'the Social Processes of Civil War the Wartime Transformation of Social Networks

g r o . s w e i v e r l . a y u l n n n o a . e s w u l w w a n o s m r o e r f p d r e o d F a . o 4 l 1 n / 6 w 0 o / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s 5 : r e 1 i v 1 . n 8 U 0 0 k 2 s . n i c a S d . d y t i l S o y P b . v e R . u n n A

The Social Processes of Civil War: The Wartime Transformation T ransformation of Social Networks Elisabeth Jean Wood Department of Political Science, Yale University, University, New Haven, Connecticut 06520; and Santa Fe Institute, Santa Fe, New Mexico 87501; email: [email protected]

Annu. Rev. Polit. Sci. 2008. 11:539–61

Key Words

First published online as a Review in Advance on March 4, 2008

violence, political polarization, political mobilization, displacement, militarization, gender roles

The Annual Review of Political Science is online at http://polisci.annualreviews.org This article’s article’s doi: 10.1146/annurev.polisci.8.082103.104832 c 2008 by Annual Reviews. Copyright  All rights reserved 1094-2939/08/0615-0539$20.00 This article is also part of the D omestic Po litical Violence and C ivil ivil War compilation.

Abstract Little att Little attent ention ion has bee been n pai paid d to the soc social ial pro process cesses es of civ civil il war war—th —thee transforma trans formation tion of social actors, struct structures, ures, norms, and practices— thatt som tha sometim etimes es lea leave ve end enduri uring ng leg legaci acies es for the pos postwa twarr per period iod.. In thi thiss article art icle,, I exp explor loree the cha change ngess wro wrough ughtt by six sixsoc social ial pro proces cesses:politi ses:political cal mobilization, military socialization, polarization of social identities, militarization of local authority, transformation of gender roles, and fragmentati fragm entation on of the local politi political cal economy. economy. Some of these social processes occur in peacetime, but war may radically change their pace, direction, or consequences, with perhaps irreversible effects. I trace the wide variation in these processes during the wars in four countries: Peru, El Salvador, Sri Lanka, and Sierra Leone. I analyze the effects of these processes as transformations in social networks. These processes reconfigure social networks in a variety of ways, creating new networks, dissolving some, and changing the structure of others.

539

INTRODUCTION Scholars have recen Scholars recently tly opened up the “blac “black k box” of civil wars, exploring the bargaining aspects of war, the logic of wartime violence, the forging of institutions likely to contribute to durable settlements, settlements, and the chall challenges enges of postwar demobilization and reconstruction. An important recent advance is the emphasis on analyzing variation in patterns of violence, participation, and institution building across civil wars. Despit Des pitee the these se adv advanc ances es in sch schola olarly rly und under er-standing, less attention has been paid to the soci so cial al pr proc ocess esses es of ci civi vill wa war— r—by by wh whic ich h I me mean an the tra transf nsform ormati ation on of soc social ial act actors ors,, stru structu ctures, res, norms, and practices at the local level—that sometimes leave profound social changes in their wake. The literature describing and analyzing these processes in their divergent patterns is dominated by case studies and policy discussi disc ussion, on, wit with h as yet ins insuffi ufficien cientt atte attenti ntion on to causal analysis and comparison across cases. In thi thiss art articl icle, e, I exp explor loree six soc social ial pro proces cesses: ses: political politic al mobili mobilization zation,, milita military ry social socializatio ization, n, the polarization polarization of social identities, identities, the militarization of local authority, the transformation of gender roles, and the fragmentation of the local political economy. These are processes that sometimes have enduring legacies. For exampl example, e, wartim wartimee polar polarization ization may lead to elector electoral al polar polarizatio ization, n, segreg segregation ation,, and a distrustful political culture in the postwar period. Civil wars differ in the extent to which these processes are at work. Just as forms of violence vary across and within civil wars, so toodo th thes esee pr proc oces esses ses.. To be su sure re,, so some me of th thee socialprocessesIdiscussareongoinginpeacetime as well, but war may radically change the pace of existing processes, redirect them, or alter their conse consequenc quences, es, with perha perhaps ps irreversible effects. In focusing on social processes cess es oth other er tha than n vio violen lence, ce, som somee of whi which ch pre pre-date the conflict, this article extends the work of anthr anthropolog opologist ist Steph Stephen en Lubkem Lubkemann ann (2007 (2007,, 2008), who argues that in their emphasis on wartime violence, scholars have neglected the

g r o . s w e i v e r l . a y u l n n n o a . e s w u l w w a n o s m r o e r f p d r e o d F a . o 4 l 1 n / 6 w 0 o / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s 5 : r e 1 i v 1 . n 8 U 0 0 k 2 s . n i c a S d . d y t i l S o y P b . v e R . u n n A

540

Wood

agency of ordinary people struggling to realize life projects in the context of war in ways shap sh aped ed by pr prew ewar ar so soci cial al no norm rmss an and d pa patte ttern rnss as well as violence. In contrast to Lubkemann, who focuses on how life projects in particular settings lead to contrasting patterns of wartime migration, I widen the focus to discuss a variety of social processes. I do not address the legacy of civil war for processes of state and regime formation. As with other ot her aspects aspect s of civil war, it is extremel tre melyy dif difficu ficult lt to mea measur suree the ext extent ent of the these se social processes. I trace the variation in these processes during the wars in four countries chosen to illustrate the wide differences in these processes, namely, Peru, El Salvador, Sri Lanka, and Sierra Leone. As we shall see, patter pat terns ns of vio violen lence, ce, pro proces cesses ses of mob mobili ilizat zation ion and recruitment, and the extent to which local authorities were militarized, gender roles transformed, and economies fragmented varied widely across these conflicts. This article draws on field research I conducted for 26 months in El Salvador as well as short trips to Sri Lanka and Peru. I analyze the effects of these processes as transformations in social networks. These processes reconfigure social networks in a variety of ways, creating new networks, dissolving some, and changing the structure of others, as when the local clients of a patron are mobilized into an armed network with a new central figure. A social network consists of persons (network nodes) linked by different kinds of relationships (edges). For example, wartime polarization may reshape friendship networks in a village, fracturing the network into two distinct networks with no edge between them. I first describe the civil wars in these countries, tri es, par partic ticula ularly rly the dif differ ferenc ences es in the str strate ate-gies of the armed actors evident in their distinct patterns of violence. I then describe the six social proce processes sses in turn, contrasting contrasting their apparent incidence in the four conflicts and making some observations about other conflicts. In the conclusion, I briefly discuss their varied consequences for the postwar period.

CIVIL WARS IN EL SALVADOR, PERU, SRI LANKA, AND SIERRA LEONE

g r o . s w e i v e r . l a y u l n n n o a . e s w u l w w a n o m s o r r f e p d r e o d F a . o 4 l 1 n / 6 w o 0 / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s 5 : r e 1 i v 1 . n 8 U 0 0 k 2 s . n i c a S d . d y t i l S o y P b . v e R . u n n A

Wartime social processes are strongly shaped by the strategies of the armed actors, particularly the patterns of violence they wield, including whether violence against civilians is carried out disproportionately by one side, whether violence is generally indiscriminate or selective (individuals are targeted for their individual behavior), the intensity of violence, and the repertoire of violence deployed by the parties to the war. By repertoire of violence, I mean the violent subset of what Tilly (1978, 2003, 2008) calls the repertoire of contention, namely, that set of violent practices that an armed group routinely engages in as it makes claims on other political or social actors. A particular group may include in its repertoire any or all of the following: kidnapping, assassinations, massacres, torture, sexual violence, forced displacement, and so on. Although Kalyvas (2006), Weinstein (2006), and other scholars have analyzed the variation in lethal violence, the variation in other forms of violence remains underanalyzed [but see Wood (2006, 2007) on sexual violence, Steele (2007) on displacement, and Hoover (2007) for a theory of armed group organization and repertoires of violence]. In particular, howdistinct repertoires of violence shape other social processes is not well understood. The social effects of a particular form of violence depend on the context; in particular,itsmeaningtosocialactorsdepends on the cultural setting (Ellis 1995, 1999). The same act of violence, e.g., the rape of a young woman by a neighbor who is a member of a local militia, may in one context be understood as part of the ethnic cleansing of the neighborhood and in another as a private act unrelated to the war aims of the parties. The pattern of violence varies sharply across our four cases in intensity, repertoire, the mix of selective and indiscriminate violence, and the degree of symmetry in the wielding of violence by the armed parties to the war.

In El Salvador’s civil war, leftist insurgents influenced by Marxist/Leninist ideology and liberation theology rebelled against an authoritarian state whose military rulers colluded with economic elites to maintain a highly unequal society based on a laborrepressive model of agriculture (Wood 2000). Violence during El Salvador’s civil war was extremely asymmetric: State agents were responsiblefor85%ofdeathswhereastheinsurgent group, the Frente Farabundo Mart ı´ para Liberaci´ on Nacional (FMLN), was responsible for 5% (the rest were unattributable; Truth Commission for El Salvador 1993). Violence against civilians by state agents was widespread, particularly early in the war: More than 50,000 civilians—in a country of five million people—were killed during the war (Seligson & McElhinny 1996).1 Violence was often indiscriminate; entire families and villages were targeted in response to proinsurgent activities by a few members. Sexual violence, though not nearly as widespread as in some other conflicts, was also committed disproportionately by state agents; indeed, the Truth Commission for El Salvador lists no cases perpetrated by the FMLN among the 150 cases in the unpublished annexes to its report (Wood 2006). The overall pattern of violence through the war was one of restraint on the part of the FMLN and increasingly selective violence on the part ofstateagents(withtheexceptionofthestate’s response to the FMLN’s 1989 offensive in San Salvador). The pattern of relative restraint in violence by the insurgents was remarkable given the group’s complicated command structure (the FMLN was a coalition of five distinct organizations).

FMLN: Frente Farabundo Mart ´ı ´ para Liberacion Nacional, insurgent group in El Salvador’s civil war

1 The

gold standard of civil war mortality estimation, namely multiple systems estimation (MSE), has not yet been done for the Salvadoran civil war, so the figure could be significantly different. MSE more than doubled the estimated number of dead and disappeared in the case of Peru, whereas the same technique halved the number in East Timor (Ball et al. 2003, Silva & Ball 2006, Lynch & Hoover 2007).

www.annualreviews.org • Social Processes of Civil War

541


In sharp contrast, lethal violence in Peru’s civil war was much more symmetric. The insurgent group Sendero Luminoso was reLTTE: Liberation sponsible for 46% of reported fatalities, and Tigers of Tamil state agents for 30% (Ball, Asher, Sulmont Eelam, secessionist group in Sri Lanka’s & Manrique 2003, p. 2). Violence was concivil war centrated in the indigenous highlands of the RUF: Revolutionary Andes (three quarters of the victims of lethal United Forces, violence spoke Quechua as their primary laninsurgent group in guage), the Amazonian lowlands, and Lima. Sierra Leone’s civil Responsibility for the cases of sexual violence war reported to the Truth and Reconciliation Commission was asymmetric, with Sendero Luminoso responsible for 11% of the reported cases of rape and state agents for 83% (CVR 2003, Vol. 6, Ch. 1, pp. 274– 79). The insurgents became increasingly abusive of civilians as the war progressed, particularly after they were pushed out of their initial strongholds (which occurred in many highland areas by 1984 or 1985). Throughout the 1980s, Sendero carried out an increasing number of massacres, while state forces became much more selective in their violence (Degregori 1999, p. 79). Sendero units forced entire communities to move to base camps to work on behalf of the insurgency; community members were not allowed to leave. In some camps, insurgent leaders forced girls and young women into sexual relationships (CVR 2003, Vol. 6, Ch. 1, pp. 287–92). Sri Lanka’s civil war of ethnic secession pits a secessionist Tamil group, the Liberation Tigers of Tamil Eelam (the LTTE), against a state dominated by two Sinhalese parties that since independence have competed for votes in a classic pattern of ethnic outbidding. Each party has used the other’s efforts at ethnic compromise to rally voters against the other. Violence in Sri Lanka’s war appears to be roughly symmetric, but I know of no credible estimate of civilian deaths disaggregated by perpetrating group.2

The LTTE commits significant violence against civilians in a complex pattern. Selective assassinations of national and local leaders, particularly rival Tamil figures, often kill significant numbers of civilian bystanders; in addition, Sinhalese villagers are targets of occasional indiscriminate violence (often in reprisal for killing of Tamils in the area), as are Sri Lankan civilians generally. State forces and their paramilitary allies engage in a wide repertoire of violence, including forced displacement, occasional massacres of Tamil civilians, disappearances, long-term detention, and sexual violence. Sexual violence by the state occurs in two patterns: opportunistic rape of Tamil women and girls at checkpoints, and a recurring, apparently systematic pattern of sexual torture of both political and criminal prisoners. In sharp contrast, the LTTE has not engaged in sexual violence (except as part of the torture of Tamils supporting rival groups), even during the ethnic cleansingofMuslimsfromnorthernSriLanka in 1990 (Wood 2009). The civil war in Sierra Leone was characterized by extremely high levels and unusually wide repertoires of violence, the forced recruitment of combatants (including the recruitment of many children by the insurgents), and what the Truth and Reconciliation Commission (TRC) of Sierra Leone termed “factional fluidity,” the frequent shifting of alliances and sides by units and groups (TRC 2004, especially Vol. 3A, Ch. 4, pp. 550–52). Violence by the insurgent group, the Revolutionary United Forces (RUF), was, unusually, largely indiscriminate. The RUF’s wide repertoire of violence included gang rape, amputations, mutilations, massacres, forced recruitment, forced displacement, and abduction for slave labor, including sexual slavery (TRC 2004, Vol. 2, Ch. 2). The RUF was also highly unusual in the proportion of combatants who were forcibly recruited: 87%,

2 See the various issues of Sri

Lanka: State of Human Rights, an annual report by the Law and Society Trust, the various reports onSri Lankaby HumanRights Watch andAmnesty 542

Wood

International, and the many publications by the University Teachers for Human Rights ( Jaffna).


according to Humphreys & Weinstein (2006). many civil wars. For example, civilian insur Much of the violence seemed to have little gent supporters carried out land invasions on purpose other than terrorizing the popula- a large scale in El Salvador (Wood 2003). tion (TRC 2004, Vol. 2, Ch. 2, p. 34). State Approximately half of civil wars since agents engaged in a wide repertoire of vio- World War II were fought as irregular wars, in lence but were responsible for less of the vio- which nonstate combatants mingle with civillence against civilians than the RUF. ians and rarely mass in significant numbers for The strategies of armed actors and their set battles (Balcells & Kalyvas 2007). There distinct patterns of violence shape other so- are some exceptions to this general pattern cial processes to varying degrees, accelerating among our cases: Insurgents in El Salvador some ongoing processes and setting in motion fought conventional battles until 1983, and others. I address each in turn. some of the most pitched battlesin Sri Lanka’s civil war have been fought along conventional lines, as in the struggle for control of Elephant POLITICAL MOBILIZATION Pass. Warfare has been predominantly irreguPolitical mobilization often precedes civil war larin all four cases,with the possible exception as contestation deepens over claims made by of Sri Lanka. The pursuit of irregular war renonstateactors(ordissidentstateactors)tore- lies on the ongoing support of at least some sources, power, and voice. In both El Salvador civilians, unless extraordinary resource availand Sri Lanka, extensive political mobilization ability renders the combatants effectivelyselfoccurred before civil war began, as nonviolent sufficient. The frequent contact with civilians social groups attempted to persuade state ac- in most cases of irregular warfare implies that tors to address theirgrievances. In the wake of everyday social processes may be reshaped by state violence in El Salvador in the late 1970s conflict processes. However, many of the soand widespread ethnic violence in Sri Lanka cial processes analyzed here occur in broadly in 1983, political mobilization was supplanted similar form during conventional civil wars, by armed conflict. In El Salvador, intense and despite the significant difference in specific indiscriminate state violence led to a rapid patterns. For example, civilians in the largely growth in insurgent ranks from 1979 to 1981, conventional Spanish Civil War were mobias many formerly nonviolent activists, disillu- lized in rear-guard areas to contribute food sioned with conventional political forms, re- and to work as volunteers in direct support of acted with moral outrage (Wood 2003). Even armed groups (L. Balcells Ventura, personal where overt social mobilization is not evident, communication). covertpolitical mobilizationalmost always ocIn irregular wars, civilians provide “cover” curs as nonstate actors attempt to expand their for nonstate combatants as well as intellinumbers, territory, and resources in anticipa- gence, supplies, transportation, and fresh retion of the struggle to come, armed or not. In cruits. Although the last three can be coerced Peru, Sendero Luminoso organized covertly, relatively effectively, as evident in the patand its goals and actors showed little continu- tern of forced recruitment in Sierra Leone, ity withearlier forms of political mobilization. the coercing of high-quality intelligence is In SierraLeone, the RUFalso recruited initial much more problematic (Wood 2003). Comembers covertly with little effort at building ercion, like torture, gives rise to perverse incivilian support networks;indeed, much of the centives to provide false information in order initial recruitment and training occurred in to satisfy the coercers, particularly if the civilLiberia and Libya (TRC 2004, Vol. 2, Ch. 1, ian does not in fact know the required data. p. 9; Vol. 3A, Ch. 3). Less analyzed than pre- State agents and actors allied with the state war social mobilization is the ongoing col- also attempt to mobilize political backing, or lective action by civilians that occurs during at least to interrupt mobilization on behalf of www.annualreviews.org • Social Processes of Civil War

543

nonstate actors. One common consequence of this pattern of warfare is increasing political polarization in areas of conflict or anticipated conflict as civilians feel themselves caught “between two fires” (Stoll 1993). The form of civilian mobilization varies significantly across conflicts, across actors within conflicts, and over time. The FMLN relied on voluntary provision of intelligence; it built up ongoing collaborative networks of covert civilian support that produced intelligence the government could not match in quality, despite various efforts to build government support through land reform and resettlement programs (Wood 2003). Mobilization in El Salvador also varied over time. Overt political mobilization re-emerged in the mid 1980s thanks to two sets of actors with varied covert ties to the insurgents: Rural residents carried out extensive land occupations, and unions mounted massive demonstrations and marches. In Peru, Sendero Luminoso initially relied on persuasion to recruit members, but the group became increasingly coercive. Cadre acted with increasing violence in communities in the Andean highlands, forcing communities to attend meetings and publicly killing community leaders and suspected informers (Starn 1995, CVR 2003). Although the Sri Lankan insurgents also attracted significant numbers of volunteers in the aftermath of the 1983 ethnic violence against Tamils, they soon developed highly coercive relations with civilians, assassinating rival Tamil elites as well as local people who favored rival groups (Tambiah 1986, Somasundaram 1998, Narayan Swamy 2004). Local elites generally mobilize existing social networks, particularly kin and clientelist ties, to counter actors encroaching on their interests—not only insurgent organizations but also rival elites seeking advantage in the disorganized context of war. Initial mobilization often includes not only the arming of these networks but also attacks on civilian groups perceived to be supportive of insurgent organizations, as in the case of the paramilitary groups in Colombia (Leal


544

Wood

Francisco 1990, Garc´ ıa 1996, Romero 2003). In Sierra Leone, the traditional kamajor societies were mobilized by local elites against the insurgents in the form of Civil Defense Forces, which carried out widespread violence in the southern part of the country against those perceived as supporting the RUF (TRC 2004, Vol. 2, Ch. 2). In El Salvador, landowning elites directed informal groups of clients to suppress protests calling for land reform with increasing violence through the 1970s, derailing a land reform initiative in 1976 (Brockett 1990, pp. 147–48). Militias often originate in such networks but may rapidly escape the control of their founders, as was true of Colombia’s paramilitary groups. State actors may also bolster their forces by drawing on clientelist networks of local allies or organizations of retired soldiers to found civilian militias or defense forces. In El Salvador, well before the outbreak of civil war, the state greatly extended its paramilitary networks. Many of the peasants in these networks were clients of powerful patrons or former soldiers, and supplied intelligence on rural troublemakers in exchange for loans, access to health services, immunity from any consequences of their paramilitary activities, and agricultural inputs such as land and fertilizer (Stanley 1996). In Peru, the state armed groups of residents—the rondas campesinas who organized local resistance to Sendero Luminoso, a form of civil defense that proved very effective against the insurgents (Starn 1995, Degregori 1999, Del Pino 2005). In Rwanda, local elites drew on kin and social networks to recruit participants in the genocidal killing of neighbors (Fujii 2009). Both the state and insurgent groups may draw combatants from these networks of civilian supporters through a variety of mechanisms. Armed organizations often draw from particular groups, e.g., a specific ethnic group, not only because incoming recruits will be much more likely to endorse a particular agenda of secession or redistribution, but also because they are likely to bring with them norms and beliefs concerning the


appropriateness of violence against other groups. Of course these norms and beliefs can be radically altered by socialization into the armed group, both through formal training and through informal small-group dynamics such as hazing (see below). The Salvadoran insurgents relied on persuasion, as did the Peruvian insurgents initially. Contributing to the recruiting of youth to both the Peruvian and Salvadoran insurgencies was the social mobility participation offered (Degregori 1999, Wood 2003). (Sendero Luminoso also recruited a significant number of university students through sexual seduction of new members by more experienced cadre.)Later in the war, Sendero Luminoso forcibly recruited members, including children—a pattern also seen in Sri Lanka, where the LTTE requires at least one member from each Tamil household in areas they control. Insurgent organizations do not always draw recruits from social networks, however. The RUF in Sierra Leone forcibly recruited members without regard for local social networks, resulting in a heterogeneous force with weak internal social ties between members when recruited (Humphreys & Weinstein 2006). The organization’s pool of recruits may also reflect its resource base: Groups without economic resources are more likely to attract recruits willing to make long-term commitments to ideological goals,whereas those with income flows fromcontrolofresourcesaremorelikelytoattract opportunistic recruits (Weinstein 2006). In contrast, state militaries often attempt to draw or conscript recruits from a wide range of subcultures in order to build national unity (Weber 1976). For example, the Guatemalan military forcibly recruited indigenous soldiers as part of its program of national integration (Black 1984). In particular, the state may recruit disaffected members of an ethnic community on whose behalf an insurgent organization is seeking autonomy or secession, a practice called “ethnic defection” (Kalyvas 2008). For example, in Sri Lanka, rival Tamil militant groups defeated by the LTTE have become paramili-

tary organizations working with the state (including the Karuna faction, a splinter group of the LTTE in eastern Sri Lanka). In Peru, the arming of the rondas campesinas led to new ties between state agents and local indigenous residents in many highland communities despite previous state violence (Starn 1995, pp.562–63).Stateactorsmayalsobuildmilitia forces by militarizing a social, political, or ethnic cleavage, as in the case of Sri Lanka, where state efforts to develop Muslim Home Guards sharply increased the tension between Tamilspeaking Muslims and Hindus in eastern Sri Lanka and led to the LTTE’s decision to expel Muslims from northern Sri Lanka in 1990 [Hasbullah 2001, International Crisis Group (ICG) 2007]. Of course, not all civilians mobilize. Where violence is intense or indiscriminate, or comprises a broad repertoire, the social implications are likely to be severe for ordinary residents. Victims of political violence, particularly sexual violence and torture, often suffer not only from physical injuries but from shame, fear, ostracism, distrust (both of others and by others), and the inability both to remember what happened and also to forget (Levi 1988, Pedersen 2002, Human Rights Watch 2003, Denov 2006, Mookherjee 2006). As a result, victims frequently retreat from social interaction, living isolated from even their family and neighborhood. Thus, many traditional forms of mutual aid weakened or disappeared in the Peruvian highlands during the years of the war; the overall effect was to demobilize, not to mobilize. Yet, some of those who suffer political violence against themselves or familymembers mobilize rather than retreat. They found civil society organizations to advocate for the return of loved ones (or at least to learn what happened and to retrieve their bodies), or they join armed organizations. Thus the political mobilization of civilian networks into support networks for armed groups reshapes social networks. In El Salvador and Peru, for example, networks of insurgent supporters came to wield significant www.annualreviews.org • Social Processes of Civil War

545

power in some areas, displacing networks that linked clients to landlords and local citizens to government authorities. In the case of Peru, in many of those areas, networks later emerged linking residents to the state through the rondas campesinas . The dissolution of traditional networks is not always achieved by the emergence of new networks, however. The pattern of dissolution may be one of increased social isolation rather than new network ties, particularly where armed groups coerce support. As discussed below, political mobilizationalso indirectly reshapes social networks through the mechanisms of polarizationand militarization of local authority.


MILITARY SOCIALIZATION When political mobilization takes the form of mobilization in support of an armed group, the result is a distinctively military form of socialization that has consequences for both combatants and civilians (Bourke 1999). The consequences of military socialization for combatants are not well documented, but they surely include the effects of recruitment and training processes as well as the effects of witnessing and wielding violence (Sofsky 2003). Whether recruits of armed groups are volunteers or have been coerced, they have to be socialized in the use of violence for group, not private, purposes, if group leaders are to control the violence deployed by their combatants, typically through the building of strongly hierarchical organizations (Huntington 1957, Siebold 2001). Training and socialization to the armed group take place both formally, through the immersion experience of “boot camp” (surprisingly similar across state militaries and insurgent armies alike), and informally, through initiation rituals and hazing. In state militaries, the powerful experiences of endless drilling, dehumanization through abuse at the hands of the drill sergeant, and degradation followed by “rebirth” as group members through initiation rituals typically meld individual recruits into a cohesive unit in which loyalties to one 546

Wood

another are felt to be stronger than previous loyalties, such as those to family (Holmes 1985, Dyer 2004). Once deployed, combatants experience (to widely varying degrees) violence as perpetrators, as witnesses, and often as victims. Combatant memoirs consistently report the traumatizing effects of watching the death or injury of fellow combatants, as well as the harrowing effects for many of using violence themselves (e.g., O’Brien 1999, Beah 2007). Among the psychological mechanisms possibly at work in these processes of socialization to group membership andthe wielding of violence are compliance, role adoption, internalization of group norms, cognitive dissonance reduction, habituation to violence, diffusion of responsibility onto the group, deindividuation, and dehumanization of the victimized group (see Straus 2007 and Lynch 2007 for analyses of these mechanisms and the evidence concerning their contribution to mass killing). Some of these processes may also support prosocial psychological transformations, such as altruistic solidarity with nonkin, as in the emergence of a new insurgent political culture of solidarity in contested areas of El Salvador (Wood 2003). The profound effects that these processes can jointly exert on combatants are illustrated most sharply by the extreme violence deployed by forcibly conscripted child insurgents in Sierra Leone. Young recruits were forced by their RUF commanders to exert lethal violence, sometimes against family members; some were forced to commit sexual violence, a few against family members (Human Rights Watch 2003, pp. 35– 42; TRC 2004, Vol. 2, Ch. 2, p. 17 and Vol. 3A, Ch. 4, pp. 486–530ff ). Violence suffered, observed, and wielded was an integral part of the process of socialization of RUF combatants (Maclure & Denov 2006, Denov & Gervais 2007; see Honwana 2005 on Angola and Mozambique). As a result of the traumatic socializationthat young recruits received, many came to view their commanders as father figures, and the armed organization as family (Maclure & Denov 2006).


Similarly, the socialization processes of ordinary life were displaced and transformed among those fighting on the other side: The traditional rite of passage to adulthood for male youth in many areas became the induction ritual for joining the Civil Defense Forces, as the traditional kamajor societies were transformed into paramilitary militias (Hoffman 2003). The frequently observed widening of repertoires of violence over the course of the war likely reflects the ongoing effects of these underlying mechanisms, particularly dehumanization, diffusion of responsibility, habituation, and deindividuation, all of which are likely to undermine constraints on violence (Hoover 2007). As noted above, Sendero Luminoso was increasingly abusive toward civilians and coercive toward its own members as the war continued. This may have reflected the weakening of communication with the leadership after the group was forced to abandon initial strongholds as well as the group’s evolving strategy, which increasingly alienated former supporters. The apparent divergence in repertoire across Sendero units is consistent with this suggestion; as central control weakened, units developed particular elements of their repertoire. A widening of the repertoire of violence—in particular, the addition of sexual violence—was also evident on the part of the African National Congress and the Inkatha Freedom Party in their contest for control of KwaZulu Natal in the final years of the conflict in South Africa (and in the years immediately following; see Bonnin 2004). The effects of participation in and witnessing of violence in the South Africa case appear to have contributed to the undermining of traditional norms restraining sexual violence. Military socialization has consequences for civilians as well as combatants. As we will see in more detail below, the recruiting of combatants has various effects on local civil society: an increased emphasis on particular political identities; the flight of youths from their homes (and sometimes their country, as in the mass exodus of Salvadorans to the United

States) due to fear of forced recruitment and violence; changes in the demographic profile of rural households, particularly the shifting of new kinds of responsibility and labor to female members; and sometimes the presence of former combatants who return after their service. In El Salvador, for example, insurgent cadre sometimes moved between combatant and civilian roles; their presence in villages reinforced the civilian insurgent support net works (Wood 2003). Similarly, the presence of former soldiers in other villages reinforced networks of state loyalists. Recruits’ socialization into military life reshapes social networks in many ways. Rather than transitioning to adult life through traditional cultural rituals of maturation, apprenticeships to particular occupations, and participation in migrant labor networks, young recruits are socialized to adulthood through their integration into armed groups and the wielding of violence. In Uganda, former child soldiers were significantly disadvantaged by their loss of schooling and skill development, and a minority was significantly traumatized as well (Blattman & Annan 2007). An overarching pattern is the substitution of complex everyday ties, shaped by multiple overlapping networks of family, employment, and community, by ties with members of the armed group.

IDENTITY TRANSFORMATION AND POLARIZATION Violence, political mobilization, and military socialization usually have the joint effect of polarizing local identities, as is widely argued for the case of ethnic identities (Kaufmann 1996; Fearon & Laitin 2000), but the process is complex. Before conflict begins, it is often the case that local cleavages are distinct from those emphasized in the rhetoric of the parties to the conflict. What appears at the national level to be the key issue—for example, class relations, constitutions, or ethnic secession—may not be salient at the local level, which may be dominated by conflicts www.annualreviews.org • Social Processes of Civil War

547

between families or clans or other social groupings concerning particular local grievances (Harding 1984, Degregori 1999, Kalyvas 2003, Lubkemann 2008). Even where local cleavages are similar to national ones, local politics reflects the specific history of the locale and its actors. For example, in El Salvador, the cleavage was fundamentally one of class,as the poor, particularlythe rural poor, organized against the long-standing privileges of rural elites. Yet, landless laborers did not always support the insurgents, and mechanics and truck drivers numbered among insurgent supporters (Wood 2003). Civilian supportersof the insurgencydistinguished between “good” landlords and “bad” landlords based on the history of past relations. As extralocal actors begin mobilizing residents or carrying out violence against them, the local cleavage may increasingly align with the national one as a result of at least three processes. The first occurs when local actors choose sides opportunistically, perhaps as a means toward local advantage, as when one local faction denounces to one national party another faction as collaborators withthe other party (Kalyvas 2003, 2006). A second process is more straightforward: Civilians may have to support one party in order to gain protection against another. This situation occurs with particular intensity when violence by one party is indiscriminate; if one’s behavior provides no safety against indiscriminate violence, joining the other side may be the safest course (Mason & Krane 1989, Goodwin 2001, Kalyvas & Kocher 2007). Violence may of course be counterproductive in other ways, as when it leads to seeking re venge through membership in the opponent organization. Revenge is said to be a common motivation for Tamils to join the LTTE voluntarily, and other Tamils join rival groups or state forces to avenge violence by the LTTE. A third, often neglected mechanism of alignment is moral outrage in response to violence. As mentioned above, in the wake of extreme state violence, some hitherto nonvi-


548

Wood

olent activists in El Salvador joined the insurgent organization. In Sri Lanka, ethnic violence against “estate” Tamils (Tamils whose families immigrated from India beginning in the mid nineteenth century to work on the tea estates) led some to support the LTTE, which previously had drawn exclusively from so-called Sri Lankan Tamils (based largely in the north). In contrast to opportunistic or protection-seeking motives, outrage aligns not just public but private loyalties with one side of the national cleavage. More generally, civilians caught “between two fires” must choose—on inadequate information and in a context of high uncertainty— which side to support, particularly when combatants force residents to declare such choices. The armed parties can attempt to influence civilian choice with incentives as well as punishment. Even in largely ethnic conflicts, loyalties do not necessarily map onto the ethnic cleavage; states often attempt to build loyalist militias and civilian support bases by offering incentives (including protection) for ethnic defection, as we saw above in the transformation of rival Tamil militant groups into state paramilitary organizations. However, such polarization does not al ways occur. In contested areas of El Salvador, beginning in the mid 1980s, it was possible to be neutral (Wood 2003). Residents had to give water and sometimes food to combatants of either side passing through the neighborhood, and both sides acted against residents if they were identified as actively collaborating with the other side. But residents were not forced otherwise to choose ´ province, sides. In several areas of Usulut an for example, about one third of residents actively supported the FMLN beyond this coerced minimum, providing high-quality intelligence as well as supplies. Two thirds did not do so but were allowed nonetheless to remain in the area as long as they complied with the coerced minimum. This unusual pattern was possible because the conflict in the later years of the war pitted an insurgency that was unusually restrained in its use of violence against


a military that was dependent on US funding, which (thanks to popular mobilization that led to substantial congressional opposition to the war) the United States threatened to cut unless human rights violations by the state were severely curtailed. Even where polarization does occur, local public loyalties may reverse polarity. In the Andean highlands, as violence by Sendero cadre exceeded local norms, some villages reversed allegiance and aligned themselves with state forces. This pattern spread quickly as state violence became less repressive and Sendero’s violence more indiscriminate in the late 1980s (Degregori 1999, Theidon 2000, CVR 2003). One result of the general pattern of increasing polarization of public loyalties—and, to varying degrees, of private identities as well—is the increasing segregation of communities. One reason is, of course, flight. The amount of displacement of civilians in three of these conflicts is staggering. In El Salvador, by 1987 10% of the population was internallydisplaced, at least 10% was in the United States, and another 5% was in Mexico and other Central American countries (Gersony 1986; Montes 1987, p. 34). In Sierra Leone, at least 800,000 people ( 20% of the population) left their homes for elsewhere in the country and 400,000 for neighboring countries (Amowitz et al. 2002, p. 514; other estimates are significantly higher, suggesting that 50%ofthepopulationwasdisplaced).InPeru, 500,000–1,000,000people were displaced, the great majority from indigenous communities in the highlands (IDMC 2007a). In Sri Lanka, the levels of displacement are much lower: 460,000 people (2%) are displaced, with another 125,000 registered as refugees in other countries (IDMC 2007b, UNHCR 2007). Of course, not all were displaced by the direct exercise of coercion or even in an immediate context of fear; some left because they judged their life projects were best realized elsewhere (Lubkemann 2007, 2008). If one side is perceived as dominant, the flight of those who feel threatened by that ∼

∼

∼

∼

group leads to increasing homogeneity of the community as those alliedto the group remain behind. An armed party may attempt to create such homogeneity, forcing everyone but its allies out of the area. Such political “cleansing” occurs at the local level in many civil wars, not just in wars of ethnic secession or genocide. Organizations may also be purged, as when local communal bodies are taken over by residents loyal to one party, forcing those loyal to the other party to leave the area. In eastern Sri Lanka, for example, an area where before the war Tamil, Sinhalese, and Muslim families lived and worked together with little acrimony, villages tended to become increasingly homogeneous overthe courseof the war, as neighbors fled the occurrence or prospect of violence against them and their coethnics (ICG 2007). In some conflicts, urban neighborhoods also segregate along political lines, either because incoming individuals and families settle preferentially in areas where residents share political loyalties or aversions, or because armed actors prohibit them from settling in particular neighborhoods (Steele 2007). Even when residents of diverse loyalties remain in the area, mobilization, violence, and polarization often lead to increasing social segregation. In El Salvador, residents of contested areas who did not support the insurgents founded and joined local evangelical churches, in large part because they perceived the Catholic Church as supporting the insurgents. This perception was due to the role of liberation theology in the mobilization of poor people before the war and the extreme state violence against that movement and its advocates, including the assassination of Archbishop Oscar Romero (Wood 2003). The polarization of political identities that frequently occursduringcivil war breaks apart prewar networks, as former neighbors are shunned and coloyalists are favored. Even communities with high degrees of intermarriage before the conflict can split along cleavage lines, as occurred in Bosnia, Croatia, and Rwanda. The newly distinct networks may www.annualreviews.org • Social Processes of Civil War

549

also move through space as those with particular identities flee or segregate.

MILITARIZATION OF LOCAL GOVERNANCE In most civil wars, local authority becomes militarized in the wake of political mobilization, violence, and polarization. By militarization I mean the supplanting of local forms of governance with new forms that reflect the influence of armed actors. In some conflicts, local elites flee from contested areas. They may be particularly targeted by armed actors, and they are more likely to have the resources to travel and the means to resettle elsewhere, including urban properties and connections, and sometimes even visas for foreign travel, as in the exodus of Tamil professionals and their families from Sri Lanka beginning in 1983 (Human Rights Watch 2006). Their flight is an opportunity for others to fill their place— most likely actors bearing arms, given the risk and uncertainty that led the old elites to flee. Insurgents and their supporters may build new local orders in a variety of patterns, depending on insurgent strategy, community structure, and local political loyalties (Arjona 2007). The militarization of governance takes place in at least some areas of nearly all civil wars, as armed actors displace civil authorities or as hard-line military or paramilitary forces displace “softer” authorities such as police when conflict intensifies. Even when civilians continue to govern, their rule is often militarized. They may rely on their ties to coercive forces more than before, or the site of effective authority may be displaced to nearby military bases or camps. In Peru, by the mid 1980s, military bases ruled surrounding territory, often with extreme abuse of power including arbitrary detention, torture, and sexual violence of residents (CVR 2003). In Sri Lanka, the LTTE strictly controls local civilian administration; those thought to collaborate with the state or to support inadequately the LTTE are punished, and are sometimes assassinated, as


550

Wood

documented in many reports by the University Teachers for Human Rights ( Jaffna). In Colombia, both the leftist insurgents and the paramilitaries insist that municipal authorities follow certain rules and norms (Arjona 2008; see also Leal Buitrago 2002). Elections or other processes of selection may be held in circumstances deemed illegitimate by some residents, as favored candidates or parties may not participate. Alternatively, elections may be deferred for years, leaving in place nominal leaders seen as illegitimate given their absence from the locale or their inadequate management of wartime challenges. In El Salvador, for example, wartime municipal elections were often held not in the municipality but in another town or city, with candidates competing who no longer lived any where near the municipality. Usually only a fraction of eligible voters participated, a fraction that disproportionatelyrepresentedthose no longer residing in the municipality. Where armed actors displace civil authorities, a common result is a generational in version of authority, by which I mean armed youth supplanting traditional elders and local authorities. This occurred in the Peru vian highlands (Degregori 1999, p. 64), areas of Sierra Leone dominated by the RUF (TRC 2004 Vol. 2, Ch. 2, pp. 34, 45), and Darfur (Flint & de Waal 2005). This inversion may mean that traditional social norms constraining violence are no longer enforced. In Medell´ ın, Colombia, young members of rival armed groups at one point controlled different neighborhoods, severely restricting civilian lives in some, as depicted in the documentary film La Sierra. The inversion may also occur along class or ethnic lines, particularly where insurgent groups govern the area, if the insurgents draw chiefly from subordinate ethnic or class groups. In areas controlled or contested by the FMLN, for example, ordinary residents came to exert unprecedented authority in the absence of both economic elites and civil authorities. Their authority was limited, to be sure, by the overarching authority of the FMLN,


and in contested areas by that of the military as well. Nonetheless, local residents made decisions about who could reside in the area and who could not, what land would be occupied (some surreptitiously, others formally), and whether to allow particular former residents, including landlords and the mayor, to return (Wood 2003). In Peru, generational conflict brought enormous violence to some areas of the highlands, as armed teenagers allied with Sendero Luminoso revolted against their parents and other traditional authorities (CVR 2003, Vol. 1, Ch. 2, pp. 95–96; Vol. 2, Ch. 1, p. 451). Insurgent governance in many areas also had ethnic and class connotations, as insurgent-allied residents who ruled some towns briefly in the early 1980s were generallybothmorepoorandmoreindigenousthan those they supplanted. Yet generational inversion does not always occur. In southern Sierra Leone, traditional militias, the kamajors , became effective civil defense forces (TRC 2004, Vol. 2, Ch. 2, pp. 76–78). The groups initially relied on traditional indoctrination rituals under the leadership of respected elders.However, those ties to tradition were increasingly disrupted by a particular elder who came to control many militias, introducing distorted, violent rituals that other elders saw as a perversion of traditional practices (TRC 2004, Vol. 2, Ch. 2, pp. 77–79). Insurgent governance takes an unusual form in Sri Lanka. Beginning in 1987 and intensifying after 1990, the LTTE developed an extensive civil administration in areas it controls (Byman et al. 2001, Wayland 2004, Stokke 2006, Mampilly 2007). The insurgent administration draws on the state administration; the group cooperates with the government civil service in providing health care and education. (The state cooperates with this arrangement because to refuse would mean that no services were provided by the state, thereby strengthening the LTTE’s argument that a separate Tamil state was necessary.) However, the insurgents maintain exclusive control over security, including new policing and

judicial agencies. They appear to place particular emphasis on education, insisting on curricula that focus on Tamil history, grievances, and aspirations. They also require participation in military training: All schoolchildren over the age of 14 are compelled to participate in exercises including security roles and mock battles (de Mel 2007). The group carries out “social cleansing,” threatening or killing local rapists and thieves, and institutes sanctions for domestic violence (Gomez 2005). The LTTE collects taxes from civil ser vants (12% of their government-paid salary) and professionals in areas it controls, and in some other areas of the country as well, and also taxes members of the Tamil diaspora, maintaining elaborate records of contributions, incomes, and promises of deferred payments (Human Rights Watch 2006). In Peru, Sendero Luminoso initially carried out locally popular acts of social cleansing but soon turned to political cleansing, often via public accountability sessions, in which local communities voiced complaints about local leaders—complaints often followed by their murder, despite calls for punishment instead (Degregori 1999). The example of civilians in El Salvador making local decisions—with the implicit backing of the FMLN—indicates that the degree of militarization of local authority varies significantly. Indeed, there are important exceptions to this general pattern. In El Salvador, the town of Tenancingo—twice deserted owing to conflict in the town itself, including bombing by the Air Force—was resettled as a zona inerme (unarmed zone) under an agreement between the FMLN and the military negotiated under the auspices of the Archdiocese of San Salvador (Wood 2003). Although both sides soon returned to carrying arms through town and conflict resumed very nearby, the town did not sufferfrom a resumption of severe conflict through the remaining years of the war. In Colombia, some communities have declared themselves comunidades de paz in an effort to protect themselves from the incursions of armed actors, with mixed results www.annualreviews.org • Social Processes of Civil War

551

(Uribe de Hincapie 2004). Some indigenous generally unmarried girls and young women, communities have succeeded in limiting the reflecting their lesser family responsibilities incursionof armed actors, includingby found- (what sociologists term biographical availabiling their own armed group (Pe ˜ naranda 1999; ity), particularly those living in refugee camps, A. Steele, personal communication). where social expectations of joining were high The militarization of local authority of- (Viterna 2006). ten displaces prewar governance networks, In Sri Lanka, women and girls comprise replacing them with new ties between some separate units, train in their own bases, and residents and the new local authority (or au- carry out missions separately as well as jointly thorities,if dual authorities arepresent—afre- with male units (Ann 1993, Balasingham 2001, quent occurrence in El Salvador and Peru). Trawick 2007). Women also carry out suicide The new governance network may be directed missions, including the assassination of Rajiv by civilians but usually relies on armed actors Gandhi (Narayan Swamy 2004). The motivafor their coercive authority and sometimes tions of girls and women to join the LTTE their legitimacy, as in El Salvador, where local appear broadly similar to those of their male civilians worked closely with the FMLN. counterparts, including revenge for state violence and nationalism. The emergence of feminist aspirations among some experienced THE TRANSFORMATION cadre apparently reflects their wartime expeOF GENDER ROLES rience rather than their motivation for join War transforms gender roles through a va- ing (Alison 2003). The presence of female riety of mechanisms. The most dramatic, of combatants does not appear, however, to discourse, is the carrying of arms by female insur- place male combatants’ notions of womangents, who comprised 30% of combatants hood; male cadre appear to prefer traditional in the Peruvian, Salvadoran, and Sri Lankan Tamil women and girls, not their fellow cominsurgencies and 25% in the Sierra Leone batants, as wives. Although female cadre dress insurgency, unusually high fractions among in guerrilla uniforms with trousers and frearmed groups (Mason 1992, p. 250; Barrig quently have short hair, the LTTE enforces 1993, cited in CVR 2003, Vol. 8, Ch. 2.1, traditional Tamil dress among noncombatant p. 56; Bouta 2005, p. 7; Humphreys & girls and women (Trawick 2007). Weinstein 2004, p. 14). Female insurgents Civilian gender roles may also change drabroke traditional social norms in these soci- maticallyduringwar. In El Salvador, Peru, and eties. Some provided household, logistic, or Sri Lanka, women became the primary intermedical services, but some served as com- locutors with the state as they sought news of batants; the mix varied across groups. Al- their detained or disappeared menfolk. Parthough women were among the members ticularly in Peru, where indigenous women of of high-level command councils in Peru, El the highlands were much less likely to speak Salvador, and Sri Lanka, in Peru and El Spanish and had little prior contact with state Salvador women generally served under male authorities, carrying out this necessary task commanders (with more exceptions in Peru was fraught with risk and suffering (CVR than El Salvador). In Peru, access to power 2003, Vol. 8, Ch 2.1, Theidon 2007). Sigappeared to depend largely on sexual relation- nificant numbers of women inquiring about ship with powerful men (including Guzm´ an; male relatives were themselves detained, and CVR 2003). In El Salvador, that was not the some suffered sexual violence while in detencase; rather, women high in the insurgent tion (CVR 2003). Some founded or joined organization were either founding members human rights groups; for most that did of the FMLN’s constituent organizations or so, this was an unprecedented role in civil advanced through the ranks. Recruits were society (Cordero 1998). In Sri Lanka, some


∼

∼

552

Wood


Tamil widows do not comply with the tradi- in the contested areas of wartime El Salvador tional social norms of widowhood, refusing (Wood 2003). The exception to this pattern to retire into the private sphere and insisting of foodstuff production and the feminization on wearing the putee, a traditional mark on of agricultural labor is the production of illicit the forehead that is usually denied to widows crops such as opium or coca in areas of Peru, (Rajasingham-Senanayake 2001, pp. 106–7). Colombia, Burma, and Afghanistan. Even in Patriarchal networks are often radically re- those cases, production occurs predominantly shaped during war because women and girls on smallholdings, whose small scale renders take on unprecedented roles as combatants them less likely to be detected by state auand interlocutors with authority, and (as dis- thorities or rival groups. cussed below) take on new forms of work. The In contrast to the general pattern of supextent of the transformation of gendered net- pression of most markets, conflict processes works varies greatly across areas and conflicts may fuel land markets. Landlords may be willbut is nonetheless frequently evident in the ing to sell if productive use of their properties high fraction of female-headed households is unlikely, if rent payments are low or nonexboth in contested areas and in areas where istent, and if buyers are willing to pay an acrefugees and displaced persons congregate. In ceptable price. In El Salvador during the war, El Salvador, female-headed households com- very few buyers were willing to purchase large prised about one quarter of displaced house- properties (those not nationalized under the holds in 1985 (USAID 1987). Many female- agrarian reform) at a price the owner would headed households are deeply impoverished accept because the buyer would face the same despite girls and women taking on new roles wartime risks as the current owner did. Some as landless laborers and farmers. land-poor and landless residents received remittances from family members who had fled to the United States. Thus, the subdivision of FRAGMENTATION OF THE a large property for local sale provided a larger LOCAL POLITICAL ECONOMY return than sale of the entire property (Wood With the displacement of local elites, the sup- 2000). planting of traditional authorities, and the deIn El Salvador, land markets flourished parture of household members to serve in in the later years of the war around small armed groups, newpatterns of production and cities and towns. There were two princilabor emerge. Particularly in highly contested pal reasons: first, the formal occupation of rural areas, household members make difficult nearby properties by insurgent supporters, choices in the wartime context of uncertainty coupled with landlord perceptions that labor relations would be difficult at the war’s and risk (Collier 1999, Humphreys 2003). One pattern in some conflicts is the “peas- end; and second, the ability of poor resiantization” of formerly commercial agricul- dents to buy, thanks to remittances from famture. In the absence of elite investment and ily members in the United States. Although management, large properties are cultivated many properties were subdivided and sold for in small plots planted by individual rural housing, others were sold for smallholdings. households. Because foodstuffs are increas- Figures 1 and 2 illustrate the changes ingly unavailable for purchase in markets, ei- wrought in the landscape around the town ther because markets do not function or be- of San Jorge, a small town on the shoulcause cash income has declined, household der of a volcano in eastern El Salvador. members typicallyplant food ratherthan mar- These maps (originally published in Wood ket crops, either with the permission of an ab- 2003) were drawn for me at the war’s end sent landlord (negotiated via some local agent by members of the Cooperativa Candeor in risky trips to urban areas), or illegally, as laria, Un Nuevo Amanecer (“a new dawn”). www.annualreviews.org • Social Processes of Civil War

553

Before the war (Figure 1), the area consisted mainly of medium-sized coffee estates. The town is shown at the lower edge of the map. Each field and grove are labeled with the name of the owner, and the crop is indicated with symbols. Figure 2 is the cooperativists’ representation of the postwar landscape.Militant campesinos founded the cooperative during the war and claimed 322 hectares of the now-rundown coffee estates at the war’s end, including a significant fraction of the coffee groves north of town. Those properties are labeled “ propiedad de la cooperatiba” [sic] (“property of the cooperative”). The abandoned coffee estates (shown at the top of Figure 2) were more forest than farm, as indicated by the serpent near the upper right corner, the brokencoffee branches near the upper left, and notations such as “ propiedad destruida” (destroyed property) and “vosque destruido” [sic] (destroyed forest). Corn was more widely planted (not evident in the figure, but easily noted when I visited the cooperative). In other settings, the large earnings of narco-traffickers may fuel markets in large properties. In Colombia, for example, illicit crops are produced on smallholdings, but earnings often go toward property acquisition (Reyes 1997, Romero 2003). Insurgentcontrolled economic activities may lead to the emergence or strengthening of markets, as in the illicit production and transport of coca in some areas of Peru (Weinstein 2006). In northern Sri Lanka, smugglers regularly cross the straits between the Indian coast and LTTE-held areas, transporting arms, cadre, and supplies (Balasingham 2001, Narayan Swamy 2004). A transformation of agrarian land and social relations may also be imposed by armed actors. Insurgents may require the production of certain crops or particular labor processes, as when the FMLN attempted to introduce some collective cultivation in areas it controlled in northern El Salvador in the early years of the war. In Peru, both the insurgents and the military repeatedly confiscated livestock, decimating the savings of ru-


554

Wood

ral households, a pattern evident in many of the testimonies presented to the Comisi´ on de ´ Sendero Luminoso Verdad y Reconciliaci on. not only attempted to impose collective production in some areas of the highlands but also attempted to forbid peasant participation in markets (Starn 1995, Degregori 1999, p. 66). The state, too, forcibly resettled some communities as a counterinsurgency measure (Starn 1995). In El Salvador, the state carried out a land reform in an effort to undermine the FMLN’s appeal. Approximately a quarter of arable land was turned over to landless workers and small tenants, including hundreds of large estates where coffee, cotton, and beef had been produced before the war. Although the initial plans called for the continuation of large-scale production under a cooperative model with access to credit from the state, within five years most properties were worked as smallholdings by cooperative members, as corruption and inadequate management quickly led to high operational debts. A distinctive part of the reform is evident in Figure 2: Two properties in San Jorge were distributed to residents in the early 1980s under the “land-tothe-tiller” phase of the agrarian reform; one is labeled “ parcelas de FINATA,” (the acronym of the administering agency) and the other simply “FINATA.” Counterinsurgency efforts may also take the form of development projects other than agrarian reform, including credit and technical support for crops to replace illicit crops whose marketing may fund armed groups, infrastructure development, and the provision of health services. The classic literature on counterinsurgency calls for such nonmilitary means to win over civilian populations (see, e.g., Trinquier 1964 and Galula 1964). In addition to the agrarian reform, the Salvadoran government repeatedly attempted to resettle displaced families in new villages at strategic sites. These efforts were unsuccessful because the families were often displaced again, or their loyalties to the state proved less reliable than assumed (Wood 2003). Beneficiaries


of the agrarian reform did not necessarily become loyal to the state; insurgent combatants used to rest and regroup on several land-reform estates in southern El Salvador. In eastern Sri Lanka, the state favors the land claims of Sinhalese farmers, often relatively recent arrivals resettled in the area through various state development programs. This policy has led to increasing conflict with Tamil families over land and ongoing displacement of Tamil families from their lands andvillages(Peebles1990,ICG2007).Incontrast, the Peruvian state initiated relatively few development projects during that country’s civil war, and those displaced by the conflict were largely left to fend for themselves in the peripheries of Lima and provincial cities (Kirk 2005). In Colombia, a renewal of conflict was triggered by the rapid expansion of palm oil production on land abandoned by civilians who left the area because of the war (IDMC 2007c). War fragments most rural markets, as commercial production stagnates, input and output transportation networks cease to function, and financial institutions withdraw services. Also fragmented, as a result, are the social networks linking owners of large properties to their workers and to providers of inputs. Similarly, the socialnetworks that undergird public and private service provision to the rural poor also wither. Residents of contested areas thus increasingly turn to family production of foodstuffs and services. In the wartime climate of uncertainty, distrust, and polarization, traditional social networks of mutual aid may likewise weaken. In some cases, new social networks with economic functions emerge, as in the case of the insurgent cooperatives in El Salvador, and the illicit networks of coca production and transport in Peru and of sea transport between the Indian coast and LTTE-held areas of northern Sri Lanka.

ways in the civil wars in El Salvador, Peru, Sierra Leone, and Sri Lanka. Political mobilization, military socialization, the polarization of social identities, the militarization of local authority, the transformation of gender roles, and the fragmentation of local political economies reshaped a wide range of local social networks, destroying some, breaking others into subnetworks, and creating new ones. What, if any, enduring legacy do these processes leave for the postwar period? The recent emergence of debate in Spain about whether and how the civil war should be memorialized, particularly the controversy over the excavation of mass graves, suggests that the legacy of mobilization, violence, polarization, and the militarization of authority may have lain dormant through the years of dictatorship and initial democratic rule. Indeed, Balcells Ventura (2007) found that patterns of lethal violence during Spain’s civil war were associated with voting patterns in elections in Catalonia four decades later. Any assessment of legacy should be very cautious about drawing firm conclusions. For example, the transformation of gender roles during war is often reversed once war draws to a close. Social norms may reassert themselves, with women leaving or being pushed from theirjobsasmenreturntocivilianroles,aswas the case in both the United States and France after World War II. Institutional actors, including those working toward reconciliation, may structure incentives or procedures such that women’s new roles are undermined, as when land transfer programs neglect to issue titles to women as well as men, and demobilization programs favor male combatants over females, whose role may not be recognized. Women may be displaced from leadership positions in political parties and civil society organizations when male combatants, particularly officers, return to civil society. Nonetheless, sometimes the social processes of civil war leave enduring changes CONCLUSION in their wake. In El Salvador, war left a The wartime social processes discussed here legacy of leftist civil society organizations occurred to varying degrees and in different and a new insurgent political culture based www.annualreviews.org • Social Processes of Civil War

555

on substantive democratic and redistributive claims in areas of strong FMLN influence (Wood 2003). Indeed, the transformation of the FMLN into a political party marked an unprecedented representation of the political left. The process was conflictual, given the group’s complex internal structure and ideological diversity. Although the party has yet to win the presidency, it performs well in municipal and legislative elections, repeatedly serving as the lead opposition party in parliament. In other cases, in sharp contrast, the legacy of wartime polarization may inhibit the reintegration of combatants and their supposed supporters. In Peru, there is significant opposition to any recognition of the suffering of family members of Sendero combatants, seen by many as terrorists whose reintegration is neither morally desirable nor politically palatable. A generation of local leaders was decimated by the war; community leaders were assassinated for their supposed political loyalties, particularly by Sendero cadre, both in the highlands and Lima neighborhoods (Burt 1998,CVR2003).InSriLanka,thesocialprocesses of war have largely eliminated any articulation of Tamil autonomy except for that embraced by the LTTE. Nascent class politics remains eclipsed by ethnic politics, as the two Sinhalese political parties repeatedly play the “Tamil threat” card to mobilize followers, rather than appeals along other lines. A common pattern of enduring change is demographic. War changes the population structure of countries,often in profound ways. In El Salvador, for example, the rural to urban ratio of 60:40 was reversed over the course of


the conflict, a significant acceleration of the prewar rate of urbanization. It appears that war may also accelerate the transition to nuclearhouseholdsfrommoreextendedformsas family networks disperse during the war. War often leaves an increased number of femaleheaded households, not only because of the higher death rate of males but also because of the dispersion of families and the disruption of stable labor migration patterns. Labor relations may be transformed, even if there is some rollback to prewar forms. It is not always the case that women and girls retire to their prewar roles; they may remain to some degree in new occupations, particularly in the absence of male partners. And women may continue in leadership roles in new civil society organizations such as human rights groups, particularly in settings where combatants of neither side are welcomed into civil society, as in Peru. This article raises many more questions than it answers. Further research on the social processes of civil war would be a valuable contribution. Under what conditions does each process occur with particular force? Under what conditions do these processes have enduring, important consequences? To what extent do answers depend on who wins the war? To what extent do they depend on the intensity of conflict, or its duration? Is it possible to distinguish the legacies of distinct processes of war? The challenge will be to develop research designs that allow the untangling of the consequences of distinct processes, a challenge that may be best addressed by research strategies that take advantage of subnational variation.

DISCLOSURE STATEMENT The author is not aware of any biases that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS I thank Ana Mar´ ıa Arjona, Laia Balcells Ventura, Regina Bateson, Samuel Bowles, Francisco Guti´ errez San´ ın, Amelia Hoover, Sashini Jayawardane, Meghan Lynch, and Abbey Steele for comments. I thank Emma Einhorn and Sashini Jayawardane for research assistance, and the 556

Wood

Harry Frank Guggenheim Foundation, the United States Institute of Peace, Yale University’s MacMillan Center for International and Area Studies, and the Santa Fe Institute for funding of research on related projects. The opinions, findings, and conclusions or recommendations expressed in this article are those of the author and do not necessarily reflect the views of the United States Institute of Peace.

LITERATURE CITED


Alison M. 2003. Cogs in a wheel? Women in the Liberation Tigers of Tamil Eelam. Civil Wars 6:37–54 Amowitz LL, Reis C, Hare Lyons K, Vann B, Mansaray B, et al. 2002. Prevalence of warrelated sexual violence and other human rights abuses among internally displaced persons in Sierra Leone. J. Am. Med. Assoc. 287:513–21 Ann A. 1993. Women Fighters of Liberation Tigers . London: LTTE Int. Secretariat Arjona A. 2008. One national war, multiple local orders: an inquiry into the unit of anal ysis of war and post-war interventions. In Law in Peace Negotiations , ed. M Bergsmo, P Kalmanovitz. Oslo: Peace Res. Inst. In press Arjona AM. 2007. Local orders and civilian collaboration in civil war . Work. pap., Dep. Polit. Sci., Yale Univ. Balasingham A. 2001. The Will to Freedom. Mitcham, UK: Fairmax Balcells L, Kalyvas S. 2007. Warfare in civil wars . Work. pap., Dep. Polit. Sci., Yale Univ. Balcells Ventura L. 2007. The effects of wars on political identities: ideological change and continuity after the Spanish Civil War (1936–39). Presented at Annu. Meet. Am. Polit. Sci. Assoc., 103rd, Chicago, IL Ball P, Asher J, Sulmont D, Manrique D. 2003. How Many Peruvians Have Died? An Estimate of the Total Number of Victims Killed or Disappeared in the Armed Internal Conflict Between 1980 and 2000. Washington, DC: Human Rights Data Anal. Group Am. Assoc. Advance. Sci. (also published in Spanish as Annex 2 to CVR 2003). Barrig M. 1993. Liderazgo feminino y violencia pol´ ıticaenelPer´ udelos90. Debates en Sociolog´ ıa 18:96–97 Beah I. 2007. A Long Way Gone: Memoirs of a Boy Soldier . New York: Farrar, Straus & Giroux Black G, with Jamail M, Chinchilla NS. 1984. Garrison Guatemala. New York: Monthly Review Press, 1984 Blattman C, Annan J. 2007. The consequences of child soldiering . Households in Confl. Netw. Work. Pap. 22 Bonnin D. 2004. Understanding the Legacies of Political Violence: an Examination of Political Conflict in Mpumalanga Township, Kwa-Zulu Natal, South Africa . London: London School Econ., Crisis States Prog. Bourke J. 1999. An Intimate History of Killing . Face-to-Face Killing in Twentieth Century Warfare. London: Granta Books Bouta T. 2005. Gender and Disarmament, Demobilization and Reintegration: Building Blocks for Dutch Policy. Netherlands: Clingendael Confl. Res. Unit Brockett CD. 1990. Land, Power, and Poverty: Agrarian Transformation and Political Conflict in Central America. Boston: Unwin Hyman Burt J-M. 1998. Shining Path and the “decisive battle” in Lima’s Barriadas : the case of Villa El Salvador. In Shining and Other Paths. War and Society in Peru, 1980–1995 , ed. SJ Stern, pp. 267–306. Durham/London: Duke Univ. Press www.annualreviews.org • Social Processes of Civil War

557

Byman D, Chalk P, Hoffman B, Rosenau W, Brannan D. 2001. Trends in Outside Support for Insurgent Movements . Santa Monica, CA: Rand Corp. Collier P. 1999. On the economic consequences of civil war. Oxf. Econ. Pap. 51:168–83 ´ de Verdad y Reconciliacion ´ (CVR). 2003. Informe Final . Lima, Peru: Comisi´ Comision on de Verdad y Reconciliaci´ on Cordero IC. 1998. Women in war: impact and responses. In Shining and Other Paths. War and Society in Peru, 1980–1995 , ed. SJ Stern, pp. 345–74. Durham/London: Duke Univ. Press de Mel N. 2007. Militarizing Sri Lanka: Popular Culture, Memory and Narrative in the Armed Conflict . Los Angeles/London/New Delhi/Singapore: Sage Degregori CI. 1999. Reaping the whirlwind: the Rondas Campesinas and the defeat of Sendero Luminoso in Ayacucho. In Societies of Fear: The Legacy of Civil War, Violence and Terror in Latin America, ed. K Koonings, D Kruijt, pp. 63–87. London/New York: Zed Books Del Pino P. 2005. Peasants at war. In The Peru Reader. History, Culture, Politics , ed. O Starn, CI Degregori, R Kirk, pp. 393–400. Durham/London: Duke Univ. Press Denov MS. 2006. Wartime sexual violence: assessing a human security response to war-affected girls in Sierra Leone. Sec. Dialogue 37(3):319–42 Denov M, Gervais C. 2007. Negotiating (in)security: agency, resistance, and resourcefulness among girls formerly associated with Sierra Leone’s Revolutionary United Front. Signs: J. Women Cult. Soc. 32(4):885–910 Dyer G. 2004. The Lethal Custom. New York: Carroll & Graf Ellis S. 1995. Liberia 1989–1994. A study of ethnic and spiritual violence. Afr. Aff. 94:165–97 Ellis S. 1999. The Mask of Anarchy—The Destruction of Liberia and the Religious Dimensions of an African Civil War . New York: NY Univ. Press Fearon JD, Laitin D. 2000. Violence and the social construction of ethnic identity. Int. Organ. 54(4):845–77 Flint J, de Waal A. 2005. Darfur: A Short History of a Long War . London/New York: Zed Books Fujii LA. 2009. Killing Neighbors: Social Dimensions of Genocide in Rwanda. Ithaca, NY: Cornell Univ. Press Galula D. 1964. Counterinsurgency Warfare: Theory and Practice. New York: Praeger a, Regi´ on, Actores y Conflicto 1960–1990 . Medell´ ın: INER-CEREC Garc´ ıa CI. 1996. Urab´ Gersony R. 1986. The Journey Home: Durable Solutions for Displaced Families . San Salvador: US Agency Int. Dev. Gomez S. 2005. A Study on Gender Based Violence in the Batticaloa District . Sri Lanka: Care Int. Goodwin J. 2001. No Other Way Out: States and Revolutionary Movements, 1945–1991 . New York: Cambridge Univ. Press Harding SF. 1984. Remaking Ibieca: Rural Life in Aragon Under Franco. Chapel Hill: Univ. North Carol. Press Hasbullah SH. 2001. Muslim Refugees: The Forgotten People in Sri Lanka’s Ethnic Conflict . Nuraicholai, Sri Lanka: Res. Action Forum Soc. Dev. Hoffman D. 2003. Like beasts in the bush: synonyms of childhood and youth in Sierra Leone. Postcolonial Stud. 6(3):295–308 Holmes R. 1985. Acts of War: The Behavior of Men in Battle. New York: Free Press Honwana A. 2005. Child Soldiers in Africa . Philadelphia: Univ. Penn. Press Hoover A. 2007. Explaining the form and extent of wartime violence against combatants: the role of armed group institutions and ideologies . Work. pap., Dep. Polit. Sci., Yale Univ. Human Rights Watch. 2003. Sierra Leone. “We’ll kill you if you cry.” Sexual violence in the Sierra Leone conflict. New York: Human Rights Watch Human Rights Watch. 2006. Funding the “Final War”: LTTE Intimidation and Extortion in the Tamil Diaspora. New York: Human Rights Watch


558

Wood

Humphreys M. 2003. Economics and violent conflict . Work. pap., Humanitarian Policy Confl. Res., Harvard Univ. http://www.preventconflict.org/portal/economics Humphreys M, Weinstein J. 2004. What the fighters say: a survey of ex-combatants in Sierra Leone, June–August 2003. Work. pap., Cent. Globaliz. Sustain. Dev., Columbia Univ. http:// www.stanford.edu/ jweinst/docs/manuscripts/humphreys combatantsurvey.pdf Humphreys M, Weinstein JM. 2006. Handling and manhandling civilians in civil war. Am. Polit. Sci. Rev. 100:429–47 Huntington SP. 1957. The Soldier and the State. The Theory and Politics of Civil-Military Relations . Boston: Belknap Press of Harvard Univ. Press International Crisis Group. 2007. Sri Lanka’s Muslims: Caught in the Crossfire. Asia Rep. No. 134. http://www.crisisgroup.org/library/documents/asia/south asia/sri lanka/134 sri lanka s muslims caught in the crossfire.pdf IDMC.2007a. Compensation Pending for Victimsof Forced Displacement . Geneva: Intern. Displace. Monit. Cent., Norwegian Refugee Counc. IDMC. 2007b. Civilians in the Way of Conflict: Displaced People in Sri Lanka . Geneva: Intern. Displace. Monit. Cent., Norwegian Refugee Counc. IDMC. 2007c. Colombia: resisting displacement by combatants and developers: humanitarian zones in north-west Colombia. A profile of the internal displacement situation . Geneva: Intern. Displace. Monit. Cent., Norwegian Refugee Counc. Kalyvas S. 2003. The ontology of “political violence”: action and identity in civil wars. Perspect. Polit. 1:475–94 Kalyvas S. 2006. The Logic of Violence in Civil War . New York: Cambridge Univ. Press Kalyvas SN. 2008. Ethnic defection in civil war. Comp. Polit. Stud. In press Kalyvas SN, Kocher M. 2007. How “free” is free riding in civil wars? World Polit. 59:177–216 Kaufmann C. 1996. Possible and impossible solutions to ethnic civil wars. Int. Sec. 20(4):136–75 Kirk R. 2005. Chaqwa. In The Peru Reader. History, Culture, Politics , ed. O Starn, CI Degregori, R Kirk, pp. 370–83. Durham/London: Duke Univ. Press Leal Buitrago F. 2002. La Seguridad Nacional a la Deriva—Del Frente Nacional a la Posguerra ´ Alfa omega-Ceso-Uniandes, FLACSO-Sede Ecuador Fr´ ıa. Bogot a: Leal Francisco DA. 1990. Clientelismo. El sistema pol´ ıtico y su expresi´ on regional . Bogot a: ´ Tercer Mundo Ed.-Iepri Levi P. 1988. The Drowned and the Saved . New York: Summit Books Lubkemann S. 2007. Refocusing the ethnographic gaze in warscapes: war as social condition. Presented at Annu. Meet. Am. Polit. Sci. Assoc., 103rd, Chicago, IL Lubkemann SC. 2008. Culture in Chaos: An Anthropology of the Social Condition in War . Chicago: Univ. Chicago Press Lynch M. 2007. Am I my brother’s killer? The social psychology of violence against civilians . Work. pap., Dep. Polit. Sci., Yale Univ. Lynch M, Hoover A. 2007. Why multiple data systems matter: evidence from Colombia . Presented at Conf. Civil Confl. and Polit. Violence, Weatherhead Cent. Int. Aff., Harvard Univ., Apr. 28 Maclure R, Denov M. 2006. “I didn’t want to die so I joined them”: structuration and the process of becoming boy soldiers in Sierra Leone. Terrorism Polit. Violence 18:119–35 Mampilly ZC. 2007. Stationary bandits: understanding rebel governance. PhD thesis. Univ. Calif. Los Angeles Mason TD. 1992. Women’s participation in Central American revolutions: a theoretical perspective. Comp. Polit. Stud. 25(1):63–89 Mason TD, Krane DA. 1989. The political economy of death squads: toward a theory of the impact of state-sanctioned terror. Int. Stud. Q. 33(2):175–98 ∼



559

Montes SM. 1987. El Salvador 1987 . San Salvador: Inst. Investigaciones e Inst. Derechos ´ Ca ˜ Humanos de la Univ. Centroam´ erica de El Salvador “Jose Simeon nas” Mookherjee N. 2006. Remembering to forget: public secrecy and sexual violence in the Bangladesh War of 1971. J. R. Anthropol. Inst. 12(2):433–50 Narayan Swamy MR.2004. Tigers of Lanka. From Boys to Guerrillas . Colombo, SriLanka: Vijitha Yapa O’Brien T. 1999. If I Die in a Combat Zone: Box Me Up and Ship Me Home . New York: Random House Pedersen D. 2002. Political violence, ethnic conflict, and contemporary wars: broad implications for health and social well-being. Soc. Sci. Med. 55:175–90 Peebles P. 1990. Colonization and ethnic conflict in the dry zone of Sri Lanka. J. Asian Stud. 49(1):30–55 Pe ˜ naranda R. 1999. De rebeldes a ciudadanos: el caso del Movimiento Armado Quint ´ ınlame.In ´ IEPRI-Tercer De Las Armas a la Pol´ ıtica,ed.RPe ˜ naranda, J Guerrero, pp. 75–132. Bogot a: Mundo Rajasingham-Senanayake D. 2001. Ambivalent empowerment: the tragedy of Tamil women in conflict. In Women, War and Peace in South Asia , ed. R Manchanda, pp. 102–30. New Delhi: Sage ıcitas en Colombia. Su Impacto Reyes A. 1997. Compra de tierras por narcotraficantes. In Drogas Il´ Econ´ omico y Social , ed. F Thoumi, S Uribe,R Rocha,A Reyes,E Garz´ on, et al., pp. 279–346. ´ PNUD Bogot a: ´ Inst. Estud. Pol´ ıticos y Romero M. 2003. Paramilitares y autodefensas, 1982–2003. Bogot a: Relaciones Internacionales, Univ. Nacional Colombia Seligson M, McElhinny V. 1996. Low-intensitywarfare, high-intensity death:the demographic impact of the wars in El Salvador and Nicaragua. Can. J. Lat. Am. Caribb. Stud. 21(42):211– 41 Siebold GL. 2001. Core issues and theory in military sociology. J. Polit. Mil. Sociol. 29:140–59 Silva R, Ball P. 2006. The Profile of Human Rights Violations in Timor-Leste, 1974–1999 . Rep. by the Benetech Human Rights Data Analysis Group to the Commission on Reception, Truth and Reconciliation. 9 Feb. http://www.hrdag.org/resources/Benetech-Reportto-CAVR.pdf Sofsky W. 2003. Violence: Terrorism, Genocide, War . London: Granta Somasundaram D. 1998. Scarred Minds. The Psychological Impact of War on Sri Lankan Tamils . New Delhi/Thousand Oaks/London: Sage Stanley W. 1996. The Protection Racket State: Elite Politics, Military Extortion, and Civil War in El Salvador . Philadelphia: Temple Univ. Press Starn O. 1995. To revolt against the revolution: war and resistance in Peru’s Andes. Cultur. Anthropol. 10(4):547–80 Steele A. 2007. Seeking safety: urban and rural displacement in civil wars. Work. pap., Dep. Polit. Sci., Yale Univ. Stokke K. 2006. Building the Tamil Eelam state: emerging state institutions and forms of governance in LTTE-controlled areas in Sri Lanka. Third World Q. 27:1021–40 Stoll D. 1993. Between Two Armies in the Ixil Towns of Guatemala . New York: Columbia Univ. Press Straus S. 2007. Second generation comparative research on genocide. World Polit. 59:476–501 Tambiah SJ. 1986. Sri Lanka. Ethnic Fratricide and the Dismantling of Democracy. Berkeley: Univ. Calif. Press Theidon K. 2000. “How we learned to kill our brother”? Memory, morality and reconciliation ´ Andines 29:539–54 in Peru. Bull. Inst. Fran¸cais d’ Etudes


560

Wood


Theidon K. 2007. Gender in transition: common sense, women and war. J. Hum. Rights 6:1–27 Tilly C. 1978. From Mobilization to Revolution . New York: McGraw-Hill Tilly C. 2003. The Politics of Collective Violence . New York: Cambridge Univ. Press Tilly C. 2008. Contentious Performances . New York: Cambridge Univ. Press Trawick M. 2007. Enemy Lines. Warfare, Childhood and Play in Batticaloa . Berkeley: Univ. Calif. Press Trinquier R. 1964. Modern Warfare: A French View of Counterinsurgency . Transl. D Lee. New York: Praeger. From French Truth and Reconciliation Commission of Sierra Leone (TRC). 2004. Final Report . http:/www. trcsierraleone.org Truth Commission for El Salvador. 1993. From madness to hope: the 12 year war in El Salvador. Report of the Truth Commission for El Salvador. In The United Nations and El Salvador, 1990–1995 . New York: United Nations UNHCR. 2007. Basic facts. UNHCR’s operations around the world and in Sri Lanka. United National High Commissioner for Refugees, Sri Lanka. http://www.unhcr.lk/basics/index.html. Accessed Feb. 2, 2008 on social en un contexto de guerra prolongada. El caso de Uribe de Hincapie MT. 2004. Emancipaci´ ´ Norma la Comunidad de Paz de San Jose de Apartado. Bogot a: USAID. 1987. The National Plan. San Salvador: US Agency Int. Dev. Viterna JS. 2006. Pulled, pushed, and persuaded: explaining women’s mobilization into the Salvadoran guerrilla army. Am. J. Sociol. 112:1–45 Wayland S. 2004. Ethnonationalist networks and transnational opportunities: the Sri Lankan Tamil diaspora. Rev. Int. Stud. 30:405–26 Weber E. 1976. Peasants into Frenchmen: The Modernisation of France, 1870–1914. Stanford, CA: Stanford Univ. Press Weinstein JM. 2007. Inside Rebellion: The Politics of Insurgent Violence . New York: Cambridge Univ. Press Wood EJ. 2000. Forging Democracy from Below: Insurgent Transitions in South Africa and El Salvador . New York: Cambridge Univ. Press Wood EJ. 2003. Insurgent Collective Action and Civil War in El Salvador . New York: Cambridge Univ. Press Wood EJ. 2006. Variation in sexual violence during war. Polit. Soc. 34:307–41 Wood EJ. 2009. Armed groups and sexual violence: When is wartime rape rare? Polit. Soc. In press


561

t a n e e s e b n a c n o i s r e v d e l i a t e d e r o m A . s s e r P y t i s r e v i n U e g d i r b m a C f o n o i s s i m r e p h t i w d e s u , ) 3 0 0 2 ( d o o W m o r f . d e / d t n o i o r p / e w s R . e r r a u w t a e e f h t / s u e / r g o r f e o . b e , r g o i d d r a b v l a m S c l a . E w , w 1 e g w r e o / / r J : u p n g i a t t F S h


C-2

Wood

/ / : p t t h

t a n e e s e b n a c n o i s r e v d e l i a t e d e r o m A . s s e r P y t i s r e v i n U e g d i r b m a C f o n o i s s i m r e p h t i w d e s u , ) 3 0 0 2 ( d o o W . / m d o r o f o d w e / t s n i e r r p e u t a R e . f / d s n u e / s ’ g r a r . w o e e g h d t i r t a b m 2 e g a r e o c . r J w u n g a w i F S w


www.annualreviews.org

Social Processes of Civil War

C-3

Annual Review of Political Science

Contents g r o . s w e i v e r . l a y u l n n n o a . e w s u l w w a n o m s r o r e f p d r e o d F a . o 4 l n 1 / 6 w o 0 / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s r 5 : e 1 i v 1 . n 8 U 0 0 k 2 s . n i a c d S d . y t i S l o y P b . v e R . u n n A

Volume 11, 2008

State Failure Robert H. Bates                                                                               1 The Ups and Downs of Bureaucratic Organization Johan P. Olsen                                                                              13 The Relationships Between Mass Media, Public Opinion, and Foreign Policy: Toward a Theoretical Synthesis Matthew A. Baum and Philip B.K. Potter                                                39 What the Ancient Greeks Can Tell Us About Democracy Josiah Ober                                                                                 67 The Judicialization of Mega-Politics and the Rise of Political Courts Ran Hirschl                                                                                 93 Debating the Role of Institutions in Political and Economic Development: Theory, History, and Findings Stanley L. Engerman and Kenneth L. Sokoloff                                          119 The Role of Politics in Economic Development Peter Gourevitch                                                                           137 Does Electoral System Reform Work? Electoral System Lessons from Reforms of the 1990s Ethan Scheiner                                                                            161 The New Empirical Biopolitics John R. Alford and John R. Hibbing                                                     183 The Rule of Law and Economic Development Stephan Haggard, Andrew MacIntyre, and Lydia Tiede                                205 Hiding in Plain Sight: American Politics and the Carceral State Marie Gottschalk                                                                          235 Private Global Business Regulation David Vogel                                                                                261 Pitfalls and Prospects in the Peacekeeping Literature Virginia Page Fortna and Lise Morj´ e Howard                                          283 v

Discursive Institutionalism: The Explanatory Power of Ideas and Discourse Vivien A. Schmidt                                                                         303 The Mobilization of Opposition to Economic Liberalization Kenneth M. Roberts                                                                       327 Coalitions Macartan Humphreys                                                                     351 The Concept of Representation in Contemporary Democratic Theory Nadia Urbinati and Mark E. Warren                                                   387 g r o . s w e i v e r . l a y u l n n n o a . e w s u l w w a n o m s r o r e f p d r e o d F a . o 4 l n 1 / 6 w o 0 / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s r 5 : e 1 i v 1 . n 8 U 0 0 k 2 s . n i a c d S d . y t i S l o y P b . v e R . u n n A

What Have We Learned About Generalized Trust, If Anything? Peter Nannestad                                                                           413 Convenience Voting Paul Gronke, Eva Galanes-Rosenbaum, Peter A. Miller, and Daniel Toffey          437 Race, Immigration, and the Identity-to-Politics Link Taeku Lee                                                                                  457 Work and Power: The Connection Between Female Labor Force Participation and Female Political Representation Torben Iversen and Frances Rosenbluth                                                   479 Deliberative Democratic Theory and Empirical Political Science Dennis F. Thompson                                                                       497 Is Deliberative Democracy a Falsifiable Theory? Diana C. Mutz                                                                            521 The Social Processes of Civil War: The Wartime Transformation of Social Networks Elisabeth Jean Wood                                                                       539 Political Polarization in the American Public Morris P. Fiorina and Samuel J. Abrams                                                563 Indexes

Cumulative Index of Contributing Authors, Volumes 7–11                           589 Cumulative Index of Chapter Titles, Volumes 7–11                                   591 Errata

An online log of corrections to Annual Review of Political Science articles may be found at http://polisci.annualreviews.org/

vi

C on te nt s

ANNUAL REVIEWS It’s about time. Your time. It’s time well spent.

New From Annual Reviews:

Annual Review of Organizational Psychology and Organizational Behavior Volume 1 • March 2014 • Online & In Print • http://o rgpsych.annualreviews.org

Editor: Frederick P. Morgeson, The Eli Broad College of Business, Michigan State University The Annual Review of Organizational Psychology and Organizational Behavior is devoted to publishing reviews of g r o . s w e i v e r . l a y u l n n n o a . e s w u l w a w n m o s r o r e f p d r e o d F a . o 4 l n 1 / 6 w o 0 / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s r 5 : e 1 i v 1 . n 8 U 0 0 k 2 s . n i a c d S d . y t i S l o y P b . v e R . u n n A

the industrial and organizational psychology, human resource management, and organizational behavior literature. Topics for review include motivation, selection, teams, training and development, leadership, job performance, strategic HR, cross-cultural issues, work attitudes, entrepreneurship, afect and emotion, organizational change and development, gender and diversity, statistics and research methodologies, and other emerging topics.

Complimentary online access to the frst volume will be available until March 2015. TABLE OF CONTENTS:

• An Ounce of Prevention Is Worth a Pound of Cure: Imp roving Research Quality Before Data Collection, Herman Aguinis, Robert J. Vandenberg • Burnout and Work Engagement: The JD-R Approach, Arnold B. Bakker, Evangelia Demerouti, Ana Isabel San z-Vergel

• Perspectives on Power in Organizations, Cameron Anderson, Sebastien Brion • Psychological Safety: The History, Renaissance, and Future of an Interpersonal Construct, Amy C. Edmondson, Zhike Lei • Research on Workplace Creativity: A Review and Redirection, Jing Zhou, Inga J. Hoever

• Compassion at Work, Jane E. Dutton, Kristina M. Workman, Ashley E. Hardin

• Talent Management: Conceptual Approaches and Practical Challenges, Peter Cappelli, JR Keller

• Constructively Managing Confict in Organizations, Dean Tjosvold, Alfred S.H. Wong, Nancy Yi Feng Chen

• The Contemporary Career: A Work–Home Perspective, Jefrey H. Greenhaus, Ellen Ernst Kossek

• Coworkers Behaving Badly: The Impact of Coworker Deviant Behavior upon Individual Employees, Sandra L. Robinson, Wei Wang, Christian Kiewitz

• The Fascinating Psychological Microfoundations of Strategy and Competitive Advantage, Robert E. Ployhart, Donald Hale, Jr.

• Delineating and Reviewing the Role of Newcomer Capital in Organizational Socialization, Talya N. Bauer, Berrin Erdogan

• The Psychology of Entrepreneurship, Michael Frese, Michael M. Gielnik

• Emotional Intelligence in Organizations, Stéphane Côté • Employee Voice and Silence, Elizabeth W. Morrison

• The Story of Why We Stay: A Review of Job Embeddedness, Thomas William Lee, Tyler C. Burch, Terence R. Mitchell

• Intercultural Competence, Kwok Leung, Soon Ang, Mei Ling Tan

• What Was, What Is, and What May Be in OP/OB, Lyman W. Porter, Benjamin Schneider

• Learning in the Twenty-First-Century Workplace, Raymond A. Noe, Alena D.M. Clarke, Howard J. Klein

• Where Global and Virtual Meet: The Value of Examining the Intersection of These Elements in Twenty-First-Century Teams, Cristina B. Gibson, Laura Huang, Bradley L. Kirkman, Debra L. Shapiro

• Pay Dispersion, Jason D. Shaw • Personality and Cognitive Ability as Predictors of Effective Performance at Work, Neal Schmitt

• Work–Family Boundary Dynamics, Tammy D. Allen, Eunae Cho, Laurenz L. Meier

Access this and all other Annual Reviews journals via your institution at www.annualreviews.org.

ANNUAL REVIEWS | Connect With Our Experts Tel: 800.523.8635 (US/CAN) | Tel: 650.493.4400 | Fax: 650.424.0910 | Email: [email protected]

ANNUAL REVIEWS It’s about time. Your time. It’s time well spent.

New From Annual Reviews:

Annual Review of Statistics and Its Application Volume 1 • Online January 2014 • http://stati stics.annualreviews.org

Editor: Stephen E. Fienberg, Carnegie Mellon University Associate Editors: Nancy Reid, University of Toronto Stephen M. Stigler, University of Chicago g r o . s w e i v e r . l a y u l n n n o a . e s w u l w a w n m o s r o r e f p d r e o d F a . o 4 l n 1 / 6 w o 0 / 4 D . 0 1 n 6 o 5 - y 9 t 3 i s r 5 : e 1 i v 1 . n 8 U 0 0 k 2 s . n i a c d S d . y t i S l o y P b . v e R . u n n A

The Annual Review of Statistics and Its Application aims to inform statisticians and quantitative methodologists, as well as all scientists and users of statistics about major methodological advances and the computational tools that allow for their implementation. It will include developments in the eld of statistics, including theoretical statistical underpinnings of new methodology, as well as developments in specic application domains such as biostatistics and bioinformatics, economics, machine learning, psychology, sociology, and aspects of the physical sciences.

Complimentary online access to the frst volume will be available until January 2015. TABLE OF CONTENTS :

• What Is Statistics? Stephen E. Fienberg • A Systematic Sta tistical Approach to Evaluating Evidence

from Observational Studies, David Madigan, Paul E. Stang, Jesse A. Berlin, Martijn Schuemie, J. Marc Overhage, Marc A. Suchard, Bill Dumouchel, Abraham G. Hartzema, Patrick B. Ryan • The Role of Statistics in the Discovery of a Higgs Boson,

David A. van Dyk

• High-Dimensional Statistics with a View Toward Applications

in Biology, Peter Bühlmann, Markus Kalisch, Lukas Meier • Next-Generation Statistical Genetics: Modeling, Penalization,

and Optimization in High-Dimensional Data, Kenneth Lange, Jeanette C. Papp, Janet S. Sinsheimer, Eric M. Sobel • Breaking Bad: Two Decades of Life-Course Data Analysis

in Criminology, Developmen tal Psychology, and Beyon d, Elena A. Erosheva, Ross L. Matsueda, Donatello Telesca

• Brain Imaging Analysis, F. DuBois Bowman

• Event History Analysis, Niels Keiding

• Statistics and Climate, Peter Guttorp

• Statistical Evaluation of Forensic DNA Prole Evidence,

• Climate Simulators and Climate Projections,

Jonathan Rougier, Michael Goldstein • Probabilistic Forecasting, Tilmann Gneiting,

Matthias Katzfuss • Bayesian Computational Tools, Christian P. Robert • Bayesian Computation Via Markov Chain Monte Carlo,

Radu V. Craiu, Jerey S. Rosenthal • Build, Compute, Critique, Repeat: Data Analysis with Latent

Variable Models, David M. Blei • Structured Regularizers for High-Dimensional Problems:

Statistical and Computational Issues, Martin J. Wainwright

Christopher D. Steele, David J. Balding • Using League Table Rankings in Public Policy Formation:

Statistical Issues, Harvey Goldstein • Statistical Ecology, Ruth King • Estimating the Number of Species in Microbial Diversity

Studies, John Bunge, Amy Willis, Fiona Walsh • Dynamic Treatment Regimes, Bibhas Chakraborty,

Susan A. Murphy • Statistics and Related Topics in Single-Molecule Biophysics,

Hong Qian, S.C. Kou • Statistics and Quantitative Risk Management for Banking

and Insurance, Paul Embrechts, Marius Hofert

Access this and all other Annual Reviews journals via your institution at www.annualreviews.org.

ANNUAL REVIEWS | Connect With Our Experts Tel: 800.523.8635 (US/CAN) | Tel: 650.493.4400 | Fax: 650.424.0910 | Email: [email protected]

'the Social Processes of Civil War the Wartime Transformation of Social Networks

Recommend Documents