State disintegration and nationalities conflict has emerged as perhaps the central concern of intellectuals and policymakers throughout the contemporary world. This article undertakes, within a theoretical and conceptual framework, a comparative analysis of the causation of state crises and nationalities conflict in two particularly stark, but apparently disparate, cases: Sri Lanka and the former Yugoslavia. The author emphatically denies that the eruption of protracted civil war in these countries is due to preexisting ethnic animosities between peoples. Instead, he takes a dynamic view of the concept of the nation and explains the formation (and transformation) of nationalist constituencies and sentiments in Sri Lanka and Yugoslavia in terms of a process of dialectical interaction between the modern state and diverse civil societies. The state crises in Sri Lanka and Yugoslavia, he argues, can be best understood through a critical focus on how centralist, unitarist strategic elites at the helm of state apparatuses in these countries have, over the past several decades, engendered growing political opposition.
STATE CRISES AND NATIONALITIES CONFLICT IN SRI LANKA AND
YUGOSLAVIA
SUMANTRA BOSE Columbia University
The unity of place is only disorder. Only the unity of problem makes a center.
—Marc Bloch (cited in Skocpol & Somers, 1980, p. 194)
A study of nationalism must follow the comparative method; it cannot remain confined to one of its manifestations. Only the comparison of the different nationalisms all over the earth will enable the students to see what they have in common and what is peculiar to each, and thus allow a just evaluation.
—Hans Kohn (1944, pp. ix-x)
This article derives its mandate from Bloch and Kohn. It sets out to examine and explain, in comparative perspective, state disintegration and nationalities conflict in two apparently dissimilar contexts: southeastern Europe and southern Asia. In so doing, it attempts to identify and interpret certain causal regularities in these processes as they have occurred in Sri Lanka and Yugoslavia.
COMPARATIVE POLITICAL STUDIES, Vol. 28 No. 1, April 1995 87-116 © 1995 Sage Publications, Inc.
87
from the SAGE Social Science Collections. All Rights Reserved.
88    COMPARATIVE POLITICAL STUDIES / April 1995
The organization of the article is as follows. The following section offers a brief explanation of the selection of Sri Lanka and Yugoslavia as critical cases for a structured, focused comparison.11 then sketch a theoretical and conceptual framework, which launches the following two sections: discussion of the specific circumstances of state crises and nationalities conflict in Sri Lanka and in Yugoslavia. In conclusion, I move briefly from the interpretive/explanatory to the prescriptive/predictive and suggest some general policy guidelines on how such conflicts might be resolved or at least contained.
WHY SRI LANKA AND YUGOSLAVIA?
Despite the contextual dissimilarities and the particularities of each case, a striking congruence in political outcomes appears to have occurred in contemporary Sri Lanka and Yugoslavia.2 Both are multiethnic societies riven by protracted civil wars exacting an immense toll of lives lost and populations displaced, and both are plagued by aggressive assertions of ethnic identities. Also, both countries have witnessed a breakdown of the state structure: an empirical and juridical disintegration in Yugoslavia and an empirical disintegration in Sri Lanka.3
Sri Lanka and Yugoslavia thus represent two great human tragedies of our times. They are also fascinating scholarly laboratories for studying mass nationalism and violent conflict. But the quest to discover what went so drastically wrong in Sri Lanka and Yugoslavia is also a profoundly practical enterprise if these two lands are ever to be extricated from their present calamitous situations and if the recurrence of future Sri Lankas and Yu-goslavias is to be prevented.
TOWARD AN ANALYTIC FRAMEWORK: ETHNICITY, NATIONALISMS, AND THE ROLE OF THE MODERN STATE
I believe that the primordialist thesis, in its many forms (Weber, Shils, Eisenstadt, Geertz), is of seriously limited value in accounting for recent and
1. For an elaboration of this methodological approach, see George (1979).
2. In this article, the term Yugoslavia is used to denote the union that existed prior to the breakup of 1991, unless otherwise specified.
3. For an explanation of the concepts of juridical and empirical sovereignty, see Jackson and Rosberg (1982).
Bose / STATE CRISES AND NATIONALITIES CONFLICT 89
contemporary nationalities conflicts. Given the continuing endurance of variants of this thesis in explaining such conflicts, however, it is worth pointing out the grave shortcomings of this approach with reference to Sri Lanka and Yugoslavia.
The evidence is compelling. Efforts to explain the Sinhalese-Tamil conflict on the basis of objective cultural differences founder on the fact that
underlying the linguistic and religious differences ... are strong cultural and racial similarities. Physically the Sinhalese and Tamils cannot be differentiated. Though the initial Sinhalese migrants were probably Indo-European language speakers who arrived over 2,500 years ago, practically all later arrivals were south Indians (mostly Tamil speakers) who were assimilated into the Sinhalese-Buddhist community. (Obeyesekere, 1984, p. 154)
As Tambiah (1986, p. 5) comments,
Although the major identity components of the Sinhalese [75% of the population] are their Sinhalese language and Buddhist religion, and of the [Sri Lankan] Tamils [13%] their Tamil language and Hindu religion, both... populations share many parallel features of traditional caste, kinship, popular religious cults, customs and so on.
There is little record of protracted armed conflict between Sinhalese and Tamils prior to the late 20th century except in the very ancient past. Moreover, Sinhalese and Tamil are by no means homogeneous, solidary categories; a vast degree of social differentiation on the basis of class, caste, gender, the rural-urban divide, and other factors is common to both groups.
Much the same can be said of the peoples of former Yugoslavia. As Aleksa Djilas (1991, p. 4) observes, over the centuries,
eastern and western influences, especially eastern Orthodox and Roman Catholic, frequently transgressed the borders [between the Ottoman and Habsburg empires]. . . [and] between the Croatian and Serbian "tribes" and . . . states, creating a pluralistic mosaic rather than a simple division between "western Croats" and "eastern Serbs."
Pan-Yugoslavist ideology originated among elements of the rising Croatian urban bourgeoisie during the mid-19th century and, in 1905, the leading Croatian and Serbian political parties allied in a historic coalition to unitedly press for the creation of a common South Slav homeland. And the resurrected Yugoslav state founded in 1945 was hardly an artificial union created through force alone. The Yugoslav Communists achieved power largely on their own strength, built against daunting odds through a popular struggle involving massive participation of South Slavs of all national communities and diverse social backgrounds between 1941 and 1945.4 4. See, for example, Tomasevich (1969).
90    COMPARATIVE POLITICAL STUDIES / April 1995
Obviously, an alternative conceptual approach is needed to provide a persuasive account of the coming of civil war to southern Asia and the Balkans.
I believe that to account for the Sri Lankan and Yugoslav outcomes, scholars of nationalism must focus on the role and impact of the modern state on identity formulation and reformulation by large collectivities of people (Sinhalese, Tamils, Serbs, Croats) and on the crucial instrumentality of state policies, actions, goals, and imperatives in the gradual, dynamic, open-ended, and transformative process of "nationality formation" by ethnic groups in Sri Lanka and Yugoslavia. I argue that the key to understanding the causation of civil war in these cases lies in studying the development of the modern state in these countries over the past four or five decades,5 and, more particularly, in critically examining the role of strategic elites6 at the helm of such apparatuses of power, control, and domination. Of course, the state is not to be treated as some abstract entity; rather, my emphasis is on the dialectical interaction of state and society in determining a certain political outcome (i.e., the rise of mass ethnic nationalisms and the breakup of the state). In particular, I demonstrate that the phenomenon of strategic state elites with centralist and unitarist imperatives and tendencies (of which the Titoist and Sinhalese-Buddhist entities are two different types) attempting to exert hegemony over highly diverse, multiethnic societies constitutes the central explanatory variable in accounting for state disintegration and nationalities conflict in southern Asia as in southeastern Europe.
My argument thus diverges sharply from, among others, that of Rabushka and Shepsle (1972). These authors contend that "the [ethnically] plural society . . . does not provide fertile soil for democratic values or stability" (p. 92) because of "incompatible, intense ethnic feelings held by members of . . . communal groups" (p. 186). I contest this claim on at least two grounds. First, it seems to imply that "plural societies" the world over are condemned to an undemocratic future simply because of their plural composition. If this is correct, it would appear that democratic aspirations are a futile fantasy for the vast majority of humankind. This seriously questionable
5. Of course, this recent phase is necessarily linked to events and processes in earlier historical periods. Thus the interwar period and World War II were especially crucial for Yugoslavia, and the period of British colonial rule (circa 1796 to 1948, especially the last few decades) was especially crucial for Sri Lanka.
6. For an authoritative elaboration of the notion of "strategic state elites," see Stepan (1978, pp. 117-157). The strategic state elite for Yugoslavia are, of course, the Communist ex-Partisans, led by Tito, who ruled the country as League of Communists of Yugoslavia between 1945 and 1990. For Sri Lanka, this elite comprises the mostly upper-class/caste Sinhalese-Buddhists who inherited the colonial state apparatus in 1948.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 91
assertion rests on an even more dubious view of the essentially static, given, and intractable nature of ethnic identifications. I argue, however, that to understand why and how such loyalties become salient and exclusive at a certain conjuncture, we must focus on the role of the modern state in such societies and, in particular, analyze the policies and actions of strategic elites who control these state-apparatuses. I hold, with Horowitz (1985, p. 684), that "there is no case to be made for the futility of democracy or the inevitability of uncontrolled conflict. Even in the most severely divided society, ties of blood do not lead ineluctably to rivers of blood."
As Rothschild (1981) argues, "politicized ethnic assertiveness is in large measure but a reflection of the contemporary state's ... crisis of legitimacy" (p. 19) and "the ethnic group's strategic decisions are in large part responses to the state's ... capacity to earn and project legitimacy" (p. 29). There are thus organic linkages between the enterprise of state building and the growth of ethnic sentiment. Twentieth-century nationalisms, however much they may seek to appropriate and exploit premodern historical and cultural symbols of the community concerned, are, in reality, entirely modern political phenomena. Historically, numerous identities have been available to serve as bases for political mobilization. We must therefore explain why one particular identity (such as Croat or Tamil) is chosen in preference to several possible alternatives at a certain temporal moment. This article argues that "nationhood," that consciousness of a collective identity, is a social and political construct, and that the choice of identity is best understood in terms of a societal response to the exercise of power by strategic elites at the helm of the modern state.
A caveat is in order here. My theoretical emphasis does not presume that strategic elites can, intentionally or unintentionally, conjure ethnic identities out of thin air. It is not as though the state can manufacture such identities out of whole cloth; clearly, stirring a pot presupposes a pot that can be stirred. For instance, what makes Tamil nationalist mobilization against the Sinhalese-Buddhist state at all possible, in the first place, is the indisputable fact that Sri Lankan Tamils do share certain objective bases for the formulation of a common identity—such as a shared language, history, and territory. What I am arguing is that such factors in themselves are simply not sufficient to explain Sinhalese-Tamil polarization and civil war. Nor are historical memories entirely irrelevant to modern conflicts. Past injustices (Ottoman subjugation, Ustasha massacres) have been frequent themes in attempts by Serbian nationalists to rationalize their military campaigns in Croatia and Bosnia. But to explain why such "memories" have apparently struck a responsive chord among Serb populations today, we must, I believe, highlight
92    COMPARATIVE POLITICAL STUDIES / April 1995
the context of state breakdown and systemic collapse. Otherwise, taking Serb nationalist rhetoric about age-old, irreconcilable differences with Muslims and Croats at face value might well lead us into serious fallacies.
It remains to elucidate a working definition of nationalism. The definition that implicitly informs my analysis has been developed by Brass (1976). Among the pluses of this definition is that it does not regard the nation as some static, rigid category that is simply assumed to exist; instead, it underlines the historicity of the formative (and transformative) processes whereby a nation typically comes into being:
At one end [of the continuum] are people clearly distinguishable from others by obvious cultural markers, but whose members do not attach value to those marks of group difference and do not pursue social, economic or political goals based on them. Such groups . . . will be characterized as ethnic categories____A second critical point on the continuum is occupied by ethnic groups
objectively distinct from their neighbours, subjectively self-conscious of their distinctness, and laying claim to status and recognition either as a superior group or as a group at least equal to other groups. Such a group will be called an ethnic community. An ethnic community has adopted one or more of its marks of cultural distinctness and used them as symbols both to create internal cohesion and to differentiate itself from other ethnic groups.... Ethnic communities may enter the political arena in two ways.... [They may] engage in interest-group politics . . . [but] when ethnic groups demand not merely enhanced opportunities for individuals but corporate recognition for the group as a whole with a right to . . . govern themselves within a federal unit in an existing state or in a separate sovereign unit, they are engaged in the politics of nationalism ... [and have] become a nationality, the third critical point on the continuum of ethnic group transformation, (pp. 225-240)7
Implicit in this definition is the notion that collective identities, including national identities, are fluid, malleable, and flexible in nature. The degree and intensity of nationalism varies across groups, as it varies across one particular group over time. There is therefore nothing inherently given about patterns of identity. Let us now proceed, in the specific contexts of contemporary Sri Lanka and Yugoslavia, to unravel the puzzle of how and why certain groups gradually develop a sense of national identity that is seen by them as being incompatible—or at fundamental variance—with an extant territorial-juridical state. In other words, let us explore how and why state and nation came to be viewed as mutually exclusive alternatives to one another.
7. Of course, the three critical points should not be taken as watertight categories. In any empirical context, the lines between them may well get somewhat blurred from time to time. Also, it is definitely possible to conceive of each point as a continuum in itself; for instance, the term nationalism may well include an uncompromisingly secessionist variant as well as a diluted federalist variant.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 93
MONOLITHIC STATES AND NATIONALITIES CONFLICT: THE TRANSFORMATION OF TAMIL NATIONALISM IN SRI LANKA, 1948-1993
My considered view is that Ceylon [Sri Lanka] has already split into two entities. At present this is a state of mind: for it to become a territorial reality is a matter of time.... Under various guises the Sinhalese elites have refused
to share power with the Tamils____The [civil] war may take several years for
a final decision. The longer it takes, the more likely it is that a separate [Tamil]
state will emerge____Compromise agreements will... not be honoured on a
permanent basis. The war will be resumed. The partition of Sri Lanka is already a fact of history.
—Alfred Jeyaratnam Wilson (1988, preface, p. 224)
Indeed, there was nothing inevitable about Sri Lanka's present predicament. Sri Lankan Tamils have been described as "reluctant secessionists" (Horowitz, 1985, pp. 243-249). As K. M. De Silva (1990, p. 32) notes,
the striking feature of the emergence of Tamil separatism in Sri Lanka ... is its late development. The transition from . . . separatist sentiment to a full-fledged separatist movement took over 25 years, and . .. separatist agitation went through several phases.
So how were the Tamils transformed from reasonably loyal citizens of the Sri Lankan state into determined proponents of its destruction? It was to take the cumulative provocation of reams of Sinhalese-supremacist and discriminatory legislation (in the fields of language, education, government employment, and religion); recurrent anti-Tamil pogroms; the dishonoring by Sinhalese regimes of successive compacts guaranteeing Tamil regional autonomy; the degeneration of a formally democratic electoral system into a farcical arena for the competitive (Sinhalese) chauvinism of the country's two main political parties; and, finally, severe military repression—in all of which the strategic Sinhalese-Buddhist elite at the apex of the postcolonial state played an infamous role—before the broad mass of Sri Lankan Tamils became convinced that there was no prospect of living with dignity and security within a unitary, Sinhalese-dominated Sri Lanka and committed themselves to the bifurcation of the island country. It was only from the mid-1970s onward that the Tamils began to define themselves as a nation entitled to an independent state, and their political organizations actively sought secession from only the 1980s on.
Between 1948 and 1956, the main Tamil political party, the All-Ceylon Tamil Congress (ACTC), favored a policy of "responsive cooperation" with the upper caste/class Sinhalese-Buddhist state elite. That changed in 1956, a landmark year for Sri Lanka, and was largely a function of competitive
94    COMPARATIVE POLITICAL STUDIES / April 1995
Sinhalese party politics. In the 1956 parliamentary elections, the major Sinhalese opposition party, the Sri Lanka Freedom Party (SLFP), decided that a strategy of aggressive Sinhalese-Buddhist nationalism (with strong anti-Tamil overtones) represented the best means of defeating the incumbent United National Party (UNP).8 This platform was encapsulated in a simple slogan—"Sinhala Only!"—which meant that on assuming office, the SLFP promised to immediately make Sinhalese the sole official and national language of Sri Lanka (the policy of the UNP regime, up to that point, had been to maintain "parity of status" between Sinhalese, Tamil, and English).9 The SLFP strategy proved successful: The UNP was defeated decisively. But in the process, the SLFP had let the genie of majoritarian nationalism out of the bottle, and it was to prove impossible, during the years and decades to come, to put it back again. The events of 1956 also produced significant changes in the Tamil political spectrum. The ACTC was permanently marginalized, and Tamils voted overwhelmingly for a new political force, the Tamil Federal Party (FP), whose program was contained in its name. Repudiating the now discredited policy of responsive cooperation, the FP spelled out its aims and ideology in very concrete terms: It declared "the Tamil people's unchallengeable title to nationhood and . . . their right to political autonomy and desire for federal union with the Sinhalese" (see Wilson, 1988, p. 24). The center of gravity of the Tamil political agenda had shifted from concern with an equitable share of power in Colombo toward a demand for autonomy for the predominantly Tamil northern and eastern regions of the country.10 The dialectic of the increasingly Sinhalese nationalist and unitary state and Tamil society had resulted in the formulation of a Tamil political charter that, for the first time, claimed the attribute of nationhood for the Tamil social formation and advanced a demand for a renegotiated federal union based on that self-definition of a collective Tamil identity.
8. As to why this was so, Wilson has probably provided the best possible answer: the Westminster-style parliamentary system Ceylon inherited from the British "depends, in the final instance, on the counting of numbers. The Sinhalese constitute the numerical majority" (Wilson, 1988, p. 32).
9. The "Sinhala Only!" campaign was, however, not merely an assertion of linguistic and cultural identity. For reasons that originate in the colonial period, Tamil speakers were somewhat disproportionately represented at this time in government employment, one of the most prestigious and lucrative sectors in Sri Lanka. At least implicit in the SLFP campaign was the message that once Sinhalese was declared the sole official and national language, the Sinhalese electorate could expect a drastic equalization of government employment opportunities. See Horowitz (1985, pp. 155-156) and Nithiyanandan (1987, pp. 111-114).
10. Approximately three quarters of the Sri Lankan Tamil population lives in the north and east of the country. This demographic concentration has supplied the territorial rationale first for the idea of regional autonomy and, much more recently, for the advocacy of an independent Tamil state called Tamil Eelam, to be carved out of these parts of Sri Lanka.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 95
Alarmed at the alienation "Sinhala Only!" had produced among the Tamil minority, the SLFP leader and prime minister, Solomon Bandaranaike, still attempted, in 1958, to effect a rapprochement with their political leadership. Thus a famous pact was concluded between him and the Federalist leader, S.J.V. Chelvanayakam, guaranteeing Tamil rights in education and government employment and providing for wide-ranging decentralization of administration and devolution of powers to an autonomous Tamil authority in the north-east. This pact represented a comprehensive federal solution to growing Sinhalese-Tamil tensions. However, the agreement was soon unilaterally abrogated by Bandaranaike. This was not the result of any intentional bad faith on his part; rather, the pact had fallen prey to the very forces that Bandaranaike himself had unleashed and exploited to capture state power. The dynamics of competitive chauvinism in Sinhalese party politics were reactivated;11 the UNP, in particular, seized on the agreement, which it depicted as capitulation to the Tamils, as a grand opportunity to retrieve some of its lost popularity. Thus the rules of competition for state power among rival segments of the elite dictated that Bandaranaike could not settle the issues hampering amicable relations between Sinhalese and Tamils.12 And exactly the same fate befell another accord, concluded in 1965 between the Sinhalese-Buddhist elite and the Federal Party (this time a UNP prime minister was involved), as a result of hysterical opposition from a coalition of the SLFP (now in opposition), UNP back-benchers, and the Buddhist clergy.
Nonetheless, the Tamil Federalists (who swept the Tamil-majority northeast in every election between 1956 and 1977) continued to be committed to a negotiated solution to the ethnic question within the framework of a united Sri Lanka. Thus their 1965 and 1970 election manifestos stated categorically that
11. This phenomenon, in an entrenched and magnified form, would recur with a monotonous if deadly regularity during the years and decades to come and would simultaneously contribute to an exacerbation of the Sinhalese-Tamil conflict and constitute perhaps the single most formidable obstacle to its resolution. For an explanation of why Sri Lanka's demographic characteristics and electoral system facilitated intra-Sinhalese competition and, consequently, a worsening of interethnic relations, see Horowitz (1985, pp. 390-393; 1989).
12. The Buddhist clergy (some 90% of Sinhalese are Buddhists) have, since the mid-1950s, consistently played the role of an ultranationalist pressure group in Sri Lankan politics and blocked any kind of settlement with the Tamils (who are Hindu and Christian), whom they consider interlopers on Sri Lankan soil. The Federal Party responded to the abrogation with a nonviolent civil disobedience campaign in the Tamil-dominated region. This was put down with use of force by the police and military. The tension generated also led to the first major anti-Tamil riots in modern Sri Lankan history; several hundred Tamils living in predominantly Sinhalese areas were killed by rampaging Sinhalese mobs, about 12,000 rendered homeless.
96    COMPARATIVE POLITICAL STUDIES / April 1995
it is our firm conviction that the division of the country... would be beneficial neither to the country, nor to the Tamil-speaking people.... The only way to regain our right to decide for ourselves our own destiny, without jeopardizing the unity of Ceylon, is the federal form of government, (cited in Wilson, 1985, pp. 85-86; emphasis added)
However, this moderate and conciliatory attitude was to undergo a sea change as the 1970s progressed. And, once again, the catalyst to change was to come from the actions of the strategic elite in command of the state.
In 1972, the SLFP government in office used its massive parliamentary majority to steamroller a new "Republican Constitution" into existence. This constitution, which replaced one bequeathed by the British, amounted to a charter of Sinhalese-Buddhist supremacy. It recognized Buddhism as de facto state religion and reaffirmed the paramountcy of the Sinhalese language. Even formal safeguards for minorities incorporated in the previous constitution were now abolished. Worst of all, around the same time, severe discriminatory measures directed against Tamils were adopted in the closely related spheres of professional higher education and government employment.13 In response, the Federalists reorganized as the Tamil United Liberation Front (TULF) and, at an extraordinary convention in 1976, resolved that
the restoration and reconstitution of the Free, Sovereign, Secular, Socialist State of Tamil Eelam based on the right of self-determination inherent in every nation has become inevitable in order to safeguard the very existence of the Tamil nation in this country, (cited in Wilson, 1988, p. 89)
The way the TULF went about acting on this program was to declare that the impending parliamentary elections, due in 1977, would be a referendum on the sovereignty of "Tamil Eelam." Thus the 1977 TULF election manifesto stated that
the TULF seeks ... the mandate of the Tamil Nation to establish an independent, sovereign, secular, socialist state of Tamil Eelam that includes all geographically contiguous areas that have been the traditional homelands of Tamil-speaking people in this country [i.e., the north and east]. . . . The only way to announce this decision to the Sinhalese state and to the world is to vote for the Tamil United Liberation Front.
The TULF proclaimed further that
the Tamil-speaking representatives who get elected through these votes, while being members of the Parliament of Ceylon, will also form themselves into a
13. The issue of discrimination in access to higher education was particularly crucial in fostering Tamil alienation. There is evidence that Tamils trying to enter universities were being systematically targeted. See Horowitz (1985, pp. 663-664), Nithiyanandan (1987), and De Silva (1984).
Bose / STATE CRISES AND NATIONALITIES CONFLICT 97
"National Assembly of Tamil Eelam," which will draft a constitution for the State of Tamil Eelam and establish the independence of Tamil Eelam... either by peaceful means or through direct action and struggle, (cited in Balasingham, 1983, pp. 27-28)
The popular response was, on the whole, impressive. It was overwhelming in the solidly Tamil-majority northern province, where TULF candidates captured every one of the 14 parliamentary seats. The picture was more ambiguous in the demographically mixed east, where Tamils comprised a plurality but not a majority of the population (a result, Tamil nationalists claim, of state-sponsored "colonization" of Tamil-dominated areas),14 but the TULF did very well in heavily Tamil pockets here as well, winning 4 of the 10 seats from the province. Federalist nationalism had finally yielded to secessionist nationalism.
Events moved very swiftly after 1977. There were four serious outbreaks of ethnic rioting directed against Tamils—in 1977,1979,1981, and 1983—all of which are alleged to have been orchestrated by agents of the state and especially by activists of the ruling party, the UNP, which had ousted the SLFP government in the 1977 polls.15 The worst of the pogroms was in July 1983, when some 3,000 Tamils were murdered in mob violence in Sinhalese-majority areas, the economic base of the Tamil minority in Colombo was practically destroyed, and more than 150,000 were made homeless. Tens of thousands of these refugees sought the relative safety of the Tamil-majority north-east, and many thousands more fled to neighboring India and the West, thereby internationalizing the Tamil question.
However, the ethnic violence cannot be seen in isolation. It went hand in hand with several other developments. First, the Sinhalese regime failed to come up with any constructive scheme to devolve power to an autonomous
14. The colonization debate has been one of the most sensitive issues in Sinhalese-Tamil relations since the 1950s. State-sponsored colonization projects have been consistently denounced by Tamil political formations as a conspiracy designed to erode and eventually destroy Tamil demographic and economic predominance in areas where they have historically constituted the majority of the population. Sinhalese regimes have steadily maintained that the sole purpose of these schemes has been to provide poor, landless Sinhalese with plots of land in sparsely populated areas of the north and east. The controversy took an especially ugly turn during the mid-1980s, when Tamil guerrillas began to attack and massacre Sinhalese "settlers" in the east and north.
15. Many newspapers reported the open involvement of leading UNP politicians and trade union officials in the 1983 violence. The government did very little to stop the violence (indeed, in many places, troops and police enthusiastically joined the rioters). President Jayewardene, in a radio broadcast at the height of the rioting, did not say a single word in condemnation of the rioters or in sympathy for the victims; instead, he blamed the Tamils for "provoking" the violence. All this irrevocably compromised the state security apparatus, as well as the president himself, in the eyes of Tamils.
98    COMPARATIVE POLITICAL STUDIES / April 1995
federal authority in the Tamil-majority region. Instead, after the 1983 riots, it passed a law excluding the elected TULF representatives from Parliament, thereby eliminating the constitutional-nationalist Tamil opposition from the political scene. This represented the proverbial last nail in the coffin of the TULF, which had already been losing ground in Tamil areas over the preceding few years to radical youth groups advocating armed struggle because of its failure to formulate any concrete program to implement the mandate for Eelam received at the 1977 elections.16
In the meantime, the situation in the north had also been deteriorating steadily as the government had, in 1979, enacted a draconian Prevention of Terrorism Act (PTA)17 and sent in large numbers of troops to flush out small groups of Tamil guerrillas staging hit-and-run attacks against police and military installations. These troops were given a mandate to wipe out "terrorism" within 6 months; as Sivanandan (1990, p. 240) notes, "at the end of the six months, the specter of terrorism the President had sent the troops to wipe out had been made flesh by the army." Counterinsurgency operations against an elusive guerrilla foe necessarily entailed large-scale victimization of the civilian population, and this only produced more alienation, more guerrillas, and more people willing to shelter and support them. Even so, it was only after July 1983 that Tamil youth began to join the Tamil Tigers and similar radical groups in really significant numbers and that the nascent insurgency spread throughout the north-east. In retrospect, it does seem that the 1983 pogrom was the incendiary spark that sent a volatile situation beyond the point of no return.
The situation continued to worsen from then on until, by the mid-1980s, the northeastern region was in the grip of full-scale civil war.18 The only
16. Radicalized young Tamils in the north had been becoming steadily disillusioned with the all-too-conciliatory parliamentarism of the TULF between 1977 and 1983. The organization that gained the most support at the expense of TULF was the Liberation Tigers of Tamil Eelam (LITE, 1990), the largest, best organized, and most disciplined of the Tamil guerrilla groups. The LTTE is the "sole spokesman" of the Tamil nationalist movement in Sri Lanka today and is also one of the most sophisticated and powerful armed opposition movements in the world.
17. This infamous legislation magnified the worst aspects of the British and South African ordinances of the same name.
18. After 1983, strategic equations in south Asia also began to impinge on the Sri Lankan conflict. Between 1983 and 1987, various Tamil guerrilla movements received significant military and logistical assistance, not to mention moral and political support, from India's central government as well as from its provincial counterpart in Tamil Nadu, home to some 60 million Indian Tamils. Indian policy during this period seems to have been motivated by a mix of perceived strategic interests (as the "great power" of south Asia), the need of Congress regimes to divert attention from compelling domestic problems, and popular pressure from sections of the citizenry in Tamil Nadu. However, the role Indian support played in boosting the Tamil insurrection should not be overstated. The civil war resulted from factors internal to Sri Lanka.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 99
policy that the Sinhalese-Buddhist elite seemed capable of was brute repression, and this was not just unproductive but counterproductive. Military offensives and atrocities (as well as, since 1990, an economic blockade against and indiscriminate aerial bombing of the north, the Tigers' "liberated" zone) only consolidated the emerging Tamil national formation and cemented the determination of the rebels to fight on. Nithiyanandan (1987, p. 150) notes "a direct correlation . . . between the growth of Tamil militancy and police and military excesses," and Tambiah warned (1986, p. 124) that "what the . . . government is achieving by its persistent military action and terrorizing of all Tamils is to drive the latter . . . towards the imagined haven of Eelam and a blanket ethnic solidarity." Indeed, a Tamil Tiger spokesman (Balasingham, 1983, p. 4) has acknowledged that "Sinhalese chauvinism and its violent manifestations have helped the coming together of the heterogeneous masses of the oppressed Tamil nation, and . .. towards a determined revolutionary struggle for political independence." A British observer visiting the war-torn North in 1991 found that the pressures of the war were severe. There was no cash economy, essential drugs could not be obtained, and fishing had drawn to a halt. Moreover, there was no electricity and people were on the brink of starvation as food shipments to the North had been either delayed or diverted. Many people looked to the promise of liberation that the Tamil Tigers offered. The aerial bombing of civilian targets such as schools, hospitals, and refugee camps had convinced those living in Jaffna that the Sinhalese state was attempting genocide (West, 1991, p. 13).
What, then, are the lessons to be drawn from the Sri Lankan case? Rothschild, in a work devoted to delineating a theoretical framework for the comparative analysis of ethnic conflicts, has unwittingly provided a remarkably accurate description of the specific case of the modern Sinhalese-Tamil conflict as it unfolded in Sri Lanka:
In sum, the state has become the decisive political vehicle and . . . arena for ethnic groups—dominant and subordinate____In the present historical era, this
Since 1987, when the Indian state made a massive military intervention to quell the Tamil uprising, the principal Tamil organization, LTTE, has been in a virtual state of war with New Delhi. But the Tigers proved themselves eminently capable of waging a successful war of attrition against the huge Indian military force sent to suppress them between 1987 and 1990. And since the resumption of hostilities with Colombo in 1990, the Sri Lankan regime has been unable to retake the Tiger-held northern province (or to eliminate the major Tiger presence in the east) despite the termination of Indian patronage of the Tamils. Similarly, whereas financial contributions from a world network of the Tamil diaspora have been exceptionally useful in sustaining the Tamil insurrection, the willingness of expatriates and refugees to contribute is really more a symptom of pervasive Tamil alienation from the Sinhalese-Buddhist state rather than a cause of the problem.
100    COMPARATIVE POLITICAL STUDIES / April 1995
means, in effect, that they must bid for exclusive or participant political control over and/or in a state. Without some such political leverage through a state-apparatus, even their cultural and socioeconomic, let alone their political, interests are jeopardized. And if an ethnic group's bid for an adequate share of political power and control within an extant multiethnic state proves unproductive, is repudiated as non-negotiable. ... It may well make a secessionist bid for a state of its own—driven by interest, as well as emotion—this secessionist option is available, of course, only to regionally concentrated, autochthonous ethnic groups. (Rothschild, 1981, pp. 232-233)
The Sinhalese-Buddhist strategic elite went for a monopoly of power over the state apparatus and sought to deny the Tamils even a "participant" share of control.19 The framework of hegemonic, unitarist rule that they created in the process slowly but steadily destroyed Sri Lanka as a unified entity. Thus the central factor that has contributed to the growth of the Sinhalese-Tamil conflict is the consolidation of a corporate Tamil identity and the gradual rise of a demand for secession as a societal response to the policies and actions of a central state elite that tried to systematically undermine, if not totally deny, that identity. This, then, is the irony of the dialectical relationship of the Sinhalese-Buddhist state and Tamil society: The elite's strenuous efforts to homogenize its domains led, ultimately, only to discord and ethnic confrontation. As a senior Tamil Tiger leader has commented on the "oppression perpetrated [on the Tamils] by successive Sri Lankan governments . . . for ... several decades":
It was an oppression ... intended to erase the Tamil national identity. Instead, it has served to consolidate ... the Tamil nation. The actions [of the state]... intended to destroy the Tamil national entity, have in reality solidified it. The discrimination [over] employment, education, language, the colonization of the Tamil homeland, and physical attacks amounting to genocide have made the Tamils resist Sinhala rule as one people, one nation. We are today a people with a deep-rooted political consciousness of our national identity. (Thilagar, 1992, pp. 1-2; emphasis added)
Indeed, the process of nation building, as attempted by those at the helm of the postcolonial Sri Lankan state, actually facilitated the emergence and development of a Tamil national consciousness and eventually precipitated a crisis of the state defined in terms of a solidary, exclusivist Sinhalese-Buddhist interest and identity. The Sri Lankan Tamils had, over a period spanning three to four decades, been gradually transformed from a mere ethnic group into a social formation acutely conscious of its national identity.
19. Incidentally, not a single elected Tamil representative from the northern province has held a ministerial portfolio in the central government since 1956.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 101
MONOLITHIC STATES AND NATIONALITIES CONFLICT: THE DIALECTICS OF NATIONALISM IN YUGOSLAVIA, 1945-1993
[The Partisan War is creating] the unity of all the nationalities of Yugoslavia, for that is the foundation that is now being built, the foundation of brotherhood, concord and unity that no one will ever be able to destroy.
—Josip Broz Tito, address to the first meeting of the Anti-Fascist Council of People's Liberation of Yugoslavia (AVNOJ), Bihac, Bosnia, November
1942 (Tito, 1972, p. 105)
Tito . . . created, I believe it will be seen, no lasting spiritual or institutional forms. . . . When the new class leaves the historical scene—and this must happen—there will be less sorrow over its passing than... for any class before it. Smothering everything except what suited its ego, it has condemned itself to failure and shameful ruin.
—Milovan Djilas (1980, p. 179; 1991, p. 106)
It would certainly be unjustified to regard the post-World War II Yugoslav experience as an unmitigated disaster. Postwar Yugoslavia has seen processes of industrialization, urbanization, and educational expansion that are, for their vast scope and rapid pace, unprecedented in the history of the 20th century.20 As Singleton (1985, p. 285) remarks,
There are... many positive factors to the balance sheet. Yugoslavia has made much progress ... in the post-war period. ... Its people are, on the whole, more prosperous and freer than in any of the post-war people's democracies.
So how does one explain the current catastrophe, except perhaps by regressing to that always tempting residual variable, the primordial enmity of nations? I believe, to the contrary, that a critical examination of the structure and development of the postwar Yugoslav state and, in particular, the role and policies of the strategic ex-Partisan Communist elite at its helm supplies the key to understanding the disintegration of the Yugoslav union.21
Zukin (1984) has noted that
any discussion of the state in Yugoslavia is forced to confront several seemingly insurmountable obstacles ... [including] the contradiction in ... socialist societies between a basic assumption of "withering away" and a continuous governing apparatus, and the striking differences between Yugoslavia and other socialist states. Because criteria for comparison usually derive from
20. For statistics, see Denitch (1976).
21. For four decades after 1945, the upper echelons of Yugoslav politics were dominated by an ex-Partisan elite of not more than some 1,200 persons, revolving around the personality of Tito.
102    COMPARATIVE POLITICAL STUDIES / April 1995
Soviet practice, the sine qua non of such a state is taken to be a centralized authority, especially in managing the economy, and a monopoly of political initiative in the hands of the communist party. At first glance, the centrifugal nature of political relations in Yugoslavia, as well as a diffusion of effective social control, constitute a departure from this model. But in reality, the Yugoslav leaders 'attempts to respond to domestic and foreign pressures merely modified the concept of a "socialist state." (p. 249; emphasis added)
Zukin's point goes to the heart of my own argument in this section. At first glance, the consolidated existence of a decentralized federal institutional structure in post-World War II Yugoslavia appears to contradict my theoretical argument. But, as I show, the Yugoslav case, although highly complex, actually serves to powerfully reinforce and confirm the validity of my thesis.
The Partisan elite that took power in Yugoslavia in 1945 were indoctrinated Communists. To them, notions such as "democratic centralism" and "the leading role of the Party" were articles of faith. Thus they were confronted with a major existential crisis when, in 1948, Stalin denounced them as "deviationists," "tools of imperialism," and so forth, and had Yugoslavia excommunicated from the Comintern. However, the severance of the Moscow connection also enabled the Yugoslav state elite to chart its autonomous course of social, political, and economic development and paved the way for ideological "innovations" such as workers' self-management (in contrast to the Stalinist command-economy model) as well as a highly decentralized political system with constituent republics of the union enjoying widespread autonomy in various fields. Some form of decentralized institutional structure probably was in the cards in any case, for the single most influential factor in winning the Partisans popular support during the war was their promise of a just solution to the festering "national question" within the framework of a federated Yugoslavia.
Despite this recognition of the diversity of Yugoslavia, the Communist strategic elite nonetheless
did not give up the idea that only a hegemonic party-state—a true vlast—could control social change. . . . Certainly differences of opinion and alternative proposals [were] expected to emerge [if only in forums monopolized by the Party] . . . but under no circumstances [were] they supposed to generate an organized opposition. (Zukin, 1984, pp. 250, 263)
Thus Edvard Kardelj, the principal Yugoslav Communist theoretician, expressly and permanently excluded the creation of multiple parties and, by extension, any dilution of the absolute power of the ruling Communists. According to him, the LCY leadership "never fell into the illusion that the role of the state [i.e., party] and of state coercion was either a subordinate or an unimportant element of the political system" (Zukin, 1984, p. 251).
Bose / STATE CRISES AND NATIONALITIES CONFLICT 103
This was the fundamental and unresolved paradox that slowly ate away at the entrails of Titoist Yugoslavia and eventually led to its apocalyptic demise. There was a basic contradiction between a clear imperative to democratize, given the inherent and ever-growing complexity and diversity of Yugoslav society, on the one hand (recent Yugoslav history, especially World War II, was rife with lessons that hegemonic rule, of whatever type, ultimately produced disastrous consequences), and the seemingly unshakable commitment of the strategic elite to a highly centralist framework (permanently enshrined via the supremacy and power-monopoly of the League of Communists), on the other.22 It was this dialectic of unitarism and federalism that facilitated the institutionalization of the Yugoslav variant of what Roeder (1991, p. 199) and Rakowska-Harmstone (1974, p. 4), writing on the former Soviet Union, have called "ethnofederalism" and "orthodox [i.e., not antisys-temic] nationalism," respectively. It was the same dialectic that, in a context of pervasive systemic crisis, led to the breakdown of the Yugoslav union. And there was no solution to this contradiction within the rigid framework of the single-party system. As Tome (1988, pp. 70,72) writes, "the party's monopoly of power continued to be an essential tenet of its definition of socialism . . . [and] one party-rule and a [truly] autonomous society are mutually exclusive."23
But why did one-party rule, however oppressive, necessarily have to lead to the disintegration of Yugoslavia along national/ethnic lines?
Many years ago, Vucinich (1969, p. 282) discerned an extraordinary preoccupation among the Yugoslav state elite with the most minute details of the institutional or organizational aspects of federalism. In fact, post-World War II Yugoslavia produced as many as four constitutions—in 1946, 1953, 1963, and 1974.24 All of the constitutions except the second guaranteed the federal republics the right to secede, and the lengthy and convoluted constitution of 1974 effectively enshrined what would appear to be a confederal institutional arrangement in Yugoslavia.
In light of this, it would seem surprising that Vucinich (1969, p. 253) wrote that "while officially described as a federal state, Yugoslavia resembles more closely a unitary state." But he explained why:
22. Milovan Djilas (1980) commented that "Yugoslav society is pluralistic and stratified. Only the party at the top is monolithic" (p. 173). Indeed, one might argue that given the far-reaching changes brought about in postwar Yugoslav society by state-sponsored industrialization and urbanization, it was becoming ever more urgent for the Titoist elite, as the years went by, to democratize the state and bring it into line with the increasing diversity and vitality of civil society.
23. The problem in Yugoslavia was not simply one-party rule but personalized one-party rule, given the mythical cult created around Tito.
24. The Basic Law of 1953 is generally regarded as a constitutional document.
104    COMPARATIVE POLITICAL STUDIES / April 1995
The Communists gave their country a federal framework in establishing ethnic republics and granting certain formal and even real concessions to individual peoples. But they never gave the peoples of Yugoslavia the opportunity to exercise self-determination, and bringing it into play ... is precluded by the Communist constitutional system and the strict application of the principle of democratic centralism. (Vucinich, 1969, p. 281)
As Rakowska-Harmstone (1974, pp. 1-2) writes of the Soviet experiment with federalism, the Soviet search for a political system "national in form but socialist in content" meant, "in practical terms, the establishment of a federal institutional structure in which the more important national groups would be accorded the forms of statehood, but their exercise of . .. autonomy would be severely circumscribed by the monopoly of power exercised by the ... CPSU."
Banac (1992b, pp. 170-173) concurs that
Tito's domestic policy . . . actually put a premium on the power of the center. ... Decentralization was enshrined in the Constitution of 1974 ... nevertheless [even] this constitution retained the majority of power for the... center and the ruling party, and was, hence, by no means the code for a confederation.
Stojanovic (1991) is even more explicit; he claims that Yugoslavia "was not a genuine federation before 1974, and did not automatically become one after the constitution of that year" (p. 94). Zukin (1984, p. 251) observes that "the distinctive forms Yugoslav socialist practice takes have all been initiated and controlled from the top." And Djordjevic (1992) has aptly described the eventual consequences for Yugoslav unity of the 1974 constitution, a document that graphically illuminates the dialectics of nationalism in socialist Yugoslavia: "Wandering between state centralism and self-managing decen-tralism, the 1974 constitution established a polycentric centralism, which resulted in the dysfunction of the whole system" (p. 334).
But how precisely did this framework of the postwar state feed into the fires of ethnically based nationalism? To answer this question, one must analyze the implications for nationalist mobilization of the decentralized administrative framework of the Titoist state.
It would be reasonable to describe the "(con)federal autonomy" guaranteed by Titoism as substantively the autonomy of the various republican LCY elites based in Belgrade, Zagreb, Ljubljana, Sarajevo, Skopje, and elsewhere. Wide-ranging republican autonomy was quite acceptable (indeed desirable) to the central elite so long as that autonomy was carefully controlled and regulated by the LCY (i.e., the party monopoly of power was not challenged).25 Thus the strategic elite delegated many of its functions and respon-
25. Zukin (1984) suggests that various elaborate institutional mechanisms at the republic level were in fact primarily intended to serve as means of manipulation for the strategic state
Bose / STATE CRISES AND NATIONALITIES CONFLICT    105
sibilities quite agreeably to its subelites. But by keeping the really decisive authority safely centralized in Belgrade and precluding all genuine political opposition, on one hand, and making the national/territorial groups (through the instrument of republican subelites) the paramount organizing principle of the postwar state, on the other, Titoist nationalities policy had the effect of conferring on not just LCY politics but on republic-based LCY politics, a virtual monopoly of "legitimate" political activity in Yugoslavia. This had two consequences. First,
by eliminating genuine alternatives in political life... the party power monopoly may have left national and religious identifications as the most salient. . . identifications. . . . Defense of localist and national interests had become the sanctioned artificial substitute for the democratic pluralism of contesting alternative programs." (Denitch, 1990, p. 78)
This was combined with the fact that
the League [of Communists] did nothing to create linkages that would cut across these [republic] boundaries, but. . . strengthened the localist tendencies. ... It created a situation that fragmented any potential opposition into a number of [republican or provincial] "reservations," and forced LCY politicians to become localists defending the interests of their own republic____The
danger always was that the situation would get out of hand and that manipulated national sentiments would assume a life of their own. At that point the local leaders were faced with the unpleasant option of either adapting to the [nationalist] forces and being attacked by the rest of Yugoslavia as nationalists, or turning against the rising nationalism in their own republics and being denounced as unitarists and traitors to their people's interests. (Denitch, 1990, p. 55)
The net effect was that Tito's replacement of political pluralism (multiparty democracy) with a peculiar "regional pluralism" (decentralization to regional communist subelites)
prevented the development of any popularly based parties that transcended ethnic lines. . . . When independent political parties finally developed at the end of the 1980s, they were, almost without exception, parties of specific nationality groups. Regional pluralization, in effect, laid the groundwork for today's ethnic mobilization and fragmentation. (Ramet, 1992, p. 251)
As Pomerants wrote of a similar process occurring during the last years of the Soviet Union, it was almost as if "nationalities . . . turned into political parties" (cited in Zaslavsky, 1992, p. 107).
elite. Stojanovic (1991) and others have argued that institutions of economic self-management were primarily meant to be transmission belts for the central state authority to penetrate as widely and deeply as possible, with the ultimate objective of securing the present and the future for one-party rule in Yugoslavia.
106    COMPARATIVE POLITICAL STUDIES / April 1995
And there were other serious failures of state policy that served to further sharpen and highlight national/regional differences. Perhaps most important was the persisting (indeed deepening) asymmetry in regional economic development—Croatia and Slovenia, on the one hand, and the rest of the republics, especially Macedonia, Bosnia-Herczegovina, and the Autonomous Province of Kosovo, on the other. To be sure, the Communists had inherited this problem when they assumed power. But during the decades of their rule, the disparity between the "advanced" and the "backward" republics widened steadily, as the relative positions of Bosnia, Montenegro, and Kosovo actually declined in comparison to Croatia and Slovenia and that of Macedonia improved only marginally, if at all.26 The political implications of this imbalance have been well stated by Singleton (1985):
This situation would be serious enough in a country with a culturally homogeneous population, but it becomes explosive in a multinational country like Yugoslavia. ... On both sides of the dividing line between rich and poor this situation has fueled the fires of nationalism, (p. 270; emphasis added)
Indeed, one of the slogans of the Croatian nationalist outburst of 1971 was "Stop the Plunder of Croatia!" reflecting the widespread Croat (and Slovene) grievance that their economies were being milked and their revenue was being siphoned off to subsidize the chronically depressed economies of the south.27 It was during the Croatian troubles of 1971 that the powerlessness of republican Communist subelites in the face of large-scale nationalist sentiment in their republics became apparent. This was hardly a surprise, for the primacy given by the LCY central elite to its republican units had ensured that the formative and socializing experiences of most party cadres during the postwar period had been wholly within the context of these units. With the range of most political careers effectively circumscribed not by the boundaries of the union but by the borders of the constituent republics and provinces, it was hardly accidental that, when push came to shove, the loyalty of republican LCY hierarchies came down solidly on the side of their ethnic brethren. This process attained its apogee during the terminal crisis of 1990-1991. But it was very much in evidence in the Croatia of 1971. As Remington (1984, p. 263) argues, "such [reformist/liberal] Croat [LCY] leaders of 1971 as Miko Tripalo and Savka Dabcevic-Kucar were captured by their republic's political constituency, not by a counter-revolutionary movement [as the official explanation made it out to bel" (emphasis added).
26. For figures, see Singleton (1985, p. 270) and Tomasevich (1969, p. 62).
27. The phenomenon of economically advanced republics being especially prone to nationalist sentiment and mobilization has been noted by several analysts of the former Soviet Union. See especially Roeder (1991) and Rakowska-Harmstone (1974).
Bose / STATE CRISES AND NATIONALITIES CONFLICT 107
As a result, the "League of Yugoslav Communists gradually dissolved into nationalistic Leagues of Communists" (Djordjevic, 1992, p. 338).
Titoist nationalities policy represented a complicated accommodation by a strategic elite of strongly ingrained unitarist inclinations with the ethnic diversity of the territory under its control. Indeed, control can be regarded as the key term here: The principal aim of this strategy was to neutralize this diversity (and minimize the possibility of conflict along national lines) in a society undergoing rapid social change and modernization. Although seemingly astute and effective in the short term, this policy, like its Soviet counterpart, "contained the roots of its own longer-term dysfunction" (Roeder, 1991, p. 196). Why was this so? To borrow once again from Roeder's work on Soviet federalism and ethnic mobilization, the Titoist elite "achieved interethnic peace not so much by eliminating the root causes of ethnic grievances as by eliminating mobilizational opportunities for independent ethnic protest" (p. 199). This was accomplished "by prohibiting all but sanctioned political entrepreneurs from mobilizing their communities"— that is, by giving party cadres, especially at the republic level, "an institutionalized monopoly on the public expression of ethnic identity" (pp. 203, 205). The problem was that in a changed context of rising political turmoil and growing economic crisis, "these federal institutions and [republic] cadres became [as in the USSR] instruments of the new ethnic assertiveness" (p. 199). The problem was compounded by the fact that more authentically "nationalist" counterelites, although generally suppressed under the Titoist dispensation, were never quite done away with altogether. As in the Soviet case, so also in Yugoslavia, ethnically based collective action inaugurated and promoted by republican party elites generated opportunities for " 'piggybacking' by [these] potential counterelites" to the point of sometimes even providing aspiring "counterelites with a readymade [nationalist] movement" (Roeder, 1991, p. 227). Croatia in 1971, as well as in the late 1980s, is one good example of this phenomenon. The dialectical interaction of rigid, single-party unitarism and decentralization to ethnically defined regional party elites is crucial to understanding both the fact and the form of the disintegration of Titoist Yugoslavia. This dialectic ensured that the quest for the Yugoslav citizen ultimately proved an illusory one.
Indeed, perhaps the ultimate irony about the Titoist state is that in the end it satisfied almost no one. Instead, every conceivable national group came to harbor some kind of grievance against it. Slovenes were left complaining about the "exploitation" of their economy, Croats continued to nurture their own bitterness at the manner in which Tito had suppressed their nationalist upsurge in 1971, and Albanians were discontented over their miserable economic condition and the denial of republic status to Kosovo.
108    COMPARATIVE POLITICAL STUDIES / April 1995
But by far the most serious sign of trouble was the renewal of an exceptionally aggressive and virulent strain of Serb nationalism, whose worldview seemed directly derived from the Serb-supremacist Chetnik movement, which had lost out badly (and, one might have thought, permanently) to the Partisans in World War II.28 Unbridled Serbian nationalism presented the gravest threat to the continuation of the fragile Titoist order because Serbs not only were the single most numerous nationality but lived in large numbers in Croatia and Bosnia-Herczegovina. As early as 1981, a mere year after Tito's death, the Serbian LC Y had been lamenting that "Serbia was the only republic not constituted as a state because of the disruptive role of the autonomous provinces [Vojvodina and Kosovo]" (cited in Banac, 1992b, p. 175). In 1983, an estimated 100,000 persons turned out in Belgrade for the funeral of Aleksandar Rankovic, posthumously elevated to the pantheon of Serb national martyrs.29 In 1986, there came the now infamous "Memorandum" of the Serbian Academy of Arts and Sciences, which alleged that Serbs had been discriminated against in the Titoist state by an "anti-Serb coalition" of Croats, Slovenes, and "Serbophobic communists" and that Tito and Kardelj (a Croat and Slovene, respectively) had conspired to divide Serbs by deliberately parceling out the Serb population among several republics while simultaneously encouraging Albanian nationalism in Kosovo, for example.30
Despite this pervasive crisis and a creeping economic arteriosclerosis,31 "until 1987 it seemed that Yugoslavia could continue to muddle along" (Ramet, 1992, p. 225). The advent of Slobodan Milosevic changed that. The role of Milosevic's ruthless opportunism and roguish demagoguery32 in precipitating the final apocalyptic demise of Tito's state has been well documented elsewhere (Banac, 1992a, pp. 141-175, 1992b; Ramet, 1992, pp. 225-269). However, it is worth reiterating that whereas Milosevic was
28. For a comprehensive study of this movement, see The Chetniks (Tomasevich, 1975).
29. Rankovic was a legendary Partisan war hero and later chief of the Yugoslav secret police, UDBa. He was also a hardline communist centralist purged in disgrace by Tito in 1966 for misuse of power and obstruction of liberalizing reforms.
30. See Ramet (1992, p. 200) and Banac (1992a, pp. 150-151; 1992b, p. 176). The Serb-Montenegrin population in Kosovo had declined from 28% in 1948 to 15% in 1981 and 10% in 1987. However, most of this decline is probably explained by spontaneous Serb migration and a vastly higher birthrate among Albanians.
31. About a million people, or 18% of the workforce, were unemployed by 1983. National product had been stagnant since 1982, and the inflation rate reached 1,000% in 1989. In 1988, 28% of Kosovo Albanians were officially without jobs.
32. Ironically, and tellingly, Milosevic had risen through the hierarchy of the Serbian LCY. The hysterical confrontation with Kosovo that he engineered was soon extended to other republics—first Slovenia, then Croatia, and finally Bosnia-Herczegovina.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 109
undoubtedly the critical catalyst and conjunctural factor that plunged Yugoslavia from confused paralysis into full-scale civil war, the processes that culminated in the systemic collapse of1990-1991 had deep structural roots. In a sense, Milosevic is more an exceptionally malignant symptom or manifestation of the terminal crisis of the superstructure of the Titoist state in relation to the infrastructure of Yugoslav society (or perhaps societies) rather than its cause. Even the revival of Serbian nationalism clearly predates the rise of Milosevic to power (although Milosevic undeniably seized on this revival with a vengeance and greatly exacerbated it through his politics). Thus I argue, with Kornai (1992, p. 394), that "the grave and long-term failures of the system must be distinguished from the ad hoc events that bring out the problems with dramatic force."
That said, it remains highly doubtful if Yugoslavia would have disintegrated so completely, and in such violence, had it not been for the destructive force of Serb nationalism. The process of disintegration, which in retrospect seems so absolute, was perhaps actually reasonably open-ended. The dominant popular sentiment in Croatia and Slovenia until at least 1988 was for a (genuine) confederation, not outright secession. It is doubtful that the Croat nationalists of Franjo Tudjman's Croatian Democratic Union (HDZ) could have won the Croatian elections in the spring of 1990 had it not been for the threat posed by Milosevic; even so, the election was closely contested between HDZ and reformist Croat Communists, campaigning on a confederal platform as the "Party of Democratic Change." Even more significantly, the Serb population in Croatia voted predominantly not for Milosevic's local surrogates, the Serbian Democratic Party (SDS), but for the Party of Democratic Change. And Bosnia and Macedonia reluctantly quit the union only when the Croats and Slovenes opted out and it became clear that if they did not follow suit, they would be left totally at the mercy of Milosevic in a rump Yugoslavia. Thus there does not seem to have been anything inevitable about the way Yugoslavia broke up until almost the very end.
Banac (1992b, p. 179) writes that whereas "Milosevic united Serbia and the Serbs," he simultaneously "destroyed Tito's party" (and, one might add, Tito's state). True enough, except that I would go considerably further in my final word on the demise of Yugoslavia. Like the Sinhalese-Buddhist strategic elite in Sri Lanka, the Titoist strategic elite in Yugoslavia went for a permanent monopoly of power over the state apparatus, even though the ethnic and social diversity of the society they sought to control and dominate militated against the long-term viability of such a project. In the process, they dug their own grave. Milosevic merely filled in the earth.
110    COMPARATIVE POLITICAL STUDIES / April 1995
POLICY IMPLICATIONS AND FUTURE PROSPECTS
Despite the frequently arbitrary and artificial manner in which it came into being, the [international] state-system is firmly anchored in contemporary reality, and the central trend appears to be its aggrandizement and reinforcement____Whether it is viewed as the ultimate framework for human fulfillment
or a constricting straitjacket of discord and division, that the state-system in ... its present form is here to stay seems one of the most durable axioms of modern politics.
—Crawford Young (1976, p. 82)
[A] critical transformation that seems imminent is the erosion of the normative force of "territorial integrity" doctrines in international affairs. The long period of state stability appears at an end. The break-up of Canada, Yugoslavia, Ethiopia, Sudan, Sri Lanka, among others, is no longer a remote contingency. Some of these will find formulas for survival, but it is inconceivable that all will persist in their present form into the next century. The international normative order will thus once again need to redefine the scope and limits of the doctrine of self-determination.
—Crawford Young (1992, p. 92)
One normative implication of the foregoing analysis has been that had the policies of strategic state elites in Sri Lanka and Yugoslavia been different, the eventual political outcomes in those countries might have been different too.33 A conclusion that would seem to emerge from my comparative appraisal of Sri Lanka and Yugoslavia is that neither a formal parliamentary democracy in an otherwise highly unitary state34 nor a partial and deeply flawed federalism (i.e., devolution of powers to regional authorities in the absence of competitive political democracy) is in itself sufficient to ensure the durability and legitimacy of the state in such multiethnic societies. In general, then, I postulate that a healthy dose of both a thriving multiparty democracy and a robust, substantive federalism represents the optimum solution and the most democratic arrangement in multiethnic countries such as Sri Lanka and Yugoslavia.35 In contrast to Rabushka and Shepsle (1972),
33. For a discussion of the role of the counterfactual in political science theory, see Fearon (1991).
34. In Sri Lanka, the electoral system, by promoting intraethnic outbidding between Sinhalese parties, seems to have worked in the opposite direction and fostered interethnic conflict. However, this is certainly no reason to conclude that multiparty democracy is in itself inimical to ethnic harmony. As Horowitz (1989) has shown, in Malaysia, another state with a fragile ethnic balance, electoral competition did promote cooperation between ethnically based Malay and Chinese parties.
35. The recent resurgence of separatist sentiment in Quebec, the Francophone province already enjoying substantial autonomy in a multiparty democracy, Canada, would appear to contradict this argument. But this is not really the case. For an explanation, see Pinard (1992). Compelling evidence is also forthcoming from other parts of the world that granting regional
Bose / STATE CRISES AND NATIONALITIES CONFLICT    111
who disagree that "the resolution of intense but conflicting preferences in the plural society [is] manageable in a democratic framework" (p. 217), I argue that any viable solution must lie in an extension and expansion of both the "procedural and substantive values" of democracy (Dahl, 1989, p. 163). A view of state power that regards it as a zero-sum game (the fatal error of strategic elites in both Sri Lanka and Yugoslavia) can never lead to the creation of the mutually supportive, complementary identities (Croat and Yugoslav, Tamil and Sri Lankan) that is essential to democracy and stability in such multiethnic states. On the contrary, it is a sure recipe for state disintegration and nationalities conflict.
Ethnic conflicts such as these can therefore be regarded, from one angle, as contests between state and nation over competing, conflictual claims to sovereignty. A conceptualization of sovereignty that emanates from the highest echelons of a monolithic state is at fundamental variance with one that claims that sovereignty is vested, above all, in the social base of that self-defining community, the nation. The Sri Lankan and Yugoslav states have either collapsed or been severely undermined36 because they viewed sovereignty as nonnegotiable, something to be monopolized by those at the apex of apparatuses of power, control, and domination. The core issue, then, is one of a fundamental democratization37 of the state and, by extension, a
autonomy to distinctive groups serves to consolidate rather than weaken the juridical state. For some fascinating findings on Basque and Catalan nationalism in Spain, see the recent research of Linz and Stepan (1992, pp. 126-130).
36. The disintegration of Yugoslavia is now a juridical as well as an empirical reality. The putative state of Tamil Eelam is still far from securing juridical status, even though the idea of such a state seems to command overwhelming support among Sri Lankan Tamils and despite the fact that the principal Tamil nationalist movement, the Liberation Tigers, now decisively controls the bulk of the territory of the putative state. The Tigers have established the framework of a de facto state in northern Sri Lanka, and that region now possesses many features of empirical sovereignty. Thus we have a situation of "fractured sovereignty" on the island, with Sri Lanka clinging to juridical statehood but deficient in empirical sovereignty and Tamil Eelam showing signs of empirical sovereignty but devoid of juridical recognition.
37.1 stress that the issue is one of democratization of the state structure as distinguished from mere liberalization. Linz and Stepan (1992) distinguish clearly between the two concepts and make the point that liberalization can occur without democratization. Halfway-house solutions such as the "constitutional engineering" advocated by Horowitz fall under the liberalization category. I feel strongly that such measures are extremely inadequate, especially in cases where state disintegration has reached an advanced stage (as in Sri Lanka), although it is possible that such measures, if adopted much earlier, might have exercised a mitigating effect on exploding conflict. But typically Horowitz-like "constitutional tricks" (a diluted form of proportional representation, a strong presidential executive) were attempted in Sri Lanka during the late 1970s but proved too little, too late. And the Yugoslav experience suggests clearly that even the most complex constitutional balancing act is liable to come apart at the seams in the absence of substantive democratization.
112    COMPARATIVE POLITICAL STUDIES / April 1995
radical restructuring of state-society relations. And, indeed, one can hardly overstate the urgency of the need to reconcile, in good time, the competing claims of state and nation to sovereignty if one remembers what has resulted in Sri Lanka and Yugoslavia, where such reconciliation was not even seriously attempted.38 The urgent need to democratize the state in multiethnic societies is thus very much a practical imperative, an issue that is (or at least should be) absolutely central to the agenda of realpolitik. It is also an imperative that wielders (and apologists) of unitary state power can ignore at their own peril. The alternative is bleak and all too familiar—repression, civil wars, and eventually de facto or de jure partition. As Lapidoth (1992, pp. 345-346) has put it,
a compromise must be found to satisfy ... the aspirations of various groups.. . . The term sovereignty can be used in a flexible manner.. . . Both the central government and regional/autonomous authorities could be the lawful bearer of a share of sovereignty, without necessarily leading to the disappearance or dismemberment of the state. . . Sovereignty . . . has to be adapted to the need for diffusion of power within the state.
And Buchanan (1992, pp. 351-352,362) has pointed us in the right direction:
Once the possibility of a variety of types of political association with differing forms and degrees of self-determination is appreciated, dissatisfied groups within existing states will not be faced with the stark choice of either remaining in a condition of total dependence within the centralized state or of taking the radical step of seceding to form their own sovereign state. Exercising the right of self-determination need not always involve secession if other degrees and forms of self-determination are available.
In other words, it is imperative to rethink our understanding of state and nation and creatively reconceptualize the notion of sovereignty to accommodate both.
What might this mean in practical terms? In line with the given argument, I would say that in the longer term, the Yugoslav successor states as well as Sri Lanka should ideally move toward some form of sovereignty association or confederation, albeit from opposite directions. In Yugoslavia, the disinte-
38. Partitionist solutions also tend to generate seriously negative side effects, as contemporary Yugoslavia clearly indicates. First, the boundaries of putative states are usually contested, requiring lengthy negotiations, arbitration, and adjudication. Second, the demographics of most of the world are such that almost any newly sovereign political units are likely to contain "minority problems" of their own. The combination of these two factors has caused havoc in Bosnia-Herczegovina and Croatia. This is not a minor consideration in Sri Lanka either; almost a fifth of the population of the territory of what would be Tamil Eelam are Muslims (the third major group in Sri Lanka, distinct from both Tamils and Sinhalese) and most Muslims do not view with enthusiasm the idea of living in an independent Tamil state.
Bose / STATE CRISES AND NATIONALITIES CONFLICT 113
gration of the state is an accomplished fact. However, given both historical past and contemporary reality, there is a strong case to be made for reviving some form of loose association that will institutionalize cooperation between the successor states in spheres such as the economy and foreign policy.39 In Sri Lanka, where the centralized state, although beleaguered, is still holding out, international pressure should be brought to bear on the Sinhalese-Buddhist elite to come to an acceptable negotiated settlement with the Tamils. To be viable, such a solution would probably now have to mean a completely renegotiated confederation of Sinhalese and Tamil regions, although special attention will have to be given to the rights of minorities in both regions, and it is most desirable to retain, as in the Balkans, mechanisms of institutionalized cooperation between the two essentially sovereign peoples on such vital questions of mutual concern as the economy, communications, defense, and foreign policy.
If this were to be achieved, we might yet witness a voluntary pooling of national sovereignties in the future in what are currently some of the world's most troubled and fractured societies. As Magas (1992, p. 112) comments about the former Yugoslavia,
There has always been a need for . . . economic and political cooperation throughout the area. ... If and when this need again finds expression, it will not take the form of a common state, for Yugoslavia's time has run out. Nevertheless . . . once the Great-Serb project has been defeated .. . this need will seek and find its own channels and forms.
Whereas the implicit optimism concerning the future prospects of the "Great Serb project" may be overstated, the basic point is very well taken. Nor is it specific to the Balkans alone.
REFERENCES
Balasingham, A. (1983). Liberation Tigers and Tamil Eelam freedom struggle. Madras: Political Committee of LTTE.
39. As Djordjevic (1992) writes: "The unity of Yugoslavs is a manage de raison. In case of a divorce all partners will lose. Too many economic, social and other ties have been established during some seventy years of living together." Bicanic (1988) argues that the economies of the former Yugoslav republics, far from being autarkic, are in fact organically interconnected. Indeed, given that a new Yugoslavia was successfully resurrected from the cauldron of multiple occupation, dismemberment, and civil war after World War II, it may be a mistake to preclude for all time to come the possibility of some kind of renewed Yugoslav union. Of course, this is a longer term prescription. The immediate priority is to restore peace; for that to happen, Serbs would have to be made to recognize and respect the borders of Bosnia-Herczegovina and Croatia.
114    COMPARATIVE POLITICAL STUDIES / April 1995
Banac, I. (1992a, Spring). The fearful asymmetry of war: The causes and consequences of
Yugoslavia's demise. Daedalus, pp. 141-175. Banac, I. (1992b). Post-communism as post-Yugoslavism: The Yugoslav non-revolutions of
1989-90. In I. Banac (Ed.), Eastern Europe in revolution (pp. 168-187). Ithaca, NY: Cornell
University Press.
Bicanic, I. (1988). Fractured economy. In D. Russinow (Ed.), Yugoslavia: Afracturedfederalism.
Washington, DC: Wilson Center Brass, P. (1976). Ethnicity and nationality-formation. Ethnicity, 3(3), 225-240. Buchanan, A. (1992). Self-determination and the right to secede. Journal ofInternational Affairs,
45(2), 347-365.
Dahl, R. (1989). Democracy and its critics. New Haven, CT: Yale University Press. Denitch, B. (1976). The legitimation of a revolution: The Yugoslav case. New Haven, CT: Yale University Press.
Denitch, B. (1990). Limits and possibilities: The crisis of Yugoslav socialism and state-socialist
systems. Minneapolis: University of Minnesota Press. De Silva, C. R. (1984). Sinhalese-Tamil relations and education in Sri Lanka: The first phase,
1971-77. In R. Goldmann & A. J. Wilson (Eds.), From independence to statehood: Managing
ethnic conflict in five African and Asian states (pp. 125-146). New York: St. Martin's. De Silva, K. M. (1990). The traditional homelands of the Tamils. In R. Premdas, S.W.R.
Samarasinghe, & A. Anderson (Eds.), Secessionist movements in comparative perspective
(pp. 32-47). London: Frances Pinter. Djilas, A. (1991). The contested country: Yugoslav unity and communist revolution, 1919-53.
Cambridge, MA: Harvard University Press. Djilas, M. (1980). Tito: The story from inside. New York: Harcourt Brace Jovanovich. Djordjevic, D. (1992). The Yugoslav phenomenon. In J. Held (Ed.), The Columbia history of
Eastern Europe in the twentieth century (pp. 306-344). New York: Columbia University
Press.
Fearon, J. (1991). Counterfactuals and hypothesis-testing in political science. World Politics, 43(2), 169-195.
George, A. (1979). Case-studies and theory development: The method of structured, focused
comparison. In P. Lauren (Ed.), Diplomacy: New approaches in history, theory and policy
(pp. 43-68). New York: Free Press. Horowitz, D. (1985). Ethnic groups in conflict. Berkeley: University of California Press. Horowitz, D. (1989). Incentives and political behavior in the ethnic politics of Sri Lanka and
Malaysia. Third World Quarterly, 11(4), 18-35. Jackson, R., & Rosberg, C. (1982). Why Africa's weak states persist: The empirical and juridical
in statehood. World Politics, 35(1), 1-24. Kohn, H. (1944). The idea of nationalism: A study in its origins and background. London:
Macmillan.
Kornai, J. (1992). The socialist system: The political economy of communism. Princeton, NJ:
Princeton University Press. Lapidoth, R. (1992). Sovereignty in transition. Journal of International Affairs, 45(2), 325-346. Liberation Tigers of Tamil Eelam. (1990). Socialist Tamil Eelam: Political programme of the
LITE. London: Author.
Linz, J., & Stepan, A. (1992). Political identities and electoral sequences: Spain, the Soviet Union
and Yugoslavia. Daedalus, 121(2), 123-139. Magas, B. (1992). The destruction of Bosnia-Herczegovina. New Left Review, 196,102-112.
Bose / STATE CRISES AND NATIONALITIES CONFLICT    115
Nithiyanandan, V. (1987). An analysis of economic factors in the origin and development of
Tamil nationalism in Sri Lanka. In C. Abeysekera & N. Gunasinghe (Eds.), Facets of
ethnicity in Sri Lanka. Colombo: Social Scientists' Association. Obeyesekere, G. (1984). The origins and institutionalization of political violence. In J. Manor
(Ed.), Sri Lanka in change and crisis (pp. 153-174). London: Croom Helm. Pinard, M. (1992). The dramatic re-emergence of the Quebec independence movement. Journal
of International Affairs, 45(2), 471-497. Rabushka, A., & Shepsle, K. (1972). Politics in plural societies: A theory of democratic
instability. Columbus, OH: Charles E. Merrill. Rakowska-Harmstone, T. (1974). The dialectics of nationalism in the USSR. Problems of
Communism, 23(3), 1-22. Ramet, S. (1992). Nationalism and federalism in Yugoslavia, 1962-91. Bloomington: Indiana
University Press.
Remington, R. (1984). Yugoslavia. In T. Rakowska-Harmstone (Ed.), Communism in Eastern
Europe (pp. 238-282). Bloomington: Indiana University Press. Roeder, P. (1991). Soviet federalism and ethnic mobilization. World Politics, 43(2), 196-232. Rothschild, J. (1981). Ethnopolitics: A conceptual framework. New York: Columbia University
Press.
Singleton, F. (1985). A short history of the Yugoslav peoples. London: Cambridge University Press.
Sivanandan, A. (1990). Race, class and the state in post-colonial societies: A case-study of Sri
Lanka. In A. Sivanandan (Ed.), Communities of resistance: Writings on Black struggles for
socialism (pp. 199-250). London: Verso. Skocpol, T., & Somers, M. (1980). The uses of comparative history in macrosocial inquiry.
Comparative Studies in Society & History, 22(2), 174-197. Stepan, A. (1978). The state and society: Peru in comparative perspective. Princeton, NJ:
Princeton University Press. Stojanovic, S. (1991). Reflections on the crisis of the Yugoslav social and state-system.
Mediterranean Quarterly, 2(2), 85-100. Tambiah, S. (1986). Sri Lanka: Ethnic fratricide and the dismantling of democracy. Chicago:
University of Chicago Press. Thilagar, L. (1992, January). A war fratricidal. Tamil Nation, pp. 1-2. Tito, J. B. (1972). Selected military works. Belgrade: Socialist Thought and Practice. Tomasevich, J. (1969). Yugoslavia during the Second World War. In W. Vucinich (Ed.),
Contemporary Yugoslavia: Twenty years of socialist experiment (pp. 59-118). Berkeley:
University of California Press. Tomasevich, J. (1975). The Chetniks: War and revolution in Yugoslavia, 1941-45. Stanford, CA:
Stanford University Press. Tome, G. (1988). Classes, party elites and ethnic groups. In D. Russinow (Ed.), Yugoslavia: A
fractured federalism. Washington, DC: Wilson Center. Vucinich, W. (1969). Nationalism and communism. In W. Vucinich (Ed.), Contemporary
Yugoslavia: Twenty years of socialist experiment (pp. 236-284). Berkeley: University of
California Press.
West, J. (1991, February 13). Civilians bear brunt of war on Tamils. The Independent (London), p. 13.
Wilson, A. J. (1988). The break-up of Sri Lanka: The Sinhalese-Tamil conflict. London: Christopher Hurst.
116    COMPARATIVE POLITICAL STUDIES / April 1995
Young, C. (1976). The politics of cultural pluralism. Madison: University ofWisconsin Press.
Young, C. (1992). National and colonial questions and Marxism. In A. Motyl (Ed.), Thinking theoretically about Soviet nationalities: History and comparison in the study of the USSR (pp. 67-97). New York: Columbia University Press.
Zaslavsky, V. (1992). Nationalism and democratic transition in postcommunist societies. Daedalus, 121(2), 97-121.
Zukin, S. (1984). Yugoslavia: Development and persistence of the state. In N. Harding (Ed.), The state in socialist society (pp. 249-276). London: Macmillan.
Sumantra Bose is Fellow of the Faculty in the Department of Political Science at Columbia University, where he is a Ph.D. candidate. He is author of States, Nations, Sovereignty: Sri Lanka, India and the Tamil Eelam Movement (New Delhi, Thousand Oaks, and London: Sage, 1994). His doctoral dissertion is titled "Democracy and National Self-Determination: A Comparative Inquiry."