Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 13 August 2020


Abstract and Keywords

: This article examines the role of individual preferences and actions in producing conflict and how conflict shapes preferences, identity, and actions. Using an analytical approach, it looks at the causes and consequences of interpersonal and group conflict, with particular emphasis on the microfoundations of conflict and peace. It also considers how conflict causes action and the formation of collective identity, much as collective identity gives rise to conflict and action. After reviewing some of the challenges raised by the study of conflict, the article discusses the link between interpersonal conflict and individual-level dynamics. It then explores ethnic and class conflict in relation to group-level dynamics, along with the endogenous dynamics of violent conflict in the context of civil war. Finally, it highlights the importance of alliance as a mechanism linking local cleavages into a conflict’s master cleavage.

Keywords: preferences, action, conflict, group conflict, peace, collective identity, interpersonal conflict, violent conflict, civil war, alliance


Consider Iraq between 2003 and 2008, the site of overlapping violent civil conflicts. There is no denying the centrality of the ‘sectarian’ division, pitting Shia against Sunni Iraqis. Nevertheless, this was hardly the only type of violence in Iraq. In the south, Shiite factions fought each other; in the west and around the capital Bagdhad, rival Sunni factions faced off; and in the north, Sunni Kurds clashed with Sunni Arabs. In fact, a multitude of political actors emerged since the 2003 US invasion, including insurgent groups of various persuasions, political parties, religious factions, local militias, semiautonomous military and police units, militarized tribes and subtribes, and criminal gangs. The political landscape is characterized by a high level of fragmentation, with many groups switching sides, most notably former Sunni insurgents who joined US forces. A journalist described a gloomy reality where ‘the lives of most Iraqis are dominated by a complex array of militias and gangs that are ruthlessly competing with one another, and whose motives for killing are more often economic or personal than religious or ideological’ (Packer 2007: 58).

Even ‘sectarian violence’ turns out to be far from an uncontested category. Journalists point to a variety of sordid motivations underlying this violence, including bad blood between neighbors or the desire to expropriate a neighbor’s house; they refer to ‘the very intimate nature of the war in Iraq—a war in which your enemies are often people you’ve known much of your life; in which your neighbors are often (p. 593) behind the crimes committed against you; in which every slight, every misdeed, every injustice is recorded and the desire for vengeance runs deep’ (Parker 2007).

Extreme fragmentation and confusion surrounding motives are hardly specific to Iraq. Many violent internal conflicts are characterized by a high level of complexity and fluidity, with a proliferation of actors and with motivations varying depending on the vantage point. Local and self-interested motivations at the micro level appear to coexist uneasily with abstract and general motivations at the macro level. In fact, violent internal conflict tends to give rise to two apparently contradictory observations.

On the one hand, much of the violence taking place in contexts characterized by intense political and social conflict is described as the outcome of abstract group-level dynamics (e.g. Horowitz 1985). In fact, group-level descriptions (e.g. ‘Serbs versus Albanians’; ‘Sunni versus Shia Iraqis or Lebanese’) inform the theoretical connection between violent action on the ground and its macro-level antecedents. This link is best encapsulated by the concept of polarization, which refers to the intensity of divisions between groups, ‘when a large number of conflict group members attach overwhelming importance to the issues at stake, or manifest strongly held antagonistic beliefs and emotions toward the opposing segment, or both’ (Nordlinger 1972: 9).

On the other hand, the magnification of one’s focus often shows specific acts of violence to be intimate, exercised between people who know each other personally—in many cases between peers. Violence in the context of civil war, more particularly, is frequently exercised among people who share everyday ties of social and spatial interaction, such as neighborhood or kinship. This explains the commonality of the term ‘fratricide’ and the frequent references to the biblical story of Abel and Cain. Furthermore, intimate violence often appears to be based on trivial everyday motivations rather than abstract enmity. In fact, there is a close connection between the intimate character of violence and the local setting within which it takes place. For example, Spencer (2000: 134) describes how the JVP (Janatha Vimukthi Peramuna: People’s Liberation Front) insurgency in Sri Lanka ‘permeated the capillary relations of everyday interaction: your political opponents would be neighbors usually, kin often, former friends sometimes.’

Are these two views, abstract enmity and intimate conflict, as contradictory as they appear to be? Are the insights derived from close observation due to an observer’s particular vantage point or do they carry broader theoretical implications? Is it possible to reconcile the intimate and local dimension of violence revealed by detailed observation with the focus on abstract group violence that informs most theorizing on this subject? Is it possible to connect individual-level and group-level actions? If yes, how?

In this chapter I argue in favor of an analytical approach to conflict that reconciles these apparently contradictory insights. This approach accepts that agency is located simultaneously at different levels of aggregation; it explains how conflict (p. 594) causes action and the formation of collective identities, much as collective identities give rise to conflict and action. Put otherwise, an analytical approach to the study of conflict demonstrates and incorporates the fundamentally interactive, bidirectional, or endogenous relation of conflict and action. Unpacking this relationship requires a clear understanding of the link between individual and group action in the production of conflict.

This chapter is divided in four main sections. In the first I review some of the challenges raised by the study of conflict, discuss popular theorizations, and outline key arguments. In the second section I turn to interpersonal conflict as a way to illustrate the application of an analytical approach to individual-level dynamics; in the third section I focus on ethnic and class conflict in order to discuss the application of this approach to group-level dynamics; and in the fourth section, I turn to civil war and discuss the links between individual-level and group-level logics.

25.1 Conflict

Conflict can be defined very broadly as an outcome of the ‘purposeful interaction among two or more parties in a competitive setting’ (Oberschall 1978: 291). In his influential analysis of conflict, Georg Simmel (1955: 14) described conflict as ‘one of the most vivid interactions’ among humans, whose causes could be traced to ‘dissociating’ factors such as ‘hate, envy, need, [and] desire.’ Although conflict has acquired a negative connotation, it is not perceived as necessarily dysfunctional. Simmel (1955) and Coser (1956) argued convincingly that conflict is functional from a social point of view, because it is an expression of the plural views and interests that exist in society, and its resolution allows the emergence of unity from conflicting multiplicity.

Conflict is as old as human history—and so the history of reflecting on conflict is also old. Simplifying what is an enormous literature, it is possible to discern a major divide between approaches stressing the role of groups, and arguments focusing primarily on individuals. This chapter argues against this dichotomy and makes the case for a theoretically and empirically explicit connection between individual and group action. Achieving such a connection allows us to grasp the fundamentally bidirectional or endogenous relation between conflict and action; it allows us to connect agency at different levels of aggregation, thus resolving the puzzle of contradictory motivations; and it finally allows us to grasp the ‘creative’ dimension of conflict which shapes individual and collective identities.

(p. 595) Conflict takes a wild multiplicity of forms, from disputes between spouses and neighbors, to wars between states. Because its forms are so varied, conflict has been divided into different types and studied accordingly. Perhaps the most common distinction is between violent and nonviolent conflict (Brubaker and Laitin 1998).

Nonviolent conflict covers, most obviously, politics as a general category. Harold Lasswell’s 1936 book carries a title that still serves as a popular definition: Politics: Who Gets What, When, How. In this vein, democracy is seen as a way of managing conflict; that is, of transforming what is potentially violent conflict into a nonviolent activity. Seen from this perspective, the entire discipline of political science is preoccupied with the study of political and social conflict and its management, with Hobbes’s Leviathan being an early, yet foundational, example.

Beyond the study of how institutions ‘tame’ politics by eliminating its potential for violence, a particularly fertile field of scholarly research on nonviolent conflict has been the study of ‘contentious action’—a term referring to nonviolent forms of collective protest such as mass demonstrations and strikes (McAdam, Tarrow, and Tilly 2001).

It would be fair to say that violent conflict has always exercised considerable fascination over researchers. Three types of violent conflict, in particular, have attracted sustained scholarly attention: criminal violence, riots and pogroms, and war in all its varied forms. ‘Deviant’ behavior has been the object of considerable study and the specialized field of criminology (Sampson and Laub 1992; Birkbeck and LaFree 1993). Research on riots and pogroms (often defined as ‘ethnic’) has taken off in recent years; in particular, studies of Hindu–Muslim riots in India by Varshney (2003) and Wilkinson (2006) have moved our understanding of this phenomenon onto a new plane.

Scale is the main distinguishing characteristic between criminal violence and ethnic riots, on the one hand, and war, on the other. Though violent criminal activity can range from individual (e.g. a single homicide) to collective (e.g. clashes between organized gangs or between gangs and state forces), it rarely reaches the level of collective mobilization usually associated with war. Likewise, fatalities caused by ethnic riots, while considerable, typically fail to reach major-war levels. War is usually defined as large-scale organized violence between political units (Levy 1983), a definition that applies primarily to international war. Internal or civil war, the most common form of armed conflict since 1945, has been defined as armed combat within the boundaries of a recognized sovereign entity between parties subject to a common authority at the outset of the hostilities (Kalyvas 2006). Operational definitions of both international and civil war rely on a fatality threshold—usually 1,000 battle-related fatalities (Singer and Small 1972).

Unsurprisingly, there has been considerable fragmentation in the study of conflict. Crime is studied by criminologists who are active in the discipline of sociology; international war has been the province of either historians or political scientists associated with the subfield of international relations; and civil war has been the (p. 596) province primarily of political-science scholars focusing on domestic politics, and hence the subfield of comparative politics.1 Additional distinctions cut through these divisions. For instance, anthropologists have focused on the local level and social psychologists on the individual level; the study of war onset has been differentiated from that of war termination; and so on. As a result, theories, methodologies, and data sources have diverged considerably, thus giving rise to a fragmented field of inquiry.

Underlining this fragmentation in the study of conflict is a fundamental methodological divide between two styles of research. One approach, dominant in sociology and political science, adopts a group-level perspective (a) taking groups as exogenously given and fixed and (b) deriving individual interactions directly from group interactions. This has been the standard approach in studies of class conflict (Dahrendorf 1959; Moore 1966), as well as ethnic conflict (Horowitz 1985; Petersen 2002). To use the example of present-day conflict in Iraq, these approaches begin by positing groups in conflict (Shia, Sunni, Kurd) and proceed to interpret individual actions as resulting from this overarching conflict. This type of analysis tends to misinterpret a great deal of the conflict and violence on the ground, missing how the violence shapes individual and collective identities and contributes to the escalation of conflict.

The alternative approach, dominant in social psychology and economics, adopts an individual-level perspective which assumes that collective-level conflict is an aggregation of individual-level tendencies or choices. Using evidence from laboratory experiments, social psychologists have argued that individuals have a natural tendency to aggregate in groups, which then become embroiled in conflicts (Tajfel and Turner[1979] 2004). In this view, conflicts emerge primarily from the stickiness of the individual association with these groups rather than past hostility or a conflict of interests. It is the omnipresent human tendency toward intergroup discrimination favoring the in-group (the so-called ‘in-group bias’) which is at the root of conflict. In contrast, economists tend to see groups as concatenations of individuals making rational choices based on their economic interests. Conflict emerges once similar individual interests, be they political or economic in nature (Downs 1957) or ethnic (Caselli and Coleman 2006), aggregate into collective ones.

Although psychological and economic approaches differ in their fundamental epistemology, they reach an understanding of group conflict that parallels that of sociological approaches. To return to the example of Iraq, psychological and economic approaches would interpret the current conflict in much the same vein as sociological ones, with the difference that they would provide a general theory for individual participation in group activities, something lacking from the latter. Hence, despite their epistemological differences, the dominant approaches to the study of conflict tend to neglect the dynamic relation between individuals and groups and, therefore, the potentially endogenous relation between conflict and action.

(p. 597) There is an alternative, however—one that specifies microfoundations and focuses on the process of interaction between individual and group action. Recent examples include the analysis by Fearon and Laitin (1996) of interethnic cooperation and the research of Fehr and Gachter (2000) on the role of norms of fairness and reciprocity for in-group cooperation. The former argue that intragroup disciplining of group members who misbehave toward outsiders explains the relative scarcity of intergroup conflict; the latter use laboratory experiments to show that individuals are willing to disregard considerable costs so as to discipline group members attempting to free-ride out of group collective action.

This recent research trend (which focuses on cooperation more than conflict) adopts an analytical strategy combining two elements. First is theoretical disaggregation or microfoundational analysis, following the seminal work of Thomas Schelling (1978); in Coleman’s (1990: 503) formulation: ‘Any theory of action requires a theory of the elementary actor. The elementary actor is the wellspring of action, no matter how complex are the structures through which action takes place.’ Second is empirical disaggregation, which entails the empirical testing of micromechanisms in order to show how individual and collective actors connect with each other.

To show the benefits of this analytical strategy on the study of conflict, I discuss three types of conflict: (a) interpersonal conflict, (b) ethnic and class conflict, and (c) civil wars. Each section is anchored in one key work: Roger Gould’s Collision of Wills, Adam Przeworski’s Capitalism and Social Democracy, and Stathis Kalyvas’s The Logic of Violence in Civil War. The objective is to illustrate diverse yet consistent strategies of theoretical and empirical disaggregation and to show how they help us come to grips with the endogenous dynamics that characterize individual and group conflict. Furthermore, I wish to show how these different types of conflict, usually treated in isolation, can benefit from a more integrated approach.

25.2 Interpersonal Conflict—Individual-Level Dynamics

In Collision of Wills Roger Gould endeavors to build a theory of interpersonal conflict and specify its links to group conflict. The originality of his undertaking is straightforward: interpersonal conflict has generally been studied as a self-contained topic. Conversely, large-scale social conflicts, articulated along macro cleavages and relating to a key political dimension such as class or ethnicity, have generally not been empirically connected to individual dynamics.2

(p. 598) Gould focuses on dynamics: his central motivating puzzle is the transformation of conflict from individual to collective and from nonviolent to violent—how to explain the fact that some interpersonal disputes escalate into violence despite arising from trivial motives such as minor debts, derisive remarks, or insults. Indeed, such violence is often described as ‘senseless.’ In other words, the goal is to account for the enormous gap between the trivial character of the dispute and the violence of the outcome. In fact, considerable research in the field of criminology suggests that criminal homicide tends to be intimate; a large part of nonpredatory common murder implicates relatives, friends, or at least acquaintances. Furthermore, the relationship between victims and assailants tends to be horizontal: people often kill their mates, friends, and acquaintances rather than their bosses (Black 1976; Katz 1988; Black 1993; Decker 1993).3

The answer to this puzzle lies in a specific mechanism: repeated interaction between individuals generates more or less formal social ranking; yet these emerging hierarchies can be highly consequential. For instance, honor is a reflection of informal, yet powerful, hierarchies. Conflict emerges when these hierarchies are challenged; and they are challenged when they appear to be ambiguous—either because previously well-established relations between individuals have changed or because new definitions of social status have emerged. The main implication is that conflict is likely to erupt among peers—people whose social status is comparable—rather than among people who are located on the opposite ends of the social spectrum. In short, Gould predicts that closeness and similarity, rather than distance and difference, are at the root of interpersonal violent conflict. This prediction challenges the social-psychological literature.

In turn, a number of testable implications are derived from this empirical prediction. They include the correlation between the outbreak of violent conflict and (a) the symmetry of relationships, (b) the ambiguity of hierarchy, and (c) (more remotely) political instability. The empirical tests take place at multiple levels of aggregation and are based on several distinct bodies of data, including a statistical analysis of homicides in two American and two Indian districts. The main causal variables are the presence of verbal arguments prior to the eruption of violence and the ambiguity of the hierarchical relations between perpetrator and victim. Gould finds that where hierarchy is ambiguous verbal disputes are more likely to escalate into homicides.

The next step consists in exploring the degree to which this argument can be generalized to group conflict. Gould reasons through analogy: groups relate to each other in the same fashion as individuals do. The analog of individual symmetry is group solidarity; that is, the willingness of individuals to sacrifice something for the group’s well-being. When solidarity within the group appears to be in question, a rival group is likely to challenge it so as to move up the hierarchical ladder. This argument is ingeniously tested with data from nineteenth-century Corsica, an island society characterized by two dimensions that rarely overlap: on the one hand, (p. 599) extensive premodern practices of blood feud or vendetta and, on the other hand, a modern police and judicial bureaucracy that were able to collect reliable data. Gould asks whether interpersonal violence always escalates into group violence, as one would expect in a blood-feuding society. Surprisingly, he finds that only a small proportion of interpersonal violent confrontations escalate into group violence. However, those that escalate into group violence do exhibit characteristics that are consistent with his theoretical expectations. An interesting and important empirical finding is that of all cases of group violence studied, only one-fifth turned out to have escalated from interpersonal violence. This finding points to the heterogeneity of conflict: some conflicts begin at the group level and then escalate among individuals while others begin among individuals and escalate up to the group level.

Though Gould’s account is not rationalist, it is analytical. Individuals are the main building block and group interactions are predicated on individual choices which are embedded in social relationships. Intergroup violence is unpacked and intragroup dynamics are connected to intergroup conflict, an insight taken successfully by recent research in different settings (e.g. Lawrence 2007). However, Gould follows the sociological tradition that takes groups to be exogenously given and fixed. Individual membership is not a choice, and the scaling from individual-level dynamics to group-level dynamics remains fuzzy, which is why the argument’s scope conditions exclude large-scale violence.

In short, Gould supplies an analytical account of interpersonal conflict and its escalation into violence based on sound empirical microfoundations, yet he limits his analysis to individuals and small groups. A final question emerges. Is it possible to expand this type of analysis to take into account large-group conflict, while also moving away from the assumption that groups are stable and fixed? Is it possible to build a theoretically and empirically sound micro-level analysis of conflict that generalizes to macro-level conflict?

25.3 Ethnic and Class Conflict—Group-Level Dynamics

Moving up the ladder from interpersonal to group conflict we encounter two important streams of research, one focusing on ethnic conflict and the other on class conflict. Let me begin with a discussion of ethnic conflict. Initially developed in the aftermath of decolonization, this literature is undergoing a renaissance following the end of the cold war and the realization that ethnic conflict is, after all, quite alive. It is impossible to summarize this very extensive body of work. Nevertheless, one can point to a number of important traditions.

(p. 600) One tradition has focused on the development of nationalism and national identity in Europe and has emphasized a variety of processes, most notably modernization and urbanization (Deutsch 1953) and urban/rural conflict in multiethnic empires (Gellner 1983). Another tradition has looked into the postcolonial world (Horowitz 1985) and, unlike the former, has approached ethnic identity as given rather than constructed. Ethnic conflict emerges either in the process of construction of ethnic and national identities or in the process of competition between groups already formed in pursuit of resources, political power, or status.

With few exceptions, this literature has focused on group-level dynamics (e.g. Petersen 2002; Varshney 2003; Wilkinson 2006). Even those trying to unpack groups have primarily looked less at followers and more at leaders or ‘elites,’ often framed as ‘political entrepreneurs’ who are stirring ethnic hatred and conflict, either in pursuit of personal political power (Figueiredo and Weingast 1999) or because they misinterpret the intentions of their ethnic rivals under conditions of heightened uncertainty, in what is known as a ‘security dilemma’ (Posen 1993). In these accounts, individuals tend to be bundled into an undifferentiated mass that responds to fear by engaging in violence. As pointed out above, Fearon and Laitin (1996) focus on individuals by reversing the standard question of ethnic conflict and asking why ethnic conflict is so rare (and ethnic peace so frequent) compared to its actual possibility given the number of extant ethnic groups. They argue that a powerful mechanism of ‘in-group policing’ exists, whereby the group punishes its members who misbehave toward members of another group; as a result cooperation tends to prevail over defection.

Class conflict occupied the center of the field of political sociology, perhaps until the end of the cold war. Influenced by either the historical analysis of cleavages which relates the type of salient political division in a country to the historical process of state- and nation-building (Lipset and Rokkan 1967), or the Marxist perspective of class struggle as a reflection of the unequal distribution of the means of production, combined with an analysis of class coalitions (Moore 1966), this body of research sought to explain the variation in intensity of class conflict across societies, as well as its outcomes. The focus was squarely placed on groups; individuals were assumed to belong to groups based on their ‘objective’ economic interests.

A key concept in explaining the emergence of group conflict, whether ethnic, class, religious, or ideological, is polarization, which comes with a variety of adjectives: it can be of an ideological, religious, class, or ethnic nature. Polarization entails an interaction between groups that simultaneously display high internal homogeneity and high external heterogeneity (Esteban and Ray 1994). Populations clustered around a small number of distant but equally large poles, this intuition goes, are likely to clash. The underlining mechanism is dislike so intense as to cancel even fraternal ties, imagined or real. Groups thus signaled by collective identities produce conflict. This is what Buruma (2002: 12) has in mind when he writes of (p. 601) ‘what gets the blood boiling, what makes people do unspeakable things to their neighbors. It is the fuel used by agitators to set whole countries on fire.’ Although polarized conflicts entail issues of material distribution, they can escalate so as to be ‘no longer over specific gains or losses but over conceptions of moral right and over the interpretation of history and human destiny’ (Lipset and Rokkan 1967: 11). They can turn, in other words, into ‘the kind of intense and divisive politics one may refer to by the name of absolute politics’ (Perez-Díaz 1993: 6). Seen from this perspective, conflict constitutes the very essence of politics. The political, as Schmitt (1976) argued, is the most extreme antagonism pitting one fighting collectivity of people against another; politics is but the reflection of the fundamental distinction between friend and enemy. Hence, war is just the natural expression of enmity; as Lenin pointed out, civil war in a class-based society is but an extension of the class struggle in it (in Martin 1995: 61); and because, as Bobbio (1992: 303) observed, the relationship between a just war and a just enemy is inverse, i.e. a war is just when the enemy is unjust, unjust enemies need no mercy. As a result, the fundamental insight of this argument is that violence is secreted from deep divisions and that barbaric violence is the inevitable result of polarized conflict which constitutes the very essence of politics.

Though plausible, this argument faces four important challenges. First, and most obviously, it wildly overpredicts violent conflict (Fearon and Laitin 1996). Second, it entails a kind of backward reasoning, from on-the-ground violence to the factors that are believed to have produced it in the first place. Linking polarization between groups on the one hand and violence on the other implies an underlying causal claim that both predicts the relevant perpetrator–victim dyad and assigns a motive to the violence. For example, the relevant dyad in a class war includes the owners of capital and the owners of labor. In an ethnic war this dyad includes members of different ethnic groups. The argument follows a two-step logic. First a group is targeted because of its position on the relevant cleavage dimension, and subsequently individuals belonging to this group are victimized because of their membership in this group. Hence, this argument explains simultaneously group conflict and interpersonal violence. The problem is that this causal link is generally assumed rather than demonstrated empirically. We observe a specific action, e.g. an Iraqi Shia victimizing an Iraqi Sunni, and conclude that these individuals are fully intercheangable (i.e. they are irrelevant qua individuals), and that this particular action is the outcome of an abstract group-level conflict—in this particular instance of sectarian conflict. Third, this argument carries an important empirical implication; namely, that the highest form of domestic political conflict, civil war, is caused by deep group divisions. However, there exists a significant body of empirical research suggesting that high levels of social, religious, or ethnic polarization fail to explain the outbreak of civil war; they appear to be neither a necessary nor a sufficient condition (e.g. Laitin 2001; Collier et al. 2003). Furthermore, the absence of a good indicator for polarization has concealed the fact (p. 602) that, as empirical evidence suggests, even in societies that are deeply polarized, only a minority of the population can really be described as holding tightly to one pole or the other; the majority tends to remain either weakly committed or uncommitted, part of a ‘grey zone’ between the two poles (Malefakis 1996: 26; Seidman 2002). When more systematic data are available, as in Yugoslavia, the link between prewar group-level polarization and war appears tenuous. From a cross-sectional perspective, there was an inverse relationship between prewar polarization and civil war, with Bosnia scoring very high on interethnic tolerance (Hodson, Sekulic, and Massey 1994); from a temporal perspective, there was little change in mass indicators of nationalist attitudes in Croatia from 1984 to 1989, suggesting that polarization increased just before and immediately after the civil war erupted (Sekulic 2005). Fourth, this argument implies that conflicts with deep pre-existing cleavages should be, ceteris paribus, more barbaric than civil wars motivated by more shallow cleavages. Likewise, variation in levels of violence within the same civil war should covary with the depth of cleavages. Existing empirical evidence points to an absence of correlation between deeply divided societies and highly intense civil wars—or between subnational variation of deep division and intensity of violence (Bosch Sánchez 1983; Casanova 1985: 59; Ortiz Sarmiento 1990: 22; Wood 2003). At the same time, recent and more rigorous empirical research uncovers a relation between high levels of prewar electoral competition and violence (but no relation between domination and subjection and violence) (Balcells 2007).

To be sure, rigorous empirical research on these issues is still in its infancy. Nevertheless, it is suggestive, especially when considered in light of the following methodological critique. Violence may be erroneously linked to prewar polarization via a variety of inference biases, including the extrapolation from the aggregate to the individual level and the privileging of target information over base-rate information. For example, Boudon (1988) has shown that even in a homogeneous society of equals it is possible to generate processes of competition (and hence violence) which would on the aggregate level appear as having been generated by deep cleavages. Likewise it has been pointed out that competition effects between groups may be merely by-products of a selection bias: even in a world where ethnicity plays no role whatsoever in defining either the likely interactions among individuals or the proclivity of individuals to engage in violence, we would still see significant violence, wrongly perceived as resulting from ethnic competition (Dion 1997).

Among the efforts to recast the study of class conflict away from unquestioned group-level assumptions, Adam Przeworski’s Capitalism and Social Democracy stands out. Przeworski sought to specify the microfoundations of class conflict in nineteenth-century Europe; he promoted an ambitious agenda, also pioneered by Elster (1985) and Roemer (1986), which consisted in reconstructing Marxist theory using modern analytical tools and in a way consistent with advances in microeconomic theory and the principles of methodological individualism.

(p. 603) Influenced by the work of the Italian thinker Antonio Gramsci, Przeworski departs from many of the standard assumptions underlying classic Marxist theory. He challenges the premise that individual identities, individual interests, and individual actions4 could be derived from ‘objective’ group interests—in this case from class. He thus elaborates a ‘constructivist’ perspective avant la lettre (and in a field other than ethnicity), by arguing that individuals have multiple options when it comes to social and political identities (1985: 100):

[T]he division of a society into classes does not necessarily result in the organization of politics in terms of class. Nor is the experience of class the only one which is objective. If ‘objective’ means experience that is inherited by individuals and is independent of individual will, then being a Catholic today in Italy is an objective experience, as is being a Black in the United States, or a woman in France. The people who perpetuate their existence by selling their capacity to work for a wage are also men or women, Catholics or Protestants, Northerners or Southerners. They are also consumers, taxpayers, parents and city-dwellers. They may become mobilized into politics as workers, but they may also become Catholic workers, Catholics, or Bavarian Catholics.

In turn, these identity options are likely to shape an individual’s self-perception of what her interests are and, therefore, her willingness to mobilize and participate in political and social conflicts. In other words, individual choices help determine the likelihood of conflict. But, then, what helps determine these individual choices in the first place? Przeworski demonstrates that rather than deriving from production relations and the class structure, these choices are shaped by the organizational choices of political parties. It is the mobilizing of parties that creates and sustains individual identities and hence both their perceptions of what their interests are as well as their subsequent actions. This is not the classic Downsian view of parties as vote-maximizing political machines that seek to position themselves as close to the median voter as possible, but the Gramscian view of parties as engaged in the construction of ‘hegemony’; that is, actively molding the perceptions and consciousness of the masses—a point that resonates with the European political history of the turn of the nineteenth and twentieth centuries. Insofar as mobilization raises the likelihood of conflict, this perspective points to a view whereby political-party activity was at the root of conflict not just in the traditional view of party competition but also in the deeper process of cleavage and identity construction and transformation. Hence a key empirical question: How did the young European democracies avert mass violent conflict?

Part of the answer is obviously related to structure: The strategic choices of political parties are constrained by structural conditions and their strategic implications, that is the ‘real conflicts of interest and values that divide society at any particular moment’ (Przeworski 1985: 129), and the strategies of their rivals. A less obvious part of the answer points to the role of democratic institutions within which political parties mobilized voters.

(p. 604) Three implications follow. First, initial choices over party strategy, such as whether to participate or not in electoral contests, ended up having long-term, and often unintended, effects on group formation and group-based action. Ideology often shifted as the result of these initial choices. Hence, participation in elections which was initially an instrumental and self-interested strategy in the context of high levels of conflict ended up reducing the likelihood of conflict.5 Second, democracy can be reconceptualized as an institutional mechanism permitting the realization of concessions by the capitalists toward the organized workers, thus preempting conflict. It is, in other words, a game that may predetermine the players’ strategies based on their resources (i.e. ‘structure’) but one that does not predetermine the exact outcome; instead it attaches probabilities to outcomes such that the range of possible outcomes varies significantly. Because democratic institutions ‘institutionalize uncertainty’ (Przeworski 1991), they succeed in drawing in new actors. Losers in elections comply because future victory remains a possibility; violent conflict is prevented and democracy becomes a ‘self-enforcing institution.’ Third, this analysis also allows the reconceptualization of legitimacy as ‘the correspondence between the uses of force and rules which specify when it can and should be used’ (1985: 141). Consent, in this perspective, is but the reflection of a deliberate set of choices under specific constraints. Repression is always ‘off the equilibrium path.’ We pay taxes in great part because we are aware of the risk of sanction if we don’t. The fact that violence is not observed does not imply that it does not matter—the contrary: it matters because it does not manifest itself. In the context of a capitalist system, it is possible to determine the a posteriori material conditions under which workers will consent, and even predict, as Przeworski does, what the levels of militancy will be. Social democracy can be, therefore, understood as the optimal answer of individual workers given a set of constraints. Consent emerges from bargains in a specific institutional context; the material bases of class-conflict mitigation are both explicit and predictable.

To summarize, by endogenizing individual identities and actions into a larger game with clear rules (democratic institutions) and a given distribution of resources (capitalism) and constraints (the possibility of repression for a certain level of worker militancy), Przeworski was able to derive some powerful implications (e.g. that democratic institutions are self-enforcing). This line of research has proven extremely fertile, as indicated by recent global-level studies by Boix (2003), and Acemoglu and Robinson (2006). To be sure, Przeworski’s analysis is primarily theoretical; his empirical tests are conducted at a highly aggregate level and are indirect. Indeed, Burawoy (1989) was correct to point out that the main weakness of this analysis (unlike Gould’s) is the absence of a direct empirical test of the theoretical microfoundations.

The contrast between Gould and Przeworski is obvious: the former focuses on small-scale conflict, takes large groups as fixed, specifies how nonviolent conflict escalates into violence, and tests his microfoundations; the latter focuses on (p. 605) large-scale conflict, shows how groups are formed, specifies how institutions contain conflict, and posits only theoretical microfoundations. Both works, however, share a fundamental similarity: they assume a unidirectional causal effect. Gould demonstrates how interpersonal conflict spills over into small-scale collective conflict, while Przeworski shows that large-scale collective conflict shapes individual action. Hence the question: Is it possible to build a theoretically and empirically sound micro-level analysis of conflict that generalizes to macro-level conflict, and vice versa, without having to resort to assumptions about exogenously fixed groups? This is the theoretical challenge taken up by Kalyvas in The Logic of Violence in Civil War (2006).

25.4 Violent Conflict/Endogenous Dynamics

Kalyvas (2006) investigates the dynamics of violent conflict in the context of civil war and attempts to resolve the apparent contradiction between the group-level theorization of the causes of large-scale violent conflict on the one hand, and the individual- and local-level dynamics revealed by close observation on the other.

Given the difficulty of empirically testing hypotheses about the relation between social cleavages and conflict intensity (due to the difficulty of measuring the depth of cleavages independently from conflict intensity), a common practice has been to derive deep group-level divisions from violence on the ground. In other words, if violence is horrific, this must mean that social cleavages run extremely deep. However, this raises a classic problem of endogeneity, for we know that violence typically causes identities to form or sharpen and conflicts to deepen. Thucydides (3.83–4) remarked, in the context of the civil conflict in the Greek polis of Corcyra, on ‘the violent fanaticism which came into play once the struggle had broken out. … As the result of these revolutions, there was a general deterioration of character throughout the Greek world. … Society had become divided into two ideologically hostile camps, and each side viewed the other with suspicion’ (emphasis added). More recently, Girard (1977) reminded us that as rivalry becomes acute, the rivals tend to forget its initial cause and instead become more fascinated with one another; as a result, the rivalry is purified of any external stake and becomes a matter of pure rivalry and prestige.

Indeed, violence may emerge from dynamics that inhere in civil wars and are either indirectly connected or totally unconnected to prewar polarization via two processes: endogenous polarization and endogenous violence. Most group-level arguments assume the distribution of popular support during a civil war to be a (p. 606) faithful reflection of (prewar) cleavages; in times of high polarization, the ‘carrying capacity’ of political actors reaches its maximum value, producing an almost total overlap between the goals of political actors and the goals of the population they claim to represent. Hence, landless peasants naturally support leftist insurgents and landowners right-wing governments; Tamils join the Tamil Tigers and Sinhalas side with Sri Lanka’s government; Catholics support the IRA and Protestants the Unionists, and so on. In other words, support is fixed and given exogenously with respect to the war; hence, determining its distribution requires only access to census data.

However, as the discussion above has pointed out, ‘groups,’ be they economic, ethnic, ideological, or religious, do not exist as such. What can be observed instead are organizations, usually armed ones, making claims on behalf of groups that may be more or less homogeneous and more or less fragmented. In fact, many of their actions are intended precisely to mobilize and homogenize these groups, a fact consistent with the observation that insurgent violence in many ethnic rebellions ‘is usually directed primarily against their own people, in order to ensure their support for the revolt, however reluctant or however passive’ (Paget 1967: 32). Although initial predispositions and preferences that flow directly from prewar politics are present at the onset of a civil war, they also tend to shift as a result of its dynamics. Sometimes the direction of the shift is toward ‘harder’ group identities, but sometimes new cleavages and identities emerge. For example, Tilly (1964: 305, 330) shows that whereas the initial configuration of parties in southern Anjou during the French Revolution ran along class lines and had crystallized ‘long before’ the outbreak of the counterrevolutionary rebellion, in 1793 ‘participation in the counterrevolution cut boldly across class lines. Therefore, no simple scheme of class alignment can account adequately for the division of forces in 1793.’

Beyond polarization, violence may also be endogenous to the war in the sense of being unconnected to the causes of war. War acquires ‘a life of its own.’ Cycles of retaliation, the emphasis on violent resolution of disputes, brutalization, a new war economy, these are all factors that produce violence independent of the initial conflict that caused the eruption of the war in the first place.

In short, the universalization of the distinction between friend and enemy implied by group-level analysis is often a consequence of the war, a by-product of its violence. The responses of political actors and individuals to the dynamics of war (and the responses to their responses) shape violence, the war, and the prospects for peace in a way that is often quite independent of the proximate causes of the conflict. Acknowledging that both violence and polarization can be endogenous to war implies an understanding of the institutional environment within which violence unfolds, very much as the analysis of class conflict in late-nineteenth-century Europe called for an understanding of democracy rather than just an understanding of class.

(p. 607) Kalyvas (2006) develops a theory of the informal institutions of civil war; his primary aim is to account for the subnational variation in violence across space and political actors—in other words, his is a theory that aims to predict where violence will erupt and by whom it will be exercised, conditional on a civil war being on. The starting point is a theory of irregular war rather than a theory of group conflict. This war, commonly referred to as ‘guerrilla war,’6 results in territorial fragmentation into geographical zones that are either monopolistically controlled by rival actors or whose sovereignty is ‘shared.’ The type of sovereignty or control that prevails in a given region affects the type of strategies followed by armed actors. These actors try to shape popular support in their favor (‘collaboration’) and deter collaboration with their rival (‘defection’). As the conflict matures, control is increasingly likely to shape collaboration, because armed actors who enjoy substantial territorial control can protect civilians who live on that territory—both from their rivals and from themselves, thus providing survival-oriented civilians with a strong incentive to collaborate with them, irrespective of their true or initial preferences. In this sense, and consistent with what Coleman (1990: 179) calls ‘power theory,’ collaboration is largely endogenous to control, though, of course, high rates of collaboration spawned by control at a given point in time are likely to reinforce it in the future. Violence is bounded by the nature of sovereignty exercised by each political actor and generally must be selective (or personalized) rather than indiscriminate (or group-level)—which tends to be counterproductive (Coleman 1990: 502).

Selective violence requires private information which is asymmetrically distributed among organizations and local civilians: only the latter may know who the defectors are and they have a choice to denounce them or not. Denunciation constitutes a key link between individuals and armed actors; this is precisely the mechanism that turns often trivial interpersonal and local disputes into violence. Given armed actors who maximize territorial control subject to specific resources and choose whether to use violence or not, and individual civilians who maximize various benefits subject to survival and choose whether to denounce their neighbors or not, it is possible to predict both the distribution of selective violence across geographical zones of control and the armed actor likely to resort to violence in a specific zone. Kalyvas provides an empirical test of these predictions.

Besides producing specific predictions about the distribution of violence, this model entails two implications that speak directly to the issues raised above. The first one extends Gould’s insights about the logic of interpersonal violence to the kind of large-scale violence associated with civil wars. The second one allows the integration of on-the-ground local dynamics and group-level politics, which are typically considered to be contradictory.

First, it is possible to connect a great deal of the violence perpetrated by armed actors and interpersonal conflict between individuals. The key lies in the observation that denunciations are often malicious and loosely related to the cleavage motivating the conflict. That is, individuals exploit the high demand for (p. 608) information to settle local and interpersonal conflicts which may bear little or no relation to the political conflict raging around them. Malicious denunciation turns out to be closely related to interpersonal conflict in contexts of ‘organic’ (rather than ‘mechanical’) solidarity: small-scale, face-to-face social settings, where people develop dense interpersonal interactions, living and working together in daily mutual dependency, rivalry, and love. These include tight-knit neighborhoods, villages and small towns, apartment buildings, family businesses, and work environments. Surprisingly, socially egalitarian settings that lack deep ethnic, religious, or class divisions appear not to be as adverse to denunciation as one would expect. Indeed, denunciation is often horizontal (Gellately 1991; Fitzpatrick 1994). Relatively homogeneous, tight-knit, and egalitarian social settings may be prone to high levels of denunciation for two distinct reasons: ‘symmetry’ and ‘concentration.’

Consistent with Gould, symmetric contexts foster denunciation, via the mechanisms of fear of status loss and envy. Social hierarchies in such contexts tend to be fluid (people are ranked in relation to others in terms of subtle gradations) and open to modification (unlike with castes or rigid strata). Competition for status (‘face’ or ‘honor’) is open, daily, and intense; it generates humiliation, shaming, and ‘loss of face’—usually experienced as among the worst things that can happen to a person. This is also consistent with Simmel’s remark (1955: 43–4) that similarity breeds intense conflict and that people who have many common features often do one another worse wrong than complete strangers do.

The other factor is ‘concentration,’ which refers to the dense overlap of social interaction in small and closed settings. Concentration is associated with denunciation through two related mechanisms. First, the number of interactions between individuals is high, hence the likelihood that interpersonal conflicts and grievances will emerge. Conflicts in dense environments tend to be intense, Simmel (1955: 44–5) has suggested, because the depth of relations between persons in intimate relationships who quarrel causes ‘not a single contact, not a single word, not a single activity or pain’ to remain isolated. It is no coincidence that name-calling, character assassination, and gossip, which Barthes (1977: 169) aptly called ‘murder by language,’ are prevalent in dense social environments. Second, dense contexts entail the presence of an audience for every interaction. The higher the number, and the greater the density of interactions among the members of a group, the greater the opportunity to directly observe each other’s behavior (Hechter 1987: 154; Gambetta 1993: 168). In short, dense communities fit the Spanish saying ‘small village, big hell.’ It is true that the existence of institutions that defuse conflicts simmering under the surface of dense environments has often led observers to interpret them as being solidaristic. However, anthropologists have pointed to their fundamentally dual nature: they are simultaneously conflictual and solidaristic (Gilmore 1987).

The second implication of the model concerns an ill-understood puzzle whereby conflicts and violence on the ground often seem more related to local issues rather than the ‘master cleavage’ that drives the civil war at the national level. As suggested (p. 609) above, many instances of violence taking place in the context of an ethnic conflict turn out to be motivated by the kind of daily interpersonal or local disputes that are unrelated to the larger conflict. This disjunction takes place even when local cleavages are usually framed in the discursive terminology of the master cleavage.

Two interpretations have been offered to explain this disjunction. The first one, inspired by group-level intuitions, views local conflicts as purely epiphenomenal of the larger cleavages. To be sure, they are often ‘contaminated’ by local elements, but they are not autonomous of the larger cleavages. The real agents are groups. This view should be qualified in light of massive micro-level evidence (Kalyvas 2006: ch. 11) and must become the object of empirical investigation as opposed to simple assertion. The second interpretation takes the opposite stand: because of the breakdown of order, all violence is personal and local. Since neither actors nor observers can possibly make sense of such wild diversity and complexity, they simplify, or ‘emplot’ them (Ricoeur 1984) by referring to core issues and cleavages which are compelling but ultimately misleading constructions unrelated to the action on the ground. Both interpretations are problematic.

This disjunction can be elucidated once it is recognized that the loci of agency spawned by civil war are multiple—hence the confusing diversity and ambiguity of motives and identities observed. In other words, the fusion and interaction between dynamics at the group level and the individual level (or the ‘center’ and the ‘periphery’) are fundamental rather than incidental to civil war, a matter of essence rather than noise. A key insight from the discussion of denunciation is that individuals and local communities involved in the war tend to take advantage of the prevailing situation to settle private and local conflicts whose relation to the grand causes of the war or the goals of the belligerents is often tenuous. Under normal circumstances these conflicts are regulated and do not result in violent conflict. We hardly think of them as connected to political violence in general and civil war in particular. Civil wars alter this reality by allowing for the exchange of information for violent action. This perspective provides a key missing link between the public and private spheres, and between master cleavages and local cleavages. Civil war can be analyzed as a process that transforms the political actors’ quest for victory and power and the local or individual actors’ quest for personal and local advantage into a joint process of violence. Departing from both practices of deriving the local from the central and ignoring the central, it is possible to identify a key mechanism linking elite and local dynamics: ‘alliance.’ This mechanism entails a set of transactions between supralocal and local actors, whereby the former supply the latter with external muscle, thus allowing them to win decisive local advantage; in exchange supralocal actors recruit and motivate supporters at the local level. Viewed from this perspective, violence is a key resource for collective action. In the long run, and with iterations, the process of alliance may produce a durable and endogenous shift in identities.

(p. 610) This perspective allows the reconceptualization of ‘master’ cleavages as symbolic formations that simplify, streamline, and incorporate a bewildering variety of local conflicts. The concept of alliance allows us to reintroduce complexity, yet in a theoretically tractable way. Civil wars are concatenations of multiple and often disparate local cleavages, more or less loosely arrayed around the master cleavage. In turn, this cleavage is often influenced by global trends that explain why the same insurgent groups may represent themselves as socialist during the cold war, ethnic nationalist after its end, or Islamist after the World Trade Center attacks.

The mechanism of alliance points to a critical dilemma for both central and local actors. On the one hand, central actors must mobilize at the local level—even when their ideological agenda is universalistic and, thus, opposed to localism. At the same time, to be effective they must also transcend particularism and localism. On the other hand, local actors resist their absorption into centralized and hierarchical structures—both organizational and discursive. Though instrumentally tied to the master cleavage, they strive to remain distinct and maintain their autonomy. The history of many revolutionary movements is a tale of the tension between these conflicting goals. Seen from this perspective, civil war is a process of political and administrative ‘normalization,’ and state-building is an externality of civil war.

Determining when and how such alliances emerge calls for fine-grained research. For instance, a recurring pattern is that losers in local conflicts are more likely to move first allying with outside forces. If this is true, civil war is particularly destabilizing, since it supplies new opportunities to losers in local power conflicts who are seeking an opportunity for revanche. Empirically, alliance is consistent with the intimate and often malicious nature of selective violence, the endogeneity of cleavages to the war, and the limited visibility of local cleavages after the war. Once a war has ended, the master narrative of the conflict provides a handy way to ex post facto simplify, streamline, and ultimately erase the war’s complexities, contradictions, and ambiguities. An interesting implication is that what differentiates modern insurgencies from premodern rebellions (but also some post-cold-war insurgencies) is the absence of sophisticated elites with a ‘global’ discourse.


The relevance of alliance as a mechanism linking local cleavages into a conflict’s master cleavage is twofold: first, it allows for a theoretical understanding of large-scale conflict that incorporates rather than ignores the puzzle of the disjunction between individual-level and group-level dynamics—and the extensive ambiguity that surrounds this process. Second, it turns the individual/group-level interface (p. 611) into a central issue and forces us to rethink precisely the modalities linking distinct actors and motivations. Recognizing the existence of a concatenation of multiple and disparate local cleavages allows the exploration of the relation between the individual and the group levels—a possibility precluded by the prevailing understanding of cleavage.7

It is possible, therefore, to reconcile within the same analytical framework, but without bundling them, two sets of dynamics that have been approached as being contradictory to each other: group-level and individual-level action. The theoretical advantage of the mechanism of alliance is that it allows for multiple rather than unitary actors, agency located in both center and periphery rather than only in either one, and a variety of preferences and identities as opposed to a common and overarching one. The main implication is that identity labels should be treated with caution and not projected on individual actions from the level of the group.

The discussion of the mechanism of alliance in civil-war contexts is illustrative. It is possible to come up with a substantial number of micromechanisms depending on the question investigated and the context studied. By adopting an analytical strategy of theoretical and empirical disaggregation of the constitutive elements of conflict, such as cleavages, groups, or identities, and by making microfoundations explicit, we can arrive at an improved understanding of the causes and consequences of conflict: one that recognizes the dynamic interaction between conflict and action, and between individuals and groups. This approach allows us to relax the usual assumptions of unitary and exogenously given groups, bypass the dichotomy of individual and group-based perspectives, integrate the interpersonal and local dimensions of violent conflict with its collective dimension, and focus on the dynamic and bidirectional relationship between groups and individuals. In short, the analytical approach to the study of conflict opens up a vast and exciting research agenda, combining abstract theorization with rich empirical evidence, while countering the disciplinary and subdisciplinary fragmentation that has undermined the effective study of conflict.


Acemoglu, D., and Robinson, J. A. (2006), Economic Origins of Dictatorship and Democracy (New York: Cambridge University Press).Find this resource:

    Balcells, L. (2007), ‘Rivalry and Revenge: Killing Civilians in the Spanish Civil War’, Estudios/working papers CEACS, no. 233.Find this resource:

      Barthes, R. (1977), Roland Barthes (New York: Hill and Wang).Find this resource:

        Birkbeck, C., and Lafree, G. (1993), ‘The Situational Analysis of Crime and Deviance’, Annual Review of Sociology, 19: 113–37.Find this resource:

          Black, D. J. (1976), The Behavior of Law (New York: Academic).Find this resource:

            (1993), The Social Structure of Right and Wrong (San Diego, Calif.: Academic).Find this resource:

              Bobbio, N. (1992), ‘Guerra Civile?’, Teoria Politica, 1–2: 297–307.Find this resource:

                Boix, C. (2003), Democracy and Redistribution (New York: Cambridge University Press).Find this resource:

                  Bosch sánchez, A. (1983), Ugetistas y Libertarios: Guerra Civil y Revolución en el Pais Valenciano, 1936–1939 (Valencia: Instituto Alfons el Magnanim).Find this resource:

                    Boudon, R. (1988), ‘The Logic of Relative Frustration’, in M. Taylor (ed.), Rationality and Revolution (Cambridge: Cambridge University Press), 245–67.Find this resource:

                      Brubaker, R., and Laitin, D. D. (1998), ‘Ethnic and Nationalist Violence’, Annual Review of Sociology, 24: 243–52.Find this resource:

                        Burawoy, M. (1989), ‘Marxism Without Microfoundations: Przeworski’s Critique of Social Democracy’, Socialist Review, 89: 53–86.Find this resource:

                          Buruma, I. (2002), ‘The Blood Lust of Identity’, New York Review of Books, 11 April, pp. 12–14.Find this resource:

                            Casanova, J. (1985), Anarquismo y Revolución Social en la Socieadad Aragonesa, 1936–1938 (Madrid: Siglo XXI).Find this resource:

                              Caselli, F., and Coleman, W. J. (2006), ‘On the Theory of Ethnic Conflict’, NBER working paper no. 12125.Find this resource:

                                Coleman, J. S. (1990), Foundations of Social Theory (Cambridge, Mass.: Harvard University Press).Find this resource:

                                  Collier, P., et al. (2003), Breaking the Conflict Trap: Civil War and Development Policy (Washington, D.C.: World Bank/Oxford University Press).Find this resource:

                                    Coser, L. A. (1956), The Functions of Social Conflict (Glencoe, Ill.: Free Press).Find this resource:

                                      Dahrendorf, R. (1959), Class and Class Conflict in Industrial Society (Stanford, Calif.: Stanford University Press).Find this resource:

                                        (p. 613) Decker, S. H. (1993), ‘Exploring Victim–Offender Relationships in Homicide: The Role of Individual and Event Characteristics’, Justice Quarterly, 10: 585–612.Find this resource:

                                          Deutsch, K. W. (1953), Nationalism and Social Communication: An Inquiry into the Foundations of Nationality (New York: Wiley).Find this resource:

                                            Dion, D. (1997), ‘Competition and Ethnic Conflict: Artifactual?’, Journal of Conflict Resolution, 41: 638–48.Find this resource:

                                              Downs, A. (1957), An Economic Theory of Democracy (New York: Harper).Find this resource:

                                                Elster, J. (1985), Making Sense of Marx (Cambridge: Cambridge University Press).Find this resource:

                                                  Esteban, J., and Ray, D. (1994), ‘On the Measurement of Polarization’, Econometrica, 62: 819–52.Find this resource:

                                                    Fearon, J. D., and Laitin, D. (1996), ‘Explaining Interethnic Cooperation’, American Political Science Review, 90: 715–35.Find this resource:

                                                      Fehr, E., and Gachter, S. (2000), ‘Fairness and Retaliation: The Economics of Reciprocity’, Journal of Economic Perspectives, 14: 159–81.Find this resource:

                                                        Figueiredo, R. De, and Weingast, B. R. (1999), ‘The Rationality of Fear: Political Opportunism and Ethnic Conflict’, in B. Walter and R. Snyder (eds.), Civil Wars, Insecurity, and Intervention (New York: Columbia University Press), 262–302.Find this resource:

                                                          Fitzpatrick, S. (1994), Stalin’s Peasants: Resistance and Survival in the Russian Village After Collectivization (New York: Oxford University Press).Find this resource:

                                                            Gambetta, D. (1993), The Sicilian Mafia: The Business of Private Protection (Cambridge, Mass.: Harvard University Press).Find this resource:

                                                              Gellately, R. (1991), The Gestapo and German Society: Enforcing Racial Policy, 1933–1945 (Oxford: Oxford University Press).Find this resource:

                                                                Gellner, E. (1983), Nations and Nationalism (Oxford: Blackwell).Find this resource:

                                                                  Gilmore, D. D. (1987), Aggression and Community: Paradoxes of Audalusian Culture (New Haven, Conn.: Yale University Press).Find this resource:

                                                                    Girard, R. (1977), Violence and the Sacred (Baltimore, Md.: Johns Hopkins University Press).Find this resource:

                                                                      Gould, R. (2003), Collision of Wills: How Ambiguitiy about Social Rank Breeds Conflict (Chicago, Ill.: University of Chicago Press).Find this resource:

                                                                        Hechter, M. (1987), Principles of Group Solidarity (Berkeley, Calif.: University of California Press).Find this resource:

                                                                          Hodson, R., Sekulic, D., and Massey, G. (1994), ‘National Tolerance in the Former Yugoslavia’, American Journal of Sociology, 99: 1535–58.Find this resource:

                                                                            Horowitz, D. L. (1985), Ethnic Groups in Conflict (Berkeley, Calif.: University of California Press).Find this resource:

                                                                              Kalyvas, S. N. (2000), ‘Commitment Problems in Emerging Democracies: The Case of Religious Parties’, Comparative Politics, 32: 379–99.Find this resource:

                                                                                (2006), The Logic of Violence in Civil War (New York: Cambridge University Press).Find this resource:

                                                                                  Katz, J. (1988), Seductions of Crime: A Chilling Exploration of the Criminal Mind—from Juvenile Delinquency to Cold-blooded Murder (New York: Basic).Find this resource:

                                                                                    Laitin, D. (2001), ‘Secessionist Rebellion in the Former Soviet Union’, Comparative Political Studies, 34: 839–61.Find this resource:

                                                                                      Lasswell, H. D. (1936), Politics: Who Gets What, When, How (New York/London: Whittlesey).Find this resource:

                                                                                        Lawrence, A. (2007), Imperial Rule and the Politics of Nationalism, Ph.D. thesis (University of Chicago).Find this resource:

                                                                                          (p. 614) Levy, J. S. (1983), War in the Modern Great Power System, 1495–1975 (Lexington, Ky.: Kentucky University Press).Find this resource:

                                                                                            Lipset, S. M., and Rokkan, S. (1967), ‘Cleavage Structures, Party Systems, and Voter Alignments: An Introduction’, in S. M. Lipset and S. Rokkan (eds.), Party Systems and Voter Alignments: Cross-National Perspectives (New York: Free Press), 1–64.Find this resource:

                                                                                              Mcadam, D., Tarrow, S., and Tilly, C. (2001), Dynamics of Contention (New York: Cambridge University Press).Find this resource:

                                                                                                Malefakis, E. (1996), ‘Aspectos históricos y teóricos de la guerra’, in E. Malefakis (ed.), La Guerra de España (1936–1939) (Madrid: Taurus), 11–47.Find this resource:

                                                                                                  Martin, J. C. (1995), ‘Guerre civile et modernité: le cas de la Révolution’, in J. C. Martin (ed.), La guerre civile entre histoire et mémoire (Nantes: Ouest).Find this resource:

                                                                                                    Moore, B., Jun. (1966), Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World (Boston, Mass.: Beacon).Find this resource:

                                                                                                      Nordlinger, E. A. (1972), ‘Conflict Regulation in Divided Societies’, Occasional Papers in International Affairs no. 29, Center for International Affairs, Harvard University.Find this resource:

                                                                                                        Oberschall, A. (1978), ‘Theories of Social Conflict’, Annual Review of Sociology, 4: 291–315.Find this resource:

                                                                                                          Ortiz sarmiento, C. M. (1990), La violence en Colombie: Racines historiques et sociales, (Paris: L’Harmatan).Find this resource:

                                                                                                            Packer, G. (2007), ‘Planning for Defeat’, New Yorker, 17 September, pp. 56–65.Find this resource:

                                                                                                              Paget, J. (1967), Counter-insurgency Operations: Techniques of Guerrilla Warfare (New York: Walker).Find this resource:

                                                                                                                Parker, N. (2007), ‘Iraq’s War Makes Intimates Enemies’, Los Angeles Times, 16 September; at <,1,7704935,full.story?ctrack=5&cset=true>, accessed 2008.

                                                                                                                Perez-díaz, V. M. (1993), The Return of Civil Society: The Emergence of Democratic Spain (Cambridge, Mass: Harvard University Press).Find this resource:

                                                                                                                  Petersen, R. D. (2002), Understanding Ethnic Violence: Fear, Hatred, and Resentment in Twentieth-century Eastern Europe (New York: Cambridge University Press).Find this resource:

                                                                                                                    Posen, B. R. (1993), ‘The Security Dilemma and Ethnic Conflict’, Survival, 35: 27–47.Find this resource:

                                                                                                                      Przeworski, A. (1985), Capitalism and Social Democracy (Cambridge: Cambridge University Press).Find this resource:

                                                                                                                        (1991), Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America (Cambridge: Cambridge University Press).Find this resource:

                                                                                                                          Ricoeur, P. (1984), Time and Narrative (Chicago, Ill.: University of Chicago Press).Find this resource:

                                                                                                                            Roemer, J. (1986), Analytical Marxism (Cambridge: Cambridge University Press).Find this resource:

                                                                                                                              Sampson, R. J., and Laub, J. H. (1992), ‘Crime and Deviance in the Life Course’, Annual Review of Sociology, 18: 63–84.Find this resource:

                                                                                                                                Schelling, T. C. (1978), Micromotives and Macrobehavior (New York: Norton).Find this resource:

                                                                                                                                  Schmitt, C. (1976), The Concept of the Political (New Brunswick, N.J.: Rutgers University Press).Find this resource:

                                                                                                                                    Seidman, M. (2002), Republic of Egos: A Social History of the Spanish Civil War (Madison, Wisc.: University of Wisconsin Press).Find this resource:

                                                                                                                                      Sekulic, D. (2005), ‘Structural Determinants of Nationalism in Croatia’, unpublished paper.Find this resource:

                                                                                                                                        Simmel, G. (1955), Conflict and the Web of Group Affiliations (New York: Free Press).Find this resource:

                                                                                                                                          Singer, J. D., and Small, M. (1972), The Wages of War, 1816–1965: A Statistical Handbook (New York: Wiley).Find this resource:

                                                                                                                                            (p. 615) Spencer, J. (2000), ‘On Not Becoming a “Terrorist”: Problems of Memory, Agency, and Community in the Sri Lankan Conflict’, in V. Das, et al. (eds.), Violence and Subjectivity (Berkeley, Calif.: University of California Press), 120–40.Find this resource:

                                                                                                                                              Tajfel, H., and Turner, J. C. [1979] (2004), ‘An Integrative Theory of Intergroup Conflict’, in M. J. Hatch and M. Schultz (eds.), Organizational Identity: A Reader (Oxford: Oxford University Press), 56–65.Find this resource:

                                                                                                                                                Thucydides (1972), History of the Peloponnesian War, trans. Rex Warner (London: Penguin).Find this resource:

                                                                                                                                                  Tilly, C. (1964), The Vendée (Cambridge, Mass.: Harvard University Press).Find this resource:

                                                                                                                                                    US Department of Justice, Bureau of Justice Statistics; at <>, accessed 2008.

                                                                                                                                                    Varshney, A. (2003), Ethnic Conflict and Civic Life: Hindus and Muslims in India (New Haven, Conn.: Yale University Press).Find this resource:

                                                                                                                                                      Wilkinson, S. I. (2006), Votes and Violence: Electoral Competition and Ethnic Riots in India (New York: Cambridge University Press).Find this resource:

                                                                                                                                                        Wood, E. J. (2003), Insurgent Collective Action and Civil War in El Salvador (New York: Cambridge University Press). (p. 616) Find this resource:


                                                                                                                                                          (*) I am grateful to Meredith Rolfe for her excellent comments on the first draft of this chapter.

                                                                                                                                                          (1.) The emerging field of terrorism studies is primarily associated with the discipline of political science as well, as is the study of genocide and ethnic cleansing.

                                                                                                                                                          (2.) Economic theories begin with theorizations of individual dynamics, yet their empirical tests tend to stay at the level of group dynamics and the mechanisms linking individuals and groups are not empirically tested.

                                                                                                                                                          (3.) For example, in 2002, 43 percent of all murder victims in the USA were related to or acquainted with their assailants, whereas only 14 percent of victims were murdered by strangers (43 percent of victims had an unknown relationship to their murderer) (US Department of Justice).

                                                                                                                                                          (4.) In this respect, Przeworski’s analysis goes beyond the standard Olsonian collective-action problem, which posits the potential impossibility of action despite interest in its favor.

                                                                                                                                                          (5.) Kalyvas (2000) has argued that a similar mechanism could be triggered in Muslim polities.

                                                                                                                                                          (6.) Keep in mind that irregular or guerrilla war is not the only type of civil war.

                                                                                                                                                          (7.) In fact, surrogate warfare has a long history in international politics, from the days of the Roman Empire up until the cold war and its aftermath. Yet we have generally failed to conceptualize civil war as a concatenation of multiple surrogate wars.