Conflict and Peace Building in Divided Societies

Responses to Ethnic Violence

Anthony Oberschall

• Publisher: Routledge
• Year: 2007
• ISBN:
  9781134128136
• Edition: 1
• Topic: Politics & International Relations
• Subtopic: Arms Control
• Index: Politics & International Relations

This groundbreaking book provides an integrated account of ethnic, nationality and sectarian conflicts in the contemporary world including the role of collective myths, the mass media and the ethnification of identities as contributors to ethnic conflicts and wars. In addition to many examples from the last two decades, Oberschall provides a comprehensive overview of the conflict and peace processes in Bosnia, Northern Ireland and the Middle East.

Oberschall analyzes:

• peace building through constitutional design

• 
  
  

• power sharing governance

• 
  
  

• disarming combatants, post-accord security and refugee return

• 
  
  

• transitional justice (truth and reconciliation commissions, war crimes tribunals)

• 
  
  

• economic and social reconstruction in a multiethnic society.

• 
  
  

In addition to many examples from the last two decades, Oberschall provides a comprehensive overview of the conflict and peace processes for Bosnia, Northern Ireland, and Israel-Palestinians. He argues that insurgency creates contentious issues over and above the original root causes of the conflict, that the internal divisions within the adversaries trigger conflicts that jeopardize peace processes, and that security and rebuilding a failed state are a precondition for lasting peace and a democratic polity.

This book will be essential reading for undergraduate and postgraduate students, researchers and academics interested in the fields of peace studies, war and conflict studies, ethnic studies and political sociology.

1 The dynamics of ethnic conflict

Ethnic conflict

In a wide-ranging review of US foreign policy for the twenty-first century, David Callahan writes that:

Ethnic conflict and quests for self-determination around the world are likely to be the most important factors . . . in the next decades . . . this phenomenon should not be seen as separate from other global problems such as terrorism, failed states, rivalry among the great powers, access to natural resources, and clashes between the modern and the traditional, or between the rich and the poor.

(2002: 02)

Most states are multiethnic. “Nation state” is a frequently misused or loosely used term for states in which a single nationality is dominant. A survey of 132 entities considered states as of 1971 showed that only 12 states (9 percent) can be justifiably termed “nation states” (i.e. with a very small minority distinct from the dominant nationality, as e.g. Japan). In 30 percent the largest nationality accounts for less than half the population, and in another 23 percent the largest is between half and three-quarters (O’Connor 1978). Twenty years later, there were about 180 states and “fewer than twenty are ethnically homogeneous, in the sense that ethnic minorities account for less than 5 percent of the population” (Welsh 1993: 45). What state an ethnic group is located in, whether it is large rather than small in relation to other groups, whether it is a majority or minority, and whether it is integrated or excluded, dominating or subordinate in mainstream institutions, are complex results of empire building and breakup, war, conquest, migration, forced expulsion, genocide, epidemics and environmental degradation, nationalist movements, state formation, assimilation, discrimination, and encapsulation, i.e. the ebb and flow of history. When the relationships of the dominant group to ethnic minorities are hostile rather than cooperative, the society can be described for short as “divided.” What determines whether the relations between these groups will be hostile rather than cooperative?

Of an estimated 700 to 800 minority groups of substantial size worldwide, the Minorities at Risk (MAR) project (Quinn and Gurr 2003) identified 285 that were politically active at some time since the 1950s. Of these, about half pursued self-determination goals such as collective rights, political autonomy, or an independent state of their own. Of the 148 that pursued self-determination goals, 78 engaged in conventional, non-violent politics, but 70 have waged some form of armed struggle at one time or another in the past 50 years. Ethnic armed movements stubbornly resist settlement. Of the 70, as of 2003, six had deescalated to conventional or militant but non-violent politics; 26 hostilities had ceased but a peace agreement remained contested (these two together are termed “contained” conflict); 25 had varying modes of violent conflict; and in only 12 cases was the conflict “settled,” i.e. there existed an uncontested peace agreement granting regional autonomy to the ethnic challenger or an independent state that is internationally recognized. The five independent states were Bangladesh in 1971, Slovenia and Croatia in 1991, Eritrea in 1993, and East Timor in 2002.

Quantitative analysts of civil wars and major civil strife of all types since World War II, including ethnic conflict, found that various measures of ethnic, linguistic, and religious divisions in a country are correlated with the risk of major violent conflict. The strength of these relationships is disputed because of differences over operationalizing civil wars, ethnic factionalism, and religious divisions. Nevertheless, after a comprehensive review of measurement issues, Nicholas Sambanis wrote that:

I have shown that there is a very strong relationship between ethnic heterogeneity [in a state] and an aggregate indicator of armed conflict and much less so with civil war. To the extent that violence escalates from minor to higher levels, we should find ethnic factionalism to be significant in a dynamic model of violence escalation.

(2004: 848)

Doyle and Sambanis (2000) found that 64 percent of 124 post-World War II civil wars are ethnic and/or religious; James Fearon (2003: 15) classifies 55 percent of civil wars as “ethnic” and another 17 percent as having an ethnic component. Roy Licklider (1995) found that 69 percent of civil wars are “identity based,” i.e. ethnic, religious, or both. Depending on various ways of defining the pivotal variables, several researchers find that between 15 percent and 30 percent only of all civil wars are settled in negotiations, the vast majority ending with unilateral military victory, and that is especially true for ethnic civil wars (King 1997; Walter 1997). Doyle and Sambanis (2000: 786) found that civil war settlements have a 65 percent failure rate, i.e. there is resumption of armed conflict after two or more years of no hostility, and write that “wars with an ethnic or religious overtone are less likely to be resolved” than other civil wars. Licklider (1995) agrees that identity civil wars settled by negotiation are less likely to be stable than those settled by military victory. Fearon (2003: 15) finds that ethnic wars last longer than other civil wars, especially when a minority fights statesponsored control of indigenous resources by immigration.

From this body of quantitative research one must conclude that, although a majority of ethnic groups have conventional, non-violent, albeit strained, political relations with their governments and other groups, there is a substantial risk that violent forms of contention will occur and that these will escalate to insurgencies and civil war. Once armed fighting starts, ethnic conflict resists negotiated settlement, lasts longer, and has a greater likelihood of renewed violence. It is therefore important to explain why ethnic groups persist, what makes them susceptible to conflict with dominant groups and one another, and what circumstances cause change from nonviolent to violent modes of conflict.

Ethnic groups

“Ethnic group” denotes a large aggregate of people who have a self-defined name, believe they share a common descent, have common historical memories and elements of shared culture (such as religion and language), and have an attachment (even if only historical and sentimental) to a specific territory (Kaufman 2001: 16). Ethnic groups tend to have solidarity and we-feeling; their members experience something of “ourselves” in each other (Horowitz 1985: 155). “Nationality” refers to a large group of people having a common and distinguishing racial, linguistic, and cultural background, and forming a constituent element of a larger group (Webster’s Third New International Dictionary 1981). The two terms, “ethnic group” and “nationality,” overlap; social scientists prefer “ethnic group” because it is more richly descriptive and includes religion as a possible focus of selfdefined identity, and because some ethnic groups are marginal rather than a constituent element of a larger group. Because “ethno-national” is more cumbersome than “ethnic,” I will simply use “ethnic.” Religious groups may not believe they share a common descent, but they share beliefs about the sacred, rituals, and organizations that set them apart from other religious groups, and they have an attachment to specific holy places. In some cases, like the designation Protestant and Catholic in Northern Ireland, the religious categories mask the underlying nationality division and conflict, which is Unionist and Nationalist, i.e. those who want the province to remain in the United Kingdom and those who want it to unite with the Republic of Ireland. Unionists and Nationalist have no conflict over the freedom of religion or matters of religious belief and doctrine. In some other cases where religion divides along lines of ethnicity or nationality, as with Bosniaks, Albanians, Muslims, Serbs, and Croats in the former Yugoslavia, or Greek and Turkish Cypriots, there may have been some peripheral religious issues on top of fundamental territorial sovereignty and minority self-determination conflicts, but freedom of worship was not contested by the adversaries. In these conflicts, religious differences carry the freight of past hostile group relations, and prejudice and justification for superior status are expressed in religious as well as ethnic labels and pejoratives. In still other conflicts – Chechen versus Russian, or Sunni and Shia in Iraq, or the military government and the Islamists in Algeria – religious division plays a more central role, but the fundamental religious issue is between a secular and a religious mode of organizing the state and the laws. Again, because “ethno-religious” is a cumbersome term, I will use “ethnic” unless the context demands more specificity.

Ethnic groups are not a fact of nature, like species, and cannot be defined by objective physiological attributes. They have been socially constructed throughout history, as the French historian Ernest Renan pointed out in his path-breaking lecture on the origin of nations at the Sorbonne in 1882, “Qu’est ce qu’une nation?” (What is a nation?):

A Frenchman is neither Gaul, nor a Frank, nor a Burgundian. Rather he is what has emerged from the cauldron in which, presided over by the Kings of France, the most diverse elements have together been simmering . . . An Englishman is . . . neither the Briton of Julius Caesar’s time, the Anglo-Saxon of Hengist’s time, nor the Dane of Canut’s time, nor the Norman of William the Conqueror’s time, it is rather the result of all these . . . Is Germany an exception? . . . That is a complete illusion. The whole of the South was once Gallic; the whole of the east . . . Slavic . . . What is the defining feature of these states? It is the fusion of their component populations.

([1882] 1996: 42–6)

What is true for nations is also true for ethnic groups. But just because nations and ethnic groups are a product of historical and social construction does not make them arbitrary or less than real in the hearts and minds of their members and for other groups. Belief systems transform what is socially constructed by human action into eternal verities by invoking God’s will and design, ancestors, history, biology, and inheritance. Such beliefs tend to legitimize ethnic differences, separation, and exclusion.

Ethnics derive benefits from group membership. Ethnic identities, solidarities, and boundaries persist because they supply a steady stream of benefits to group members. Research by Tajfel and others on social identity and the sources of group boundaries found that “a positive social identity is achieved by comparing one’s own group with other groups to establish a positively valued psychological distinctiveness for the in-group vis-à-vis the outgroup” (Hewstone and Cairns 2001: 321). Self-esteem, a sense of belonging, and an identity (who am I?) are not private acts but produced and validated in group interactions. Individuals seek support for their beliefs and values, especially if these cannot be empirically verified, like religious beliefs and political values, and get it with social validation in like-minded groups. That in turn increases loyalty to the support group (Hinde 1997: 10). Preference for the group one belongs to is caused by these benefits of membership and is not motivated by hostility towards other groups.

Why would a group that confers such psychological benefits be an ethnic group rather than some other group? Why would ethnic groups have an in-group preference for neighboring, for schooling, and for associations of all sorts with their fellow ethnics and, by extension, transacting selectively within rather than across ethnic boundaries? Ethnic groups are composed of people who share more in common with one another, on average, than with other ethnic groups and the population at large. They share (or believe they do) ancestors and history, often language and religion, salient values and beliefs, and lifestyles, and they believe that these attributes distinguish them from other groups. There are benefits, not just emotional we-feeling, from such commonalities in transacting with others.

The theory of collective action (Coleman 1990) has demonstrated the advantages of transacting with partners who expect a long-term relationship over an episodic one, who have much rather than little information about one another, who share overlapping social networks rather than weak links, and who count on third-party assistance for social control, and reciprocity rather than absence of obligation. These attributes of transacting make for cooperation rather than opportunism, long-term investment in relationships and the nurturing of social capital rather than a short-term instrumental stake, open-ended rather than instant reciprocity, and personalized rather than stereotyped interaction. In-group relationships build on trust and also create more trust than stranger interactions. Many ethnic trading communities throughout the world, called middlemen minorities, owe their success to group business organization and a distinctive in-group morality. A classic anthropological study of a Hausa trading community in Ibadan attributes its success to the sectarian boundaries it erected for checking assimilation into Yoruba society, and sums it up thus: “ethnicity . . . is not an archaic survival arrangement carried over into the present by conservative people . . . ethnic groups are . . . interest groupings” (Cohen 1969: 190, 192).

Because of in-group preferences, establishing and maintaining shared institutions with outsiders can be problematic. Thomas Schelling (1984) has demonstrated how micro preferences, i.e. slight in-group preferences for fellow ethnics in mixed institutions, can have unexpected macro consequences in the long run, i.e. totally segregated institutions like neighborhoods and schools. Evidence from studies of black–white residential segregation and school integration in the US support that analysis (Giles 1978; “New Survey” 1979). Schelling’s insight is that, without discrimination and with a fair amount of acceptance for group mixing, in-group preference makes stability of mixed institutions problematic.

What rapidly unravels mixed institutions is not in-group preference alone but communal violence and the security concerns it engenders. Meron Benvenisti (1986: 86–8) recalls that during his youth there was no residential segregation in Jerusalem. Then came the Arab riots of 1929 and later the Arab revolt of 1936–9 against the British mandate and the Jewish presence. Jews fled to Jewish districts and the result was almost total segregation: “segregation is a mechanism for coping with physical threat . . . for preserving group identity from alien influences, and to conserve cultural heritage and life-styles.” Studies of increasing Catholic–Protestant segregation in Belfast from 1840 on show that the sectarian riots of 1857 and 1886, the civil strife in the early 1920s, and the sectarian rioting followed by paramilitary violence in 1969 led to sharp, discrete jumps in residential segregation (Boal 2002). Nor did public policy in housing reverse the trend. The Northern Ireland Housing Executive had a policy of color-blind allocation, but the residents “reproduced Belfast sectarian geography by self-selection” into green and orange housing estates (“green” for Irish and “orange” for British) “such that public sector housing became more segregated than private sector housing” (Bollens 1998: 97–8). In a wide-ranging analysis of communal riots, Donald Horowitz concluded that they reduce ethnic heterogeneity and lead to separation and sharper ethnic boundaries (2001: 424).

Ethnic groups and collective goods

Well-being depends not only on person-to-person transactions but on collective goods. Mancur Olson (1976) reflected on what would be an optimum social unit whose members would satisfy each other’s wants and preferences through interpersonal exchange and sharing collective goods. The answer turns out to be a social unit composed of people who have similar preferences for collective goods but are diverse in their preferences for individual goods and in their talents and capacities. Consensus on collective goods (e.g. language, group identity, collective symbols of shared values and group dignity as expressed in the celebration of holidays, the naming of streets and public monuments, the role of religion in public life) reduces contentious conflicts in politics and makes for low-cost governance. Variety in private tastes, resources, and talents allows for beneficial exchange in fulfilling individual wants. Ethnic groups satisfy the two Olson dimensions of a viable social unit that deliver well-being to members.

Social movement theory and research have shown that ethnic groups are capable and willing to undertake collective action in defense of their interests and way of life. Ethnic groups tend to have a viable communal organization with dense interpersonal networking and cultural, social, and religious associations embedded in the ethnic community. Political mobilization of ethnic groups does not start from scratch. Ethnic activists redirect a preexisting robust ethnic infrastructure of associations and networks to political pursuits, adapting ethnic symbols and loyalties to political goals. This mode of political mobilization bypasses cumbersome recruitment of individuals a few at a time by recruiting pre-existing groups, termed “bloc mobilization,” and overcomes the cost of organization by federation of the blocs into an overarching organization. It is a rapid, low-cost mode of mobilization (Oberschall 1993: chapter 1). Because the blocs are small face-to-face groups, solidarity and internal networking overcome the usual free-rider tendencies in large groups against collective good attainment. Last but not least, solidarity and we-feeling create a “multiplier effect” for grievances: wrongs and injustices to some members of the ethnic group are framed and experienced by others as wrongs and as injustice.

The conclusion of a substantial body of social science is that ethnic groups are viable social units that benefit their members, that they provide advantages for maintaining boundaries and separation with other groups, and that they have high capacity for collective action. Avoidance of other groups in “live and let live” arrangements does not necessarily follow from ethnicity, but tends to be the legacy from past inter-ethnic violence. What might weaken the viability of ethnic groups?

Ethnic groups and social change

In conventional wisdom, social change erodes ethnic groups and boundaries. Social changes produce internal differentiation from migration, work, and education, and ultimately intermarriage between ethnic groups. In the labor force, one has no choice of work mate, but has to accept the employers’ assignments. In industries, mines, offices, laboratories, sports teams, armies, and other pursuits, members of various ethnic groups learn cooperation and develop shared interests across ethnic groups, such as trade union or professional association. In time, the more affluent and the less affluent, the more and the less educated, and the religious-minded and the secular will form identities and relationships that breach ethnic encapsulation. In Olson’s analysis, internal differentiation produces new wants and preferences for collective goods, e.g. collective bargaining for all workers in an industry. Individual wants, goods, services, and transaction partners can be satisfied just as well or better by associating with outsiders as within one’s ethnic group. When ethnic barriers weaken, the pool for associates, friends, and spouses expands. Social psychologists have shown that integration works best when the groups interact as equals and when the contribution of each group is necessary to achieve goals that benefit all (Sherif 1966).

Ethnic majorities and minorities typically resulted from conquest and colonization, labor migration, trading opportunities for immigrants, the diffusion of religions, wars and forced population movements, the formation of states, and the drawing and redrawing of international boundaries. In these processes, relationships of domination and subordination, of superior to inferior, prevailed over ethnic equality. Charles Wagley and Marvin Harris put it thus:

Numerous underprivileged groups of people, ineptly called minorities . . . are singled out by their societies in which they reside and . . . are subjected to economic exploitation, segregation, and discrimination . . . [they] are disliked and ridiculed because they speak a different tongue, practice a different religion, or because their skin is a different color, their hair a different texture, or simply because their ancestors emigrated from a different country.

(1958: 1)

Unless public policy is opposed, ethnic relations of domination and subordination, privilege and exclusion, attraction and rejection in ethnic relations will reproduce along historical and cultural fault lines.

The pre-modern state built on the viability of ethnic groups and on ethnic and religious separation. An ancient principle of empires was that laws and rights were not territorial but group based, each group maintaining their separate institutions, though accepting the overall authority of the imperial rulers. Thus monogamy and polygamy (and more broadly different family law) could coexist in the same city, but practiced in separate religious communities, as in Indian cities for Hindus and Muslims, and in Jerusalem for Muslims and Jews. When empire, colonial rule, and the dynastic principle for state and government were successfully challenged with the democratic principle of government by the people and the right to selfdetermin...