previous section
next chapter



Puzzles of Violence

In the midst of Israel's hotly contested 1996 election campaign, Palestinian militants launched a series of deadly bomb attacks in Israel, killing dozens of Jews in downtown Tel Aviv. The deaths came at a particularly inopportune time for the then-ruling Labor Party, preoccupied as it was with convincing Jewish voters it could be as tough as the political right on national security. Some Labor ministers proposed dramatic acts of retaliation, including expelling entire groups of Palestinians or destroying Palestinian villages. The government vetoed those suggestions as too drastic, however, preferring instead to intensify ongoing policing measures such as arrests, coercive interrogations, and restrictions on Palestinian travel and movement.[1]

In a separate incident soon after, Lebanese Islamist guerrillas fired a handful of rockets toward northern Israel, causing no casualties and only limited physical damage. This time, however, the Israeli government did not hesitate, ordering the Israeli army to mount Operation Grapes of Wrath, a prolonged bombardment of southern Lebanon that displaced 400,000 civilians, killed and wounded hundreds, and destroyed homes, roads, and bridges.[2] With this spectacular display of violence, Israeli officials signaled Lebanese guerrillas that more rocket attacks would trigger disproportionate Lebanese suffering, and responded to calls for vengeance by some Israeli-Jewish voters.

The difference between Israel's methods in Lebanon and Palestine, the term I use here for the West Bank and Gaza, is remarkable. Both were

Israeli-occupied Arab lands bordering on Israel, but the Israeli state distinguished clearly between them when choosing its repertoires of violence. Israel defined Lebanon as an object of war but saw Palestine, in those years, as an object of policing. What explains this difference?[3]

One answer might point to different levels of threat, suggesting that Israel's methods were shaped by the magnitude of the security challenge it faced. The Lebanon-based guerrillas were more dangerous, and thus Israel dealt with them more harshly. Upon closer investigation, however, that argument fails to persuade. If intensity of threat alone determined Israel's methods, the government should have ordered the army to bombard Palestine, not southern Lebanon, since it was Palestinian militants who were able to explode bombs in the center of Israel. The West Bank challenge, moreover, posed a far greater threat to Zionism than did Lebanon. For many Jewish nationalists, the West Bank was an integral part of Greater Israel, while for many military strategists, Israeli control over the area was vital to national security. Lebanon, by comparison, was both ideologically and strategically less important. Were the Israeli army to have behaved in accordance with objective levels of threat, it should have treated Palestine more harshly than Lebanon.

This puzzle is further complicated by historical variations in Israel's treatment of Palestinians. Prior to its 1967 occupation of the West Bank and Gaza, for example, Israeli forces mounted large-scale raids on West Bank and Gaza villages, killing many in what was then Jordanian- or Egyptian-held territory.[4] During the 1947–49 Israeli-Arab war, moreover, Jewish troops forced hundreds of thousands of Palestinians from their homes and then bulldozed villages to prevent their return.[5] Ever since Israeli troops took the West Bank and Gaza in the 1967 war, however, Israel ceased using intensely destructive violence, relying instead on harsh, police-style tactics. The more Israel consolidated its control over Palestinian lands and populations, in other words, the less dramatic its methods of coercion became. Strangely enough, this occurred during the 1970s and 1980s, when anti-Palestinian sentiment in Israel rose sharply.[6]

Similar puzzles appear in repertoires of Serbian violence during 1992–93, the first year of the Bosnian war. Ethnic Serb paramilitaries based inside Serbia and its smaller ally, Montenegro, launched crossborder sorties into Bosnia-Herzegovina and expelled Muslim and Croat populations. In the Bosnian town of Visegrad, for example, fighters from the Serbian White Eagles paramilitary reportedly massacred many Muslim Slavs in full view, dropping their bodies from the town's central bridge.[7] Strangely enough, however, those same militias seemed reluctant

to kill Muslim Slavs living in Serbia proper. This was true even in the Sandžak, a Muslim-majority region of Serbia and Montenegro just over the border from Višegrad. There, White Eagles and others maintained rear bases amid hundreds of thousands of Muslims similar in every way to their Bosnian co-nationals, save for their geographic location. Serbian paramilitaries used despotic violence in Bosnia, but did not bring those methods back home.[8] Like Israel's security forces, Serbian paramilitaries seemed surprisingly sensitive to geography and borders.

An even broader puzzle emerges when we compare the Serbian record of 1992–93 to that of Israel in 1988, the first year of the Palestinian uprising. Both states were overtly nationalist in their orientation, perceiving themselves as defenders of a persecuted people threatened by powerful neighbors. Both were prone to ethnocentrism, partly as a result of World War II traumas. While the Nazis were killing Jews en masse in Eastern Europe, their Croat allies were doing the same to ethnic Serbs in Yugoslavia, albeit with less efficiency. The legacy of those horrors, combined with domestic politics and regional tensions, transformed both Serbia and Israel into nationalist states intent on securing contested lands. Serbia hoped to ensure ethnic Serb hegemony in mixed areas such as the Sandžak, Vojvodina, Kosovo, and parts of Bosnia, while Israel promoted Jewish rule over the West Bank and Gaza. Both Serbia and Israel had a national political core that sought to expand into adjacent areas, persevering in the face of bitter opposition from ethno-national rivals such as Palestinians, Croats, Bosnian Muslims, and ethnic Albanians.

Yet while Israel and Serbia shared many characteristics, their armed forces responded differently to challenges from Bosnia and Palestine. The Bosnian government's 1992 demand for independence prompted Serbian-backed ethnic cleansing, but the 1988 Palestinian bid for sovereignty prompted Israeli ethnic policing, a pernicious but less dramatic effort. Ethnic cleansing involved the forcible removal of unwanted populations through violence and terror. Ethnic policing included corporal punishment, mass incarceration, and administrative harassment, but left the unwanted population in place. Why did these two similarly constructed states respond in such different ways? Why, after a century of often violent colonization, did Israel use policing rather than expulsion, despite a groundswell of popular support for pushing Palestinians out?[9]

Matters are further complicated when we note differences within Serbian and Israeli zones of influence. Serbia treated Muslim Slavs differently depending on whether they lived in the Sandžak or Bosnia, while

Israel differentiated between Arabs in Lebanon or Palestine. How should we account for these within-case variations?


If we were to focus solely on Bosnia and Palestine, we might argue that Serbian and Israeli policies differed as a result of fundamental differences in regime type. Israel was a democracy in the late 1980s, whereas Serbia in 1992 had a quasi-authoritarian, populist regime. Wouldn't Israeli democracy explain its more subtle methods of control? Wouldn't Serbian authoritarianism explain its resort to an unabashedly brutal regime of domination?

There are difficulties with this argument, however. First, the designation of Israel as a democracy is problematic, since its military was absolute ruler over some 1.8 million Palestinians.[10] Although the occupation was officially transitional, it had endured for over three decades, and a generation of Palestinians had grown up under Israeli occupation. Within Israel proper, moreover, some 3.5 million Jews enjoyed a broader range of political rights and social respect than the country's 800,000 Palestinian citizens. Like the grossly imperfect democracy of postcommunist Serbia, in other words, Israel combined both authoritarian and democratic features.

Second, variations within the Serbian and Israeli cases suggest that the nature of each country's regime cannot, on its own, explain patterns of state violence. How can Israel's regime type explain its different styles of violence in Palestine and Lebanon? How can Serbian authoritarianism explain the Sandžak/Bosnia variation? Why, moreover, did Serbian authoritarianism not translate into greater tyranny at home? Why were Muslims safer in Serbia than in Bosnia?

The drawbacks of regime-based arguments emerge more generally from the tarnished record of democracies and semi-democracies worldwide. France, for example, waged vicious wars, replete with forced displacement, torture, and indiscriminate terror, against rebellious colonized peoples in Algeria and Vietnam. The world's largest democracy is India, but its war with Kashmiri separatists is an entirely brutal affair. Turkey is democratic in many ways, but has forcibly depopulated large swathes of its Kurdish-majority southeast. Regime type, in and of itself, is too blunt an explanatory tool to account for an individual state's varying repertoires of violence.


What, then, of the notion that Jewish and Serbian nationalisms were profoundly different in content? If Zionism, for example, was fundamentally kinder than Serbian nationalism, wouldn't that explain Israeli restraint? Regardless of this claim's validity, arguments of this sort encounter the same difficulties as regime-based explanations. How can the supposed moderate nature of Zionist ideology explain both the Lebanese and Palestinian experiences? How can Serbian radicalism explain both Sandžak moderation and Bosnian extremism? Nationalism may explain why states use discrimination and violence in the first place, but it cannot explain divergent repertoires of coercion by the same state in the same general time period.

A third explanation—objective threat—is also unpersuasive.[11] As noted above, Israelis might have viewed Palestine as a far greater threat than Lebanon, but it was in Lebanon, and not in Palestine, that Israeli artillery had free rein. In Serbia, similarly, Kosovo's 1.8 million ethnic Albanians presented the most powerful threat of all to Serbian national interests, but it was Bosnian Muslims, who in fact presented the least acute threat to Serbian national security, who were first targeted. If national security was the guiding logic, then Kosovo should have been ethnically cleansed long before Bosnia. The Sandžak poses a similar puzzle. Serbian officials saw a Muslim presence in the Sandžak as a strategic nightmare, and if objective levels of threat were determinate, they would have ethnically cleansed the area along with Bosnia. Perceptions of national security matter enormously, of course, but interpretations of what constitutes a "threat" are always mediated by other factors.


Despite intense media interest in Israel and Serbia, these are not particularly unique cases. Instead, they are members of a larger group of states that define their communities more narrowly than their actual populations, relying on ascribed characteristics such as nationality, religion, or ethnicity. In such cases, a dominant group captures the state apparatus, using the bureaucracy, legislature, and armed forces to promote in-group privileges. Consequently, such states are wracked by struggles over collective dignity, identity, and resources. These disputes turn especially bitter when out-groups seek territorial autonomy or independence. Examples include Kurdish rebels in Turkey, Kashmiri separatists in India, Chechen insurgents in Russia, and indigenous peoples in Mexico. Like Serbia and Israel, these states all enjoy some measure of internal democracy

and accountability, but they are also discriminatory in allocating resources, public services, and social respect. In seeking to capture both the democratic and discriminatory aspects of such states, scholars have used terms such as "semi-democracies" or "ethnocracies," with the latter describing polities where exclusion is constructed along ethnic lines.[12]

All states seek to monopolize the use of force in their territory, and the rulers of semi-democratic states are no different. As a result, they feel compelled to use substantial violence to efficiently dispatch physical challenges to their rule. At the same time, however, these rulers encounter pressures from domestic and international audiences urging greater restraint. These audiences are influenced by local and international laws and norms, which cumulatively require states to subject their use of force to scrutiny and regulations. Domestic constituencies urging the state to play by the rules of the game are strengthened by a dense network of international human rights activists, nongovernmental groups, United Nations (UN) bodies, and bilateral agencies.

Although human rights critics cannot halt excessive or illegal state violence, they can raise popular awareness and impose modest penalties on some human rights abusers. International tribunals are prosecuting war crimes in Rwanda and the former Yugoslavia, and individual governments, from Chile to Spain, Belgium, Ethiopia, Senegal, and Chad, have launched investigations of generals and politicians suspected of abuses, including some unrelated to their own country's experience. Human rights terminology is increasingly prominent in foreign news reporting, often rivaling economic and political interpretations of ongoing events.[13] Human rights–inspired intervention by Western militaries in Bosnia, Kosovo, East Timor, Congo, and Sierra Leone, along with international human rights pressure in dozens of other conflicts, attests to the theme's growing salience.

The effect of international human rights oversight is greatest on small or moderately powerful states such as Serbia and Israel, dependent as they are on international flows of aid, trade, and legitimacy. When excluded populations resist, these states discover that repression is an increasingly complex affair, especially in an era of instantaneous global communications. How can one both suppress insurgencies and at the same time project a legitimate image to domestic and international observers? Countries such as Israel, Mexico, Turkey, India, and others constantly wrestle with this dilemma, seeking to evade criticism while simultaneously conducting effective repressive campaigns. Their dilemmas are exacerbated by a recent wave of global democratization, which has

made small and moderately powerful states increasingly vulnerable to domestic human rights pressure. Negotiating the contradictory imperatives of repression and legitimacy, states are trapped in what Isaac Balbus has called the "dialectic of legal repression."[14] Across time and space, coercive forces negotiate this dialectic in different ways, leading to dissimilar and often unanticipated outcomes.

During counterinsurgency operations in the early 1990s, for example, Turkish security forces burned Kurdish villages to crush the Kurdish Workers Party, or PKK, but did not kill large numbers of civilians or drive them across international borders.[15] Forced dislocation within Turkey without large-scale massacres was Turkey's de facto compromise between its contradictory cravings for both security and legitimacy. As a result, the fate of Turkey's Kurds has been very different from those of Iraq, who were killed in large numbers during the Iraqi Anfal campaign of the late 1980s. Iraq was not a semi-democracy and was relatively indifferent to international pressures because of its oil wealth. Unburdened by the need to cater to domestic or international critics, the Iraqi regime, unlike that of Turkey, had few constraints on its behavior.

Serbia and Israel, like other medium-sized semi-democracies, were trapped within the dialectic of legal repression during the late 1980s and early 1990s. On the one hand, both states had constructed discriminatory systems privileging one group over another, especially in times of war and crisis. Israel was organized as the state of the Jews, rendering the position of Arabs quite precarious, while Serbia was increasingly organized as the state of the Serbs, threatening the welfare of Muslim Slavs, ethnic Albanians, ethnic Croats, and others. Given nationalist politics in each country, it would have been hard for either government to ignore resistance from ethno-national outsiders, especially when those same groups appealed to international powers for support. In both cases, leaders saw violence as a necessary response to pressing security threats.

At the same time, both countries had domestic critics and international obligations, forcing them to consider norms governing the use of force. Although ruled by a populist and authoritarian regime, Serbia had, ever since 1990, enjoyed vigorous elections, as well as a moderately free press. And while Israel dominated Palestine through its military, it was also a democracy of sorts within its de jure borders, granting full rights to Jews, and many rights to Palestinians with Israeli citizenship. Israel's democratic elements had a longer pedigree than those of Serbia, but Israel did not have Serbia's forty-year legacy of multicultural communism. Both Serbia and Israel, moreover, were exposed

to international human rights pressures, since both saw themselves as part of the West, and both sought access to economic, political, and cultural flows from the wealthy, trend-setting global core.

In both cases, the dilemmas created by the dialectic of legal repression were profound. Neither Serbia nor Israel was entirely committed to any one violent repertoire, adopting different methods in different geographic regions, often at one and the same time. The Serbian and Israeli coercive apparatuses, like most complex organizations, did not present a single, unified face to the world; instead, they were often bundles of diverging policies.[16]


Repertoires of state violence are shaped by pressures for repression and restraint, both of which come together in different ways over time and space. Repression is deployed in discrete institutional settings that vary in terms of visibility, level of state control, and degree of state regulation.[17] These settings are specific in terms of both conflict and geography, and each has its own rules of the game. The notion of institutional setting is borrowed from organizational sociology, and refers here to a clearly defined social or geographic space where organizational action is shaped by notions of appropriate and legitimate behavior.

Two settings of particular importance here are what I call "frontiers" and "ghettos." Bosnia became a frontier vis-à-vis Serbia in 1992, facilitating Serbia's resort to ethnic cleansing, whereas Palestine became a ghetto within Israel, prompting ethnic policing. Both Serbian and Jewish nationalism contained radical and more moderate strains, but actual repertoires of domination were determined by institutional setting. The impact of institutional settings is nicely illustrated by the within-case variations discussed in this book. The Sandžak, for example, was a ghetto of a certain type within Serbia (and Montenegro) during the 1990s, and therefore experienced less extreme forms of nationalist violence than Bosnia. And Lebanon, which served as a frontier of sorts vis-à-vis Israel, experienced more dramatic repertoires of Israeli violence than did Palestine. Serbian state behavior in the Sandžak was not identical to Israeli conduct in Palestine, and the Lebanese experience is not an exact replica of Bosnia's trajectory. Still, these comparisons do highlight the ability of institutional settings to shape repertoires of state coercion.

The crucial difference between frontiers and ghettos is the extent to

which states control these arenas and feel a bureaucratic, moral, and political sense of responsibility for their fate. States enjoy an unrivaled level of control over the ghetto's borders and territory, suppressing challenges to their monopoly over force. Although this grants states some distinct advantages, it also implies important responsibilities. Ghetto residents are despised members of society, but both local and international rules stipulate that the state bears substantial responsibility for their welfare. Frontiers, by contrast, are perched on the edge of core states and are not fully incorporated into their zone of control. States do not dominate frontiers as they do ghettos, and they are not bound by the same legal and moral obligations. In times of crisis and uncertainty, frontiers more easily become sites of ethnic cleansing.

By applying the notion of frontier to Bosnia, we can better understand why Serbia resorted to ethnic cleansing in 1992–93. On their own, the breakdown of the Yugoslav state, Serbian nationalism, and Bosnian demands for independence might not have prompted ethnic cleansing. It is only when Serbian nationalism interacted with the Bosnian frontier that expulsion became a viable option. When we view Palestine as a ghetto within Israel, moreover, the reasons for Israel's reliance on ethnic policing become clear. Institutional context promotes some policies over others, with poorly regulated environments selecting out more radical strands of nationalist thought, and more heavily institutionalized arenas promoting police-like regimes of domination.


Chapter 1 discusses relevant theoretical issues in greater depth, using a modest amount of academic terminology. Then, Part I, comprised of Chapters 2 through 5, argues that the interaction between Serb nationalism, which pushed Serbian officials to promote Bosnian ethnic cleansing, and Western recognition of Bosnian sovereignty, which prohibited Serb cross-border activity, created two distinct institutional settings: a Bosnian frontier and a Serbian core. The core was Serbia, senior partner in the new rump Yugoslavia, [18] while the frontier was Bosnia, situated to the west of the newly created international border. Bosnia became a frontier in 1992 because the new, Muslim-led Bosnian government was enfeebled, and the new Bosnian Serb entity, later known as Republika Srpska, was just emerging. As a result, the eastern and northwest parts of Bosnia were largely controlled by local Bosnian Serb fighters working with roving, semi-private paramilitaries from Serbia proper. Both were

classic frontier agents, belonging officially to no legally constituted authority and enjoying considerable local autonomy. Together, these actors were responsible for much of the initial wave of Bosnian ethnic cleansing.

Chapter 3 includes a discussion of one of the most hotly debated issues of the Bosnian war, the links between the Belgrade government and ethnic Serb fighters in Bosnia. Human rights activists and journalists have made considerable efforts to prove Belgrade's connection to Serb forces in Bosnia, a task complicated by the dearth of relevant documentation. These intensive legal investigations, however, have helped obscure the broader sociological importance of Serbia's clandestine links. The lack of public chains of command-and-control between Belgrade and Bosnia indicate the extent to which Belgrade's cross-border activities were driven underground by Western recognition of Bosnia's sovereignty. Covert linkages allowed Serbia to remain involved in Bosnia, but ensured that the region was not officially Serbia's responsibility. Once forced into an underground, illegitimate social space, ethnic Serb fighters encountered new opportunities and constraints. Secrecy helped them conduct ethnic cleansing in defiance of state and international norms, but illegitimacy prevented them from laying official claim to their conquests once the fighting ended.

Chapters 4 and 5 test my argument by examining patterns of nationalist violence inside the Serbian core. Here, the Serbian political elite's responsibilities for human rights abuses were clear and the setting was more heavily institutionalized. During the early part of the 1990s, when the Bosnian war was at its height, the state prevented Serbian paramilitaries in Kosovo, the Sandžak, and Vojvodina from using Bosnia-style methods against non-Serb populations. As Chapter 5 explains, however, Kosovo's institutional setting changed in 1998–99 from ghetto to frontier through a combination of Kosovo Albanian and international actions. The result was a full-scale Serbian ethnic cleansing effort.

In Part II, Chapters 6 to 8 discuss the emergence of the Palestinian ghetto and patterns of Israeli violence, including its 1988 policing campaign in the West Bank and Gaza. The introduction to Part II briefly surveys the rise of radical Jewish nationalism in the late 1970s and 1980s, while Chapter 6 traces the emergence of a Palestinian ghetto enclave during those same years. Palestinian militants tried and failed to disrupt Israel's ghetto-formation policies through armed rebellion, and then also failed to gain international recognition of their sovereignty. Chapter 7 analyses Israel's ghetto policing tactics, while Chapter 8 probes two alternatives

to the ethnic policing model: Jewish vigilantism and Israeli operations in Lebanon. As was true in Serbia, some semi-private Jewish paramilitaries wanted to expel Palestinians, but failed to gain state support because of the West Bank and Gaza's ghetto status. In Lebanon, by contrast, Israel encouraged allied paramilitaries to use intense violence and deployed a range of despotic methods itself. Both Palestine and Lebanon were not part of Israel's de jure territory, but Palestine had become a ghetto, while Lebanon retained some frontier-like qualities, leading to varying repertoires of violence.

previous section
next chapter