situations, territorial autonomy, though imperfect and potentially hazardous, may be the only available or feasible compromise. Where no armed clashes have occurred between ethnically defined groups, however, it is both desirable and practicable to avoid the ethnicization of territory through the institution of autonomy. Where there is still a possibility of supporting cross-cutting identities and discouraging the linking of territory to ethnicity, this should be done, and ethnofederal solutions should be avoided.

Autonomy and Rival Explanations

Before moving to a deeper analysis of the regional developments, it is necessary to address several critics' potential challenges--that the analysis of autonomy in the former Soviet context is misplaced, given that true autonomy did not exist; a credible argument could even be made that for most practical purposes, the Soviet Union differed little from a unitary state. However, this would miss the point--one of the mechanisms through which autonomy operates with relation to conflict is in the realms of institutional structure and symbols. Moreover, in certain realms such as education and, equally important, the creation of national cadres and elites, federal structure was functional throughout [End Page 256] the Soviet era even though real political autonomy was absent. As Carol Skalnik Leff has argued with reference to the union republic level in the former Yugoslavia and USSR, ethnofederal institutional arrangements during the time of transition took on in practice the role they were earlier accorded on paper: "It is in that context that the bargaining environment for ethnonational disputes clearly differs from that of unitary multinational states: in the course of political opening, federal structures provide republic-level political bases for challenges to the existing political order and offer distinctive opportunities to key actors in the transition." ³² This statement, although written in the context of union republics under a nonterritorial federal center, is equally valid for the case of autonomous regions.

Another objection that can be preempted would be the assertion that the minorities that were initially granted autonomy in the first place were those minorities with greater grievances and in a higher degree of conflict with their central government. After all, autonomy is normally granted in response to ethnic demands; groups that have expressed such demands are arguably more likely to experience renewed secessionism than minorities that have not voiced such claims in the past. However, the more than thirty autonomous regions that were created in the Soviet Union in the 1920s and 1930s were not established as a result of ethnic demands. The very structure of the Soviet state was built on ethnic federalism; minority groups were mapped, evaluated, and assigned a certain status, often according to the whims of the highest decision makers, notably Stalin himself. The full explanation as to why certain minorities received autonomy and others did not may never be available; it is relatively safe to argue, however, that the decisions had little to do with actual ethnic demands.

At the end of the Soviet era in the three South Caucasian republics of Armenia, Azerbaijan, and Georgia, there were nine compactly settled minorities. ³³ Four had an autonomous status: the Armenians of Nagorno-Karabakh in Azerbaijan and the South Ossetians in Georgia held autonomous regions; and the Abkhaz and Ajars in Georgia held autonomous republics with a higher level of self-determination within the Soviet state structure. The nonautonomous minorities were the Azeris of Armenia, the Azeris of Georgia, the Armenians of Georgia, [End Page 257] and the Lezgins and Talysh of Azerbaijan. ³⁴ In the upheaval of the late Soviet years and during the transition to independence of Armenia, Azerbaijan, and Georgia, three violent ethnic conflicts broke out, all between central governments and autonomous minorities. This outcome, upon first reflection, seems counterintuitive because in a region where both autonomous and nonautonomous minorities exist, one logically would assume that nonautonomous minorities would have more grievances and therefore would be more likely to challenge the central government than those enjoying autonomous status. Nonautonomous minorities would tend to demand a status similar to that enjoyed by those already endowed with autonomy, whereas autonomous minorities would likely tend to be more satisfied with their current status. However, the pattern of conflict in the South Caucasus of the late 1980s and early 1990s did not follow such logic. From the start of the process of political liberalization under Mikhail Gorbachev in 1986, no broad-based, well-organized, or credible separatist movement emerged among any of the five nonautonomous minorities, whereas all four autonomous minorities displayed high levels of separatism, with all but one case (Ajaria) ending in armed conflict. Is this apparent correspondence spurious or is it indicative of a causal link between autonomy and conflict? The literature on the causes of ethnic conflict suggests that a wide variety of factors play a role in explaining the occurrence of conflict or the lack thereof. To prove that autonomy was indeed decisive, it must be isolated from other factors--which range from cultural differences over discrimination and geographic, topographic, and economic conditions to external factors.

Amid this multitude of explanatory factors, it should be noted that the South Caucasus displays certain specific characteristics that facilitate the isolation of autonomy as a root cause of conflict. First, the Caucasian republics' membership in the Soviet Union brought with it similar levels of both political freedom and discrimination. This also held true for the different minority populations in the individual republics. Although minorities such as Armenians in Azerbaijan or Abkhaz in Georgia claim to have suffered more discrimination than other groups, this allegation and its political implications must be put in proper perspective. The Soviet Union was a totalitarian state that showed little respect for the human or political rights of its citizens, thereby making it difficult to determine whether specific groups or the entire population were the targets of state abuse. Who was to blame for [End Page 258] discrimination--the republics or the central Soviet state? Blaming republican governments would actually be somewhat illogical if major decisions were taken in Moscow; regardless, discrimination is largely a matter of perception. It is ultimately irrelevant in assessing the likelihood that conflict will ensue whether discrimination has actually taken place; what matters is whether there is the perception of discrimination. While there have been exceptions to this, most notably in the case of the "repressed peoples" deported under genocidal conditions during the Second World War, no population group present in the South Caucasus today was ever subjected to such treatment. ³⁵ And perceptions of discrimination were entertained by elites in autonomous regions, whereas no comparable elite existed to do so for nonautonomous minorities. In terms of geography, all minorities in this study were located in border regions of their respective republics, contiguous with ethnolinguistically related peoples living across the border. All minorities are comparable in terms of size and are all vastly inferior numerically to the titular population of the state they inhabit. Moroever, given the relatively small geographic size of the South Caucasus and the existence of equally "porous" Soviet military installations in the three republics during the period of transition from communism, the availability of weapons was comparable for all minorities. Indeed, the entire region is saturated with arms, another factor that can be treated as a parameter in the study. Hence, the specifics of the Caucasian situation permits the elimination from the study of those factors on which all cases have equal scores.

Yet given the diversity of factors cited in the literature that have a bearing on the likelihood of ethnopolitical conflict, there are many factors besides autonomy that can potentially account for the variation in outcome. The factors pertinent to the study include the extent of cultural differences between the minority group and the titular population of the state; whether the national conception of the state was civic or ethnic in character (the latter being less conducive to accommodation); the intensity of past conflict and mythification thereof; the topography of the minority's region, that is, the existence of rough terrain, particularly mountains; ³⁶ whether the minority demographically dominated the region it inhabits; whether minority populations had ethnic kin in [End Page 259] neighboring countries; whether the minority-populated region was economically viable; whether a radical leadership existed among the minority population; and finally, whether there was external support for the minority. Several of these factors interplay with autonomy; indeed, for reasons described earlier, autonomy increases the likelihood of mythification of past conflicts as well as of radical leadership. All the factors listed above can be formulated as propositions that are expected to correlate positively with the level of ethnic mobilization and hence also with the likelihood of a minority challenge to the central government--and hence a priori with the level of conflict. ³⁷

One can obtain a rough indication of the role played by autonomy by comparing its correlation with conflict with its correlation with other factors. Although the number of cases and number of independent variables in this study precludes any statistically significant outcome, contrasting the presence of various factors, including autonomy, with the existence or not of conflict in the given case, provides a measure of the explanatory value of the possible causal factor in question. In statistical terms, this corresponds to the Fisher Exact Probability Test, a technique for analyzing discrete data with small samples. This is done, first, by assigning a straightforward score of "yes" or "no" to each case on the occurrence of conflict--Abkhazia, Nagorno-Karabakh, and South Ossetia are cases in which an armed conflict has occurred, whereas the other six are not. Then, in the same manner, a score of "yes" or "no" is assigned to each case for each possible causal factor. The explanatory value of each factor is then assessed by viewing its correlation with the occurrence or nonoccurrence of conflict. In case of a perfect covariation, all areas of "no conflict" would have a "no" score on the factor in question, and cases of conflict would all have a "yes" score.

Table 1 summarizes the findings of this overview of causal factors. A case is considered as supporting the proposition if a "no" score on the factor is matched by a "no" score on the occurrence of conflict, or if a "yes" score on the factor is matched by a "yes" score on the occurrence of conflict. ³⁸ [End Page 260]

Of the ten possible causal factors surveyed, none displays a full correlation. The explanatory value of three factors is supported in seven of nine cases, but only one indicator--autonomy--is supported by eight of the nine cases. Only the case of Ajaria (in Georgia) does not support the proposition that autonomy is a factor leading to conflict; it represents a case of an autonomous region that had not experienced armed conflict with its central government. This finding does not jeopardize the argument made in this study; as mentioned earlier, autonomy is neither a necessary nor a sufficient factor for conflict to take place. Indeed, of the more than one dozen autonomous republics in the Russian Federation, only one--Chechnya--has engaged in armed conflict with Russia since the political liberalization of the USSR began in the late 1980s.

This brief survey strengthens the basis of the theoretical argument previously presented, since autonomy more than any other factor mentioned in the literature accounts for the emergence of ethnic conflict in the Caucasus. This, however, does not prove that autonomy indeed acts as a cause of ethnic conflict. Perhaps, given the insufficient number of cases for a statistically satisfactory result, the correlation could be spurious. Moreover, this survey does not add to our understanding of the mechanisms that in practice make autonomy a conflict factor; nor does it explain under which circumstances, and in interrelation to which other factors, autonomy becomes a contributing factor to ethnopolitical conflict. It is therefore important to further investigate the available empirical record. [End Page 261]

Autonomy in Practice: Conflicts in Georgia, 1987-2000

The comparability of the nine cases in the brief survey of Armenia, Azerbaijan, and Georgia is hampered by the divergent political and economic developments in these three countries. General governmental policies, turmoil at the central level, and the foreign relations of any individual state may have affected its policies toward minorities. While a detailed study of all nine cases is beyond the scope of this article, it is both expedient and useful to focus on Georgia, which displays the full range of variations found in this study. Georgia still includes five compactly settled minorities; the Ajars, South Ossetians, and Abkhazians have held autonomous areas since the 1920s, whereas the Armenians and Azeris have never had any autonomy. The following analysis will cover events from 1987--when the first movements toward dissociation with the Soviet Union emerged in Georgia--to 2000. During this period, armed conflict occurred in Abkhazia and South Ossetia. Ajaria has maintained a high level of autonomy, involving a sometimes high level of political but not armed conflict with the government of Georgia. The Armenian minority in the Javakheti region has occasionally expressed dissatisfaction with its situation, but has not seen any major ethnic mobilization. Finally, the Azeri minority has been almost completely quiet during this period.

The most violent armed conflict in Georgia took place in Abkhazia, where the separatist leadership managed to secure control over the entire territory, even though the ethnic Abkhaz formed less than 2 percent of Georgia's population and only 17 percent of the population of their own autonomous region. A similar phenomenon, although to a lesser degree, occurred in the South Ossetian Autonomous Region, where the Ossetians constituted two-thirds of the population but only numbered sixty-seven thousand; yet their separatist leadership still managed to gain control of half the territory. How could such developments have taken place in these two regions especially as the Georgian government was adamantly opposed to secession and fought it aggressively?

By contrast, even though political observers for a decade warned of armed conflict involving the Javakheti Armenians, no such conflict occurred, in spite of existing tensions. In fact, among the six cases in which no conflict took place, Javakheti's scores indicated the highest propensity for conflict (seven indicative factors, compared with five indicators for Ajaria, four indicators in three other cases, and three factors in the case of the Talysh of Azerbaijan). Javakheti's scores differed from those of Abkhazia or South Ossetia only by its lack of autonomy. [End Page 262] Why did a credible and strong separatist movement not emerge in Javakheti, as indicators would show?

Finally, Ajaria represents the only case of an autonomous region in the South Caucasus not engaging in violent conflict with its central government. Though few other indicators pointed to a high risk of conflict there, Ajaria had experienced many of the same circumstances that led to conflict in Abkhazia and South Ossetia. How then did the Ajars avoid armed conflict?

These four cases, as well as the Azeris, will be examined in further detail.

Abkhazia: The Impossible Happening

The conflict in Abkhazia occurred against all odds. ³⁹ In 1989, the ethnic Abkhaz formed only 17 percent of the ASSR population of half a million, while Georgians accounted for 45 percent, Armenians 14 percent, and Russians 12 percent. Interethnic tensions had erupted briefly in 1978 and 1988 but remained limited. In June 1989, however, ethnic clashes in the capital Sukhumi left a dozen dead and hundreds wounded. ⁴⁰ Despite these incidents, Abkhazia was relatively calm during the rule of the nationalist politician Zviad Gamsakhurdia in 1990-92, whereas all other minorities in Georgia--including the Ossetians, Armenians, Ajars, and Azeris--had uneasy relations with the center. It was after Gamsakhurdia's fall from power in early 1992 that tensions began heating up between Tbilisi and Sukhumi.

Historian Vladislav Ardzinba was elected chair of the Abkhaz Supreme Soviet in December 1990. Soon after, a new electoral law was adopted, providing for a sixty-five-seat parliament. Twenty-eight seats were reserved for the Abkhaz, twenty-six for Georgians, and the remainder were distributed among the Armenians, Russians, and Greeks. Hence, despite only constituting 17 percent of the population, the Abkhaz controlled 43 percent of the parliamentary seats. A parliament was elected along these lines in fall 1991 amid unrest in Tbilisi that eventually brought down the Gamsakhurdia regime, but it soon split into two factions--an Abkhaz-led group composed mostly of non-Georgian deputies and the Georgian group. The main dispute between the factions occured over the need for two-thirds majority on "important" issues that had been defined only vaguely in the electoral code. It was resisted by the Abkhaz-led group but insisted upon by the Georgians, who saw the measure as a guarantor of their position. [End Page 263]

Prior to the election of the parliament, tensions arose over a Moscow-sponsored referendum on a new treaty to restructure the Soviet Union. Whereas the Georgian government, seeking to secede, refused to hold the referendum, it was held in Abkhazia and South Ossetia, where the Georgian population loyal to Tbilisi boycotted the vote. Developments of a regional nature also took place in Abkhazia, when in November Sukhumi hosted a congress of Mountain Peoples of the Caucasus (where Ossetians and North Caucasian peoples including the Chechens were represented) that adopted a document establishing a "Confederative Union of Mountain Peoples of the Caucasus". Throughout 1991, Abkhazia continued to distance itself increasingly from Georgia by building up a political system enshrining the dominance of the Abkhaz ethnic group and by forging ethnic coalitions both in Abkhazia (with Armenians and Russians) and regionally (with north Caucasian peoples). These centrifugal developments drew attention in Tbilisi, and in early summer 1992, a high-level Georgian delegation traveled to Sukhumi to discuss the division of powers between Tbilisi and Sukhumi, but the talks led nowhere. ⁴¹ As tensions rose between the two sides, the Abkhaz leader, Vladislav Ardzinba, soon declared Abkhazia "strong enough to fight Georgia," a somewhat surprising statement given the Abkhaz's demographic position and lack of military equipment or training. ⁴² Nonetheless, that same summer, Abkhazia reinstated its 1925 constitution defining it as an independent state. ⁴³

This ethnopolitical activity at the helm of the Abkhaz ASSR would have been impossible without ethnic Abkhaz domination of political life in the autonomous republic. As the titular nationality, the Abkhaz benefited from affirmative action policies that ensured full control over republican institutions despite their vast numerical inferiority. In addition to the quota of seats in the republican parliament reserved for ethnic Abkhaz, in practice, more than two-thirds of government ministers and local communist party department heads were also ethnic Abkhaz. ⁴⁴ Hence by forming alliances with segments of the Russian and Armenian populations, guaranteeing control over the parliament, the Abkhaz could dominate the political development of the republic and [End Page 264] guide policy toward the central government in Tbilisi against the wishes of the Georgian plurality. The existence of autonomous structures was a sine qua non for the Abkhaz to control the political institutions and thereby the territory of Abkhazia, compensating for their weak demographic status.

Despite controlling these republican structures, the challenge of establishing an independent Abkhazia was daunting. Achieving a peaceful secession appeared unlikely, particularly given the violent Georgian response to Ossetian separatism. Moreover, the Abkhaz not only had to contend with the resources mobilized by the entire Georgian state, but also had to deal with the many ethnic Georgians loyal to Tbilisi living in Abkhazia. Given that the prospect of achieving independence by arms must have seemed far-fetched, the confidence displayed by Abkhaz authorities during summer 1992 looks perplexing.

The events that followed, however, shed some light on the calculations that may have underlain their confidence. Poorly controlled and disciplined Georgian paramilitary forces attacked Abkhazia in mid-August, occupying Sukhumi and driving back the Abkhaz formations to the Russian border. Yet the Abkhaz counterattacked in early October, suddenly equipped with heavy armaments, help from North Caucasian volunteers, and Russian air support. Sukhumi was eventually recaptured by Abkhaz forces in September 1993, and virtually all Georgians living in Abkhazia were evicted. An unstable cease-fire has essentially held since late 1993, interrupted in early 1994 and during May 1998. ⁴⁵ The sudden increase in ethnopolitical assertiveness in Abkhazia in 1992 may have been partly conditioned by existing knowledge that outside support would be forthcoming in the event of conflict. The close relations between Abkhaz leaders and Russian military forces in the North Caucasus are fairly well known, ⁴⁶ so it is likely that the heavy military equipment supplied to the Abkhaz was in keeping with existing agreements.

A number of factors clearly contributed to the ethnic mobilization and conflict in Abkhazia. Existing grievances with Georgia--particularly over the state's policies toward minorities in the 1990-92 period--played a significant role in creating a tense situation that could foster ethnic mobilization among all minorities in the country. Moreover, the availability of external support for the secessionist cause was especially important in guiding the Abkhaz toward confrontation. But the crucial [End Page 265] factor was the existence of autonomous structures that enabled the ethnic Abkhaz to form a political elite that took control over the territory and administration of Abkhazia. Without autonomy, the Abkhaz elite would not have had the necessary institutions--such as the Supreme Soviet of the Abkhaz Autonomous Republic--with which to legitimately decide on secession from Georgia. Such institutions also enhanced the Abkhaz elites' ability to win external support. Through the linkages and channels inherited from Soviet Communist Party structures, Abkhaz elites had access to contacts in the former Soviet military forces that were crucial in securing support for the struggle. There can be little doubt that the existence of autonomy was a sine qua non for the Abkhaz secession from Georgia.

South Ossetia: The Power of a Parliament

The chance of South Ossetia seceding from Georgia was initially not much greater than Abkhazia's. While ethnic Ossetians, unlike the Abkhaz, did form a majority (of just over two-thirds) of their autonomous region's population in 1989, their numbers were diminutive--roughly sixty-seven thousand out of a population of only ninety-eight thousand. However, almost a hundred thousand Ossetians lived scattered in other regions of Georgia. Like the Abkhaz, South Ossetians were a comparatively small minority within Georgia, but they also had ethnic brethren in the North Caucasus--the Autonomous Republic of North Ossetia in Russia. A November 1988 law strengthening the position of the Georgian language in South Ossetia led to disturbances the following year. ⁴⁷ This was the first step in what has been termed a "war of laws," which began in earnest in the fall of 1989. ⁴⁸ With perestroika, an Ossetian popular front called Ademon Nykhas emerged, and in spring 1989 it addressed an open letter to the Abkhaz people, supporting their secessionist claims. Isolated instances of violence started occurring in South Ossetia, and guerrilla attacks by both Ossetian and Georgian armed bands were reported throughout the summer. In August, Tbilisi took measures to make Georgian the sole official language for use in public life. ⁴⁹ Such a provision would have affected South Ossetia--where only 14 percent of Ossetians knew Georgian--to a higher degree [End Page 266] than Ajaria or Abkhazia, given South Ossetia's lower status in the hierarchy of autonomy. This fueled an emerging movement for unification with North Ossetia, and accordingly Ademon Nykhas sent a petition to Moscow in support of that effort.

By late September, tensions had grown to the point where additional Interior Ministry troops had to be brought in to secure order. Interethnic clashes began to erupt despite attempts to calm the situation, and in early November the South Ossetian Supreme Soviet demanded that South Ossestia be upgraded to the status of an autonomous republic. Within a week, Georgia affirmed its right to secede from the Soviet Union, exacerbating the tensions in South Ossetia; resulting interethnic clashes throughout the fall left several dead. ⁵⁰ Meanwhile, a refurnishing took place within the South Ossetian elite, moving the Supreme Soviet closer to the position of Ademon Nykhas. The Gamsakhurdia government responded in late November by organizing a "March on Tskhinvali" ⁵¹ that was attended by over ten thousand people. The march was billed as "a peaceful march for reconciliation," but Ossetians perceived it as a show of force and blocked the marchers, leading to an armed clash contained only by armored forces of the Soviet Interior Ministry (clashes, nonetheless, continued until January 1990). Georgian legislation in August 1990 banned regional parties from the upcoming elections, and the South Ossetian Supreme Soviet immediately riposted by unilaterally upgrading its status to that of an "Independent Soviet Democratic Republic." After elections to the Supreme Soviet of the "new" South Ossetia were held in early December 1990, the Georgian Supreme Soviet abolished the South Ossetian Autonomous Oblast. ⁵² Only the presence of Soviet troops was now preventing armed conflict.

The change of government in Tbilisi after the ouster of Gamsakhurdia and the accession to power of Eduard Shevardnadze temporarily eased tensions, but the collapse of the Soviet Union--resulting in the removal of Soviet peacekeeping troops--led to conflict by April 1992 as artillery duels accelerated and North Caucasian volunteers amassed in North Ossetia, much as they would later to support Abkhazia. The Russian government also openly sided with the Ossetians and by late [End Page 267] spring 1992 there was a substantial risk of the conflict turning into a Georgian-Russian war. ⁵³ This threat nevertheless forced Georgia's new government to submit in late June 1992 to a Russian-led peacekeeping force that effectively removed Georgia from approximately half of South Ossetia's territory.

The Georgian-Ossetian conflict initially involved only the sixty-five thousand Ossetians within the borders of the South Ossetian Autonomous Region. The conflict escalated as a result of a "war of laws," in which a formerly rubber-stamp Soviet parliament in South Ossetia was transformed into a vehicle for Ossetian political aspirations that within a month decided to make Ossetian the state language of South Ossetia and later unilaterally declared independence. The role of the autonomous region's institutions in the development of the conflict was critical. In September 1989, before Gamsakhurdia's march on Tskhinvali, Ademon Nykhas had already petitioned for the unification of North and South Ossetia within the Russian Federation, adopting an extreme stand following the enunciation of the Georgian language laws. The South Ossetian Supreme Soviet, by contrast, was content with announcing Ossetian as the region's official language and asking Moscow to raise its status to that of an autonomous republic like Abkhazia. As the situation deteriorated during the winter, the stream of decisions from the South Ossetian Supreme Soviet slowed dramatically. Ademon Nykhas apparently gained control over the institution during these months, and by September the body abandoned compromise and declared total independence from Georgia. Clearly, the preexistence of a legislative body for use by the South Ossetian leadership was a key element in the escalation of conflict. As will be discussed, the Armenians of Javakheti possessed no such organ, and their nationalist organization was unable to assume a mobilizing role. In South Ossetia, autonomy provided the Ossetian leadership with a decision-making mechanism for responding to Tbilisi, thereby heightening tensions. As in Abkhazia, external support certainly played a role in the development of the conflict. By the time external actors became seriously involved in late 1991, however, the conflict had already escalated out of control. While the existence of North Ossetia surely spurred the southerners to commit to action, it was autonomy that provided the permissive conditions for the development of the conflict in South Ossetia. The case of Javakheti, sharing many similarities with the Ossetian case save its lack [End Page 268] of autonomy and absence of conflict, serves as a further illustration of this fact.

Javakheti: Escape from War?

The main concentration of Armenians in Georgia is found in the province of Samtskhe-Javakheti, bordering Armenia. Ethnic Armenians form a compact majority there, in the Akhalkalaki and Ninotsminda districts, ⁵⁴ numbering about 150,000. In many ways, the Javakheti Armenians exhibit numerous similarities with both the South Ossetians and the Armenians of Nagorno-Karabakh in Azerbaijan. With their compactly settled Armenian minorities near the present Armenian border, both Javakheti and Nagorno-Karabakh had been bones of contention between the short-lived Armenian Democratic Republic and its neighbors, the Georgian and Azerbaijani Democratic Republics, between 1918 and 1920. ⁵⁵ In fact, wars were fought over both regions, although the Armenian-Azerbaijani war of the time was significantly more severe. Relations between these Armenian minorities and their host nations have been uneasy. Because their Armenian identity remains very strong, Armenians in both Azerbaijan and Georgia feel excluded from even the most civic interpretation of these two nations' national conceptions. Unlike Ajars in Georgia or Lezgins in Azerbaijan, Armenians are not considered part of the majority nation and do not consider themselves as such. Both groups also reside in mountainous terrain--Javakheti's capital, Akhalkalaki, is located at approximately 3600 feet above sea level. Of course, differences exist as well. Cultural differences are less marked; Armenians and Georgians are both Christian peoples (although of different rites), whereas Azeris are Muslims. Most pronouncedly, the historical relationship between Armenians and Georgians has occasionally been characterized by weariness and suspicion but seldom by overt conflict--the brief war in 1918-1919 being the exception. By contrast, Armenians widely equate the Azeris with Turks, ⁵⁶ a group widely perceived by Armenians as perpetrators [End Page 269] of genocide against them. Armenians and Azeris fought wars in 1905-06 and 1918-20. As a result of these surviving historical antagonisms and myths, Armenian-Azeri relations are incomparably more complicated than are Georgian-Armenian relations.

The Javakheti regional center of Akhalkalaki is the site of a Russian military base where local Armenians comprise over two-thirds of the soldiers and noncommanding officers and a third of the officers--suggesting that Javakheti Armenians enjoy the support of a foreign patron and have ample access to arms and military training. Indeed, Javakheti has been a thorn in the side of the Georgian government since independence. Armenian ethnic entrepreneurs have nonetheless remained relatively powerless in their relationship with the Georgian government, largely unable to make their voices heard in Tbilisi or to extract any concessions of significance. In spite of these frustrations, they have not managed to stir up a large-scale popular movement.

Much like other minorities, specifically the South Ossetians, Javakheti Armenians were alarmed by the nationalist movement in Georgia led by Gamsakhurdia in 1989-91, and there emerged at roughly the time of the formation of Ademon Nykhas a political organization named Javakhk, which campaigned for the creation of an Armenian autonomous region at par with other autonomies in Georgia. The issues at stake for Javakheti Armenians during Gamsakhurdia's rule were similar to those in South Ossetia--with language issues, in particular, at the top of the agenda. Minorities in Georgia generally spoke both their mother tongue and Russian, the language of interethnic communication in the Soviet Union; however, few spoke Georgian. The 1988 law strengthening the position of the Georgian language was hence perceived as a threat to the minorities. Meanwhile, the Armenian population of Javakheti had enjoyed substantial cultural autonomy even in the absence of political or territorial self-rule. Most schools in the area were Armenian, and the Armenians were equally disturbed by the development of Georgian legislation in the period. Javakhk has not managed to sustain a permanent level of active popular support and has been plagued by internal disputes, so no clear chain of command exists in the organization. Its position is at times contradictory; certain statements from the organization seem conciliatory toward Tbilisi whereas others are more militant in their demands for Armenian self-determination. ⁵⁷

Another contentious issue has been that of centrally appointed prefects, which the Gamsakhurdia regime introduced. Through large-scale [End Page 270] demonstrations, Javakheti Armenians physically prevented three different appointed prefects of ethnic Georgian origin from assuming office. ⁵⁸ The ensuing lack of legitimate governing authorities was solved through the creation of a provisional council of representatives for the Akhalkalaki region, with twenty-four elected representatives. In other words, Javakheti Armenians took a first unilateral step toward the creation of institutions of self-rule. The popular legitimacy of these institutions, which might have been served as an embryonic autonomist or secessionist movement, nevertheless remains doubtful as they basically self-dissolved even before the ascent of Eduard Shevardnadze to the head of the Georgian state. Basically, the Armenian activists were unable to create legitimate institutions for their struggle, and no preexisting institutions were present. In 1995, Georgian authorities successfully merged the Javakheti region with the region to its west, Meskheti (which has a clear Georgian majority), to create the province of Samtskhe-Javakheti. This move was interpreted by Javakhk as an attempt to artificially dilute the Armenian demographic position in the administrative units of southern Georgia. Although the move did not lead to large-scale protests, it did increase Armenian resentment and suspicion of Tbilisi.

The absence of conflict in Javakheti must be associated with the lack of a strong and legitimate nationalist leadership, especially in comparison with South Ossetia, which had a comparable conflict potential. Ademon Nykhas did not have a higher degree of initial popular legitimacy than Javakhk; the key difference in the development of the two organizations was that the autonomy enjoyed by South Ossetia facilitated the cohesion and strengthening of the nationalist/separatist movement around the governmental institutions of the region. In Javakheti, the Javakhk movement needed to build up its position on its own, including the creation of provisional administrative structures; in South Ossetia, such institutions already existed, with a rigid hierarchy and an accepted decision-making process. The legitimacy of national leaders was determined not only by their personality and achievements but also by the posts they held. Moreover, when unfavorable laws were introduced by the Georgian parliament, the Javakheti Armenians had little to respond with except petitions or popular demonstrations. Ossetians, by contrast, possessed a legislative body, the Supreme Soviet of the autonomous province, which provided them with an institutional channel for the struggle against Georgian actions. Likewise, the internal [End Page 271] dissension within Javakhk and what amounts to an incapacity to take, let alone enforce, decisions reflected the lack of a clear hierarchy of decision making. Autonomous structures, as the case of South Ossetia proves, possess a chain of command that facilitates the decision-making process and the enforcement of decisions taken at a higher level.

Other factors have played an important role in Javakheti's relatively peaceful development. First, the Georgian government already having been defeated by two other secessionist movements, has been cautious not to provoke the Javakheti Armenians. Furthermore, the Armenian government, mindful of the importance of its relations with Georgia, has been careful to defuse potential problems in the region, intervening several times to dissuade Javakhk from holding referenda on autonomy or secession. The external support provided by the Russian military base, therefore, is mitigated by the calming effect of the Armenian government. In the final analysis, however, the lack of autonomy in Javakheti must still be noted as a significant reason for the weakness of ethnic mobilization and the absence of armed conflict there.

Azeris: The Silent Mass

Little political activity, let alone unrest or separatism, has been observed among Azeris in Georgia, who are concentrated in the southern and southeastern regions of Georgia. However, the Azeris were a target of Georgian nationalist groups fearful of the rapidly increasing birth rates of Azeris and other Muslim peoples in Georgia. In 1989, Georgian informal groups forced several hundred Azeri families in the Bolnisi region to migrate to Azerbaijan. Another incident worth mentioning is a 1990 case of information failure between Georgian nationalists and Azeris, in which a false rumor spread that the Azeris were about to secede and join Azerbaijan. Upon receiving this news, a group of Georgian nationalists gathered followers and marched on Azeri areas. Georgian authorities were able to defuse the situation before any blood was shed, after having coordinated their position with the Azerbaijani Popular Front and ascertaining that no steps toward secession were being taken. In the postindependence era, the close and improving relations between Azerbaijan and Georgia have ensured stability in the region. ⁵⁹ The Azeris live relatively scattered across southern and southeastern Georgia, while forming majorities of over 70 percent in the Marneuli, Bolnisi, and Dmanisi districts. The remainder of the population is composed of Georgians, Armenians, and Kurds. There is no [End Page 272] clearly demarcated Azeri area in Georgia, and given their lack of autonomy, it would be difficult for Azeris to outline which areas they would consider "theirs". The good relations between Georgia and Azerbaijan have also eliminated any tangible external support for nationalism among Azeris in Georgia; nor has a nationalist leadership been able to emerge. It should be mentioned, however, that the Azeris in Georgia live mainly in rural areas, and though their economic conditions are fairly good, they are rather isolated from Georgian social and political life. Very few among them speak Georgian, a proportionally very low number have access to higher education, and the Georgian government has done little to integrate them. A certain amount of resentment and feeling of alienation has been reported, but it has failed to find an effective channel for expression; a national leadership did not exist "by default," as it did in autonomous areas. In sum, the low level of ethnopolitical activity, the role of the government of Azerbaijan, and the lack of autonomy have made Azeri areas among the calmest in Georgia's turbulent post-perestroika history.

Ajaria--Regionalism Enforced

The case of Ajaria, as mentioned earlier, represents the only case of a South Caucasian autonomous region not involved in armed conflict with its central government. ⁶⁰ This is not surprising, given that few indicators ever pointed to a high risk of ethnic conflict between Ajaria and Georgia. Most importantly, Ajars are in fact ethnic Georgians, differing from the majority population on account of their Muslim religion. The majority of its inhabitants adopted Islam and much of Islamic culture (during the centuries of Ottoman rule that ended in 1878) while retaining strong cultural similarities with Christian Georgians. This may make it questionable as to whether classifying them as a minority is appropriate. Christianity certainly constitutes an important part of the Georgian national identity, yet after seventy years of Soviet atheism Ajarian Islam has a comparatively weak hold on the population. Ajars therefore remain accepted in the predominant definition of the Georgian nation, which can be said of no other minority in Georgia.

Nevertheless, since the early 1990s Ajaria has been dominated by a [End Page 273] local potentate with significant political ambitions. Aslan Abashidze, descending from an influential family of the local nobility, acceded to the leadership of Ajaria in 1991 and has since imposed an increasingly authoritarian rule. During the wars in South Ossetia and Abkhazia as well as in intra-Georgian feuds, Abashidze skillfully achieved wide self-rule by maintaining neutrality. Moreover, Abashidze managed to turn Ajaria into an important economic region through the development of trade links with Turkey, using the asset of the port of Batumi, the regional capital on the Black Sea coast. In turn, the Georgian government, desiring no further trouble with its provinces, left Ajaria on its own in most matters. Although Abashidze endorsed Eduard Shevardnadze's bid for the Georgian presidency, Tbilisi's subsequent attempts to rein in Ajaria within the hierarchy of the Georgian state has led to the deterioration of relations between Batumi and Tbilisi; for example, Ajaria refuses to harmonize its laws with national guidelines. Although the Ajarian case displays clear elements of regionalism, these do not have a significant ethnic character. Indeed, Abashidze has established himself as a politician on the national Georgian level, in fact emerging as the most serious challenger to Shevardnadze and his Citizen's Union of Georgia Party even though his political life is based heavily in Ajaria. Charles H. Fairbanks summarizes the Ajarian situation well:

The local boss, Aslan Abashidze, has never raised any question of secession from Georgia. He wants simply to do what he wants and to enjoy the profits of vacation hotels, tropical products, and smuggling across the border with Turkey. There seems to be nothing public in Abashidze's motives; he is operating essentially like a small businessman. The Russian garrison on the border, whose main occupation seems to be smuggling, gives Abashidze the protection to defy the central Georgian government; the Moscow government approves this arrangement because it limits Georgian independence from Russia. ⁶¹

In the final analysis, the independent-minded rule of Abashidze in Ajaria would have been impossible without the institution of autonomy. The weakness of the separate Ajarian identity does not provide a base for a vigorous nationalist movement or for excessively strong regionalism; in this context the population of the province of Mingrelia arguably has stronger regionalist attitudes. ⁶² The institutions of Ajarian autonomy brought Abashidze to his position of power and enabled [End Page 274] Ajaria to remain largely isolated economically and politically from the rest of Georgia; likewise, the political institutions and financial resources of the autonomous republic have provided a base for Abashidze's bid for a national political role in Georgia. Given the historical and ethnic framework, there were no conditions under which autonomy would lead to ethnic conflict in Ajaria. However, autonomy did provide a ground for the emergence of a local potentate who succeeded in keeping his region free of the problems experienced by other Georgian regions while establishing a rigid autocratic rule.

Conclusions

The provision of institutionalized, territorial autonomy for an ethnic minority may cause the opposite of its intended effect--it may augment rather than reduce the potential for conflict between a minority and a central government. As stated earlier, autonomy is neither a sufficient nor a necessary cause of conflict. Yet, it has a strong causal relationship with both a minority's willingness and especially its capacity to revolt. It is reasonably clear that within the Caucasian context, autonomy has been a source of conflict and not a solution to it. This preliminary conclusion suggests that ceteris paribus, secessionism is likely to be significantly higher among autonomous minorities than among nonautonomous minorities. This study has been empirically limited to the former Soviet space, and it should be acknowledged that the specific history and characteristics of this area leave open the question as to whether the findings here would be replicated in similar studies of other political settings. Nevertheless, a number of factors inherent in the institution of territorial autonomy are likely to have the same consequences in other areas of the world as they have had in the former Soviet Union.

The practical implications of these findings are nevertheless significant. When confronted with the results of this study, one disheartened practitioner of diplomacy raised the obvious question: if autonomy, bearer of much hope for the management and resolution of ethnopolitical conflicts, is not actually a solution but is rather inherently problematic, what is the way to manage ethnic tensions? The answer does not lie in any general and easily applicable model, and that may itself be one of the most important consequences of realizing the pitfalls of autonomy. Where there has been a tendency to view ethnofederal solutions in one form or another as cure-all prescriptions, this study points to the merits of devising political structures that cut across ethnic [End Page 275] and other communal divisions, encourage civic identities, but discourage the use of ethnicity in the political sphere. That does not mean that all autonomy solutions are necessarily destined to collapse or to lead to war. It does mean that whenever the ethnicization of territory can be avoided, it should be avoided.

This study has attempted to show that the advocacy of resolving or preventing ethnic conflict through solutions based on the devolution of power along ethnic lines is at best a questionable and at worst a disastrous enterprise. The little publicized pitfalls of ethnofederalism hence need to be kept in mind while formulating policies in and toward multiethnic societies.

 

Svante E. Cornell is Editor of the Central Asia-Caucasus Analyst, Researcher at the Central Asia-Caucasus Institute of the Johns Hopkins University's School of Advanced International Studies, and Lecturer in the Departments of Peace and Conflict Research and East European Studies at Uppsala University, Sweden. He is the author of Small Nations and Great Powers: A Study of Ethnopolitical Conflict in the Caucasus (2001).

Notes

* The author would like to thank Anna Jonsson, Regine Spector, Erik Melander, Magnus Öberg, and S. Frederick Starr for their valuable comments and suggestions. Research for this article was made possible by a grant from the Swedish Research Council for the Humanities and Social Science.

1. Ted Robert Gurr, "Ethnic Warfare on the Wane," Foreign Affairs 79 (May-June 2000), 53.

2. Significant works on nationalism and ethnic conflict include Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism, 2d ed. (London: Verso, 1991); Ernest Gellner, Nations and Nationalism (Oxford: Blackwell, 1988); Anthony D. Smith, The Ethnic Origins of Nations (Oxford: Blackwell, 1986); Donald Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985); Ted R. Gurr, Why Men Rebel (Princeton: Princeton University Press, 1970); Gurr, Peoples versus States: Minorities at Risk in the New Century (Washington, D.C.: usip Press, 2000).

3. In other instances, however, ethnic demands are not for "exit"--autonomy or secession--but for greater participation in the government of the central state, particularly when settlement patterns overlap.

4. It must be noted, however, that ethnic conflict in such situations is possible, and when it does occur, it is likely to be significantly more severe than in cases of less intermingled settlement patterns. Conflict in intermingled states would tend not to be over a part of the state's territory and its affiliation but over the control of the state apparatus, that is, the entire territory of the state. The emergence of violent conflict in such situations would be significantly more likely to lead to large-scale ethnic cleansing and/or genocide: a geographic partition line being much more difficult to draw, the conflict is likely to take place not on a warfront between two organized military formations but in civilian-inhabited areas over a much larger tract of territory. Moreover, the knowledge that a clean territorial break is impossible or very difficult encourages the urge to displace or eliminate members of the other group and even the perception that it is necessary. A political solution would also imply that one would continue to live intermingled with members of the other group. This is in turn interpreted as a security threat to the own group and again increases the urge to expel or eliminate the other group, actions that are even conceptualized as defensive and indispensable for the own group's well-being. This situation is referred to as the "security dilemma". See, for example, Barry Posen, "The Security Dilemma and Ethnic Conflict," Survival 35, no. 1 (1993).

5. James D. Fearon and David D. Laitin, "Weak States, Rough Terrain, and Large-Scale Ethnic Violence since 1945" (Paper presented at the annual meetings of the American Political Science Association, Atlanta, September 2-5, 1999), 16, emphasis in original.

6. Ted Robert Gurr, "Peoples against States: Ethnopolitical Conflict and the Changing World System," International Studies Quarterly 38 (Fall 1994), 366.

7. Kjell-Åke Nordquist, "Autonomy as a Conflict-Solving Mechanism: An Overview," in Markku Suksi, ed., Autonomy: Applications and Implications (The Hague: Kluwer, 1998), 59.

8. Interestingly, in certain instances the central government, in particular in states trying to build a civic national identity, argues that granting autonomy to a minority population would be tantamount to defining that population as second-class citizens. The Turkish government, for example, sticks to its

refusal for special rights to citizens of Kurdish origin, on the grounds that they are already enjoying all existing rights as first-class citizens of the Turkish republic; any special rights would imply their segregation from the rest of the population and by extension their diminishment to second-class status.

9. See Ruth Lapidoth, Autonomy: Flexible Solutions to Ethnic Conflict (Washington, D.C.: usip Press, 1996), 203. By the same token it can be argued that the refusal to grant autonomy could be an even stronger incentive for a state affiliated with the minority to intervene.

10. David J. Meyer, "A Place of Our Own: Does the Ethnicization of Territorial Control Create Incentives for Elites to Conduct Ethno-Political Mobilization? Cases from the Caucasus in Comparative Perspective" (Paper presented at the Fifth Annual Convention of the Association for the Study of Nationalities, New York, April 2000). See, for example, Daniel J. Elazar, Federalism and the Way to Peace (Kingston: Queens University, 1994); also Lapidoth (fn. 9).

11. See, for example, Henry Hale, "Ethnofederalism and Theories of Secession: Getting More from the Soviet Cases" (Paper presented at the annual meeting of the Association for the Study of Nationalities, New York, April 1999); Robert Dorff, "Federalism in Eastern Europe: Part of the Solution or Part of the Problem?" Publius: The Journal of Federalism 24 (Spring 1994).

12. See, for example, Rogers Brubaker, Nationalism Reframed: Nationhood and the National Question in the New Europe (Cambridge: Cambridge University Press, 1996); Philip G. Roeder, "Soviet Federalism and Ethnic Mobilization," World Politics 43 (January 1991). For a recent treatment of the subject, see Carol Skalnik Leff, "Democratization and Disintegration in Multinational States: The Breakup of the Communist Federations," World Politics 51 (January 1999).

13. Hans-Joachim Heintze, "On the Legal Understanding of Autonomy," in Suksi (fn. 7), 7.

14. Henry J. Steiner, "Ideals and Counter-Ideals in the Struggle over Autonomy Regimes for Minorities," Notre Dame Law Review 66 (1991), 1542.

15. Ibid., 1542.

16. Heintze (fn. 13), 7.

17. UN Information Organization, Documents of the UN Conference on International Organization, vol. 6 (New York, 1945), 296, cited in Heintze (fn. 13), 9.

18. Steiner (fn. 14), 1550.

19. Ibid.

20. Ibid., 1552.

21. See Douglas Sanders, "Collective Rights," Human Rights Quarterly 13 (August 1991).

22. See M. Brems, Die Politische Integration Etnischer Minderheiten (The political integration of ethnic minorities) (Frankfurt am Main: Lang, 1995), 142.

23. L. Lyck, "Lessons to be Learned on Autonomy from the Faeroese Situation since 1992," Nordic Journal of International Law 64 (Fall 1995), 481-87.

24. Ildus G. Ilishev, "Russian Federalism: Political, Legal and Ethnolingual Aspects--A View from the Republic of Bashkortostan," Nationalities Papers 26 (Fall 1998), 724-759. Reproduced as an appendix to the article is the Treaty on the Mutual Derogation of Powers between the State Organs of the Russian Federation and the State Organs of the Republic of Bashkortostan, in Stanovlenie Dogovornykh Otnosheniy Respubliku Bashkortostan i Rossiyskoy Federatsii, 1990-1996 gg., Sbornik Dokumentov (Ufa, 1997).

25. Group cohesion, willingness to act, and capacity to act have been identified as the major categories of factors leading to conflict in the literature on ethnopolitical conflict. See, for example, Gurr (fn. 2, 2000).

26. Thongchai Winichakul, Siam Mapped: A History of the Geo-Body of Siam (Honolulu: University of Hawaii Press, 1994).

27. Anderson (fn. 2), 175.

28. Gurr (fn. 2, 2000).

29. Dmitry Gorenburg, "Nationalism for the Masses: Popular Support for Nationalism in Russia's Ethnic Republics," Europe-Asia Studies 53 (January 2001), 74.

30. Meyer (fn. 10), 2.

31. Ibid.

32. Leff (fn. 12), 210.

33. Here minority is defined as an ethnically defined population group under the jurisdiction of a territorial unit with a distinct titular nationality. Hence, in the South Caucasus, the Nakhjivan Autonomous Republic is left outside the scope of the study since the population of Nakhjivan was and remains homogeneously Azerbaijani-populated (93 percent) and under the jurisdiction of Azerbaijan.

34. For a detailed overview of the conflicts in the Caucasus, see Svante E. Cornell, Small Nations and Great Powers: A Study of Ethnopolitical Conflict in the Caucasus (Richmond, UK: Curzon Press, 2001)

35. The Meskhetian Turks of Southern Georgia were deported in 1943 but have not yet been allowed to return to their native lands.

36. Fearon and Laitin find that "mountain groups were six times more likely to see large-scale fighting with the state following the Soviet collapse." As they note, moreover, rough terrain is a useful tool to explain how minorities with small numbers can "sustain significant guerrilla conflicts with the state." Fearon and Laitin (fn. 5), 18-20.

37. Ethnic mobilization does not automatically carry with it ethnic conflict; conflict only occurs if the central government decides to answer the minority's challenge by force. Secession, of course, can take place peacefully, if the government simply lets go of the province in question; however, cases of peaceful secession are diminutively few. Another option is the cooptation of the minority elite either through integration in the central government or simply through bribery. In the final analysis, it is nevertheless the norm and not the exception that a minority challenge on the subject of territory is answered by force on the part of the central government.

38. The full supporting information and the coding of factors and cases relevant to this study, including tables for each factor, is available at http://www.cornellcaspian.com/autonomy.html.

39. For a detailed overview of the conflict, see Cornell (fn. 34), 142-96.

40. See, for example, Current Digest of the Soviet Press 41, no. 29 (1989), 14-16.

41. BBC Monitoring Service, June 18, 1992, quoting Russian television "Rossiya," June 16, 1992.

42. "Georgia: Abkhazia 'Strong Enough to Fight Georgia,'" BBC Monitoring Service, July 30, 1992.

43. BBC Monitoring Service, July 25, 1992. The 1925 constitution did stipulate that Abkhazia was tied to Georgia by a special union treaty, but de facto it amounted to a secession from Georgia and was certainly perceived as such in Tbilisi.

44. See Darrell Slider, "Democratization in Georgia," in Karen Dawisha and Bruce Parrott, eds., Conflict, Cleavage and Change in Central Asia and the Caucasus (Cambridge: Cambridge University Press, 1991), 170.

45. See Cornell (fn. 34), chap. 4.

46. Ibid., 142-96.

47. See Elizabeth Fuller, "Draft 'State Program' on Georgian Language Published," Radio Liberty Research Report no. 559/88, December 12, 1988.

48. See Catherine Dale, "Abkhazia and South Ossetia: Dynamics of the Conflicts," in Pavel Baev and Ole Berthelsen, eds., Conflicts in the Caucasus, Report no. 3 (Oslo: International Peace Research Institute, 1996), 13-26.

49. See Elizabeth Fuller, "South Ossetia: Analysis of a Permanent Crisis," Report on the USSR, February 15, 1991, 21.

50. Ibid.

51. See Julian Birch, "The Georgian/South Ossetian Territorial and Boundary Dispute," in J. Wright et al., eds., Transcaucasian Boundaries (London: soas, 1995), 182. Tskhinvali is the capital of the South Ossetian Autonomous Region.

52. See Elizabeth Fuller, "Georgian Parliament Votes to Abolish Ossetian Autonomy," Report on the USSR, December 21, 1990, 8; Rachel Denber, Bloodshed in the Caucasus: Violations of Humanitarian Law and Human Rights in the Georgia-South Ossetia Conflict (New York: Helsinki Watch, 1992), 8.

53. See Alexei Zverev, "Ethnic Conflicts in the Caucasus, 1988-94," in Bruno Coppetiers, ed., Contested Borders in the Caucasus (Brussels: vub Press, 1996), 46.

54. For overviews of the issue, see Voitsekh Guretski, "The Question of Javakheti," Caucasian Regional Studies 3, no. 1 (1998) (poli.vub.ac.be/publi/crs/eng/0301-05.htm, accessed October 2001); Ugur Akinci, "Javakhetia: The Bottle-Neck of the Baku-Ceyhan Pipeline," Silk Road: A Journal of West Asian Studies 1 (December 1997); Igor Rotar, "Tbilisi Has Only Partial Control over Georgia's Armenian Regions," Jamestown Prism 4 (May 15, 1998); Levon Sevunts, "Squeeze Play in the Caucasus: Russia Could Tighten Its Grip as Georgia, Armenia Break Out of Bear Hug," The Gazette (Montreal), November 29, 1999.

55. See Firuz Kazemzadeh, The Struggle for Transcaucasia, 1917-1921 (Oxford: George Ronald, 1951).

56. The Azerbaijanis are a Turkic people speaking a language closely related to Anatolian Turkish, but for the overwhelming part of their history they have been politically separated from Turkey and have had closer relationships to Iran.

57. Annotated Headlines of the Georgian Press, September 21-22, 1998.

58. Guretski (fn. 54).

59. Author interviews in Tbilisi and Marneuli, 1998.

60. See Judith Hin, "Ajaria. Authoritarian Governance, Favorable Economic Location, and Minor Ethnic Tensions: The Interests of the Local Potentate in Keeping Violent Conflict at Bay" (Paper presented at the Fifth Annual Convention of the Association for the Study of Nationalities, New York, April 2000). The paper was based on unpublished data from the Brown University research project, "Can Deadly Conflicts Be Prevented," funded by the Carnegie Corporation.

61. Charles H. Fairbanks, Jr., "Party, Ideology and the Public World in the Former Soviet Space," in Arthur M. Meltzer, Jerry Einberger, and M. Richard Zinman, eds., Politics at the End of the Century (Lanham, Md.: Rowman and Littlefield, 2001), 252.

62. Cornell (fn. 34), 184-85.