The Sociology of Ethnic Conflicts: Comparative International PerspectivesAuthor(s): Robin M. Williams, Jr.Source: Annual Review of Sociology, Vol. 20 (1994), pp. 49-79Published by: Annual ReviewsStable URL: http://www.jstor.org/stable/2083359Accessed: 29/05/2009 15:37

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=annrevs.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review ofSociology.

http://www.jstor.org

Annu. Rev. Sociol 1994. 20:49-79 Copyright C 1994 by Annual Reviews Inc. All rights reserved

THE SOCIOLOGY OF ETHNIC CONFLICTS: Comparative International Perspectives

Robin M. Williams, Jr. Department of Sociology, Cornell University, Ithaca, New York 14853

KEYWORDS: competition, inequality, class, states, nationalism (see also: rebellion, collective action)

Abstract Oppositions and deadly conflicts among ethnic collectivities are important around the world. Ethnies (our term for ethnic groups) also strongly affect interstate relations. Both interethnic and ethnic-state conflicts tend to be severe, protracted, and intractable. At the extremes, the stakes are total: survival versus genocide.

Competition and rivalry for individualized economic and political goods are important, but the most intense conflicts are to be expected when the stakes are collective goods, including categorical claims to prestige and political authority.

States are major actors in creating, accentuating, or diminishing ethnic identities. States are both arenas of rivalry and conflict and resources for ethnic mobilization and counter-mobilization.

Because both ethnies and states are diverse, careful specification is required for fruitful analysis. The same dictum applies for the diverse types of opposi- tions and of conflicts.

Ethnic conflicts arise from complex combinations of ethnic strength, class, inequality, political opportunity, mobilization resources, interdependence, and international interventions. Frequent but nonviolent protests, for example, are most likely by organized collectivities with substantial resources, operating in relatively open political systems. International aid to parties in domestic con- flicts appears to prolong and intensify ethnic struggles.

Research in this field contends with many difficulties, and one-sided theories do not fare well. Yet abundant descriptive materials are available, statistical

49

0360-0572/9410815-0049$05.00

50 WILLIAMS

techniques are improving, conceptual clarification continues, and substantive knowledge does accumulate. Accordingly, there is hope for better understand- ing of some of the most destructive and tragic conflicts of our times.

INTRODUCTION

The current significance of this topic is not in doubt. First, the world-wide prevalence of ethnic diversity is indicated by the presence of about 6000 languages (Grimes 1988) and somewhere between 900 (Murdock 1967) and 1600 (Levinson 1991-1993) major cultural groupings. For example, there were about 130 officially recognized "nationalities" in the former Soviet Union. Resurgent ethnic movements have appeared in many Western countries. Sec- ond, of the 183 states in the United Nations (as of June 1993), only a handful are ethnically homogeneous; multiethnicity is the rule. Third, in a warring world (Brogan 1990), ethnic-related conflict is frequent and often deadly. One half of the world's states have experienced significant ethnic conflict since World War II. About 80% of deaths in warfare during that period have been internal to national states (Russett & Starr 1989:171), and much of that total has come from collective ethnic violence. Estimates of the number of deaths attributable to ethnic violence since 1945 vary widely; a minimum estimate of 11 million has been given by Topor (1992), but an upper figure of 20 million fatalities does not appear to be an excessive guess. In the ultimate case of genocide, as Leo Kuper (1985:161) has noted, the majority of deaths in do- mestic genocides result from struggles for power by ethnic, racial, or religious groups. (Of the vast literature on genocide we mention here only Hovannisian 1986, Conquest 1990, and IL Horowitz 1992.)

Over the past quarter-century scholars have paid rapidly increasing (even if belated) attention to these remarkable phenomena. Ethnopolitics, including nationalistic developments, is a crucial global force (cf Ragin 1983:1317, Moynihan 1993). Prior reviews of research on ethnicity have attributed this growing interest to the near-universality of multiethnic states, the persistence of strong ethnies in complex modern societies, ideological and policy disagree- ments, and the prevalence of severe ethnic conflicts (Olzak 1983, Yinger 1985). Meanwhile, at long last, military strategists are paying close attention to terrorism (Gibbs 1989) and so-called low-intensity conflicts, which often are ethnically based (cf Schultz 1991, Hoffman 1992).

Because it has become quite impossible in this field to "review all the literature," this chapter has been forced to omit the enormous assemblage of data and interpretations dealing with religious fundamentalism, e.g. the six volumes emerging from The Fundamentalism Project of the American Acad- emy of Arts and Sciences (cf Marty & Appleby 1993a,b). Similarly we give only passing mention to many important historical studies of particular soci-

ETHNIC CONFLICTS 51

eties (e.g. Suny 1985). Prior reviews make it possible to omit a survey of recent literature on ethnicity in the United States: Williams (1975, 1977, 1978), Porter & Washington (1979) Lambert (1981), Willie & Cunnigen (1981), Olzak (1983), Yinger (1985), Longshore & Prager (1985), Portes & Truelove (1987), Jaynes & Williams (1989). Finally, to observe space limits we must omit the large literature on resolution and management of ethnic conflicts.

Our survey is illustrative rather than comprehensive. The central problem is how and under what circumstances interactions among ethnic collectivities produce strong oppositions and overt conflicts (Rule 1988:9-10). Much of the relevant research has been done by nonsociologists; we happily include that work wherever the content is suitably sociological.

Before entering upon the main task, however, we must briefly set the review in a world-context and then examine the central concepts that are essential for the later discussion.

THE WORLD CONTEXT

In the last decade of this century, the world is increasingly interdependent, economically, politically, culturally. Transnational and international organiza- tions have become more and more important. Meanwhile national states mul- tiply, armaments increase to unprecedented levels, and political instability is endemic. At the same time, economic inequalities among countries are great and are increasing. Ethnic, religious, and class oppositions place severe stresses on weak states. The heritage of the great wave of decolonization through the 1960s (Boswell 1989, Strang 1990) was a new population of multiethnic states, racked by domestic conflict.

A heavily armed and politically unstable universe of sovereign states regu- larly produces some 30 or more wars per year. As of 1992 about 40 violent collective conflicts were being fought in 29 countries (Hoffman 1992:26).

Ethnic conflicts since World War II have been facilitated by the rise in the number of new multiethnic states, by state-building activities, increased re- sources for mobilization, mass communication, diffusion of ideologies-and by external interventions. These developments have helped to instigate ethnic conflicts over control of central polities, conflicts over cross-national claims to territory, and struggles over autonomy or secession (Nagel 1980:280-82). The primary conditions thus are ethnic distinctiveness, geographic concentra- tion, interethnic inequality, "alien" state penetration (and fear of exclusion), ethnic organization, and external support.

Ethnic divisions and ethnic-state relations are created, accentuated, reduced, or destroyed by the increasing internationalization of political economies, both in the division of labor and in the internationalization of state security (Enloe 1980, 1986).

52 WILLIAMS

Rapid population growth characterizes the less-industrialized world regions where there is great poverty and economic inequality within fragile states. World-wide increases in highly destructive weaponry accompany extensive militarization. And the (partial) end of the Cold War has created new interna- tional uncertainties and has released many formerly suppressed domestic con- flicts.

The enormous flows of international migrants continually reshape ethnic relations. The United Nations estimates that in 1993 as many as 100 million individuals live outside their country of birth or citizenship; as many as 18 million are refugees. Numerous guest-workers and other migrating workers elicit tensions and some violence in many places, e.g. Germany, France, the United States. (We note that Frankfurt is the third largest Turkish city, behind Istanbul and Ankara.)

The structure of intrasocietal cleavages and the history of relationships among social formations-ethnies, classes, regions-constitute a global con- text of constraints upon and opportunities for collective action. Within such contexts, purposive actions by ethnic entrepreneurs and political leaders influ- ence subsequent events, leading toward or away from ethnic conflict. Accord- ingly, a major dimension of ethnic conflict or cooperation, world-wide, is the relation of ethnies to territorial states that claim unitary sovereignty over a territory and its population.

CONCEPTS AND THEORIES

Problems of definitions just will not go away, so controversies continue re- garding how best to characterize such terms as ethny, ethnic group, competi- tion, rivalry, conflict, solidarity, mobilization, nation, nationalism, state. Many of these terms have been defined and discussed in prior volumes of the Annual Review of Sociology, e.g. Williams 1975, Hirschman 1983, Olzak 1983, Yinger 1985. Accordingly, this chapter treats only the most salient recent discussions.

Structural Units

The awkward and potentially misleading term "ethnic group" is so deeply entrenched in the professional literature that it may be folly to resist it. Yet there is the better word, "ethny," proposed by van den Berghe (1981:22, cf Dutter 1990) and used routinely by AD Smith (1981); we shall use it here whenever convenient. The term captures the core idea: "The prototypical ethny is ... a descent group bounded socially by inbreeding and spatially by territory" (van den Berghe 1981:24). From this core, more extended ethnies develop by preferential endogamy, extended nepotism, fictive kinship, descent myths, and extension into large territories. At one extreme is the small, cohesive, closed

ETHNIC CONFLICTS 53

local ethny; at the other, the vastly extended "symbolic" boundaries of modem ethnic identities (Yinger 1985:161). The moral: there are many different kinds of ethnies; analysis needs to make specific distinctions. Numerous concrete types of ethnic collectivities obviously exist: indigenous native populations, refugees, guest workers, ethno-territorial groupings, diaspora, ethnoclasses, and so on (Thompson & Rudolph 1986:29-36). "Race" is a special case of ethnicity.

Our editorial comment can be brief: let us dispense with arguments as to what definition is "correct," recognize the inherent diversity, and move on. It is essential only to recognize that ethnicity is qualitatively different from such markers as ideology or occupation (Dutter 1990:312).

Scholarly controversies are endemic concerning the relationships between class and ethny. The definitional problems do seem open to clarification. Social stratification refers to inequalities in the distribution of any scarce value. A social class is an aggregate of social units similarity positioned with relation to any or all of such scarce values, e.g. control of the means of production (van den Berghe 1983:222). To the extent that such aggregates are marked by special densities of interaction and by exclusion of nonmembers, they become collectivities. Collectivities defined by birth and cultural or physical distinc- tiveness are ethnies-which may or may not be ranked, and may or may not be internally stratified. A ranked collectivity enforcing endogamy becomes over time an ethnoclass or caste. Large ethnies typically are both ranked and internally stratified.

How is ethnicity related to "nation" and "nationality"? A useful character- ization: "A nation (but see Gellner 1983) is a politically conscious ethny, claiming statehood rights on the basis of common ethnicity" (van den Berghe 1983:222, cf Smith 1989:342). Nationalism is an ideological movement in support of a nation. Note, however, that the term "nationalism" is used to refer to two quite different conceptions: (i) identification with and loyalty to the state structure, regardless of ethnic composition of the population (American, Argentinian); (ii) identification with and loyalty to an ethnic/religious "nation- ality" that may or may not coincide with a state's jurisdiction (Scottish, Breton, Tamil) (Connor 1981:201). The importance of languages for both ethnicity and nationalism is now widely recognized (cf Conversi 1990, Balibar 1990: 350, Eriksen 1990).

The bases of politically significant collective solidarities thus include kin- ship group (family , clan), region, class, language, religion (or sect), tribe, ethny, and state (country). Within national, centralized states, cleavages based on substate affinities constitute an explosive component, world-wide (cf Esman & Rabinovich 1988). Since few territorial (national) states are populated by a single ethny, the term "nation-state" is muddled and intellectually dangerous (Connor 1981). In a long line of devastating critiques, Ra'anan's 1990 article

54 WILLIAMS

shows why this semantic monstrosity should no longer be casually used (in Montville 1990:5-20).

Since Eugen Weber's crucial study of French nationalism (1979), it has become increasingly clear that nationalistic unity of modem states is a very late and partial historical development (Connor 1990). For example, "France" was not a single linguistic or cultural unit among the masses until after World War I, at earliest (Weber 1979, cf Watkins 1991). As Connor (1990:92) says, nationalism as a mass phenomenon cannot be directly inferred from elite claims.

Processes

These, then, are our central structural units: ethnies, nations (nationalities), classes, and states. What are the processes within and among these units to which our review must be addressed?

First, there is the notion of competition, usefully defined as "striving for scarce objects under rules which limit the damage competitors can inflict on each other" (Thompson & Quets 1990:249, cf Williams 1947:43). (Rivalry is competition among specific, identified actors.) It is crucially important to distinguish between peaceful individual competition and regularized collective opposition-i.e. political, economic, and cultural contention-and overt con- flict. It is correct that almost always ethnic conflict has been based in part upon fear of competition (or rivalry). But an essential difference exists between rule-constrained individualized striving for scarce values and collective actions aimed at displacing, neutralizing, injuring, or destroying opponents.

Conflict of course, has been given numerous definitions-psychological, cultural, and social. Social conflict was defined some decades ago as "a struggle over values (distributive or nondistributive) in which the immediate aims of the opponents are to neutralize, injure, or eliminate their rivals" (Williams 1947: 43).

This characterization points to a general area of interest. For particular research purposes, however, it is necessary to identify a complex range of more specific phenomena. The main forms of collective ethnic conflict are (i) turmoil (strikes, demonstrations, mutinies (Rose 1982), protests, sabotage, communal rioting, and terrorism); (ii) internal war (coups d'etat, secessionistic rebellions, civil wars and revolutions); (iii) genocide. Each of these involves further distinctions, e.g. there are many complexities in defining terrorism (Gibbs 1989), rebellion (Boswell & Dixon 1990:540), or even coups d'etat (Collihan & Danopoulos 1993:435, DL Horowitz 1985: Ch. 12, cf Hintzen 1989).

Such distinctions among types of conflict involve (i) the extent of mass mobilization, (ii) the extent of centralized organization and control, and (iii)

ETHNIC CONFLICTS 55

the amount and types of violence. 'Turmoil" generally is low on all three criteria; revolutions maximize all (see Zimmerman 1980:168-71).

Recent Developments in Research and Theories The Preface to Volume 18 (1992) of the Annual Review of Sociology has said, "The essays in this volume reflect the increasingly comparative nature of sociology, whether international or historical" (p. v). Our own survey indicates that in the field of ethnic studies, comparative sociology still has room to grow. Until very recently the sociological literature in the United States was very heavily weighted toward studies of domestic populations, especially blacks (Lavender & Forsyth 1976). A similar survey of political science journals found (i) very little attention to any ethnic topics, and (ii) major emphasis on blacks (70% of articles dealing with ethnic groups) (Rlood 1980).

Correspondingly, "theories" of ethnic conflict have been primarily focused upon intrasocietal processes: split labor markets, (Bonacich 1976), economic competition (Banton 1983, Olzak 1983, 1992, Olzak & Nagel 1986), internal colonialism (Hechter 1975), assimilation (Gordon 1964, Hirschman 1983, 1991). Yet such domestic processes manifestly are strongly affected by eco- nomic and political rivalries among national states and by many other trans- national and international influences. Recognition of such influences since the 1970s has led to greater attention to the part played by state power and resources and international conflict and cooperation (Gurr 1980, Greenwood 1985, Brass 1985, DL Horowitz 1985, Gurr 1993).

A rich set of descriptive studies of varied ethnic situations around the world has provided a sort of inventory of conditions affecting interethnic relations (cf Horowitz 1985, Montville 1990, Portes & Rumbaut 1990.). This inventory includes: type of state and regime, political resources of ethnies, number and size of ethnies, relations of ethny and class, regional inequalities, relative deprivation, demographic "balance" and changes therein, cultural content of ethnic identities, extent and kind of "modernization," political party structures, sequences of interactions among ethnies and between them and the state. For example, dependency, world system and global hegemony theories of domestic conflict have paid special attention to weak authoritarian or elitist states in the world-system periphery; the primary mechanisms posited are class polarization and capitalistic intrusions upon "traditional" societies (Jenkins & Schock 1992: 176-79).

A tendency among theorists in this field is to focus upon a single factor or set of conditions and to ignore, downplay, or reject others. So, it is said, the primary factor in ethnic conflict is resource mobilization, or competition, or state actions, or internal colonialism, or economic inequalities, or labor market segmentation, or ethnic inequalities, or social strains and anomie-and so on. Fashions abound-at one time socially induced motivations such as relative

56 WILLIAMS

deprivation are in favor-only to be rejected out of hand at a later time. A major exception to the unidimensional theorizing is the work of Pinard and associates (see Pinard & Hamilton 1984, 1986), who propose a multifactorial scheme of six components: deprivations, aspirations, moral obligations, col- lective incentives, selective incentives, and expectations of success (Pinard & Hamilton 1986).

For several decades the academic industry created by allegedly antithetical models has flourished: sociobiological perspectives (van den Berghe 1981, 1990, Shaw & Wong 1989, van der Dennen & Falger 1990), primordial solidarities (Geertz 1973, Stack 1986), competition (Olzak 1983); group mo- bilization (Klandermans 1984, Tarrow 1988); internal colonialism (Hechter 1975, 1987); cultural content (Prager 1982, B Anderson 1983, Smith 1988), rational actors (Muller & Weede 1990).

Efforts to reformulate and integrate the diverse approaches have been few but not wholly lacking, e.g. Williams (1978), McKay (1982), Scott (1990), and Gurr (1993).

Research methods used in studies reviewed here cover a very wide range, indeed. Case studies predominate in the descriptive work, although increas- ingly there are explicit comparisons (cf again, DL Horowitz 1985, Wickham- Crowley 1992). Ragin (1991) has given a thorough introduction to the prob- lems of emphasis on variables. On the whole, comparative social science in the past has been preoccupied with case studies, but recent cross-societal studies have shifted the focus to multivariate statistical analysis of many cases. New methodologies-e.g. event history analysis and qualitative comparative analysis-hold some promise of combining analytic generalizations with con- textual specification.

Recent years have brought a marked convergence between studies of ethnic- ity and studies of nationalism (Smith 1992). Increasingly also we are seeing cross-state statistical analyses of large data sets. By 1980, Gurr could say that modern research on conflict is characterized by the use of systematic methods, comparing phenomena across settings or across time, and by widespread reli- ance on quantification and statistical techniques (1980:4).

Although the great majority of ethnic studies rely on cross-sectional data, some now analyze sequences of conditions or events (Hannan 1979, Olzak 1983, Strang 1990) either in narrative accounts or by event-history techniques. Case studies actually reinforce the advocacy of sequential approaches. Thus, in a review of nine major works on Ethiopia, Colburn (1991) has urged that an extended sequence of "misgovernment" must be given an important place in analysis of revolution and violent ethnic separatism. Not just prior structural conditions but active choice, he contends, must be part of an adequate under- standing (Colburn 1991:585-86).

The comments by Jenkins & Kposowa (1990:862) concerning hypotheses

ETHNIC CONFLICTS 57

and theories that attempt to explain military coups apply to ethnic conflicts more generally: "They have been treated as single propositions evaluated by case analysis or conventional regression techniques. We believe these are complex theories with multiple causal links that should be evaluated with complex causal models."

THE CHARACTER OF ETHNICITY: CONTROVERSIES AND CONVERGENCES

The repetitive arguments between advocates of primordial and instrumental or situational conceptions of ethnicity can and should be superseded. Ethnies are both primordial and circumstantial-in different ways under different condi- tions. Some ethnies persist for many centuries (Smith 1992); some are newly emergent; some have fluid boundaries, others are rigid entities. It is important only to adequately specify the character of whatever units we make the object of analysis. Modem nationalisms are analogous to older forms of ethnic elec- tion (cf Armstrong 1982, van den Berghe 1981). From time to time, studies of ethnicity rediscover the distinction between category and collectivity-eth- nic populations or categories versus ethnic communities or groupings (ethnies). Categories turn into ethnies through the development of membership and identity arising in social interactions. Ethnic networks develop and ethnic elites and leaders can then seek to advance or defend collective interests through collective action (Martiniello 1993:238-43).

Both van den Berghe (1981) and DL Horowitz (1985:52-70) indicate that ethnies involve a belief in common descent (affinity through ascription), some notion of distinctiveness, and a membership transcending face-to-face interac- tions. It is central to ethnicity that there is always an important emphasis upon descent; it is this ascriptive quality, above all, that renders ethnic cleavages so often intractably conflictful.

In the long term, there are both major changes in ethnic classifications and considerable mobility across ethnic boundaries. Thus, ethnicity may be seen historically as arbitrary, variable, and ambiguous. Bqt in the short term, ethnic boundaries often are strongly correlated with social, economic, and political status, and such inequalities are often reinforced by categorical discrimination and public policies. Hence, ethnic divisions can become the focus of strong passions and of collective conflicts (Hirschman 1986). The "construction" (or "emergence") of new ethnies out of previously diverse cultural groupings or the subdividing of previously unitary collectivities is described in numerous studies: of Trinidad (Brereton 1979), "Hispanics" an&"white ethnics" in Uni- ted States (Alba 1990, Waters 1990, Jiobu 1990), US Indian tribes (Barsh & Henderson, 1980), Cote d'Ivoire (Vogel 1991), Punjabis in California (Leonard 1992), Sierra Leone (Cohen 1981), India (Schermerhorn 1978)-to name only

58 WILLIAMS

a few. Works that emphasize ethnic change or fluidity include Eller & Cough- lan (1993), Nagel & Snipp (1993), Heisler (1990), Fugita & O'Brien (1991), B Anderson (1991) [1983]. More complex or integrative views are represented by van den Berghe (1981), DL Horowitz (1985), Greenwood (1985), Scott (1990). It is clear that present-day ethnic populations have been constructed over time by conquest, religious movements, migration, biological blending, acculturation, and absorption of ethnic-linguistic groups. Thus, ethnic bound- aries shift historically, and many individuals have multiple ethnic identities. But, as Smith has stressed, ethnicity can be and is, as Smith notes, historically perennial without being a fixed essence-rather ethnic continuity resides in a varying but basic boundary perception (cf Smith 1989, 1992). Other studies that point to the persistence of ethnies and to their resistance against forces of change include Spicer (1971), Smith (1981, 1989, 1992), Douglas (1988), Kimmel (1989), and Petrissans (1991).

Assertions that many modem ethnies are artificially created, fabricated, instrumentally formed, or imagined are mostly irrelevant for the conclusions to be drawn from ethnic revivals and upsurges. The intense passions manifest in modern cases of ethnogenesis make pointless the constructionist skepticism: the solidarities clearly do exist and are accompanied by strong sentiments and collective actions (cf Fishman 1981:239). It is striking testimony that a work entitled Creating Ethnicity reaches this conclusion:

Ethnic groups and their cultures are not merely a completely arbitrary construct: there is always a minimum of incontestable and noninterpretable facts necessary to win something from the opponent. ...The reality is very elastic but not totally arbitrary. (Roosens 1989:156)

Of course, there are social identities and formations below ethnies, e.g., families, lineages, villages, clans, and there are identities above the ethnies, e.g. religious organizations, regions, states, interstate organizations, transna- tional organizations and social movements. Ethnicity is not everything, but its place in social conflict can now be well defined.

THE STAKES IN ETHNIC OPPOSITIONS

Grievances and Objects of Contention The notion of grievance is often conflated with related concepts of dissatis- faction or deprivation. But to say "grievance" is to say that someone is aggrieved, not merely deprived, disappointed, frustrated, or dissatisfied. A grievance is "a wrong considered as grounds for complaint" (Random House Dictionary of the English Language)-that is, a claim is made that an injustice has occurred. Just as a claim is a socially legitimated demand, not just a desire or want, so a grievance is a normative protest, not just a dissatisfaction. The

ETHNIC CONFLICTS 59

conceptual distinction has deep social roots. The special intensity so evident in many ethnic conflicts frequently arises from a sense of victimization (Zas- lavsky 1992) arising from loss of autonomy, loss of historically claimed ter- ritory, infringement of prior rights, or generally, treatment thought to be unfair- ly discriminatory (Harris 1977). These considerations suggest a specific hy- pothesis: The likelihood of conflict is higher when disagreement and opposition concern collective goods, e.g. language rights, religious beliefs and symbols, civil and political rights and privileges, regional-ethnic power, or regional-eth- nic parity in the economy. The more nearly indivisible the goods and the less the access of the "disadvantaged," the greater is the resentment and the more likely is ethnic mobilization, followed by overt conflict (see Fraser et al 1990).

Conflict is favored when there is interethnic rivalry and at the same time the ethnies have low positive interdependence (complementarity), are unequal in socioeconomic status, and when the subordinated ethny perceives the su- perordinate collectivity as violating norms of fairness (Belanger & Pinard 1991). Anything that greatly increases ethnic salience is likely to produce increased attention to collective goods and-given ethnic inequality-to in- creased apprehension of discrimination. What then generates perceptions of unfairness is competition/rivalry when an ethny is subordinated or disadvan- taged in economic opportunity, social status, political voice and rights, or cultural expression. These conditions stimulate grievances that acquire a tinc- ture of moral outrage (cf Belanger & Pinard 1991:449).

Control of Territory Some of the intractable ethnic conflicts of recent years have involved disputed claims to land: "territory" has not lost its significance as an object of contention (cf Kamen 1991). Many ethnoregional conflicts are directed against the state in efforts to gain or restore control over a homeland: as in Quebec, Scotland, southern Sudan, Eritrea, Tamil Eelam (Sri Lanka), Kosovo, Biafra, Sikh Khali- stan (India) (Esman 1985:438). Other conflicts are directed toward rival ethnies (often in-migrants) that take land or are thought to threaten to reduce an indigenous ethny to a demographic minority in danger of losing control of the state and the economy, e.g. Malays, Fijians, Assamese. In such situations a central fear is that of being overwhelmed and dispossessed (Esman 1985, Horowitz 1985:166-81). A frequent pattern is for violent conflicts to erupt in societies in which several ethnies claim priority of indigenous status, as in Lebanon and Yugoslavia.

Other Stakes The stakes in ethnic rivalries include participation in political decisions, cul- tural status, economic opportunities (Esman 1990:58-59). The great difficulties of resolving or managing many of the world's ethnic confrontations reside in

60 WILLIAMS

the ascriptive character of ethnicity and its high symbolic content (e.g. the indivisible good of collective prestige) (DL Horowitz 1990:115-16). The ev- idence is convincing that social respect and prestige are real and important foci of ethnic rivalries and of violent conflicts.

The overwhelming impression created by the research literature and the events to which it refers is the extreme fragility of economic interdependence in confrontations involving polarized ethnies, as in Northern Ireland (See 1986), Cyprus, Yugoslavia, the Soviet Union, Rwanda (Newbury 1988), Sri Lanka (Tambiah 1986).

ETHNICITY AND NATIONAL STATES

The Context Ethnic relations affect and are influenced by relations among states. This basic and important truth has been somewhat obscured by the concentration of much scholarship in history and political science upon states and by the parochial fixation of much sociological research upon domestic ethnic relations. A step forward is Esman's (1990) description of six categories of transnational or interstate influences involving ethnies. Ethnic ties do not stop at state bound- aries: nearly two thirds of the 227 communal groups included in Gurr's analysis have kindred in another country (1993:175). There are numerous international networks of communication, travel, political support, and material assistance; these networks are especially prominent among Muslims and indigenous peo- ple. Some 42% of major ethnies are split among two or more states, as witness the familiar cases of Kurds, Basques, Tamils, Hungarians, Bakongo, Balochi (Nielsson 1985). Among some of the upwards of 100 million persons who live outside their countries of citizenship (Teitelbaum 1992:208), there is a dis- placed localism that can lead to unrestrained nationalism of a militant kind (B Anderson 1991 [1983]). And communal ethnic conflict is itself often conflict about the nature of the state and its authority-as in the classic cases of Belgium and Quebec (Covell 1985:233, Breton 1988).

More generally there is evident a world-wide tension between the principle of status based on individual achievement and that of status based on ascribed collectivity-membership. Striking examples are found in India (Beteille 1991) and in Malaysia (Horowitz 1985, Hirschman 1986, Marty & Appleby 1993a,b).

Evidently sociological analysis needs to consider ideological factors. "Na- tional self-determination" or "self-determination of peoples" is a variant of doctrines of popular sovereignty. If sovereignty of a state rests upon the will of the people it claims to encompass, then the legitimacy of states can be challenged by the claims of subnationalities. If at the same time it is claimed that sovereignty is absolute, so that states treat as illegitimate all external

ETHNIC CONFLICTS 61

humanitarian efforts or human rights interventions, then ethnic issues can become crucial not only for regime and government stability but also for state autonomy and international relations. That "foreign" rule is illegitimate is indeed an explosive idea in a world of multiethnic national states (cf Connor 1990, 1992).

In the crucial test-case of genocide, the conspicuous failures of the United Nations to intervene have been analyzed in detail by Kuper (1985). Not only do heads of states fear intervention elsewhere as a precedent that may later be used against them, but reluctance may be heightened also by the circumstance that political mass killing has been a means to power on the part of many governments represented in the United Nations (Kuper 1985:128).

SOURCES OF OPPOSITION AND CONFLICT: HYPOTHESES AND FINDINGS

Introduction Given then, that there are ethnies and other relevant social formations, what are the sources of the many different kinds of oppositions, protests, and con- flict? First of all, the literature tells us that the objects of ethnic contention include such valuables as use of residence areas, control of territory, credit, jobs, markets, capital, annaments-and collective goods such as language use, access to and control of education, civil rights, practice of religion, cultural expression, collective respect and prestige, and political rights. Any or all may become the focus of confrontations between ethnies and states or among ethnies themselves. Second, many studies focus upon particular types of con- flicts-attempting explanations of communal violence, riots, coups d'etat, secessionistic rebellions, or revolutions. Third, many other studies reverse the perspective to single out causal variables (e.g. demographic changes, class inequalities) of particular explanatory models alleged to hold across diverse settings and conflicts: rational choice, economic competition, cultural content, political mobilization, modernization, and so on.

To begin with, we examine some general processes of opposition and con- flict. There are marked differences in the conditions associated with different types of ethnic rivalry and discord. For example, Beissinger (1992) analyzes data on 2347 nonviolent protest demonstrations and 379 violent mass actions in the USSR, 1965-1989, to show sharp differences in prior conditions and accompanying contexts. Violent protests varied independently of political op- portunity structures; nonviolent protests were less frequent when opportunities were either open or closed. Violent protests were characteristic of rural or small town settings (in contrast to US experience); nonviolent protests were more typically urban. The latter protests represented relatively high levels of

62 WILLIAMS

social mobilization among relatively well-educated urban populations with substantial organizational resources (cf Taagepera 1992, on Estonia). Violent outbreaks come from populations with fewer independent resources for col- lective action but with a strong sense of victimization. (For several major ethnies, violent events actually preceded nonviolent demonstrations.) A prior history of nonviolent protest helps to predict nonviolent demonstrations at a later time, but no significant relation was found for violent events. Not a history of ancient animosities, but rather, current conditions best explain these out- breaks of collective violence.

Processes of Violent Conflict It has long been recognized that violent conflicts, once underway, develop their own internal dynamics. Conflict is a vortex into which are drawn new elements and within which there emerge numerous transformations (Williams 1965:25). The self-sustaining and transformative nature of violent ethnic con- flicts has been repeatedly noted, e.g. Azar & Burton (1986), Azar (1990), Panda (1993). These processes are chaos-like-full of probablistic and near- random events, not predicted by the actors. A little-understood phenomenon is the phase-change in which collective violence produces otherwise unthink- able behavior.

Violence is not limited to challenging collectivities that are seeking inclusion in, control of, or exit from a polity. It may be evoked or initiated when a collectivity already nominally legitimized as an insider regards its position as basically threatened. Although initiation of violence by a challenger often carries high risks-increased stakes, possible disunity, harsh repression-nev- ertheless violence may emerge during a process of conflict as a viable "last resort" (Grant & Wallace 1991).

Grievances must be taken into account; they are not constructed out of the air: objectively ascertainable conditions do matter. Gurr's 1993 analysis (e.g. p. 188) finds that grievances are created by poverty and political and economic differentials among communal groups, and that restricted political access and a history of lost autonomy are important conditions for separatist demands and rebellion.

Not only diffuse ethnic oppositions and tensions but also separatist move- ments draw strongly upon fears of being 'swamped' or permanently subordi- nated by a dominant ethny or ethnic coalition (Horowitz 1985, Mayall & Simpson 1992:15, Pfaffenberger 1990). "Backward" groupings frequently fear cultural extinction, but even advantaged ethnies can be aroused to see dire threats of subordination.

The fact that ethnic solidarity may have been roughly constant while overt ethnic conflict has fluctuated does not mean that the solidarity is not important in explaining conflict-only that other conditions, of threat and opportunity,

ETHNIC CONFLICTS 63

are necessary to evoke the more militant/violent confrontations. Indeed, soli- darity in the sense of mutual recognition of collectivity membership is a necessary basis for most forms of collective conflict (Darby 1990).

Because collective violence typically occurs along pre-existing lines of nonviolent oppositions (Rule 1988:266), to account for collective conflict one needs to account for collective solidarity. Here several major theoretic issues are joined. For example, Hechter has rejected what he calls "normative" and "structural" accounts of group solidarity in favor of an individualistic "rational choice" model (Hechter 1987). In that model solidarity is created when "the group" [collectivity] produces exclusive collective goods for its members, who depend upon the collectivity for diverse, multiple goods, and when the collec- tivity exerts sanction ("control") against nonconforming members.

But numerous cases show that structural processes can create and solidify ethnic identities. Periodic violent conflicts intensify ascribed membership, and the greater the violence the more intense become the pressures for solidarity, as noted by Darby (1990:157) for Northern Ireland. In general, "civil violence is more likely the greater the solidarity among the participants" (Rule 1988: 266). And in a cross-national study, Boswell & Dixon (1990:556) note that the data show that ethnic conflict, measured as "separatism", is the variable most highly correlated with "rebellion."

Studies that closely examine the interactive processes of mobilization and conflict are rare. McAdam has described the black insurgency of the 1955- 1970 period in terms of a process of tactical interaction between civil rights advocates and their opponents-a continuous process in which tactical inno- vations are met by counter-measures (McAdam 1983). Morris emphasized the pre-existing bases and internal structure of the civil rights movements as important determinants of both initial mobilization and the innovations and spread of strategies and tactics (Morris 1981).

The concept of a repertoire of collective action, as developed by Tilly (1986), proposes that particular forms of protest and conflict are historically contin- gent-they are used and they "work" only in particular settings. Voluntary associations using mass demonstrations to prod political change arose in so- cieties with electoral systems. Systematic nonviolent mass protests rest upon complex sets of understanding, skills and tactics, typically developed in mod- em societies with urban elites experienced in electoral politics (cf Beissinger 1992, Zaslavsky 1992, and the prescient work of Amalrik 1969 and d'Encausse 1978 for the Soviet Union).

"Political process" theories are said to argue that "collective action stems from political struggles over the entry of new groups and is determined by the mobilization of contenders and the structure of political opportunities" (Jenkins & Schock 1992:170). Groups excluded from effective voice in the polity may become unruly-but under what conditions? A major clue is that the intensity

64 WILLLAMS

of rebellions is higher when potential separatism (primarily ethnic) is great and so is political and economic discrimination (Jenkins & Schock 1992:171, citing Gurr 1968, Gurr & Duvall 1973, Gurr & Lichbach 1979, Boswell & Dixon 1990).

Class, Economy and Competition Recurrent debates as to the relative importance of class position and interests as over against ethnic identities and commitments have not produced-and cannot be expected to produce-a decisive verdict. To dismiss ethnicity as false consciousness ignores the clear evidence that ethnies often sacrifice economic interests in favor of symbolic gains, e.g. Sinhalese and Tamils in Sri Lanka (Pfaffenberger 1990:249). The relevant question is how the two aspects of social reality affect actions in specified contexts (Goyden 1983, Norton 1984).

An instructive example is the collective violence in India (more than 7000 deaths in five years, 1978-1983) between Assamese Hindus and Bengali Muslims. Bengali immigration clearly led to competition for land, jobs, and other economic and political resources. But cultural issues (especially of lan- guage and religion) become salient in political processes, transforming a sec- ular movement into one of ethnic and religious oppositions. The state was not a neutral arena but rather both a resource and a major influence. Conventional monofactorial theories fail to account for the violent and protracted conflicts (Darnell & Parikh 1988). Appropriate interpretation involves class, demo- graphic change, ethnic identities, religious affiliations, state actions, and po- litical processes.

Analyses of labor unrest and ethnic conflict in the United States (Olzak 1989) have led to the following conclusions: (i) Periods of economic contrac- tion show high levels of ethnic conflict. (ii) Rates of ethnic conflict decline sharply with time elapsed since the last event of this kind. (iii) Waves of strikes and antiblack events are strongly correlated. (iv) But the effect of strikes disappears when strength of union organizations are taken into account. (v) Yet growth of unions coincides with rates of attacks on blacks-but is nega- tively associated with rates of attacks on nonblack groups. (vi) Immigration intensified labor market competition among foreigners, native whites, and blacks, but two thirds of all collective attacks were against blacks.

Thus, the targets of ethnic conflict were not predominately the white im- migrants (who provided much of the direct job competition), but rather the black Americans-who were already objects of systematic discrimination.

Ethnic rivalry, whether or not extending into violent conflict, takes different forms in ranked, in contrast to unranked, systems. When ethnies are ranked, as in caste systems, class and ethnic membership coincide, hierarchies are locally based, direct interactions of subordinates and dominants are frequent.

ETHNIC CONFLICTS 65

So long as dominance is stable, struggle takes such forms as malingering, sabotage, evasion, work stoppages, and sporadic violence. If collective vio- lence eventually occurs, it is likely to mean a social revolution, e.g. Rwanda in 1959 (Horowitz 1985:30). In contrast, unranked ethnies within a single state tend toward a politics of inclusion/exclusion-seeking power-sharing or con- trol, and minimizing intra-ethnic class divisions. If ethnic conflict then be- comes overt and polarized, the hostile ethnies seek autonomy and either dominance or separation.

Ethnicity, clearly, often coincides with class. Equally clearly, neither class nor ethny is reducible to the other (cf Rothschild 1981). Increased scholarly attention to ethnic conflict does not lessen the importance of class and eco- nomic sectorial bases for other forms of collective rebellion, e.g. peasant-based revolts and revolutions (cf Paige 1975, Jenkins 1982). Yet an emphasis on class struggle can lead to a neglect or downplaying of ethnic cleavages as a factor in peasant uprisings. Detailed studies of particular cases, not surpris- ingly, show complex interactions of class, state, ethnicity, and local social structures (e.g. on southern Peru: Orlove 1990). Evidently, local and provincial (state) structures may create legal and other constraints that shape the nature and outcomes of labor market competition. Thus, market competition as a determinant of ethnic conflict rests in part upon prior political controls (cf James 1988).

The accumulated historical studies repeatedly tell us that large-scale collec- tive conflicts are enormously complex mass phenomena. Thus Markoff notes that in rural France in 1789 there was not a unified peasant rising; rather, there were hundreds of varied local insurrections-"an amalgam of quite diverse conflicts" (Markoff 1985:761).

Under conditions of extreme scarcity, informal economies develop-smug- gling, black markets, hoarding, diversion of resources from the public arena, official collusion. Such parallel systems tend to cohere around ethnic and regional social networks. One result is a sharpening of ethnic boundaries- within a class system shaped by the state (Chazan 1986:142-43).

The respective parts played by class and ethnic factors in political opposi- tions can change through time. In the Quebec independence movement, early class differences were marked: lowest support among managers and propri- etors, farmers and workers; higher among the highly educated and intellectual strata. Over time, support increased in all classes, but the intelligentsia con- tinued to lead. These findings are consistent with evidence for the high im- portance of the intelligentsia in nationalistic movements in countries as dif- ferent as Nigeria, France (Brittany), Spain (Basques), and Trinidad (Pinard & Hamilton 1984).

Ethnic activists include a high proportion of intellectuals-school teachers, mass media people, writers, publicists. They recently have been upwardly

66 WILLLAMS

mobile. They are articulate and ambitious and at the same time often occupied with a search for identity (cf Beer 1980).

In the well-studied case of Belgium, ethnic opposition was preceded by struggles that were first defined in religious terms and then as class conflicts. In each period, the political solutions were constitutional, but the resulting arrangements were subsequently objects of ethnic confrontation. Analysis by Covell (1985: see Brass, ed.) suggested once more that it is when class and ideological or religious differences coincide with ethnic cleavages that the most severe conflicts will appear (Covell 1985:256-57).

In a skeptical review of research on political protest, Zimmerman (1980:77) was willing to say that if there is one basic finding, it is that economically developed societies have less intense forms of political violence. It happens, of course, that the economically developed societies differ from others in many other potentially decisive ways, especially in the structure of political oppor- tunities.

Ethnoregional studies offer puzzles. Recall that Beissinger (1992) presented data for the Soviet Union showing greatest violence in poorer and less urban- ized regions. But Beer's major work on ethnic activism in France (1980) reported that extra-electoral dissent was higher in the economically less de- veloped areas, whereas ethnic electoral political activity is more typical of "developed" regions. Horowitz (1985: chapter 6) developed the distinction between "backwardness" (disadvantage) of ethnies and of the region of con- centrated residence. His inventory of diverse cases shows that secessionistic movements tend to be frequent and early for backward groups in backward regions, somewhat frequent but late for advanced groups in backward regions, rare but early for backward groups in advanced regions, and, finally, rare and late for advanced collectivities in advantaged regions (Horowitz 1985:258). Special threats and opportunities help to explain such cases as Basques and Catalans in Spain and Slovenes and Croats in Yugoslavia (cf Aslund 1992:85).

If ethnies are geographically concentrated in different regions of a sizeable country, the unevenness of environment is enough to create inequalities and thus potentials for conflict. Since ethnies are ascriptive, "ethnic homelands" favor boundary persistence and inequality (Lambert 1981, Ayoade 1986:110). Such ethnicity becomes especially salient when there is a perceived external threat.

Demographic Factors Demographic shifts can powerfully affect ethnic relations. Between 1979 and 1989, the following rates of population growth prevailed in these contrasting regions of the Soviet Union: Estonians 0.7%, Latvians, 1.4%, Tajiks, 45.5%, Uzbeks, 34.0%. Central state redistribution of resources under these conditions came to be seen as subsidizing rapidly growing groups, creating potential

ETHNIC CONFLICTS 67

migrants into the richer, slow-growing regions, and generating a long-term threat to cultural survival (Zaslavsky 1992:115).

The 1990s have brought many examples in which rapid and large in-migra- tion of ethnically distinctive populations arouses fear and hostility. Thus, the surges of Bengalis into Assam waked fears of being "swamped" and led to demands that Assamese be the language of education and administration. The resulting complex developments eventuated in the violence of the 1983 elec- tions (WK Anderson 1990).

The internationalization of labor supply in the long period of economic expansion from the end of World War II to the mid-1970s produced tensions in country after country between the status of citizenship and the status of in-migrant populations that have distinctive cultural or physical characteristics (cf Axford 1987:374-75). The large-scale international labor migration of ethnically identified populations into democratic polities raises political con- tention and episodes of violence, as in Germany, France, and Great Britain, among others (Esman 1986, Brubaker 1989, Horowitz & Noiriel 1992).

High rates of population growth in poorer countries ensure pressures for international migration; when migration occurs, it results in diaspora and ethnic enclaves-which in turn increase the likelihood of ethno-political conflicts in receiving states and between exporters and importers of populations (Esman 1992). There is evidence of a growing consensus in the European community in favor of ethnically restrictive immigration policies (Allen & Macey 1990: 378).

The movement toward an integrated European economy increasingly raises broader questions concerning immigration, labor mobility, ethnicity, and citi- zenship. National states still claiming sovereignty face increasing ethnic con- flict (Miles 1992). A typical response of national states is to impose restrictions on immigration.

Very large international movements of population are increasingly volatile and unpredictable, and both economic inequalities and violent conflicts gen- erate great pressures for migration (Teitelbaum 1992). Accordingly, the likely prospects are for continuing ethnic tensions in receiving societies.

State Versus Ethnies The accumulated research findings show not only how state structures affect and are affected by ethnic formations, but also how political changes react upon ethnies. Thus, the politicization of ethnicity has been a prominent feature of transitions, since 1970, from authoritarian rule in some 25 countries. Periods of transition relax suppression and open new opportunities for ethnic entrepre- neurs. Ethnies or ethnoregional groupings with advanced economies and great cultural differences vis a vis the dominant political groups are most likely to seek autonomy or national sovereignty (Frye 1992:623, cf Horowitz 1985 on

68 WILLIAMS

"advanced" and "backward" groups). Examples are: Basques and Catalans in Spain; Estonia, Lithuania, Latvia, Moldavia, Armenia, and Georgia in the Soviet Union; Slovenia and Croatia in Yugoslavia. Of 12 early seekers after autonomy, 10 were in advanced regions (exceptions: Slovakia and Kosovo).

In general it appears that successful ethnic challenges to central states are more likely to follow a state crisis than to succeed because they produced a state crisis. Putative causes of rebellions against the state include income inequality, economic decline, state repression, international economic and political dependency, wars, invasions and interventions (Boswell & Dixon 1990). Boswell & Dixon claim that their analysis of 63 countries points to three clusters of factors inducing violent rebellions; (i) relative deprivation (inequality and/or economic decline), (ii) resource mobilization (economic development; inverted U for state repression), (iii) dependency (indirect influ- ence through inequality, repression, and slow economic growth).

We note that results of multivariate analyses of relationships between in- come inequality and deaths from political violence often show high sensitivity to unusual cases as well as to measurement and analysis problems (Weede 1985, Muller 1985). Also, it is clear that rapid social change does not lead directly to political upheaval; the accompaniments are quite variable (Zimmer- man 1980:279-83). Nor do ethnic tensions and conflicts invariably lead to rebellions or separatist movements. Alternatives include: stable dominance of one ethny, together with cooperation of others; multiple power-sharing; and communal conflict and civil war. In some cases, however, persisting violent ethnic conflict signifies "endemic secessionism" (Mayall & Simpson 1992). Major instances are north and south Sudan and Eritrea and Ethiopia. Mayall and Simpson identify five major contextual conditions that help to predict whether ethnic cleavages will lead to secessionist movements: colonial poli- cies, post-colonial efforts by states to erase ethnic diversity, political economy of dissidence, religious (confessional) allegiances, and a regional environment of support or opposition (1992:9-10). In general, modernization appears to be much less important as a source of ethnic conflict than previously thought (Leifer 1981).

Numerous studies have examined the proposition that collective protests and/or rebellions are greatest at intermediate levels of political opportunities and lowest under full access or extreme repression (see Jenkins & Schock 1992:172-75). Conditions favoring collective action have been variously iden- tified as elitist regimes, moderate political freedom, moderate repression, ex- treme but incomplete repression over an extended period, elite divisions and political instability. Several of these circumstances are likely to be found in weakly institutionalized ethnocracies in multiethnic societies (cf DL Horowitz 1985). There is considerable scholarly agreement that the type of political system strongly affects the frequency and intensity of protest behavior, i.e.

ETHNIC CONFLICTS 69

protests are more intense in elitist and autocratic than democratic states (Zim- merman 1980:173, Tiryakian & Rogowski 1985, Beissinger 1992, Gurr 1993: 184-85). In general democracies have more extensive but less deadly protests.

Once a society has been through a series of ethnically based political con- frontations and violent changes in control of the state, it typically has a narrowly based authoritarian regime-an ethnocracy. Such regimes are diffi- cult to displace, as witness Hafez Assad in Syria or Saddam Hussein in Iraq. As DL Horowitz notes (1985:499): "Only an escalation of the violence to the dimensions of civil war or invasion-and both require external assistance-is likely to dislodge an ethnocratic regime that has gone through a series of coups and purges."1

In any case, regime changes, whether through coups or otherwise, are seldom a matter of indifference to other states that are either nearby or linked by economic or political ties. Thus in the coup-prone areas of West Africa, interstate conflicts typically follow the "radical behavior of emergent regimes" (Agyeman-Duah 1990:566).

The need for explanatory schemes tailored to particular kinds of conflicts is well-illustrated in the case of military affairs, to which we now turn.

Ethnicity, Military Forces, and Coups d'Etat Military and paramilitary forces have played central parts in Third World countries for several decades in the efforts of authoritarian regimes to maintain their power (Janowitz 1977). Like states as total institutions, military forces are both resources for, and objects of, ethnic conflict; in state after state ethnically linked military groupings intervene or threaten to intervene in ci- vilian politics. Seldom is the military ethnically neutral-whether because of the legacy of biased recruiting by colonial regimes, a cleavage between the military and political parties and the civilian government, or interpenetration of civil and military ethnic fractions (see the comprehensive treatment in DL Horowitz 1985: chapters 11-12).

The importance of ethnicity in the recruitment, deployment, and political role of military forces has been well-recognized for several decades (cf Enloe 1980, Kirk-Green 1980, Janowitz 1977, 1981). Military service, sometimes thought to be a homogenizing, even an integrative, process (e.g. in nineteenth century France) has proven in many cases to be ethnically divisive and a source of major conflicts.

The importance of ethnic divisions as a factor in military rule varies greatly among regions of the world, e.g. very great in Africa and South Asia, joint with religious cleavages in the Middle East, much less in Latin America (Tilly

1(Reprinted by permission of the author and the University of California Press. Copyright 1 1985, The Regents of the University of California.)

70 WILLLAMS

1992:219). Independence in Latin America came primarily through military revolts; in Africa, through peaceful means. In the latter, colonial forces had been ethnically skewed to make them as a-political as possible, with the consequence of unrepresentative militaries after independence (Welch 1986: 321).

The great difficulties of understanding military coups are highlighted in the recent turbulent history of South American countries (Markoff & Baretta 1986). Of the many plausible but empirically dubious explanations proffered, few give any attention to ethnicity. For Africa, however, numerous studies have examined militarized ethnicity in detail (e.g Jackman 1978).

Case studies identify important context and variables and generate evocative hypotheses (Barua 1992). To appraise causally important structures and pro- cesses, however, it is necessary to have more systematic multivariate compar- isons. For military coups, increased understanding has come from recent work.

Of the 45 independent black states in Africa, from 1960-1982, nearly 90% recorded a military coup, attempted coup, or plot (Johnson, Slater & McGowan 1984:627). There were 56 successful coups, 56 other attempts, and 102 plots. Political conflicts have made these sub-Saharan states among the most unstable in an unstable world. In the late 1980s, the central executive in some 25 of them was a military regime, and military influence was important in most of the others. The most complete available analysis of this extraordinary situation has reviewed four major explanations: political development (weak states and heavy demands), military centrality, ethnic cleavages or antagonisms, and economic dependency (Jenkins & Kposowa 1990). Although the results of detailed factor and regression analyses were highly complex-and differ for plots, attempts, and successful coups-the strongest predictor of military coup activity was military centrality, followed by ethnic cleavages and ethnic po- litical competition (Jenkins & Kposowa 1990, Kposowa & Jenkins 1993). Social mobilization (or "modernization") did not fare well as an explanatory condition: it did not affect coup activity through rising levels of political participation or domestic turmoil, although it was associated with greater likelihood of coups, possibly through enhanced communicative and organiza- tional resources. Economic dependence was not a strong or consistent factor. The net conclusion is that military centrality and ethnic tensions are the primary sources of irregular military interventions. Ethnic plurality and competition predict intervention whereas ethnic dominance lowers the likelihood of coup activity. Ethnic factors have this impact primarily in elite struggles rather than in mass turmoil. Cultural cleavages and rivalry between the two largest ethnies were primary contexts for military coups. The likelihood of coups was greater when there also was a strong military establishment, along with substantial political mobilization.

Of course, coups constitute a limited set of events and are overshadowed in

ETHNIC CONFLICTS 71

consequences by large-scale communal violence, rebellions, and civil wars. Still, the regularities apparent in such events illuminate the continuing signif- icance of ethnies as actors in political arenas.

In the larger, more deadly, conflicts there is by now abundant case material to support the conclusion that unilateral international involvements in conflicts between ethnies and states often, perhaps typically, result in lengthening, militarizing and escalating the struggle, in strengthening ethnic boundaries, and in reducing the likelihood of negotiated settlements (Nagel & Whorton 1992). Among many cases are Angola, Iraq, Eritrea, Sudan, Lebanon, Sri Lanka, Iran, and several countries in Central America.

Great variations in issues and in the extent of violence are noted in a close examination of ethnic conflicts, peace negotiations, and peace accords in several salient cases-Canada, Cyprus, India (Punjab), Sri Lanka, and Sudan (de Silva & de A Samarasinghe eds. 1993). With the exception of Canada, all the conflicts were substantially internationalized, reached high levels of vio- lence, and posed formidable difficulties for mediation or durable peace accords. Indeed, Richardson & Wang (1993) suggest that violent ethnic conflicts tend to be more protracted and more difficult to resolve than wars between states. Not surprisingly, also, a civil war seems more likely to develop into an interstate war when one party has ethnic links to another country (cf Blaine 1973).

DISORDER AND PROMISE IN RESEARCH

The continuing chaotic state of general theories of collective violence still besets analyses. Thus, Rule's (1988) historical review of theories-rational choice, irrational collective outbreak, mass society, structural, resource-mobi- lization and others-fails to find central convergence, although particular models are deemed useful.

Accounting-for-variance studies are notoriously difficult, and those that wager all upon a single variable typically are inconclusive. Thus, Lichbach cites "diverse and contradictory" findings from some 40 studies that support every conceivable relationship between economic inequality and political con- flict. This disconcerting circumstance has been variously attributed to problems of research design or statistical modeling and to deficiencies in fonnal models and theories. Eclectic work had failed to resolve the major puzzles. Remedies presumably must be sought in more rigorous studies that test crucial differences in predictions (Lichbach 1989).

Further, we must acknowledge that satisfying explanations of collective conflict cannot be confined to multivariate "accounting" in which a cluster of conditions are shown to precede or accompany the outcome of interest. Timing and sequence must be analyzed as well-"system-level outcomes depend not

72 WILLIAMS

only on the configuration of actors, motives, and resources but also crucially on the sequence in which action occurs..." (Sewell 1987:171).

What we now have in stock can be briefly summarized. First, there is a vast set of richly descriptive case studies, which suggest hypotheses, record phe- nomena ("what is possible"), and clarify concepts. Second, there are numerous models (or, conceptual schemes), often of considerable evocative value. Third, many studies usefully compare a limited set of case studies. Fourth, new and promising analyses are appearing that use statistical techniques to analyze large data-sets. From all these resources, gains are being made. An example is in order.

Defining a universe of 227 large "communal groups"-those with culturally distinct and persistent identities that do not have recognized states or institu- tionalized political status, Gurr (1993) has produced a massive analysis of conditions associated with political mobilization, protest, and rebellion against the state. The study identifies five overlapping types of communal groups: ethnoclasses, ethnonationalists, militant sects, indigenous peoples, and com- munal contenders. Although having common characteristics, each of the types is distinctive in important ways. The complex findings defy brief summary, but contain such surprises as the weak effects on mobilization of such "re- sources" as size, concentration, and coherence (Gurr 1993: 179), and the significant apparent influence of contagion and diffusion on protest and rebel- lion among militant sects and indigenous peoples (pp. 180, 189). A general explanatory model is proposed that combines the perspectives of "grievance" and "political interest" (mobilization). Both sets of factors are seen to interact in situations of protest and rebellion.

The general scheme proposes that "objective" conditions such as economic deprivation, discrimination, and loss of political or cultural autonomy provide the basis for grievances, which are then used by entrepreneurs to mobilize collective action. The outcomes of mobilization, in turn, depend upon op- portunity structures including international influences. Gurr has summarized the main findings as follows (1993:161): "Statistical analysis shows that cultural identity, inequalities, and historical loss of autonomy all contribute to their grievance. Political mobilization, grievances, and the international diffusion and contagion of communal conflict jointly explain the extent of political action in the 1980s. Democracy, state power, and institutional change help determine whether conflict [opposition] takes the form of protest or rebellion." (Reprinted with permission of Butterworth-Heinemann Ltd and the author.)

This study along with hundreds of others allows us to infer the most lethal configurations for ethnic conflict. The recipe for this Devil's Brew would include: multiethnic population, centralized state with redistributive powers,

ETHNIC CONFLICTS 73

substantial military forces, a few large ethnies of nearly equal political re- sources, territorial concentration, marked inequality and/or rapid change in ethnic positions, and ethnic political struggles for collective goods. Mix well, and run for shelter.

Space limits foreclose a discussion of strategies for managing such conflicts. Considerable knowledge does exist.

CONCLUSION

The difficulties of research in this field are well-known. The phenomena of central interest-ethnicity, states, conflict-are of stunning complexity, requir- ing both disaggregation and multivariate analysis. The available data typically are cross-sectional, whereas repeated observations over time are needed for the dynamic analyses that could begin to disentangle causal sequences. Data often are crude and of uncertain validity, e.g. many conflicts go unreported and accounts may be incomplete and distorted. Information on microprocesses is scarce. Sequential-narrative analysis is in an early state of development. As we have just seen, a profusion of concepts, models and dominant-factor the- ories hinder a coherent research strategy and impede the organization of cumulative knowledge.

None of these difficulties is insurmountable. Indeed, the work here reviewed has produced much new substantive knowledge and conceptual clarification.

In spite of numerous controversies and confusing empirical findings, the same basic conditions for ethnic conflict repeatedly are identified in the works here reviewed. Most important are:

(i) strong ethnic identities/boundaries; (ii) grievances; (iii) opportunity structures; (iv) resource mobilization.

An implication is that, at the very least, any research that seeks valid explanations should attend to all these simultaneously. Future studies should carefully specify the type of conflict that is in question-whether nonviolent protest, riots, pogroms, mutinies, coups d'etat, communal violence, low-in- tensity conflict (sabotage, kidnapping, assassinations, bombings, extortion), major armed combat. Type of ethny and extent and kind of grievances must be similarly identified. Linking the configurations of structural conditions to processes of mobilization and counter mobilization will open new possibil- ities for understanding some of the most intractable and tragic conflicts of our times.

74 WILLIAMS

Any Annual Reviw chapter, as well as any article cited in an Annual Review chapter, may be purchased from the Annual Reviews Preprints and Reprints service.

1-800-347-8007; 415-259-5017; email:arpr@class.org

Literature Cited

Agyeman-Duah B. 1990. Military coups, re- gime change, and interstate conflicts in West Africa. Armed Forces & Soc. 16:547-70

Alba RD. 1990. Ethnic Identity: The Transfor- mation of White America. New Haven/Lon- don: Yale Univ. Press. 374 pp.

Allen S, Macey M. 1990. Race and ethnicity in the European context. Br. J. Sociol. 41:375- 93

Amalrik A. 1969. Will the Soviet Union Survive Until 1984? Amsterdam: Herzen

Anderson B. 1991. [1983]. Imagined Commu- nities: Reflections on the Origin and Spread of Nationalism. London: Verso

Anderson WK. 1990. Multiethnic conflict and peacemaking: the case of Assam. See Mont- ville 1990, pp. 327-39

Armstrong J. 1982. Nations Before National- ism. Chapel Hill, NC: Univ. N. Carolina Press

Aslund A. 1992. Russia's road from Commu- nism. Daedalus 121:77-93

Axford B. 1987. United Kingdom. See Sigler 1987, pp. 369-93

Ayoade JAA. 1986. Ethnic politics in Nigeria: a conceptual reformulation. See Thompson & Ronen 1986, pp. 105-18

Azar EE. 1990. The Management of Protracted Social Conflict. Hampshire, England: Dart- mouth

Azar EE, Burton JW, eds. 1986. International Conflict Resolution: Theory and Practice. Brighton, UK: Wheat-sheaf Books

Balibar E. 1990. The nation form: history and ideology. Review 13:329-61

Banton M. 1983. Racial and Ethnic Competi- tion. Cambridge: Cambridge Univ. Press

Barsh RL, Henderson JY. 1980. The Road: Indian Tribes and Political Liberty. Berke- ley/Los Angeles: Univ. Calif. Press. 301 PP.

Barua PB. 1992. Ethnic conflict in the military of developing nations: a comparative analy- sis of India and Nigeria. Armed Forces & Soc. 19:123-37

Beer WR. 1980. The Unexpected Revolution: Ethnic Activism in Contempory France. New York: New York Univ. Press. 150 pp.

Beissinger MR. 1992. [1991]. Protest mobiliza- tion among soviet nationalities. Presented at Annu. Meet. Am. Sociol. Assoc., 86th, Cin- cinnati; rev. 1992. Peace Stud. Program, Cor- nell Univ.

Belanger S, Pinard M. 1991. Ethnic movements

and the competition model: some missing links. Am. Sociol. Rev. 56:446-57

Beteille A. 1991. Society and Politics in India: Essays in a Comparative Perspective. Lon- don: Athlone. 316 pp.

Blaine G. 1973. The Causes of War. London: Macmillan

Bonacich E. 1976. Advanced capitalism and black/white race relations in the United States: a split labor market interpretation. Am. Sociol. Rev. 41:34-51

Boswell T. 1989. Colonial empires and the cap- italist world economy: a time series analysis of colonization, 1640-1960. Am. Sociol. Rev. 54:180-96

Boswell T, Dixon WJ. 1990. Dependency and rebellion: a cross-national analysis. Am. Sociol. Rev. 55:540-59

Brass P, ed. 1985. Ethnic Groups and the State. London & Sydney: Croom Helm. 341 pp.

Brereton B. 1979. Race Relations in Colonial Trinidad 1870-1990. New York: Cambridge Univ. Press. 251 pp.

Breton R. 1988. From ethnic to civic national- ism: English Canada and Quebec. Ethn. Ra- cial Stud. 11:85-102

Brogan P. 1990. The Fighting Never Stopped: A Comprehensive Guide to World Conflict since 1945. New York: Vintage

Brubaker WR, ed. 1989. Immigration and the Politics of Citizenship in Europe and North America. Lanham MD: Univ. Press of Amer- ica

Chazan N. 1986. Ethnicity in economic crisis: development strategies and patterns of eth- nicity in Africa. See Thompson & Ronen 1986, pp. 137-58

Cohen A. 1981. The Politics of Elite Culture: Explorations in the Dramaturgy of Power in a Modern African Society. Berkeley/Los An- geles: Univ. Calif. Press. 257 pp.

Colburn FD. 1991. The People's Democratic Republic of Ethiopia: Masking and unmask- ing tragedy. World Polit. 43:570-86

Collihan KM, Danopoulos CP. 1993. Coup d'etat attempt in Trinidad: its causes and fail- ure. Armed Forces & Soc. 19:435-50

Connor W. 1981. Nationalism and political illegitimacy. Can. Rev. Stud: Nationalism 8: 201-28

Connor W. 1990. When is a nation? Ethn. Ra- cial Stud. 13:92-103

Connor W. 1992. The nation and its myth. Int. J. Comp. Sociol. 33:48-57

ETHNIC CONFLICTS 75

Conquest R. 1990. The Great Terror: A Reas- sessment. New York: Oxford Univ. Press. 570 pp.

Conversi D. 1990. Language or race?: The choice of core values in the development of Catalan and Basque nationalisms. Ethn. Ra- cial Stud. 13:50-70

Covell E. 1985. Ethnic conflict, representation and the state in Belgium. See Brass 1985, pp. 228-61

Darby J. 1990. Northern Ireland: The persis- tence and limitations of violence. See Mont- ville 1990, pp. 151-59

Darnell AT, Parikh S. 1988. Religion, ethnicity, and the role of the state: explaining conflict in Assam. Ethn. Racial Stud. 11:263-81

d'Encausse HC. 1978. Decline of an Empire: The Soviet Socialist Republics in Revolt. New York: Newsweek

de Silva KM, de A Samarasinghe SWR, eds. 1993. Peace Accords and Ethnic Conflict. London/New York: Pinter (dist. St. Martin's Press). 208 pp.

Douglas WA. 1988. A critique of recent trends in the analysis of ethnonationalism. Ethn. Ra- cial Stud. 11: 192-206

Dutter LE. 1990. Theoretical perspectives on ethnic political behavior in the Soviet Union. J. Confl. Resolut. 34:311-34

Eller JD, Coughlan RM. 1993. The poverty of primordialism: the demystification of ethnic attachments. Ethn. Racial Stud. 16:183-202

Enloe CH. 1980. Ethnic Soldiers: State Security in Divided Societies. Athens: Univ. Georgia Press. 276 pp.

Enloe CH. 1986. Ethnicity, the state, and the new international order. See Stack 1986, pp. 25-42

Eriksen TH. 1990. Linguistic diversity and the quest for national identity: the case of Mauritius. Ethn. Racial Stud. 13:1-24

Esman M. 1986. Diasporas and international relations. In Modern Diasporas in Interna- tional Politics, ed. G Sheffer, pp. 333-49. London: Croom-Helm

Esman M. 1992. The political fallout of trans- national migration. Diaspora 2:3-42

Esman M, Rabinovich I, eds. 1988. Ethnicity, Pluralism, and the State in the Middle East. Ithaca/London: Cornell Univ. Press. 304 pp.

Esman MJ. 1985. Two dimensions of ethnic politics: defense of homelands, immigrant rights. Ethn. Racial Stud. 8:438-40

Esman MJ. 1990. Ethnic pluralism and interna- tional relations. Can. Rev. Stud. Nationalism 17:83-93

Fishman JA. 1981. Language maintenance and ethnicity. Can. Rev. Stud. Nationalism 8: 229-47

Flood LG. 1980. Ethnic politics and political science: a survey of leading journals. Ethnic- ity 7:96-101

Fraser NM, Hipel KW, Jaworsky J, Zuljan R.

1990. A conflict analysis of the Armenian- Azerbaijani dispute. J. Confl. Resolut. 34: 652-77

Frye TM. 1992. Ethnicity, sovereignity and transitions from non-democratic rule. J. Int. Aff. 45:599-623

Fugita SS, O'Brien DJ. 1991. Japanese Amer- ican Ethnicity: The Persistence of Commu- nity. Seattle: Univ. Wash. Press. 207 pp.

Geertz C. 1973. The Interpretation of Cultures: Selected Essays. New York: Basic Books

Gellner E. 1983. Nations and Nationalism. Ith- aca, NY: Cornell Univ. Press

Gibbs J. 1989. Conceptualization of terrorism. Am. Sociol. Rev. 54:329-40

Gordon MM. 1964. Assimilation in American Life. New York: Oxford Univ. Press. 276 pp.

Goyden JC. 1983. Ethnicity and class identity: the case of French- and English-speaking Ca- nadians. Ethn. Racial Stud. 6:72-89

Grant DS II, Wallace M. 1991. Why do strikes turn violent? Am. J. Sociol. 96:1117-50

Greenwood DJ. 1985. Castilians, Basques, and Andalusions: an historical comparison of na- tionalism, 'true' ethnicity, and 'false' ethnic- ity. See Brass 1985, pp. 202-27

Grimes BF, ed. 1988. Ethnologue: Languages of the World. Dallas: Summer Inst. Linguist. 11th ed.

Gurr TR. 1968. A causal model of civil strife: a comparative analysis using new indices. Am. Polit. Sci. Rev. 62:1104-24

Gurr TR, ed. 1980. Handbook of Political Con- flict: Theory and Research. New York: Free Press. 566 pp.

Gunf TR. 1993. Why minorities rebel: A global analysis of communal mobilization and con- flict since 1945. Int. Polit. Sci. Rev. 14:161- 201

Gurr TR, Duvall RD. 1973. Civil conflict in the 1960s: a reciprocal theoretical system with parameter estimate. Comp. Polit. Stud. 6: 135-69

Gunf TR, Lichbach MI. 1979. A forecasting model for political conflict within nations. In To Auger Well, ed. JD Singer, MD Wallace, pp. 153-93. Beverly Hills: Sage

Hannan MT. 1979. The dynamics of ethnic boundaries in modern states. In National De- velopment and the World System, ed. JW Meyer, MT Hannan, pp. 253-75. Chicago: Univ. Chicago Press. 334 pp.

Harris GS. 1977. Ethnic conflict and the Kurds. Ann. Am Acad. Polit. Soc. Sci. 433:112-24

Hechter M. 1975. Internal Colonialism: the Celtic Fringe in British National Develop- ment 1536-1966. Berkeley: Univ. Calif. Press. 361 pp.

Hechter M. 1987. Principles of Group Solidar- ity. Berkeley/Los Angeles: Univ. Calif. Press. 219 pp.

Heisler MO. 1990. Ethnicity and ethnic rela-

76 WILLIAMS

tions in the modem West. See Montville 1990, pp. 21-52

Hintzen PC. 1989. The Costs of Regime Sur- vival: Racial Mobilization, Elite Domination and Control of the State in Guyana and Trin- idad. Cambridge: Cambridge Univ. Press. 240 pp.

Hirschman C. 1983. America's melting pot re- considered. Annu. Rev. Sociol. 9:397-423

Hirschman C. 1986. The making of race in co- lonial Malaya: political economy and racial ideology. Sociol. Forum 1:33061

Hirschman C. 1991. What happened to the white ethnics? Contemp. Sociol. 20:180-83

Hoffman B. 1992. Current research on terror- ism and low-intensity conflict. Stud. Confl. Terrorism 15:25-37

Horowitz DL. 1985. Ethnic Groups in Conflict. Berkeley/Los Angeles: Univ. Calif. Press. 697 pp.

Horowitz DL. 1990. Ethnic conflict manage- ment for policymakers. See Montville 1990, pp. 115-116

Horowitz DL, Noiriel G, eds. 1992. Immigrants in Two Democracies: French and American Experience. New York: New York Univ. Press

Horowitz IL. 1992. Taking Lives: Genocide and State Power. New Brunswick, NJ: Transac- tion. 240 pp. 3rd ed.

Hovannisian R, ed. 1986. The Armenian Geno- cide in Perspective. New Brunswick, NJ: Transaction Books

Jackman RW. 1978. The predictability of coups d'etat: a model with African data. Am. Polit. Sci. Rev. 72:1262-75

James DR. 1988. The transformation of the Southern racial state: class and race determi- nants of local-state structures. Am. Sociol. Rev. 53:191-208

Janowitz M. 1977. Military Institutions and Co- ercion in the Developing Nations. Chicago: Univ. Chicago Press. 211 pp.

Janowitz M, ed. 1981. Civil-Military Relations: Regional Perspectives. Beverly Hills: Sage

Jaynes GD, Williams RM Jr, eds. 1989. A Com- mon Destiny: Blacks and American Society. Washington, DC: Natl. Acad.

Jenkins JC. 1982. Why do peasants rebel? Structural and historical theories of modem peasant rebellions. Am. J. Sociol. 88:487-514

Jenkins JC, Kposowa AJ. 1990. Explaining mil- itary coups d'etat: black Africa, 1957-1984. Am. Sociol. Rev. 55:861-75

Jenkins JC, Schock K. 1992. Global structures and political processes in the study of domes- tic political conflict. Annu. Rev. Sociol. 18: 161-85

Jiobu RM. 1990. Ethnicity and Inequality. Al- bany: SUNY Press. 208 pp.

Johnson TH, Slater RO, McGowan P. 1984. Explaining African military coups d'etat, 1900-82. Am. Polit. Sci. Rev. 78:622-40

Kamen CS. 1991. Little Common Ground: Arab Agriculture and Jewish Settlement in Pales- tine, 1920-1948. Pittsburgh: Univ. Pittsburg Press. 327 pp.

Kimmel MS. 1989. Defensive revolutionaries: The moral and political economy of Basque, Breton and Quebecois nationalism. In Re- search on Social Movements, Conflicts and Change, ed. L Kriesberg, 11:109-28. Green- wich, Conn: JAI. 339 pp.

Kirk-Green AHM. 1980. 'Damnosa Herditas': ethnic ranking and the marital races impera- tive in Africa. Ethn. Racial Stud. 3:393-414

Klandermans B. 1984. Mobilization and partic- ipation: social-psychological expansions of resource mobilization theory. Am. Sociol. Rev. 49:583-600

Kposowa AJ, Jenkins JC. 1993. The structural sources of military coups in postcolonial Af- rica. Am. J. Sociol. 99:126-63

Kriesberg L, ed. 1990. Research in Social Movements, Conflicts and Change. Green- wich, Conn: JAI. 344 pp.

Kuper L. 1985. The Prevention of Genocide. New Haven/London: Yale Univ. Press. 286 PP.

Lambert RD. 1981. Ethnic/racial relations in the United States in comparative perspective. Ann. Am. Acad. Polit. Soc. Sci. 454:189-205

Lavender A, Forsyth J. 1976. The sociological study of minority groups as reflected by lead- ing sociological journals. Who get studied and who get neglected? Ethnicity 3:388-98

Leifer EM. 1981. Competing models of politi- cal mobilization: The role of ethnic ties. Am. J. Sociol. 87:23-47

Leonard KI. 1992. Making Ethnic Choices: California's Punjabi Mexican Americans. Philadelphia: Temple Univ. Press. 333 pp.

Levinson D, ed. 1991-1993. Encyclopedia of World Cultures, Vols. I-IV. Boston: Kittall

Lichbach MI. 1989. An evaluation of "Does economic inequality breed political con- flict?" studies. World Polit. 41:431-70

Longshore D, Prager J. 1985. The impact of school desegregation: a situational analysis. Annu. Rev. Sociol. 11:75-91

Markoff J. 1985. The social geography of rural revolt at the beginning of the French Revo- lution. Am. Sociol. Rev. 50:761-81

Markoff J, Baretta SRD. 1986. What we don't know about the coups: observations on recent South American politics. Armed Forces & Soc. 12:207-35

Martiniello M. 1993. Ethnic leadership, ethnic communities' political powerlessness and the state in Belgium. Ethn. Racial Stud. 16:236- 55

Marty ME, Appleby RS, eds. 1993a. Funda- mentalisms and Society: Reclaiming the Sci- ences, the Family and Education. Chicago: Univ. Chicago Press

Marty ME, Appleby RS, eds. 1993b. Funda-

ETHNIC CONFLICTS 77

mentalisms and the State: Remaking Polities, Economies, and Militance. Chicago: Univ. Chicago Press

Mayall J, Simpson M. 1992. Ethnicity is not enough: Reflections on protracted secession- ism in the Third World. Int. J. Comp. Sociol. 33:5-25

McAdam D. 1983. Tactical innovation and the pace of insurgency. Am. Sociol. Rev. 48:735- 54

McKay J. 1982. An exploratory synthesis of primordial and mobilizationist approaches to ethnic phenomena. Ethnic & Racial Stud. 5: 395-420

Miles R. 1992. L'Europe de 1993: l'etat, l'immigration et la restructuration de l'exclusion. Sociol. Soc. 24:45-57

Montville JV, ed. 1990. Conflict and Peace- making in Multiethnic Societies. Lexington, Mass: Heath. 558 pp.

Morris A. 1981. Black Southern student sit-in movement: an analysis of internal organiza- tion. Am. Sociol. Rev. 46:744-67

Moynihan DP. 1993. Pandaemonium: Ethnicity in International Politics. New York: Oxford Univ. Press. 221 pp.

Muller EN. 1985. Income inequality and polit- ical violence: the effect of influential cases. Am. Sociol. Rev. 51:441-45

Muller EN, Weede E. 1990. Cross-national variation in political violence: a rational actor approach. J. Confl. Resolut. 34:624-51

Murdock GP. 1967. Ethnographic Atlas. Pitts- burg: Univ. Pittsburg Press

Nagel J. 1980. The conditions of ethnic sepa- ratism: the Kurds in Turkey, Iran, and Iraq. Ethnicity 7:279-97

Nagel J, Snipp CM. 1993. Ethnic reorganiza- tion: American Indian social, economic, po- litical, and cultural strategies for survival. Ethn. Racial Stud. 16:203-35

Nagel J, Whorton B. 1992. Ethnic conflict and the world system: International competition in Iraq (1961-1991) and Angola (1974- 1991). J. Polit. Mil. Sociol. 20:203-35

Newbury C. 1988. The Cohesion of Oppres- sion: Clientship and Ethnicity in Rwanda, 1860-1960. New York: Columbia Univ. Press. 332 pp.

Nielsson GP. 1985. States and "nation-groups": A global taxonomy. In New Nationalisms of the Developed West, ed. EA Tiryakian, R Rogowski, pp. 27-56. Boston: Allen & Un- win

Norton R. 1984. Ethnicity and class: a concep- tual note with reference to the politics of post-colonial societies. Ethn. Racial Stud. 7: 426-34

Olzak S. 1983. Contemporary ethnic mobiliza- tion. Annu. Rev. Sociol. 9:355-74

Olzak S. 1989. Labor unrest, immigration, and ethnic conflict in urban America, 1880- 1914. Am. J. Sociol. 94:1303-33

Olzak S. 1992. The Dynamics of Ethnic Com- petition and Conflict. Stanford: Stanford Univ. Press

Olzak S, Nagel J. 1986. Competitive Ethnic Relations. New York: Academic

Orlove BS. 1990. Rebels and theorists: an ex- amination of peasant uprisings in southern Peru. See Kriesberg 1990, pp. 139-87

Paige J. 1975. Agrarian Revolution. New York: Free Press

Panda KK. 1993. The Punjab imbroglio: The prelude to the accord of 1985. See de Silva & de A Samarasinghe 1993, pp. 83-98

Petrissans CM. 1991. When ethnic groups do not assimilate: the case of Basque-American resistance. Ethn. Groups 9:61-81

Pfaffenberger B. 1990. Ethnic conflict and youth insurgency in Sri Lanka: The social origins of Tamil separatism. See Montville 1990, pp.241-57

Pinard M, Hamilton R. 1984. The class basis of the Quebec independence movement. Ethn. Racial Stud. 7:19-54

Pinard M, Hamilton R. 1986. Motivational di- mensions in the Quebec independence move- ment: a test of a new model. In Research in Social Movements, Conflicts and Change, ed. K Land, GE Lang, 9:225-80. Greenwich, Conn: JAI. 280 pp.

Porter JR, Washington RE. 1979. Black identity and self-esteem: a review of studies of black self-concept, 1968-1978. Annu. Rev. Sociol. 5:53-74

Portes A, Rumbaut RG. 1990. ImmigrantAmer- ica: A Portrait. Berkeley/Los Angeles: Univ. Calif. Press. 300 pp.

Portes A, Truelove C. 1987. Making sense of diversity: recent research on Hispanic minor- ities in the United States. Annu. Rev. Sociol. 13:359-85

Prager J. 1982. American racial ideology as collective representation. Ethn. Racial Stud. 5:99-119

Ra'anan U. 1990 [1980]. The nation-state fal- lacy. See Montville 1990, p. 20

Ragin C. 1983. Book review. Am. J. Sociol. 88:1317-19

Ragin CC. 1991. Introduction: The problem of balancing discourse on cases and variables in comparative social science. Int. J. Comp. Sociol. 32:1-8

Richardson J Jr, Wang J. 1993. Peace accords: Seeking conflict resolution in deeply divided societies. See de Silva & de A Samarasinghe 1993, pp.173-98

Roosens EE. 1989. Creating Ethnicity: The Process of Ethnogenesis. Newbury Park, Calif: Sage. 186 pp.

Rose E. 1982. The anatomy of mutiny. Armed Forces & Soc. 8:561-74

Rothschild J. 1981. Ethnopolitics: A Concep- tual Framework. New York: Columbia Univ. Press. 290 pp.

78 WILLIAMS

Rule JB. 1988. Theories of Civil Violence. Berkeley/ Los Angeles/London: Univ. Calif. Press. 345 pp.

Russett B, Starr H. 1989. World Politics: The Menu for Choice. New York: Freeman. 2nd ed.

Schermerhorn RA. 1978. Ethnic Plurality in India. Tucson: Univ. Ariz. Press. 369 pp.

Schultz RH. 1991. The low-intensity conflict environment of the 1990s. Ann. Am. Acad. Polit. Soc. Sci. 517:120-34

Scott GM Jr. 1990. A resynthesis of the primor- dial and circumstantial approaches to ethnic group solidarity: towards an explanatory model. Ethn. Racial Stud. 13:147-71

See KO. 1986. First World Nationalisms: Class and Ethnic Politics in Nothern Ireland and Quebec. Chicago: Univ. Chicago Press. 215 Pp.

Sewell WH Jr. 1987. Theory of action, dialec- tic, and history: comment on Coleman. Am. J. Sociol. 93:166-72

Shaw RP, Wong Y. 1989. Genetic Seeds of Warfare: Evolution, Nationalism, and Patri- otism. Boston: Unwin Hyman

Sigler JA, ed. 1987. International Handbook on Race and Race Relations. New York: Green- wood. 483 pp.

Smith AD. 1981. The Ethnic Revival. Cam- bridge: Cambridge Univ. Press. 240 pp.

Smith AD. 1989. The origins of nations. Ethn. Racial Stud. 12:340-67

Smith AD. 1992. Chosen peoples: why ethnic groups survive. Ethn. Racial Stud. 15:436-56

Spicer EH. 1971. Persistent identity systems. Science 174:795-800

Stack JF Jr, ed. 1986. The Primordial Challenge: Ethnicity in the Contemporary World. New York: Greenwood. 231 pp.

Strang D. 1990. From dependency to sover- eignty: An event history analysis of decolo- nization 1870-1987. Am. Sociol. Rev. 55:846-60

Suny RG. 1985. The Making of the Georgian Nation. Bloomington and Stanford: Indiana Univ. Press/Hoover Inst. Press

Taagepera R. 1992. Ethnic relations in Estonia, 1991. J. Baltic Stud. 23:121-32

Tambiah SJ. 1986. Sri Lanka: Ethnic Fraticide and the Dismantling of Democracy. Chicago: Univ. Chicago Press

Tarrow S. 1988. National politics and collective action: recent theory and research in Western Europe and the United States. Annu. Rev. Sociol. 14:421-40

Teitelbaum MS. 1992. Advocacy, ambivalence, ambiguity: immigration policies and pros- pects in the United States. Proc. Am. Philos. Soc. 136:208-25

Thompson DL, Ronen D, eds. 1986. Ethnicity, Politics, and Development. Boulder, Colo: Lynne Rienner. 222 pp.

Thompson JLP, Quets GA. 1990. Genocide and

social conflict: a partial theory and a com- parison. See Kriesberg 1990, pp. 245-66

Thompson RJ, Rudolph JR Jr. 1986. See Thompson & Ronen 1986, pp. 25-63

Tilly C. 1986. The Contentious French: Four Centuries of Popular Struggle. Cambridge, Mass: Harvard Univ. Press. 456 pp.

Tilly C. 1992. Coercion, Capital, and European States, AD 990-1992. Cambridge, MassJOx- ford, UK: Blackwell. Rev. ed.

Tiryakian EA, Rogowski R, eds. 1985. New Nationalisms of the Developed West: Toward Explanation. Boston: Allen & Unwin. 394 Pp.

Topor G. 1992. Review of JG Kellas (1991). The Politics of Nationalism and Ethnicity. J. Int. Aff 45:645-48

van den Berghe PL. 1981. The Ethnic Phenom- enon. New York/Oxford: Elsevier 301 pp.

van den Berghe PL. 1983. Class, race and ethnicity in Africa. Ethn. Racial Stud. 6:221- 36

van den Berghe PL, ed. 1990. State Violence and Ethnicity. Niwot: Univ. Press of Colo- rado

van der Dennen J, Falger V, eds. 1990. Sociobiology and Conflict: Evolutionary Perspectives on Competition, Cooperation, Violence and Warfare. London: Chapman & Hall

Vogel J. 1991. Culture, politics, and national identity in Cote d'Ivoire. Soc. Res. 58:439- 56

Waters MC. 1990. Ethnic Options: Choosing Identities in America. Berkeley: Univ. Calif. Press. 197 pp.

Watkins SC. 1991. From Provinces into Na- tions: Demographic Integration in Western Europe, 1870-1960. Princeton, NJ: Prince- ton Univ. Press. 235 pp.

Weber E. 1979. Peasants into Frenchmen: The Modernization of Rural France, 1870-1914. London: Chatto & Windus

Weede E. 1985. Income inequality and political violence reconsidered. Am. Sociol. Rev. 51: 438-41

Welch CE Jr. 1986. Ethnic factors in African armies. Ethn. Racial Stud. 9:321-33

Wickham-Crowley TP. 1992. Guerrillas and Revolution in Latin America: A Comparative Study of Insurgents and Regimes Since 1956. Princeton: Princeton Univ. Press. 424 pp.

Williams RM Jr. 1947. The Reduction of Inter- group Tensions. New York: Soc. Sci. Res. Council

Williams RM Jr. 1965. Social change and social conflict: race relations in the United States, 1944-1964. Sociol. Inquiry 35:8-25

Williams RM Jr. 1975. Race and ethnic rela- tions. Annu. Rev. Sociol. 1:125-64

Williams RM Jr. 1977. Mutual Accommoda- tion: Ethnic Conflict and Cooperation. Min- neapolis: Univ. Minn. Press. 458 pp.

ETHNIC CONFLICTS 79

Williams RM Jr. 1978. Competing models of multiethnic and multiracial societies: an ap- praisal of possibilities. In Major Social Is- sues: A Multidisciplinary View, ed. JM Yinger, SJ Cutler, pp. 50-65. New York: Free Press. 575 pp.

Willie CV, Cunnigen D. 1981. Black students in higher education: a review of studies. Annu. Rev. Sociol. 7:177-98

Yinger JM. 1985. Ethnicity. Annu. Rev. Sociol. 11:15140

Zaslavsky V. 1992. Nationalism and demo- cratic transition in postcommunist societies. Daedalus 121:97-121

Zimmerman E. 1980. Macro-comparative re- search on political protest. See Gurr 1980, pp. 167-237