Control of Violence: Historical and International Perspectives on Violence in Modern Societies

Control of Violence Wilhelm Heitmeyer · Heinz-Gerhard Haupt · Stefan Malthaner · Andrea Kirschner Editors Control of...

Author: Wilhelm Heitmeyer | Heinz-Gerhard Haupt | Stefan Malthaner | Andrea Kirschner

57 downloads 3331 Views 6MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Control of Violence

Wilhelm Heitmeyer · Heinz-Gerhard Haupt · Stefan Malthaner · Andrea Kirschner Editors

Control of Violence Historical and International Perspectives on Violence in Modern Societies

123

Editors Wilhelm Heitmeyer University of Bielefeld Bielefeld 33615, Germany [email protected]

Stefan Malthaner University of Bielefeld Bielefeld 33615, Germany [email protected]

Heinz-Gerhard Haupt Department of History and Civilization European University Institute Florence 50133, Italy [email protected] Andrea Kirschner University of Bielefeld Bielefeld 33615, Germany [email protected]

ISBN 978-1-4419-0382-2 e-ISBN 978-1-4419-0383-9 DOI 10.1007/978-1-4419-0383-9 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010937925 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface

Are modern societies capable of controlling violence effectively? Media coverage of terrorist attacks, rampage shootings, endless conflict in Iraq and Afghanistan, and relentless everyday violence in “fragile” states like Somalia or Colombia tends to suggest a pessimistic answer. Events such as these are disconcerting and upsetting not only because of their overwhelming media presence, but also because they give the impression of being completely new forms of violence that are both unpredictable and uncontrollable. Are we therefore to conclude that the societies in which such acts occur have lost control over violence? Or is this merely a false perception, fed by fear and threat discourses, while in reality our options for controlling violence are greater than they have ever been in history? What are the real problems, and what mechanisms of violence control do we have in the twenty-first century? How do the various social, cultural, and historical contexts differ? A group of researchers from various academic disciplines and different countries spent a year at the Center for Interdisciplinary Research at Bielefeld University studying these issues, between October 2007 and September 2008, discussing their findings in a series of conferences with their colleagues in the field. This volume is the result of their work. The book reaps many of the benefits of working in a research group. The collaborative environment created a unique, long-term context for discussions centering on the question of violence control, which in turn gave rise to a common perspective that deeply informs the papers collected in this volume and that is of crucial importance because the spectrum of the contributions is remarkably wide. The papers deal not only with fundamental questions and aspects of violence control—such as the transformation of control mechanisms during processes of modernization; the significance of public discourses and the framing of violence; the relationship between religion and violence; and forms of self-control—but also with three more specialized aspects of violent phenomena which the researchers deem paradigmatic for the discussion of a loss of control over violence, namely school shootings, terrorism, and violence in states in crisis. The cross-disciplinary discussions that arose during the research year are reflected in the integrative thrust of the research topics, which explore the shifting forms, mechanisms, and actors involved in the control of violence. Another common characteristic of the papers, which has been possible not least because of the researchers’ constant and

v

vi

Preface

constructive “confrontations” with colleagues from other disciplines and research fields, is their critical reflection on entrenched opinions and viewpoints. Thus, the research project demonstrated once again that cooperation across disciplinary, geographical, and cultural frontiers is not only possible, but also effective and extraordinarily productive. It is our hope, therefore, that the working relationships and networks that were formed during the research year and in the course of the completion of this book will not only “informally” outlast the project, but will also continue to serve as a source of cooperation, inspiration, and intrepid projects. We would like to thank all the authors represented in this volume for the generosity and patience with which they submitted their contributions to several reviews both by the editors and external reviewers. We also thank all the reviewers for their support during the review process. We are indebted for important arguments and valuable suggestions to all our colleagues who presented papers or participated in discussions at our conferences and workshops, which were funded by the Thyssen Foundation, the German Foundation for Peace Research, the Volkswagen Foundation, and the German Research Foundation. We thank them for their support. The book could not have been completed without the outstanding translation and editing services of Meredith Dale and his team, nor without the tireless work of Larissa Appelt, Nils Böckler, Judith Scherer, and Martin Winands, who formatted the individual contributions. Our heartfelt thanks to them all. Finally, special thanks are due to our former colleague Barbara Kaletta and to the staff of the ZiF for their dedicated, patient, and kind support during the research year, without which this book would not have been written. We hope that this volume will provide food for further thought about the interrelationships between violence and control, even if—in our own experience—such thought may entail a loss of control of some sort There is not a single idea of any quality that does not stem from an intoxication, a loss of control, an ability to err and thereby renew oneself. (Emil Cioran, The New Gods)

Bielefeld, Germany Florence, Italy Bielefeld, Germany Bielefeld, Germany

Wilhelm Heitmeyer Heinz-Gerhard Haupt Stefan Malthaner Andrea Kirschner

Contents

Part I

Introduction

Control of Violence—An Analytical Framework . . . . . . . . . . . . . Andrea Kirschner and Stefan Malthaner Part II

3

Mechanisms and Strategies of Violence Control

An End to Violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michel Wieviorka Cross-National Homicide Trends in the Latter Decades of the Twentieth Century: Losses and Gains in Institutional Control? . Steven F. Messner, Benjamin Pearson-Nelson, Lawrence E. Raffalovich, and Zachary Miner Self-Control and the Management of Violence . . . . . . . . . . . . . . Charles R. Tittle Self-Control, Conscience, and Criminal Violence: Some Preliminary Considerations . . . . . . . . . . . . . . . . . . . . . . . . . Helmut Thome

47

65

91

121

Reading Religious Violence in Terms of Theories of Social Action . . . . Hans G. Kippenberg

145

Religion and Control of Violence . . . . . . . . . . . . . . . . . . . . . . Levent Tezcan

165

Gun Violence and Control in Germany 1880–1911: Scandalizing Gun Violence and Changing Perceptions as Preconditions for Firearm Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dagmar Ellerbrock Controlling Control Institutions: Policing of Collective Protests in 1960s West Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . Klaus Weinhauer

185

213

vii

viii

Part III

Contents

The Micro-level: School Shootings

School Violence and Its Control in Germany and the United States Since the 1950s . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dirk Schumann School Shooting: A Double Loss of Control . . . . . . . . . . . . . . . . Nils Böckler, Thorsten Seeger, and Wilhelm Heitmeyer Explaining and Preventing School Shootings: Chances and Difficulties of Control . . . . . . . . . . . . . . . . . . . . . . . . . . Rebecca Bondü and Herbert Scheithauer Masculinity, School Shooters, and the Control of Violence . . . . . . . . Ralph W. Larkin Media and Control of Violence: Communication in School Shootings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glenn W. Muschert and Massimo Ragnedda Part IV

233 261

295 315

345

The Meso-level: Terrorism

Terrorism as Performance: The Assassinations of Walther Rathenau and Hanns-Martin Schleyer . . . . . . . . . . . . . . . . . . . Bernd Weisbrod

365

Party Politics, National Security, and Émigré Political Violence in Australia, 1949–1973 . . . . . . . . . . . . . . . . . . . . . . . . . . . Mate Nikola Toki´c

395

Control of Terror—Terror of Control . . . . . . . . . . . . . . . . . . . Jitka Maleˇcková

415

Terrorism: Conditions and Limits of Control . . . . . . . . . . . . . . . Friedhelm Neidhardt

431

Fighting for the Community of Believers: Dynamics of Control in the Relationship Between Militant Islamist Movements and their Constituencies . . . . . . . . . . . . . . . . . . . . . . . . . . . Stefan Malthaner Baseless Jihad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Khaled Al-Hashimi and Carolin Goerzig Part V

445 467

The Macro level: Violence in States in Crisis

Ethnic Riots in Situations of Loss of Control: Revolution, Civil War, and Regime Change as Opportunity Structures for Anti-Jewish Violence in Nineteenth- and Twentieth-Century Europe . . Werner Bergmann

487

Contents

Control and Chaos: Paramilitary Violence and the Dissolution of the Habsburg Empire . . . . . . . . . . . . . . . . . . . . . . . . . . . Robert Gerwarth Failed States in Theoretical, Historical, and Policy Perspectives . . . . . Jean-Germain Gros Putting Out the Fire with Gasoline? Violence Control in “Fragile” States: A Study of Vigilantism in Nigeria . . . . . . . . . . Andrea Kirschner

ix

517 535

563

Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heinz-Gerhard Haupt and Wilhelm Heitmeyer

593

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

599

Contributors

Khaled Al-Hashimi Technical University Berlin, Berlin, Germany, [email protected] Werner Bergmann Centre for Research on Antisemitism, Technical University of Berlin, Berlin, Germany, [email protected] Nils Böckler Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany, [email protected] Rebecca Bondü Department of Educational Science and Psychology, Freie Universität Berlin, Habelschwerdter Allee 45, 14195 Berlin, Germany, [email protected] Dagmar Ellerbrock Department of History, Bielefeld University, Bielefeld, Germany, [email protected] Robert Gerwarth College of Arts & Celtic Studies, School of History & Archive, University College Dublin, Dublin, Ireland, [email protected] Carolin Goerzig EU Institute for Security Studies, Paris, France, [email protected] Jean-Germain Gros Department of Political Science, University of Missouri, Saint Louis, MO, USA, [email protected] Heinz-Gerhard Haupt Department of History and Civilization, European University Institute, Florence, Italy, [email protected] Wilhelm Heitmeyer Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany, [email protected] Hans G. Kippenberg School of Humanities and Social Sciences, Jacobs University, Bremen, Germany, [email protected] Andrea Kirschner Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany, [email protected]

xi

xii

Contributors

Ralph W. Larkin John Jay College of Criminal Justice, New York, NY, USA, [email protected] Jitka Maleˇcková Institute for Middle Eastern and African Studies, Charles University, Prague, Czech Republic, [email protected] Stefan Malthaner Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany, [email protected] Steven F. Messner Department of Sociology, University at Albany, S.U.N.Y., Albany, NY, USA, [email protected] Zachary Miner University at Albany, S.U.N.Y., Albany, NY, USA, [email protected] Glenn W. Muschert Department of Sociology and Gerontology, Miami University, Oxford, OH, USA, [email protected] Friedhelm Neidhardt Social Science Research Center, Berlin, Germany, [email protected] Benjamin Pearson-Nelson Indiana University-Purdue University, Fort Wayne, IN, USA, [email protected] Lawrence E. Raffalovich Department of Sociology, University at Albany, S.U.N.Y., Albany, NY, USA, [email protected] Massimo Ragnedda Dipartimento di Economia Istituzioni e Società, Università degli Studi di Sassari, Sassari, Italy, [email protected] Herbert Scheithauer Department of Educational Science and Psychology, Freie Universität Berlin, Habelschwerdter Allee 45, 14195 Berlin, Germany, [email protected] Dirk Schumann Department of Medieval and Modern History, Georg-August University, Göttingen, Germany, [email protected] Thorsten Seeger Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany, [email protected] Levent Tezcan Department of Culture Studies, Tilburg University, Tilburg, The Netherlands, [email protected] Helmut Thome Institute for Sociology, Martin Luther University, Halle-Wittenberg, Germany, [email protected] Charles R. Tittle Department of Sociology and Anthropology, North Carolina State University, Raleigh, NC, USA, [email protected] Mate Nikola Toki´c The American University in Cairo, Cairo, Egypt, [email protected]

Contributors

xiii

Klaus Weinhauer Department of History, Bielefeld University, Bielefeld, Germany, [email protected] Bernd Weisbrod Department of Medieval and Modern History, Faculty of Humanities, Georg-August University, Göttingen, Germany, [email protected] Michel Wieviorka Ecole des Hautes Etudes en Sciences Sociales, Paris, France, [email protected]

About the Editors

Wilhelm Heitmeyer is professor of socialization and director of the Institute for Interdisciplinary Research on Conflict and Violence at Bielefeld University. His research interests concentrate on violence, social disintegration, right-wing extremism, and ethnic-cultural conflicts. He is editor of the International Handbook of Violence Research (co-edited with John Hagan), and editor-in-chief of the International Journal of Conflict and Violence (with D. Massey et al.). He is organizer of the international research group “Control of Violence” at the Center for Interdisciplinary Research (ZiF), Bielefeld University (with Heinz Gerhard-Haupt). Heinz-Gerhard Haupt is professor of social history at Bielefeld University and head of the Department of History and Civilization at the European University Institute in Florence, Italy. His work focuses on social history and political history of modern Europe. Recent publications include Neue Politikgeschichte [New political history] (with Ute Frevert, eds.) (2005); Terrorismus in der Bundesrepublik: Medien, Staat und Subkulturen in den 1970er Jahren [Terrorism in the Federal Republic of Germany: Media, state, and subcultures] (with K. Weinhauer and J. Requate, eds.) (2006). He is organizer of the international research group “Control of Violence” at the Center for Interdisciplinary Research (ZiF), Bielefeld University (with Wilhelm Heitmeyer). Stefan Malthaner is a researcher at the Institute for Interdisciplinary Research on Conflict and Violence at Bielefeld University. He received his Ph.D. in sociology from Augsburg University. His work focuses on terrorism and insurgent violence, social movements, and religious communities. He was a member of the Junior Research Group “Micropolitics of Armed Groups” at Humboldt University, Berlin, and fellow of the research group “Control of Violence” at the Center of Interdisciplinary Research (ZiF). Andrea Kirschner was a researcher at the Center for Interdisciplinary Research, Bielefeld University, before she became member of the DFG Research Training Group “Self-Making. Practices of subjectivation in historical and interdisciplinary perspective” at Oldenburg University. She received a diploma in social sciences xv

xvi

About the Editors

and a master of arts in peace and conflict research. Her research interests concentrate on violence, migration, and development with particular attention being paid to post-structural research approaches.

Part I

Introduction

Control of Violence—An Analytical Framework Andrea Kirschner and Stefan Malthaner

1 The Social Order and the Problem of Controlling Violence One of the core challenges faced by societies in all cultures and ages is that of limiting, and if possible preventing, destructive violence. The scientific and public debates about the possibility of controlling violence through the framework of the social order provide a particularly striking illustration of the contrasts between optimistic and pessimistic visions of social conditions and development potentials. Violence as such can be employed both to preserve existing conditions and to bring about social change, and its use is not confined to individuals and social groups. Rather, it has been a fixed component of all forms of state order both in the past and in the present. The link between state power and violence has been indissoluble since the early modern era, even in situations where violence is regulated by law. In the words of Walter Benjamin (2007, 281): “Lawmaking is power making, and, to that extent, an immediate manifestation of violence.” The relationship between violence and order, therefore, is a reciprocally constitutive one. According to Popitz (translated from 1992, 63), “Social order is a necessary precondition for containing violence; violence is a necessary precondition for maintaining the social order.” Thus, the concept of controlling violence is not one which has exclusively positive connotations; rather, it is an extremely ambivalent category arising as an element of strategies for enforcing power and maintaining order. These strategies contain elements of force and frequently also of violence. However, control of violence is also a mechanism for structuring social and political movements, within which it serves to defend and assert leadership. Control is based on superior power, which may be political, legal, physical and military, or economic, or may take the form of cultural hegemony (Gramsci 1991). Those who define the types and conditions of control are in a position to control others and to enforce certain standards of behavior—not infrequently by threatening or employing violence. Thus, A. Kirschner (B) Institute for Interdisciplinary Research on Conflict and Violence, University of Bielefeld, Bielefeld, Germany e-mail: [email protected] W. Heitmeyer et al. (eds.), Control of Violence, DOI 10.1007/978-1-4419-0383-9_1, C Springer Science+Business Media, LLC 2011

3

4

A. Kirschner and S. Malthaner

the control of violence is by no means a neutral aspect of society. Rather, it is formed and influenced by society’s structures, perceived threats, and discourses about violence—which it in turn influences. Changes in social conditions and the awareness of new threats cause changes in the controlling regime, i.e., in the totality of the rules, forms, mechanisms, and objectives of control within a society. At the same time, changes in the control regime have an impact on social structures and affect the framework within which the forms and mechanisms of control are constituted (Singelnstein and Stolle 2008, 111). The mechanisms and institutions that a society employs in its attempts to limit violence are indicative of fundamental social structures and relationships. In this sense, as Spierenburg points out, control is always a “sensitizing concept:” “It draws attention to the relationships between various mechanisms inducing people to act in a way that is desirable according to a certain standard or ideal. In all societies social control constitutes a major key to understanding violence, conflict, and problems related to the formation and acceptance of social norms.. . .. this crucial function of social control is operative regardless of the period one examines or a country’s political system. It works in the early modern period as well as in recent times and in both democratic and authoritarian societies” (2004, 10; following Herbert Blumer). The modern world, with its faith in technical progress, long believed that social processes were completely controllable. This belief was supported and reiterated in theories of modernization and civilization (Elias [1939] 2009). Against the background of growing prosperity, these theories held that modern societies had been pacified through an increasingly successful combination of control by the state and emotional self-restraint by the individual, and supported this argument by pointing out that, in the western world at least, homicide rates had been declining for centuries (Eisner 2003; Thome 2004; Spierenburg 1994; Dinges 1998). But the twentieth century’s record on violence is not unambiguous. Considering the unparalleled number of casualties caused by wars, civil strife, genocide, and totalitarian and dictatorial regimes in the twentieth century, there are good reasons to doubt the assumption that social development and the consolidation of governmental control mechanisms result in a general decrease in violence (Hobsbawm 1994; Mazower 1998). The role of the rational, bureaucratic state in controlling violence proves to be ambivalent: on the one hand, the state regulates social affairs and limits violence, but on the other hand, it acts as a protagonist of force—indeed violence—both at home and abroad. The organized power of its institutions enables the state to exercise force to a hitherto unprecedented degree. Its violent “interior face” was revealed most vividly in the Nazi, fascist, and communist states of Europe and showed that ideologies such as these are in no way incompatible with modern, rational principles of organization and order in society, industry, and technology, but may in fact have a strong affinity to such principles (Mazower 1998; Bauman 1989). However, this affinity is also apparent in the case of collective acts of violence in non-European contexts, which western societies like to dissociate from their own cultures by interpreting them as an expression of “pre-modern,” unbridled barbarism. The genocide in Rwanda, for instance, was anything but an uncontrolled,

Control of Violence

5

intoxicated, “pre-modern” outbreak of ethnic violence, even though this interpretation was strongly suggested by the machetes that came to symbolize the events in the media. Rather, the genocide was the product of a modern, systematic, rationally planned, and highly institutionalized and organized form of mass violence that absolutely presupposes the existence of modern governmental structures, colonial control instruments (ethnic categorization), media infrastructures, and an ethno-nationalist ideology (Uvin 1998).

2 Violence and Its Ambivalence Like control, violence can be attributed to that field of ideas, which the philosopher William Gallie (1956) once described as “essentially contested concepts.” What these concepts have in common is that they are judgmental and complex; they cannot be categorized within clearly definable fields of general usage that would point to an unambiguous meaning. The concept of violence has been described by means of a wide spectrum of definitions and explanations that vary in breadth and validity that refer to different subjects, and that involve controversy about the authority to define concepts at all—especially with regard to the question of what ought to be described as violence and what coping strategies ought to be derived from the definition (Heitmeyer and Hagan 2003). But ambiguity is one of the characteristics of violence (Heitmeyer and Soeffner 2004, 11). Accordingly, the different definitions of violence can vary extremely widely and may encompass psychological, cultural, and physical violence and even the concept of structural violence. At the same time, there are numerous other systematologies that may focus on its covert or overt nature or its target, which may be persons or objects (Imbusch 2003, 13– 39). The content of the definition of violence is constantly subject to historical and cultural change as well as social and cultural debate (Liell 2002, 6). The decision whether to apply the concept of violence to social conditions, a protagonist, or an action invariably involves a conflict about the legitimacy of violence as well: a conflict that remains insurmountable because social conventions and rules that govern the legitimate use of violence may change at any time in the course of political conflicts (Weller 2003, 11). In the present discussion, we will follow Heinrich Popitz in taking violence to be “violent action” (Aktionsmacht) that results in deliberate physical injury to others (Popitz 1992, 48). Compared to more wide-ranging concepts of violence, this narrow definition, which limits itself to violence committed against persons with the intent to harm their physical integrity, has the advantage of clearly delineating a manageable field of analysis. The concept of structural violence, which Galtung (1969, 114) defines as a form of violence that “is built into the structure, and shows up as unequal power and consequently as unequal life chances,” widens the scope of the subject to the point where it becomes too vast to handle, both from a normative and from an empirical perspective (Liell 2002, 7; Daase 1996, 46). In this way, the concept of violence acquires an enormous potential for mobilization, but it simultaneously loses its distinctiveness. Although even a narrowly circumscribed

6


concept of violence cannot completely eliminate the problem of imprecise definitions nor the fact that violent actions always take place within social, cultural, and historical contexts, it nevertheless succeeds in limiting the empirical object of study and reduces questions involving normative evaluations of social conditions.

3 Toward an Understanding of Control 3.1 Etymological Origins The etymology of the term “control” is rooted in the Old French contre-rôle, which denotes a “counter-register” that is used to verify an original or an object. The German term Kontrolle/kontrollieren, the French contrôle/contrôler, and the English control/to control have all preserved the sense of control as verification. All three languages, however, also use the term in the secondary, derivative sense of control as mastery or command over something, or as the power and authority to direct and govern. In this sense, “to control” is to rule, guide, and govern. To control something is to regulate and limit it.1 Both meanings are closely interlinked, as the ability to monitor and verify something and to impose penalties is also a mechanism of domination. Nevertheless, each of the two levels of meaning employs a different conceptual logic. Control in the sense of domination refers to the effective action and its result, i.e., the successful supervision and regulation of a situation with the result that it comes “under control.” In contrast, control in the sense of verification and monitoring is initially an open-ended activity that entails either checking whether a behavior or a state of affairs is as it should be (verification), with the option to respond by imposing penalties if it is not, or observing a situation (monitoring) in order to intervene when undesirable developments or actions are noted. The penalties or interventions may or may not be successful. It is clear at this juncture that the two concepts of control also differ in their temporal dimension. Control in the sense of domination is a condition, generally designed to be permanent, in which something can be supervised (controlled) at any time, whereas control in the sense of verification (and penalization) refers to an action that occurs at a subsequent point in time. Monitoring, in contrast, connotes a preceding, preventive activity, where the responses to specific developments come later in the chronological sequence. As we will see later on, the distinction between an initial dimension of control that takes the form of actions by certain persons or groups of persons (actors of control) who aim to monitor, verify, and regulate the actions of other persons (objects of control)—in other words, a dimension that represents an interactive process—and a second dimension of control that takes the form of a condition that governs processes and behaviors is a central issue for the paradigm of the control of violence 1 Etymologisches

Wörterbuch der deutschen Sprache, 24th ed., s.v. “Kontrolle”; The Oxford English Dictionary (1989), 2d ed., s.v. “control”.

Control of Violence

7

with which we are concerned here. A conspectus of both dimensions must focus not only on the objectives of the control of violence, but also on its efficacy.

3.2 Sociological and Historical Perspectives on Control “Control of violence” is not a clearly defined concept in sociology and historical scholarship.2 Any attempt to develop a definition inevitably brings one up against the established concepts of the social norm and the social order, of power and rule, and of approaches to social control. 3.2.1 Control as State Repression or Social Self-Regulation: Classical Models of the Social Order (Hobbes, Scottish Enlightenment) Limiting violence has always been one strand of the scientific discourse about the state and forms of social order. Two fundamentally opposite models of an order that controls violence must be mentioned here: on the one hand, forced, repressive pacification by a state wielding superior power; and on the other hand, the selfregulation of social relations by mechanisms of socialization, social control, and market forces. Both models were definitively formulated in Great Britain during the seventeenth and eighteenth centuries. The political theory of Thomas Hobbes (1588–1679) was developed under the influence of the English Civil War and the Thirty Years’ War on the European continent. Its point of departure was the nightmare scenario of violent chaos in which fear reigns supreme and social life is paralyzed by the ever-present danger of armed assault. Hobbes holds that fundamentally insatiable human desires represent one of the main reasons why conflicts arise in the absence of a state of political order. The citizens’ perpetual feelings of individual insecurity, however, are an even greater factor in the perpetuation of violence. When the individual does not know whether the other represents a violent threat, all must behave as though the threat were real, arm themselves or strike first (Hobbes [1651] 1970, 63–66, 87–90).3 Even banding together in groups cannot eliminate the danger, as the groups themselves are always under threat from other groups of similar strength. According to Hobbes, the only possible solution lies in collective submission to a single, all-powerful state with a monopoly on the use of force. The state disempowers and disarms all other protagonists of force and guarantees the safety of individual citizens by making and enforcing laws internally and by defending its borders externally. Here the control of violence appears as the sole task of the all-powerful state, whose main function

2 The same ultimately applies to “control” in general, although one must take into account hypothe-

ses on forms of control and influence in social relationships (e.g., Oppenheim 1961). However, the term “control” is sometimes used as a synonym for “social control.” 3 It is worth noting that, according to Hobbes, the primary motive in a war of all against all is not malice or greed, but fear. The compulsions of insecurity do not stem from the evil nature of all; in the view of Hobbes, the existence of a few truly evil people is sufficient to force all the others to act as though all were evil (Hobbes [1651] 1970, 63–66, 87–90).

8


toward its subjects is coercive. Individual and collective violence is prevented and adherence to norms, agreements, and laws is guaranteed only by the threat and enforcement of punishment. From this perspective, loss of control constitutes the loss of the state’s repressive efficiency, behind which the danger of new civil wars is eternally present. Assuming that human nature is largely unchangeable and that the resultant compulsions arising from insecurity remain largely the same, Hobbes sees no hope that this state of affairs could improve or that the danger of violent chaos could be gradually averted. The containment of violence calls for unceasing vigilance, and it is always provisional and eternally vulnerable to loss of control. Another model of order was developed in the mid-eighteenth century by the authors of the Scottish Enlightenment, most notably by Adam Smith (1723–1790). Here the social order appears not as something imposed from the outside, but as something that comes into being spontaneously in the form of cooperation and normative systems (“moral feelings”) as a result of the social interactions themselves. It acquires efficacy through socialization, habit, self-control, and informal social control (Smith [1759] 1976). It must, however, be borne in mind that Scottish moral theory was developed against a very different historical background than the political theory of Hobbes. The constitutional monarchy in Great Britain had already been consolidated and was going through a period of peace and affluence. Thus, Smith views the market as an initial form of order in which the passions of individuals seeking personal profit are automatically regulated and converted into beneficial cooperation. Governed by an “invisible hand,” the market requires no external sanctions (Smith [1776] 1991). Of particular note with respect to the question of violence control is Smith’s analysis of a second mechanism of social regulation: the creation and enforcement of behavioral norms in social interactions. One of the recurrent themes of the Scottish Enlightenment is that people are born into families and societies whose habits and moral feelings they learn through processes of socialization. In these processes, affect control and conformity to norms develop as the outcome of a self-control that is based on the internalization of norms (moral feelings) and subtle forms of social control. The individual accepts the expectations of others regarding his behavior and internalizes them as an “impartial observer” who acquires an almost absolute authority not only through the fear of losing reputation and being ostracized by the community, but also through the fear of economic disadvantage through being shunned in business dealings (Smith [1759] 1976, 134–156). Hirschman (1977) expresses the logic of these two models of order as the repression of human passions, on the one hand, and the harnessing of these passions, on the other hand; in the latter case, it is precisely the interests of the individual that serve to harness the passions to cooperation (Hirschman 1977, 14–20, 31–42). However, state-imposed order and social self-regulation are not mutually exclusive. The thinkers of the Scottish Enlightenment stressed the necessity of laws enforced by the state. But law enforcement by the state only becomes necessary when forms of social self-regulation reach their limits. Here the state is not so much an all-powerful, external coercive apparatus as an entity that arises from the institutionalization of social mechanisms of order and that is driven by an interest in

Control of Violence

9

guaranteed civil and economic rights. Conflicts and competing interests represent a potentially useful element that powers social development. Thus, it is possible to derive an optimistic developmental perspective from this theory: if people are able to learn, then it is possible to increasingly pacify social relationships by stabilizing cooperative relationships and self-control. In the logic of this perspective, loss of control—even over violence—means primarily the disintegration of social relationships and the loss of efficacy of social mechanisms of socialization and self-control. 3.2.2 Power and Rule, Punishment and Discipline: Control of Violence in Modern Societies (Weber, Elias, Foucault) The concept of control is closely linked to the concept of power, which Weber described as the opportunity to enforce one’s own will within a social relationship (Weber [1922] 1968, 53). As in the case of control, power entails possibilities for influencing other actors. The two concepts are thus very closely related. Nevertheless, the relationships they reference are different: Power describes a potential, while control describes an effect that actually occurs. Control aims specifically to guide or limit the behavior of others, while power focuses on ensuring the dominance of one’s own side—enforcing one’s own will even against the will of others. Finally, control connotes a more regulated and stable form of influence, while power is more unstable and thus appears more uncontrolled. Rule, too, implies control, as it represents institutionalized power that has been transformed into a stable relationship between command and obedience. The ruler controls the conduct of his subjects, from whom he is able to demand obedience. Thus, rule is the more narrowly defined term for a special form of control, namely politically institutionalized control. However, this too goes far beyond the mere control of violence and applies to a wide range of different behavior patterns. In being transformed into rule, however, control acquires a specific extension, which is simultaneously a limitation: control as rule—at least in its ideal version as defined by Weber—always presupposes the willingness of the subordinates to obey. This willingness, in turn, is rooted in their acceptance of the order on which rule is based (Weber [1922] 1968, 212–217). Thus, Max Weber’s sociology of rule is only superficially beholden to the logic of repressive order as formulated by Hobbes. It is true that political associations employ physical coercion by means of a dedicated administrative staff in order to secure their own existence and the validity of the order they impose. Thus, the state is, first and foremost, a coercive institution, and the concept of law necessarily presupposes the existence of law enforcers (Weber [1922] 1968, 54). However, the coercive element is supplemented by the willingness of the subordinates to obey—in other words, to acknowledge rule because it is based on an order they accept as exemplary and obligatory ([1922] 1968, 212–217). As a result of this legitimacy of rule, the relationship of control differs in two ways from that described by Hobbes: First, Weber postulates that control is based on an abstract order that places constraints even on the ruler and limits the scope and applicability of his controlling authority. Rule represents the opportunity to receive obedience

10


for a specific command by a specific target group (1968, 53). At the same time, the functionality of rule is transformed from external coercion into a mechanism of self-control: it is not the command that causes obedience, but the individual’s compliance with a system of order (reinforced by coercive mechanisms that support it) that is regarded as obligatory and imposes a duty to obey (see also Lemke 2001). Thus, legitimacy relieves rule “of the necessity to commit ever more acts of coercion and violence” (translated from Tyrell 1980, 90), so that coercion and violence can fade into the background and become “latent” without reducing the controllability of rule. Losses of control arise when the legitimacy and obligatory nature of rule are called into question, causing the mechanism of self-control to lose efficacy and once again requiring the social order to be guaranteed through physical coercion. In this case, both state violence and the violent protests of social protagonists against an order they perceive as illegitimate are indications of a loss of control over violence by the state. Second, another control mechanism that is immediately linked to rule in Weber’s view is discipline. Weber cites bureaucratic administrative bodies, standing armies, and economic processes as examples of discipline. Discipline also involves the following of orders, which individuals must carry out in a well-planned and prepared fashion with unconditional disregard for whether or not they personally agree with the order. However, what characterizes the control mechanism of discipline is the development of cognitive structures through practice and training to foster an “undeterred objectivity” whose purpose is to promote rational, uniform obedience in large numbers of people. The emphasis here is not on the acceptance of a social order, but on practice through repetition—in other words, on a specific form of socialization (Weber [1922] 1968, 53, 1148–1157). Norbert Elias, in his theory of the civilizing process, further refines Weber’s subjective-individual and objective-structural elements of rule and control and links them more explicitly by referring to the theories of Sigmund Freud. The development of increasing social control as described by Elias takes place on two parallel planes. The psychogenetic dimension of the civilizing process involves the gradual “civilization” not only of individual behavior, but also of the personality structure of the individual and growing control of instincts and affects. In a complementary sociogenetic process, the modern state’s monopolization of financial and executive resources causes structures of rule and apparatuses of control to develop into ever more differentiated and ever more effective institutions and mechanisms: “The controlling agency forming itself as part of the individual’s personality structure corresponds to the controlling agency forming itself in society at large” (Elias [1939] 2009, 373). External coercion is supplemented here by internalized behavioral norms and self-control. This, however, is not simple adaptation to external expectations; rather, it represents part of a comprehensive remodeling of the personality (Lemke 2001). Similarly to Weber—who, however, associates his concept of discipline most immediately with certain structures (the military, bureaucratic administration) rather than with society as a whole—Elias describes this type of behavior modification as a form of conditioning that causes individuals to internalize social constraints through socialization processes, with the result that these

Control of Violence

11

processes take place almost automatically (Elias [1939] 2009, 415–416). This kind of self-control is far removed from the autonomous self-restraint of rational selfinterest in the market. Rather, it takes the form of an incorporated constraint and the penetration of the individual by the outer social order. In the process of civilization, control, as the reduction of arbitrariness in individual behavior, is affected less and less frequently by exterior constraints and increasingly takes the form of selfconstraint. The increase of affective control is not, however, a linear one, as Elias pointed out in The Germans (1996). It goes hand in hand with phases of decivilization that once again highlight the precarious nature of control and the impossibility of entirely removing violence from human interactions. According to Elias, therefore, acts of violence come about through losses of control over violence both on the part of the individual (affective control) and the government (monopoly on violence). A counter-thesis was formulated by Zygmunt Bauman (1989), who held that the enabling factors for the Holocaust lay not in the collapse of civilization, but rather in elements that are integral to civilization and the modern age. In his view, therefore, it is not a loss of control, but an excess of control in the form of rationalization and bureaucratization, as well as the concomitant high degree of individual affective control that enabled the mass violence of the twentieth century. The relationship between state rule, self-control, and the formation of the modern subject reappears in a central position in the thought of Michel Foucault, where it is linked to a comprehensive analysis of different forms of control over deviance and violence. In Discipline and Punish (1977), Foucault traces the development of punitive practices since the seventeenth century. He holds that the excessive punitive violence of torturing and executing lawbreakers, which was intended as a demonstrative restoration of the sovereignty of the monarch, was succeeded by the reform-minded concept of punishment as a means of education and of restoring individuals to the status of obedient entities, and by imprisonment as the predominant technique of correction. This in turn gave way to disciplinary measures as a way to subjugate bodies and make them useful (Foucault 1977, 135–141). In the course of this process, the forms, mechanisms, and objectives of control changed along with their function within the state and society. The ruler’s revenge for violations of his sovereignty gave way to the defense and protection of society, which feels attacked as a whole by infringements of the law. According to Foucault, punishment becomes a “generalized function, coextensive with the function of the social body and with each of its elements” (1977, 90). The figure of the ruler as a visible and identifiable agent of control and punishment recedes behind the anonymous character of punishment. At the same time, the violence of the punishment—which in previous times was frequently inflicted in public—is reduced and regulated and finally becomes transformed into a prison sentence that is served away from the public eye. Prisons evolve into a disciplinary system in which individuals are regulated and monitored in every detail of their daily lives and become the “objects of a supervised transformation” (Foucault 1977, 245) and the “development” and “disciplination” of the body (1977, 130–131). In this connection, Foucault points out the central significance of surveillance and of the close relationship between surveillance and domination in control: The “panoptic” mechanism of power is effective primarily

12


through continuous visibility, which causes the objects of control, who are aware of their visibility, to control themselves and thus to adopt the “coercive measures of power” and use them against themselves (1977, 200–209). In Discipline and Punish, Foucault paints a gloomy picture of modern societies as “disciplinary societies” in which surveillance and discipline extend beyond the penal system and turn into ubiquitous but largely invisible forms of control (1977, 264–291) whose economic effectiveness makes more violent forms of power obsolete. Foucault here shows how the state safeguards its own existence by disciplining bodies and minds, by intervening in many different public spheres and institutions such as schools, military barracks, factories, and prisons, and even in the ostensibly “state-free” sphere of its citizens’ private lives. Here control becomes more subtle and presupposes a certain degree of freedom of its subjects. In his later lectures about “governmentality,”4 Foucault (2004) refines this perspective, arguing that the control of social relationships and the control of the modern subject in today’s “liberal” European societies are grounded mainly in mechanisms of self-regulation based on individual freedom. While surveillance and discipline remain present, especially in the penal system, as control mechanisms, they do not dominate the overall logic of governmental control. In Foucault’s view, the form of government that is currently dominant is one that is based on the observation and regulation of social relations, one whose concern is not so much to exercise a confining influence as to allow the positive organization of “circulations,” i.e., productive social interactions (Foucault 2004, 36–38, 100). The core of this technique of government are “security dispositives:” relationships of (self-)control that can be initiated or enabled but are based primarily on mechanisms where the phenomena themselves give rise to the developments by which they are regulated (2004, 58–68). Control of violence figures here as the management of harmful developments that adapts scientifically and pragmatically to circumstances and aims to reduce harmful phenomena to acceptable proportions. This form of control may employ indirect mechanisms by acting on remote factors in order to regulate a phenomenon (Foucault 2004, 110). While Foucault’s analysis of disciplinary systems stressed the relationship between mechanisms of power and subjectivation as the subjective dimension of subjugation, his concept of governmentality highlights the freedom of the individual (2004, 101). In a conspicuous echo of the Scottish Enlightenment, he describes individual desires as productive forces whose interplay serves the collective interest (Foucault 2004, 112). 3.2.3 Control, Deviance, and Violence: The Concept of Social Control In general, the term “social control” is applied to structures, mechanisms, and strategies whose purpose it is to cause society’s members to adhere to its valid norms and standards. Thus, “social control” differs from “control of violence” in that it is the 4 Foucault

defines governmentality as “The ensemble formed by the institutions, procedures, analyses, and reflections, the calculations and tactics that allow the exercise of this very specific albeit complex form of power, which has its target population, as its principle form of knowledge political economy, and as its essential technical means apparatuses of security” (1991, 103).

Control of Violence

13

wider term that encompasses all kinds of norms and behaviors and is not limited to violence. In its original sense as used by Ross (1896, 1916), the concept of social control was closely linked to the question of how to reconcile individual behavior with the overall interests of modern society, and a wide range of control mechanisms and forms of social integration was studied for this purpose. Put another way, social control in this wide sense of the term denoted a society’s capacity to regulate itself. As a form of control that employed mechanisms of social influence, it was contrasted with forms of state control that applied coercion. In the first half of the twentieth century, the concept was adopted by US sociologists who were studying socialization processes and mechanisms of internalization (Mead 1959) and social control in small groups (Homans 1950) and in face-to-face interactions (Cooley 1922). Subsequently, however, the concept was increasingly adopted in sociological and criminological deviance research. Here the concept of social control was used in a narrower sense to denote those mechanisms and structures with which a society causes its members to comply with its norms, and special emphasis was placed on its formal control mechanisms and instruments (the police, the prison system, schools, etc.). Two strands of theory in the sociology of deviant behavior merit special mention in this context. Both these strands analyzed and developed the question of control in unique ways. The first strand encompasses the social control theories of Hirschi (1969) and Gottfredson and Hirschi (1990) and Tittle’s control-balance theory (1995); the other comprises approaches to conflict theory and interactionist perspectives in criminology (Lemert 1967; Becker 1963). The strand of social control theories is based on a human understanding and model of order which, as Jensen (2003) points out, are not unlike those of Hobbes. People are regarded as being fundamentally, or “by nature,” prone to criminal acts, while successful behavior regulation is seen as a product of internal and external control that limits this potential. Thus, losses of control or the weakening of originally effective control mechanisms become the central causative factor in explaining criminality and violence. This approach gained prominence through Hirschi’s Causes of Delinquency (1969), in which he developed a typology of four main forms of social bonds which ensure conformity to norms and which, when weakened or absent, may cause criminal behavior: emotional ties (attachment), individual goals and hopes (commitment), membership of and investment in conventional sequences of action (involvement), and the acceptance of moral and normative concepts (belief). Whereas these bonds represent a combination of internal and external control, the General Theory of Crime published by Gottfredson and Hirschi (1990) focused mainly on forms of self-control. According to this theory, individuals develop different levels of self-control as a result of interactions and processes of socialization—in other words, they learn to varying degrees to consider the consequences of their actions and to regulate their behavior accordingly. Patterns of high or low selfcontrol solidify into relatively stable personality types, which are more or less prone to criminal and violent conduct. While the first approach interprets violence as the

14


result of a loss of control due to weakened mechanisms of normative behavior regulation, the second approach explains violence as the result of the failed or inadequate development of self-control strategies. Tittle’s control-balance theory (1995) deals with control not only in the form of the external or self-control of an individual, but also examines the individual’s control over others. His central argument is that the probability of deviant behavior is affected by the balance between the control to which individuals are subject and the control that they exert over their environment (Tittle 1995, 142). Deficits in an individual’s control over his or her environment lead to predatory, defiant, or submissive forms of deviance, while an excess of control increases the probability of exploitative or decadent forms. The relationship between social control and deviant behavior is a complex one. Overly powerful social control over an individual and the individual’s resultant powerlessness do not, in Tittle’s view, result in a decline in criminal behavior; on the contrary, they may even trigger such behavior. According to this model, social control as a strategy for producing conformity may thus prove to be counter-productive and to result in a loss of control in the sense of the effective regulation and containment of individual behavior. The cause for this loss of control is an individual, emotional reaction to an external control which is perceived as excessive and which may give rise to an increased motivation to commit acts of violence. In a slightly different form, this ambivalent relationship between social control and violence has been the subject of conflict theory since the 1960s, in approaches stressing the role of interactive processes between deviant behavior and society’s reactions and control strategies. These approaches counter the definition of social control as resulting from social self-regulation by pointing out that it is always the powerful strata of society that have the capacity to control and to define norms and deviations. At the same time, the dominant question is no longer that of how mechanisms of social control can effectively prevent deviant behavior. On the contrary, the object of study is the ways in which definitions of norms and deviance, stigmatizing processes, and control strategies can actually contribute toward causing deviance and violence or to entrenching criminal careers (Lemert 1967; Becker 1963). The term “secondary deviance” coined by Lemert refers to this relationship, in which the labeling of certain persons as criminal and the application of certain control strategies by the state (institutional reactions, prison sentences) help to socialize the affected individuals into criminal roles and to trigger corresponding behavior patterns. Thus, this approach does not proceed from the assumption that control is a desirable condition in which criminality and violence can be contained and limited. Rather, it studies certain strategies and mechanisms of social control and their counter-productive effects, the results of which help to produce violence and, therefore, contribute to a general “loss of control.”5 5 Similar

interrelationships between control measures and violence can also be observed in the development of protest movements and violent conflicts. In general, as Neidhardt (1981) points out, the escalation of violence is a circular, interactive process in which both sides incite each other toward increased levels of deviance (and violence). This process results in compulsions, which are difficult for either the state or its adversaries to evade, so that both sides lose control over the process to some extent.

Control of Violence

15

The approaches described here examine social control from very different perspectives. In one case, social control is the only guarantor of conduct that conforms to social norms. In the other, agencies of control contribute significantly to the genesis of deviant behavior. What both approaches have in common is that—like the majority of research into deviant behavior in general—they are concerned primarily with types of formal state control or with control on the individual level (self-control). Horwitz criticizes this tendency in The Logic of Control (1990) and stresses the significance of informal norms and informal control in communities, families, and other groups. These relationships and social structures contain forms of social control that are at least as effective, if not more so, in regulating deviant behavior and violence.6 In general, formal control mechanisms only intervene in a subsidiary capacity and only to enforce certain norms, while the majority of the rules of conduct are guaranteed by informal mechanisms of control and intervention. In the case of modern societies, however, Horwitz diagnoses a weakening of these informal control mechanisms as a result of individualization processes and the dissolution of traditional social structures and value systems, although these tendencies vary considerably between individual states and social groups. Historical studies, too, tended to focus on informal methods of control. However, the concept of “social control” was not adopted until 1977, when A.P. Donajgrodzki published a study of Social Control in Nineteenth Century Britain. This work concentrated on charitable organizations, educational systems, the Salvation Army, and public markets, arguing that primary social control is exercised not by the police and the legal system, but by many different social institutions. Initially, this concept attracted little interest among historians. It was not until 2004 that Pieter Spierenburg sought once again to draw the attention of researchers to the social significance of social control mechanisms in different societies, stressing that research into social control had to focus not only on the classical field of crime and prosecution, but also on the churches and their normative controlling influence as well as control rituals and controlling institutions that are rooted in popular culture. Like Horwitz, Spierenburg’s chronology of social control postulates a decline in informal types of control and an increase in the significance of formal state mechanisms in the period between 1950 and 2000.

3.3 The Paradigm of Control: Fields of Action, Forms, and Mechanisms of Control When one studies the sociological and historical perspectives on control, it becomes clear that control in modern societies is a multi-layered, complex phenomenon. Control arises as a combination of self-control and external control, of coercion, discipline, and self-regulation, and it is exercised by various different protagonists. Its direct or indirect objective may be the control of violence. And control itself is subject to many different transformative processes, which change the forms, 6 Horwitz

(1990, 5–15); similarly, Garland points out the substantial role of informal control even for the functioning of formal control mechanisms (2001, 5, 124, 170).

16


strategies, techniques, and mechanism of violence control along with its functions and the spaces in which it is exercised. How can we define the concept of violence control? 3.3.1 Two Dimensions of Control Within our paradigm of control, we must distinguish between the two dimensions of the concept of control (as outlined above). Control includes state and private— formal and informal—mechanisms and techniques with which social protagonists, social groups, and individuals attempt to limit or prevent certain forms of violent activity or to eliminate such activity (or its causes). In this sense, controlling violence encompasses the various reactive or preventive actions by one or more protagonists that aim to monitor, verify, and intervene in accordance with the normative standards that prevail in a society or a group. These include general rules of conduct and rules governing the settlement of conflicts as well as certain margins for tolerance, i.e., ideas about the extent and the forms of violence that are “normal,” acceptable, and legitimate in a given society and what levels of violence go beyond the level that is culturally and discursively defined as “normal.” Thus, violence control refers to an interactive process of regulation between control agencies and objects of control. This terminology reflects the remarks we made earlier about the significance of control as activities of verification and surveillance. It is possible for these kinds of control strategies to fail or to have unintended, counter-productive consequences such as an escalation of violence. The second dimension of the concept of control describes an objectively or subjectively perceived state of mastery over processes and behavior patterns on the individual, collective, or social level. In this sense, control of violence means to analyze and observe a situation that encourages violence and to control potential or actual protagonists of violence in order to prevent, stop, or reduce violent acts. Thus, control in this sense refers to successful regulation or limitation—in other words, violence, a situation, or certain protagonists are under control. This implies not only effectiveness, but also a certain permanent quality of the regulating and limiting influence. If violence has been reduced by isolated measures but could break out again at any moment, it is not possible to say that the situation is under control. This form of control over social processes can also be guided by the statistical probability of violent acts rather than responding to specific acts of violence that have already occurred. Accordingly, losses of control arise when protagonists or groups of protagonists lose their original relationships of influence or when knowledge about the cause-and-effect relationships of violent acts is called into question, so that they lose control over their own or other protagonists’ behavior or over a situation. 3.3.2 The Field of Action of Control According to our definition, “control” refers primarily to specific acts by individuals or agencies embedded in specific institutional and structural contexts, not to

Control of Violence

17

forms of control that are directly derived from anonymous social structures or a topdown “controlling state.” Societies and states do not control violence; policemen do. So do private security services, councils of elders, and national representatives in the United Nations General Assembly. So what we are interested in are governmental and social, private, public, and commercial actors of control who have a considerable range of formal and informal strategies and forms of control at their disposal. As we saw above in our overview of historical and sociological perspectives, control of violence is a process that depends on many factors. Accordingly, the strengths and weaknesses of the actors of control and the efficacy of their strategies—i.e., their “control gains” in the sense of the successful containment or prevention of violence—cannot be evaluated in the abstract, but are dependent on various factors and on the concrete context in which violence occurs. The efficacy of control may depend on the means available (coercion, incentives, etc.). In certain contexts, therefore, it may also depend on the threat of potential penalization by state or social protagonists or by a coercive apparatus. It may also depend on the extent to which a group or an individual acknowledges or resists the prevailing norms and the legitimacy, which is ascribed to the controlling actor and the underlying social order. State protagonists who claim to have legal jurisdiction and who are, in general, able to employ powerful coercive measures may superficially appear to have a distinct advantage controlling violence by informal protagonists. This advantage is by no means inevitable, however, and does not apply to all occurrences of violence. Thus, a state that is perceived as over-powerful and illegitimate may trigger protest and even resistance from those it controls, while informal protagonists in a local community may be able to secure compliance by virtue of their high-status positions and moral authority. Although certain protagonists may achieve an increase in control at the expense of losses of control on the part of others, this is not inevitably the case either. For example, Spierenburg shows that, during the nineteenth century, the police, as the central authority for curbing violence, was not alone in gaining influence; rather, informal means of control in the workplace acquired increasing significance in the course of industrialization, while, community sanctions and ecclesiastical discipline remained alive in many regions of Europe. Conversely, however, some modern cities are experiencing simultaneous losses of control on the part of both the state and the informal sector; these are places where the presence of state institutions is little more than ephemeral and where the moral authority of community leaders is in decline (Spierenburg 2004, 9). In certain situations, therefore, there are simultaneous gains and losses of control on the part of numerous different state and informal actors of control. The object of control may take the form of individuals, groups, or whole societies and states. It may also comprise physical spaces, such as public squares that are controlled by video surveillance, or certain time periods. The selection of the object of control depends on the objectives and the temporal dimension of the control. If the aim is to end or contain violence, the acute violence of an actor is the decisive criterion for the application of reactive control strategies. In contrast, if the aim is

18


to prevent an act of violence, the selection of the object and the measures of control follows a logic of prevention or anticipation. In the former case, the police responds to a crime that has already been committed in order to prevent a repeat crime. In the latter case, a “puzzle” of specific parameters is assembled that will help to construct a scenario delineating the presumed future behavior of individuals or collectives that have hitherto maintained a low profile (Bigo et al. 2007, 7). In some contexts, for example, adolescents, migrants, the urban poor, members of certain organizations or associations, and even whole states that are deemed to be particularly susceptible to violent internal conflicts (“fragile states”) or to initiating external violence (“rogue states”) are construed as high-risk groups and thus become the object of measures to control violence. It becomes clear, therefore, that anticipatory strategies directed against “potential violent actors” are highly ambivalent because they are invariably based on the fragmentation of the social space into objects requiring protection and “dangerous ‘others’”—a process that involves labeling processes (Cohen and Scull 1983; Cohen 1985) and stereotype formation: “suspicion is selective and contains a large part of preconception as to who and what is suspect” (Lianos 2003, 421). Here, then, control is based not only on physical or legal power, but also on the power of definition. This phenomenon also shows that the field in which control is exercised cannot be limited to objects of control and immediate controlling entities such as the police or the legal system. Third parties too, who may function as mediators or in other capacities for resolving conflicts without violence, must be included in the field of control. Even protagonists from science and the media exert an influence on the control of violence by producing certain forms of knowledge. By trivializing or dramatizing different forms of violence they help to cause public insecurity and denial, influence public expectations of the actions to be taken by the political administration, and legitimize certain strategies of violence control. Thus, state and social protagonists in the media society must deal with the threat scenarios created by the media and must also counteract alternative interpretations of the situation and justify their own control strategies—by, for example, employing measures commensurate to the situation and making astute use of symbols. For example, current scientific discourse is dominated by neuroscientific explanations for violence in which the causes for violence and deviant behavior are traced to the neurochemical personalities of criminals. Like the discursive and political figurations of the late nineteenth century, in which discontent with existing institutions and measures for preventing violence and crime resulted in a shift from social to biological explanations for violence, these new “power-knowledge complexes” can contribute to the legitimization of new control regimes (Becker et al. 2008). Finally, it must be stressed that the interactive relationships between objects of control and actors of control are always characterized by a specific relationship between control and freedom. While interactions of control can result in control in the second sense of the term, i.e., in a relatively stable state of ruling or being ruled, we can nevertheless conclude, following Foucault (1999, 191 f.) that any interaction of control also includes “practices of liberty.” Garland (2001, 146) correctly notes that each punitive form of control “carries a price in terms of the erosion of civil liberties and the reduced power of the

Control of Violence

19

citizens vis-a-vis the state.” At the same time, however, control is based on certain degrees of liberty that are granted to the objects of control, and on their ability to resist and behave in non-conforming ways: “social control—even in its most formal expressions—has always been based on the ‘regressiveness’ of conformity, that is to say, on the possibility of negotiating via degrees of behaviour seen as ‘suspicious’, ‘dubious’ to the norm” (Lianos 2003, 420). In our view, therefore, the objects of control are not merely the passive recipients of control. Rather, they may offer a greater or lesser degree of resistance to control, depending on the social situation and the possibilities for resistance that it offers, and they may question rules and forms of control and even negotiate them anew. Thus, the relationship between control and liberty is not a simple trade-off except on a very superficial level. 3.3.3 Forms and Styles of Control Different actors of control may employ a wide variety of forms and styles of control. Horwitz identifies penal, therapeutic, conciliatory, and compensatory styles of control, all of which may be employed in conjunction with different forms of control. “Style and form of social control are two separate dimensions of control efforts. Each style can proceed through inaction, resort to the extralegal network or mobilization of social control agents” (Horwitz 1990, 97). Thus, even passivity, i.e., the failure to react, may represent a form of control. As we have seen, forms of formal and informal control may be distinguished on the level of societies and social groups. In liberal democracies, formal controls, which are generally exercised by state authorities, are usually regulated by laws. When a violent act is punishable under an existing law code, it results in sanctions that are intended to have an educational effect on the violent actor and to serve as a deterrent for others. The aim of such sanctions, therefore, is to control violence by preventing (recurrent) acts of violence. In contrast, informal controls are enforced not by invoking the authority of legally defined rules and procedures, but by the authority of persons in society who serve as guides and role models for certain types of behavior (Imbusch 2008, 463). Informal controls may also be exercised by groups that employ rituals and measures to enforce adherence to group norms within a specific cultural context. In eighteenth and nineteenth century villages, “rough music” served as a means of delivering a social rebuke—publicizing grievances and denouncing conditions or actions that were perceived as violations of the norm. Formal and informal types of violence control may strengthen and supplement one another. However, especially under conditions of “hybrid statehood” in postcolonial societies, they may also compete with one another by stipulating different penalties or by regarding a specific form of violent behavior—for example, violence against adulterous wives—as illegitimate within the legal order of the state but as legitimate in a non-governmental order. Here, violence is perceived as “control violence” in the form of a penalty for non-conforming behavior. As a counterpart to external forms of control exercised by others, the concept of self-control refers to forms of internal control on the level of the subject. Here it is once again necessary to distinguish between forms of self-discipline and

20


self-guidance. In contrast to self-discipline, forms of self-guidance “modify behavior autonomously and ostensibly in accordance with the individual’s own wishes to conform with anticipated standards, without requiring the presence of an explicit or current threat or a promise of specific benefits” (Singelnstein and Stolle 2008, 70). Thus, self-guidance represents a form of control over liberty. As we have seen, this concept appears in Foucault as governmentality. Here, unlike in Ulrich Beck’s “risk society,” (Beck 1986) modern individuals are not the passive victims of uncontrollable risks, such as becoming the victims of violence or losing their jobs. Instead, since the modern view holds that all risks are essentially controllable, they have the task of autonomously managing such risks. Thus, control in the form of self-guidance is both a promise and a threat: “It is impossible to invoke personal responsibility without victim blaming; the happy news that you are the master of your fate implies that it is your own fault if your fate is not a happy one” (translated from Bröckling 2000, 156). 3.3.4 Strategies and Mechanisms of Control There are many different strategies and mechanisms of control, such as penalties and discipline as informal or legal control strategies; strategies of persuasion and public opinion; mechanisms of socialization, integration, and compensation; and strategies of exclusion. In the emergent welfare states of the second half of the twentieth century and in pre-colonial societies, control through compensation and integration into social groups and associations was of primary importance. Control in the sense of adherence to norms was guaranteed through the incentives of integration and support during short-term difficulties and exclusions. These incentives might take the form of state welfare or of systems for self-help and informal compensation in the event of illness or inability to work. Thus, violent behavior that infringes a legally defined system of norms or a set of informal rules must be evaluated against the consequences of exclusion from the social context. The mechanism of control that is at work here is based simultaneously on punishment (exclusion) and incentive (compensation). The control strategies of exclusion may be temporary or may culminate in the permanent exclusion of “dangerous superfluous members” who are deemed incapable of self-guidance (Singelnstein and Stolle 2008, 112). In this context, then, losses of control consist in the failure of social integration (Imbusch 2008, 465). However, integration and exclusion must always be considered side by side. For example, the concept of citizenship is based on the inclusion of certain groups who are granted rights and promised protection, while it simultaneously involves the exclusion of other groups who are not granted this status (Schwinn 2008). On the aggregate level, the various control actions performed by different protagonists coalesce into a control regime that encompasses the totality of the dominant rules, forms, styles, mechanisms, and goals with which control is exercised within a society. Thus, societies in different epochs and different geographical regions employ different control regimes. While every society is likely to exhibit similar forms and mechanisms of control (socialization and integration, and coercion and

Control of Violence

21

exclusion), there are variations in the dominant patterns of control in different state and social orders (Janowitz 1975, 84). In authoritarian or totalitarian dictatorships, for example, forms of repressive state control are often more strongly developed. At the same time, there are fewer limitations on violence-controlling state authorities such as the police, the military, and the legal system, which are subject to weaker control than their counterparts in liberal democracies. In democratic states, the legitimacy of state rule is also a more powerful prerequisite for the efficient exercise of violence control by the state, while comparatively extensive use is made of forms of social self-regulation via market processes.

4 Relationships of Control and Violence and the Ambivalence of Violence Control Control and violence may be interrelated in many different, complex ways.7 An understanding of the causes of violence and the relationships between different factors is always a necessary, though by no means sufficient, precondition for controlling violence (Tittle and Böckler et al. this volume). While an understanding of the causes and manifestations of violence improves the possibilities for eliminating its causes and thus increases the likelihood of preventing certain acts of violence, and while the definition and categorization of violent phenomena may be an important precondition for effective control, control—such as that exercised by a coercive apparatus with deterrent effect—may be effective even in the absence of reliable theories to explain the occurrence of violence. Conversely, adequate explanations and an understanding of the causes of violence and the relationships in which it occurs do not necessarily result in effective control. Even when adequate explanations and efficient strategies for controlling certain kinds of violence exist, it may be impossible to implement them effectively, for example, if certain control strategies are seen as unethical and illegitimate or are irreconcilable with existing laws. Thus, violence control is contingent on numerous preconditions and is strongly dependent on normative criteria.

4.1 Losses of Control as Prerequisites for the Genesis of Violence One preliminary way of approaching the relationship between control and violence consists in a theoretical study of the preexisting conditions that may impede or favor violent acts. Subsequently, one might investigate whether, and if so, in what way, the controllability of such conditions has changed. If one assumes that such factors as the processes of socialization and distribution, an effective state monopoly on violence, and international alliances and treaties are preconditions which, in certain

7 Thanks

are due to Barbara Kaletta for contributing many important ideas and profound preparatory work for this section.

22


contexts, are able to impede the genesis of specific forms of violence or to contain acute acts of violence, then the ineffectiveness of formerly efficacious control mechanisms—in other words, a loss of control—may be conceptualized (independently of the construct of violence) as a precondition for the genesis of certain forms of violence. However, the variables of loss of control, on the one hand, and violent acts, on the other hand, are not in an explicitly causal relationship here. In other words, such losses of control do not determine violent acts; they merely limit or open up opportunities for violence. As we have seen, control as “acting on actions” (Foucault) aims to influence decisions to act, and protagonists of violence always have a moment when they are able to decide for or against a violent act (Oppenheim 1961; Tyrell 1980). Additionally, certain forms of violence are driven by reasons that have nothing whatever to do with the actions of violence-controlling authorities (Neidhardt 1989, 234), while violent interactions may develop their own internal dynamics that may resist all forms of control (Eckert and Willems 2003).

4.2 Dilemmas of Control: Control Measures as Triggers or Escalators of Violence Another possible relationship between control and violence can be observed where control measures and strategies, rather than merely failing or being ineffective, themselves become triggers of violence or contribute to the escalation of a violent situation. We have already seen that control, in the context of different control regimes and depending, for instance, on whether it is constitutionally legitimate or exercised by an authoritarian/dictatorial regime, may either subjugate the objects of control or provoke opposition and resistance that can culminate in violence. In this case, the object of the violence is to evade an actor of control that is perceived as excessively powerful and illegitimate. As Neidhardt (1989) shows, control of violence, especially by state authorities, always involves the weighing up of different factors: “The problems in dealing with violence are problems of balance, and solving them is a matter of finding the right dosage” (translated from Neidhardt 1989, 243). Thus, agents of control may either overreact or underreact to problems of violence. However, it is not the nature and extent of the control—“hard” punitive measures on the part of the state or “soft,” therapeutic or conciliatory control styles—that determines the success or failure of violence control and its paradoxical effects, i.e., the triggering or escalation of violence. Rather, it is the perceived appropriateness of the control measures that determine their effect. Whether or not measures to control violence will themselves contribute to an escalation of violence, therefore, depends to a large extent on the prevalent perceptions about whether the form and intensity of certain control strategies is just and appropriate to the observed violent behavior of individuals or groups and can, therefore, be considered legitimate (Neidhardt 1989, 241). Measures to control violence that are perceived as underreactions, however, may not only weaken the deterrent effect on potential perpetrators of violence, but also give rise to a different control dilemma. Garland shows that the under-enforcement of law in Britain contributed to a deep-seated public anxiety and a perceived control

Control of Violence

23

deficit that culminated in demands for harsher punishments and thus ultimately led to “excessive” reactions on the part of the controlling authorities. However, control dilemmas may be caused not only by a failure to apply the control measures expected by the public, but also by the continual invention of new control strategies and technical measures for preventing violence. Early-warning systems, surveillance technologies, and the continual identification of new risk factors inevitably also give rise to new threats and affect society’s discourse on violence and its expectations of control. As new control formations are never exclusively the result of rationalized procedures, but are always partly a symbolic and emotional response to perceived threats, control always includes “an unintended ‘recursive’ logic of expansion” (Innes 2001). The failure of control strategies and the experience of unintended, escalating consequences is not infrequently followed by extended, refined, and deepened measures and the inclusion of additional actors of control who are connected to one another in increasingly complex networks (Cohen 1985). The desire for violence control—which, unlike social control, is related directly to the basic, existential human need for physical integrity—is almost insatiable and thus contains a tendency toward progressus in indefinitum. At the same time, however, control measures are de facto limited and generally provoke the resistance of the controlled if they exceed a certain level. As violence itself is a “wave phenomenon” (Baacke 1993, 30) that does not feature continually in public discourse, but is cyclically spotlighted in the mass media and in scientific discourse (Butterwegge 1997, 163), our methods of dealing with violence, too, are linked to “control cycles” that reflect political developments, media attention, changed patterns of perception, and public demands for safety. And this gives rise to another control dilemma. Actors and institutions who are given a mandate to control violence or who volunteer to control violence invariably also pursue their own individual or institutional goals, so that they always follow an inherent logic dedicated to preserving the status quo—that may run counter to the goal of controlling violence. This can be seen particularly clearly in the commercial security sector, where actors of control abide less by social norms and standards of order than by their clients’ specifications, and are also less easy to bring to account for their actions (Leander 2006). But even the field of scientific knowledge production which we discussed above, which aims to contribute toward explaining the causes and contexts of violence and thus ultimately also toward controlling violence, cannot fully avoid this dilemma: it is increasingly adapting itself to the capitalist markets and has thus become dependent to some extent on the obligation to follow research trends that receive third-party funding. Accordingly, this field may generate an overly dramatized sensitivity to certain forms of violence and a lucrative range of ostensible solutions that may contribute toward the legitimization of counter-productive control strategies.

4.3 Violence as the Cause of Loss of Control Finally, the third dimension of the issue at hand is that violence may represent one of the causes for a loss of control over violence. This happens when certain actors

24


are faced with a reduced range of options for effectively containing or preventing violence by means of available control mechanisms. Forms of collective violence in particular, such as wars, revolts, or civil strife may contribute toward the disintegration of groups or whole societies and thereby destroy not only the social foundations for socializing and normativizing mechanisms of control, but also the economic prerequisites for control through social redistribution and security systems. Similarly, collective violence can weaken the social and political power of violence-controlling agencies and thus prevent access to apparatuses of repression. In extreme cases, the result of this may be anomic conditions in which neither state nor societal instances are capable of effectively containing violence and exercising influence on an increasingly complex network of competing and cooperating actors of violence driven by a wide range of different motives. Violent situations thus become uncontrollable, and a comprehensive, far-reaching, and long-term loss of control over violence manifests itself. More frequently, however, constellations appear in which severe violence results both in processes of reorientation for mechanisms and strategies of control and in transference and postponement on the level of the actors of control. Thus, state authorities may respond to an actual or perceived increase in violence or to the appearance of a new kind of violence by stepping up their repressive control mechanisms as a response to the perceived shift in appropriate control levels that makes such measures appear legitimate. At the same time, the population or the social groups in violent contexts may attempt to satisfy their need for safety and protection by means of increased vigilantism, civil defense societies, or cooperation with warlords and violent actors, so that either violence control is simply transferred to different authorities, or a “battle for control” ensues that further escalates the violence. Studies in cultural anthropology in Latin America, for example, have shown that in societies in which the state was unable to assert its monopoly on legitimate violence, a permanent confrontation developed “between equal or near-equal” actors of violence (Deas 1997). Additionally, Waldmann has shown for Colombia that, in situations in which violent actors and violence are ubiquitous, violence may become a mundane phenomenon in that norms of prohibition and affective barriers erode and the use of violence becomes a normal assertive strategy that no longer attracts any degree of public attention (Waldmann 2003, 157–165). Thus, a high frequency of violence is not necessarily always an indicator for actual or perceived individual and state or societal losses of control over violence.

4.4 The Ambivalence of Violence Control As our remarks so far have shown, violence control is a highly ambivalent process. The ambivalence is rooted in the fact that control itself may be a source of violence, for example, when individual or collective behavior is deemed by the actors of violence control to deviate from the norms that are conducive to serving the interest of the government. Thus, the boundary between legitimate control of violence and illegitimate “controlling violence” fluctuates. This was shown by Rosenbaum and

Control of Violence

25

Sederberg in their concept of “establishment violence,” which “consists of acts or threats of coercion in violation of the formal boundaries of an established sociopolitical order which, however, are intended by the violators to defend that order from some form of subversion” (Rosenbaum and Sederberg 1974, 542; see also Black 1983). It is hard for the analytical gaze of “neutral” observers to grasp these boundaries; their definition depends on subjective perceptions and on evaluation according to normative yardsticks. If violence in the first case is a deliberately applied method of violence control, the ambivalence lies in the fact that, in a second dimension, violence may escalate as an unintentional side effect of efforts to control it. A third dimension highlights the ambivalence of control as the power of definition and as an activity involving the drawing of boundaries. Thus, measures of preventive violence control in particular are always connected to the labeling and stereotyping of certain persons, groups, or whole societies. This may result in the development of prejudices with disintegrative effects that are difficult to deal with both on the individual level and on that of society as a whole. In this context, control is ambivalent because, although it is based on integration, it invariably also produces exclusion and disintegration.

5 Shifts in Perception Patterns and the Social Conditions of Violence Control The development of processes and mechanisms of violence control must be examined simultaneously on the levels of the state, society, and the subject. On all these levels, historical analyses and present-day assessments point out the growing requirements and problems of violence control as well as changes in perception patterns about the controllability of violence. First, on the social level, informal forms of control aimed at socialization, such as those exercised by families, neighborhoods, associations, or traditional communities appear to have been losing their significance as behavior-forming institutions since the mid-twentieth century. Many studies (such as Rojek 2001; Shrivastava 1992) point out that, in Europe and the United States as well as in post-colonial societies, rapid urbanization and modernization processes have frequently triggered a crisis for informal institutions of control. This is particularly true of traditional extended families, but also affects neighborhoods and church communities. Although the loss of significance of informal protagonists and control mechanisms has not been ubiquitous or uniform, it may be assumed that, as a result of these developments, formal control strategies of state institutions and protagonists are becoming increasingly important. Against the background of the crisis of social institutions or “milieus of enclosure” (prison, hospital, factory, school, and the family), which form the basis of control in disciplinary societies, Deleuze (1995) postulated a transition from the disciplinary society to the “society of control” and thus predicted a fundamental change in the nature of prevalent control mechanisms: “Enclosures are molds, distinct castings, but controls are a modulation, like a self-deforming cast that will

26


continuously change from one moment to the other, or like a sieve whose mesh will transmute from point to point.” The sources and mechanisms of control themselves become increasingly opaque and unpredictable from the perspective of the controlled. At the same time, they are also more comprehensive and more accurate, thanks to the advent of new technological possibilities for surveillance and monitoring. The electronic tag becomes the symbol of permanent control in an open milieu (Deleuze 1995, 178). Second, on the state level there appears to be a simultaneous crisis in formal regimes of state violence control both in OECD countries and in the post-colonial states (where the presence of a legitimate state monopoly on violence is the exception rather than the rule). Assessments of the gravity of this crisis vary. The literature draws attention to the decline in state resources for control in the context of globalization (Strange 1996). Some authors hold that this decline results in an additional increase in the significance of alternative government structures—usually referred to as “global governance”—on the supra-national level (for example, Schuppert and Zürn 2008; Nuscheler 2000; Brand et al. 2000; Duffield 2001). Additionally, other writers point out that competing protagonists and forms of violence control frequently become more powerful on the sub-national, local level in this context (von Trotha 2005). The general consensus is that almost every part of the world is showing signs of a crisis of the state monopoly on violence and a “denationalization,” privatization, and commodification of violence control (Schlichte 2000). In many countries, areas of limited statehood are developing in which the monopoly on violence begins to erode and the legitimacy of state rule is increasingly questioned (Risse and Lehmkuhl 2006). Against this background, we can also observe the simultaneous existence of different modern and traditional codes of law and violence in post-colonial states, while the monopoly on violence in western societies, although not explicitly called into question, is likewise being transferred to a noticeable extent from public to private control in the form, for example, of commercial security companies (Schlichte 2000, 162; Leander 2006). While some researchers view these phenomena as a more or less linear process in which the state monopoly on violence continuously loses its civilizing force and thus loses the ability to exercise the effective control on intra-societal violence that it was long believed to hold (von Trotha 1995; Thome 2004), historical studies on the eighteenth, nineteenth, and twentieth centuries suggest that it is unwise to overestimate the penetrative ability of state methods of violence in any era. They also pose the question whether the effectiveness of the state monopoly on violence is cyclical and dependent on specific processes of social and political change (Knöbl 2006). Additionally, it has frequently been pointed out that relationships between state and sub-state protagonists of violence control as well as their ideas of order and the rationales that govern their actions may be characterized both by competition and by cooperation as well as by the awareness of substitutive and complementary functions: “Non-state violence can attack, ignore, undermine, or complement the state. There are no clear lines of distinction” (Translated from Zinecker 2009, 1). Finally, on the level of the modern subject, theories of individualization and civilization have long held that social control is increasingly transferred into

Control of Violence

27

the individual in the form of various self-control strategies (Elias [1939] 2009). According to this view, extreme, sanctioning modes of control decline in importance while there is an increase in the significance of internal, norm-oriented mechanisms of control that are disseminated by agents of socialization (Eichener 1989, 356). If this thesis of a crisis of the core agents of socialization is correct, however, then the mode of action of self-control too must be questioned, and accordingly, increasing doubt has been cast on the assumption that the growing importance of self-control mechanisms and the declining significance of outside control represents a long-term “historical automatism” as suggested by Elias’s theory of civilization (Franz 2000, 72). Studies that reach the opposite conclusion, especially in the field of youth studies, point out a shortage of norms and values that could serve as an aid to orientation for the younger generations and predict that the result will be an increase in the application of outside control, since “the transformation of outside coercion into internalized self-coercion no longer occurs with sufficient reliability” (translated from Eisenberg 2002, 23). This perspective suggests that there is a crisis of control on the level of the modern subject as well. This crisis results in a “de-internalization of social control” (27). Against this background, James Coleman drew the following conclusions: “If some of these predictions are confirmed by research, they will be of great significance for social control in the future. They suggest declining levels of internalization with respect to norms in future generations, assuming that the family continues to function as the central agent of socialization. This means either that more systems of external control will be applied or that social control will weaken” (1991, 388). While these developmental trends of social control as self-control are fairly general, they nevertheless appear to be very closely linked to the question of violence control. Numerous studies indicate that aggressive behavior and the use of violence frequently originates in feelings of powerlessness (Sutterlüty 2002). These feelings result in a lack of individual control skills, and these skills can only be acquired if the individual is able, to some extent, to control and influence his social and natural environment. Thus, individual perceptions of controllability, i.e., of being able to control circumstances and social processes to a certain degree, and the converse perception of being powerless against one’s environment, are fueled by internal and external capacities for control that are reciprocally related to one other (Thome and Birkel 2007, 46). Given the developments we have outlined here, the mechanisms of violence control on the levels of the state, society, and the subject face a variety of challenges. They are confronted by new forms of individual and collective violence and by the unleashing of violence of types and magnitudes previously unknown. These factors increasingly cast doubt on the control capacities of modern states as well as those of social protagonists and individuals. However, what we are observing appears to be not so much a clear trend to more violence—i.e., a linear and global process of an increase in violence—as various phenomena of “shifting violence” (“bewegliche Gewalt:” Neidhardt 1997, 7): a tendency to violence that is difficult to localize and that is characterized by increases in certain areas and decreasing violence levels in others. This makes it difficult to combat the causes of the violence and restricts the possibilities for predicting future developments.

28


Additionally, there have been changes even in the ways in which we perceive and interpret safety and social dangers. David Garland (2001) points out the effects of this perceptual change in the United States and Britain, which he believes is closely linked to general perceptions of a lack of security that is, in turn, characterized by social, economic, and cultural factors in the late-twentieth century. In particular, a general sense of insecurity has permeated not only the working and lower middle classes, but also the educated middle classes in the face of changes and breaks in family structures and of the individualization of job market risks and social security. In Garland’s view, growing concern about crime, which had previously been an issue predominantly affecting the working and lower middle classes, and the weakening of natural controls in neighborhoods and communities, coincided with a pattern of the “under-enforcement of law” and dramatic media reports eroding the psychological and emotional detachment of their recipients. The result was the spread of a deep-seated perception of a “control deficit,” now expanding to the professional middle classes (156).

6 Empirical Fields of the Relationship Between Violence and Control If one takes violence seriously as an exceedingly complex interactive phenomenon that permeates the lives of individuals, that may be functionalized by collectives, and that represents the foundation of state institutions, then any engagement with issues of violence control must consider all three of these levels—the individual, social groups, and the state. In our view, three fields of violence lend themselves particularly well to analysis: school shootings, terrorism, and states in crisis. First, they appear to be particularly incalculable as to when and where they occur and/or as to the identity of the potential perpetrators and the targets of the violence, i.e., the selected victims. Second, some observers hold that these phenomena indicate a new dimension of violence due to the large number of casualties and the general destructive effect caused by these kinds of violence. Third, the causes and triggers of these phenomena appear particularly unclear, and there is a dearth of knowledge about the interplay of different factors. Fourth and final, the phenomena in question are perceived by the public as particularly unsettling, an impression that at times directly contradicts the actually small probability of them actually occurring.

6.1 School Shootings On the micro-level, the phenomenon of school shootings has attracted a high degree of public attention since the end of the 1990s, most notably in the United States, but also in some European countries. Although crime rates at US schools have been dropping since the 1990s along with most other types of crime, including killings of juvenile (Blumstein 2003, 659; Newman 2004), there has been a marked increase in

Control of Violence

29

the incidence of school shootings (Böckler, Seeger, and Heitmeyer in this volume). School shootings are a type of violence that occurs primarily in highly developed industrial nations. The two countries with the highest incidence—the United States and Canada—are two of the world’s wealthiest nations. In Europe, too, school shootings occur most frequently in countries in the wealthier north and west of the continent, with Finland and Germany leading the field. Although school shootings remain an extremely rare phenomenon, the discrepancy between the publicity given to school shootings by the media and society and the actual threat represented by this kind of violence indicates the presence of a deep-seated insecurity about its controllability. There are many factors that might lie at the root of this insecurity. As the victims at the selected school are chosen more or less at random, school shootings are not only an extremely incalculable phenomenon, but also represent a symbolic attack on the institution of the school as a whole. While Goffman (1972) and Foucault (1977) viewed the school as a “total institution” with permanent and uninterrupted surveillance capabilities, school shootings seem to provide a particularly graphic illustration of the gaps in school control systems (Newman 2004). Additionally, the places where school shootings occur do not conform to the general perceptions and images of typical “violent places.” While larger cities are perceived almost everywhere as places of chaos, crime, and violence, school shootings occur predominantly in smaller suburban or rural communities that are generally viewed as cohesive and strongly family-oriented with a high degree of social control. Additionally, school shootings represent a form of violence with a strong expressive dimension which, thanks to continual advances in electronic media technology, can easily be idealized and communicated to increasingly wide audiences and may thus serve as templates for imitation. Against this background, it is possible to ask a number of questions about the relationship of control and violence and about the possibilities and limitations of control in the case of this form of violence. Was it the schools and families, the central agents of socialization in western societies that failed in these cases? Could such acts of violence have been averted by more attentive teachers, parents, and peers, or by professional early-warning systems and stricter media and gun laws? Or are school shootings a phenomenon that is fueled only by the intrinsic motives of its usually male perpetrators and by a lack of self-control, so that external attempts to control it are fundamentally limited? Even more importantly, is it possible that when this is the case, initiatives to influence the behavior of a “conspicuous” student achieve nothing but to strengthen his or her urge to evade all such attempts at control and heighten the student’s resolve to demonstrate his or her superiority by violently transgressing every available normative boundary?

6.2 Terrorism On the meso-level, we focus on the phenomenon of terrorism—not only because both society and researchers are under enormous pressure to explain it, but also

30


because there is a widespread conception in the aftermath of 9/11 that these forms of violence are becoming increasingly unpredictable and uncontrollable. However, instilling fear and uncertainty in one’s enemy by creating a perception of unpredictability and uncontrollability has always been one of the typical characteristics of terrorist violence. Terrorist groups are comparatively weak protagonists who use a strategy of unexpected attacks from the underground to spread fear and uncertainty (Waldmann 1998; Hoffman 2006). Accordingly, “terrorism’s most important resource is the diffuse fear of its opponent” (translated from Vasilache 2006, 155). Thus, terrorism presents special challenges to actors of control not only by its specific forms of organization (cells, networks) and violent practices (suicide attacks); as a strategy of deliberate public taboo breaking, it also creates an obligation to act that governments and agencies of control are unable to avoid. More importantly still, terrorism is based on provoking the ruling powers into overreaction and thereby playing into the terrorists’ hands (Waldmann 1998, 27–40). In this sense, state control measures are part and parcel of the terrorist strategy. And indeed, institutions forced to respond with violent counter-measures always run the risk of generating sympathy for the terrorists and escalating the situation. However, certain forms of current terrorist violence seem to be particularly conspicuous for their randomness and unpredictability. Religious terrorism and transnational networks in particular appear to be beyond the power of national and international state institutions to control. Although even this kind of terrorism has local roots and causes (Coolsaet and Van de Voorde 2008, 23), the network structure of transnational terrorism creates the impression of an absence of local points of reference, as though the phenomenon were “homeless” and “nomadic” (Schneckener 2006, 49). Since the 1990s the number of international terrorist attacks has decreased, but the number of victims has risen steadily, with most of the perpetrators belonging to religiously motivated violent groups (Hoffman 2006). This development, however, is not a linear one. The number of international terrorist attacks dropped by half between 2004 and 2006, while the number of victims decreased by 60%. During the same period, however, acts of domestic terrorism increased by 175% (Coolsaet and Van de Voorde 2008, 20). Because of problems in measuring violence and clearly delimiting definitions and concepts, quantitative statements about terrorist violence must be read with some caution. This makes it all the more important to examine the new challenges for violence control that arise from these issues. Do the police and other internal security authorities still have efficacious measures at their disposal for countering the forms of present-day terrorism, or are we dealing with a new quality and new organizational methods of violence that are completely impervious to police control strategies of the nation-state and that can be contained only through international cooperation and the deployment of the secret service or even the military? Or has the international “war on terror” already created a situation where control of terrorist violence is nothing but an illusion that provides a false sense of security, while the terrorists are increasingly gaining control of future events and are able to dictate the next steps taken by their “adversaries?” What other control methods are available outside of state intervention, and under what conditions can they be effective?

Control of Violence

31

6.3 Violence in States in Crisis On the macro-level, finally, we examine violence in the context of statehood in transformation, with a view to identifying its impact on control methods. As we have seen, almost all parts of the world exhibit tendencies toward the denationalization, commercialization, and privatization of violence control. It is not in the postcolonial societies alone that the incomplete monopolization of violence control by the state is becoming more and more apparent in the aftermath of the changes in the global system of states that have occurred since the end of the Cold War. Even in the OECD states, the legitimacy of state rule and its ability to impose control is increasingly being questioned. Nevertheless, it is the post-colonial societies that appear to be especially affected by losses of control over violence. Depending on the extent of this loss of control, these societies are described as “weak,” “fragile,” or “collapsed” states (see, for example, Zartman 1995; Clapham 2002; Rotberg 2003; Milliken 2003). These problems are not empirically new, but the discourse “consolidates well-known and observable conditions into an image of growing threats” (translated from Schlichte 2005, 75). In particular, the phenomenon of what are termed “new” wars (Kaldor 1999) has prompted many observers to question the controllability of violence. This form of violence has three essential characteristics: asymmetric warfare, the centrality of sub-national/semi-private protagonists, and the demilitarization of targets, i.e., an extremely high number of civilian casualties. Individually, none of these factors are empirically new. In combination, however, they have resulted not only in changes to the processes of warfare, but also and most importantly—like the so-called fragile states—to a new perception of threat (Münkler 2002). Thus, civil wars have become much more visible and the subject of much more public discourse since the end of the East-West conflict, even though their number has been declining since the early 1990s (Wallensteen and Sollennerg 2000; Schreiber 2002; Marshall and Gurr 2005). This new perception of threats has led to an increase in external interventions that are only partly authorized by international law. Not infrequently, these attempts to regain control produce negative side effects and exacerbate existing escalation dynamics; every form of intervention has an impact on local power structures and, invariably, also serves the intervening power’s other—often competing—interests. Whether and to what extent this means that we are faced with an increasing loss of control over violence or, conversely, whether internal wars and the collapse of states must be interpreted as the battlefields on which various protagonists contend for a monopoly on violence control (Schlichte 2005, 79) is a question that must be examined in a historical context. Are violent conflicts and civil wars, then, signs of a far-reaching loss of control over violence on the part of the state, or are our modern ideals of states with monopolies on violence simply inadequate in the face of real-world conditions—in other words, could it be that state control has never been strong enough to allow us to speak of a loss of control over violence on the part of state authorities? Could another—and under certain circumstances, perhaps a better—way to control violence within societies be found by employing non-state protagonists and informal

32


procedures and strategies? Could violent conflicts be halted if only we had an exact idea of their causes and were able to eliminate those causes, or are we dealing with dynamics that are so complex and so autonomous that all interventions are doomed to fail, or even to have counter-productive effects, if they attempt to influence the power balance between the participating actors?

7 Structure of the Book 7.1 Topics and Objectives of This Volume The present publication examines the issue of controlling violence as a reflection of society’s fundamental social and political characteristics and one of the core challenges of modern societies. It sets out to examine the complexity and multidimensionality of different forms of control, their possibilities and prerequisites, and the interrelationships between control and violence, as well as to analyze the social and political contexts in which they occur. Who controls violence in modern societies? Which actors and institutions function as societally authorized or selfappointed controllers of violence? How is violence controlled, what strategies and mechanisms are employed to control it, and what ambivalences and dilemmas do they entail? How is violence and the controllability of violence perceived and discussed in different societies and cultures? Are we really facing new challenges from “uncontrollable” forms of violence? To what extent do control regimes change as debates about violence progress, and how are they affected by new technological options? These questions—which give some indication of the multi-layered complexity of the topic—are examined from an international, comparative standpoint, tackling the subject from the perspective of history, sociology, political science, and criminology. It is our view that this approach is particularly appropriate when examining the question of how to control violence in modern societies, as comparisons between different historical scenarios are indispensable for determining which particular violent phenomena and control mechanisms are specific to contemporary societies and which phenomena already existed in more or less comparable forms in other historical periods. Thus, our task is to assess the specificity of current constellations of violence control and to place them in a context of longterm causal relationships. Including observations about conditions in different parts of the world allows us to make international comparisons of forms and mechanisms, perception patterns and discourses, and of the significance of different protagonists. Although we strive on the conceptual level to identify fundamental mechanisms and overarching tendencies in violence control in modern societies, an empirical analysis of changes in the regimes and forms of violence control naturally requires us to focus on specific violent phenomena. It is only through a precise cultural, social, and historical contextualization of violent phenomena, violence awareness, and control regimes that we can arrive at an improved understanding of violence

Control of Violence

33

and violence control in different societies both in the present and in the past. Accordingly, this volume is divided into four sections. The contributions in the first section, “Mechanisms and Strategies of Violence Control,” deal with fundamental aspects, forms, and development patterns of violence control in modern societies. On the basis of the conclusions of the first section, the authors of the following three sections examine—from different perspectives, but always on an empirical basis— various aspects of violence control in the three selected phenomenological fields of school shootings (micro-level), terrorism (meso-level), and violence in states in crisis (macro-level).

7.2 Mechanisms and Strategies of Violence Control The first two contributions in this section deal with long-term development patterns and with the extent to which modern societies may be able to effectively contain violence on a permanent basis. Michel Wieviorka explores the fundamental preconditions of violence and postulates that violence is subject to ongoing change. In his view, violence has substituted conflict and thus stems from a principle more fundamental than the social. As a result, present-day forms of violence can no longer be described by means of classical sociological explanations. Wieviorka argues that society must resocialize violence in order to escape cycles of violence and to contain violence in the long term. To achieve this, society must once again place the motives of different types of subject of violence into a social context and must also consider the perspective of the victims. Wieviorka holds that an engagement with the trauma and suffering of the victims and their desire for reparations and compensation can help to improve our understanding of violence in general and can also serve as a necessary prerequisite for abandoning violence. The long-term developmental patterns of violence and control mechanisms are also the theme of the study by Steven Messner, Benjamin Pearson-Nelson, and Lawrence E. Raffalovich. Based on a cross-national sample, they analyze the development of murder statistics in the second half of the twentieth century, focusing especially on the role of the family—an institution that is typically believed to play a key role in mechanisms of informal social control. Building on LaFree’s general thesis that trends in crime at the national level reflect the operation of the basic institutions of a society, the authors argue that profound changes in traditional family arrangements have been responsible, at least in part, for a weakening of institutional control and rising levels of crime and violence. Charles Tittle and Helmut Thome highlight forms of control on the level of the individual. From different perspectives, both authors examine forms of self-control and study the conditions under which they can be effective. Processes of individualization and the weakening of traditional institutions of informal control—a weakening that was diagnosed by Messner, Pearson-Nelson, and Raffalovich in the case of the family—impart a special and growing significance to the concept of selfcontrol. Charles Tittle’s contribution takes a critical look at the self-control theory of Gottfredson and Hirschi (1990) and, in particular, its suitability as a theoretical

34


basis for control measures. He identifies several weaknesses, which not only suggest that it would be problematic to extrapolate criminological prevention or intervention programs from the theory, but also suggest ways of eliminating these weaknesses. In Tittle’s view, effective control strategies must be based on a more comprehensive understanding of deviance and control. His main argument is that self-control theory holds promise as a beginning for policy formation and effective control strategies but it must be elaborated and integrated with other theories before it is likely to provide effective guidance. Helmut Thome’s contribution, too, deals with theories of self-control, with special emphasis on the concept of the conscience, which has been neglected or deemed irrelevant in many theories. The multidimensional concept of the conscience suggested by Thome encompasses several functions of human activity that are related to the necessity of expressing personal identity, safeguarding long-term personal interests, and sustaining cooperative relationships with others. Additionally, Thome discusses the role of positive and negative self-evaluation (shame and guilt) and their influence on violent actions. Thome argues that conscience as defined in this sense not only has an analytical potential for explaining (violent) crime, but that it can also contribute toward an analysis of more advanced social structures and relationship patterns. The two following studies focus on socio-cultural conditions and challenges of control. The contributions by Hans G. Kippenberg and Levent Teczan deal with the significance of religion for the genesis and control of violence. Religion—not just Islam—is widely believed to play an increasing role in violent conflicts, which may pose specific problems for violence control. One must also ask, however, to what extent religion can help to control violence. Kippenberg stresses that, while religions may prescribe actions that can legitimize violence, such prescriptions do not necessarily lead to a violent result. Rather, the results depend on how the protagonists define their own situation. Violent reactions must be expected only if the autonomy, or the very existence, of the religious community is under threat; in general, however, its powers of collective self-control and its willingness to make concessions are strong enough to prevent violence. Thus, Kippenberg cautions that religions in general should not be unilaterally demonized, emphasizing that misguided strategies for controlling religious groups and inadequate concepts on the side of the organs of control can have unintended consequences and may cause situations to escalate. Levent Tezcan takes up this point in his study of the ambivalent relationship between religion—in this case, Islam—and violence control on three relevant social levels. On the level of the individual, behavior is regulated by religious ideas of morality that aim to ensure harmony within the religious community, but that can also lead to violence against dissenters. On the group level, Tezcan studies the role of the religious establishment, which can exercise its influence on violent groups in discourses about legitimacy but which also feels that these groups undermine its authority. Finally, he describes the ambivalence of state control strategies that attempt to prevent violence by acting on Muslim communities. The contributions by Dagmar Ellerbrock and Klaus Weinhauer deal with discourses about and perceptions of violence control and with the emergence of new control regimes. Dagmar Ellerbrock examines the implementation of state controls

Control of Violence

35

on firearms in the early twentieth century. Faced with widespread violence that was rooted partly in a historical tradition of weapons ownership in Germany, the authorities were able to implement control measures only after a fundamental perceptual change had occurred. This change was reflected in newspaper reporting of gun violence as shocking and scandalous crimes and by the politicization of gun control in parliamentary debates. Klaus Weinhauer, too, examines forms of violence control in Germany, but he does so in a very different context, focusing on the dynamics of police behavior and police strategies for controlling demonstrations and political protests in the 1960s. He shows that police measures frequently had the effect of exacerbating the situation rather than containing it. The aggressive behavior of individual police officers was partly the result of an organizational culture based on aggressive masculinity and a code of professional ethics rooted in the time of the Weimar Republic. This organizational culture proved to be resistant to the attempts of political and police leaders to reform the system, prompting Weinhauer to discuss not only the contingency of control styles on cultural and perceptual patterns, but also the problem of how to control the controlling institutions themselves.

7.3 The Micro-level: School Shootings Following on these general considerations about fundamental mechanisms and strategies of violence control, the subsequent sections deal with the control of violence in three selected phenomenological areas. Section 2, on violence control on the micro-level, focuses on school shootings and juvenile delinquency, and opens with Dirk Schumann’s comparative historical study of interpretations of school violence and control strategies in the United States and Germany since the 1950s. Invoking the concept of “moral panic,” Schumann shows the ways in which the perception of school violence was related to fears about broader social developments. While the American and German debates differed until the 1970s—America viewed school violence primarily as a school problem, while Germany ascribed it to the effects of developments in society—the discourses converged in the 1990s. The growing sensitivity to school violence was the basis for a growing differentiation of increasingly sophisticated and complex control instruments. The following four contributions focus on the more specific phenomenon of rampage school shootings. The section starts with a contribution by Nils Böckler, Thorsten Seeger, and Wilhelm Heitmeyer, whose analysis centers around the theory of double loss of control. Differentiating between the individual level of the perpetrator and his loss of control over his own life, on the one hand, and society’s loss of control over the shooting, on the other hand, the authors show that school shootings cannot be systematically prevented through control measures. They argue that school shootings should not be regarded as a purely psychological problem. Instead, their analysis of the dynamics of escalation between control, loss of control, and the violent quest for control focuses more strongly on the problems of social disintegration and losses of recognition.

36


Next, Rebecca Bondü and Herbert Scheithauer turn to the explicit question of how to prevent incidents of school shooting. In their analysis of the preconditions and limits of control, they identify the phenomenon of “leaking,” in which the future perpetrators allow their violent fantasies or plans to “leak” out in advance, as a crucial early-warning sign and examine possible preventive strategies that may arise from this leaking behavior. They cite the example of the Berlin Leaking Project, in which researchers work in close collaboration with police experts and school psychologists. The authors stress that school shootings can, in principle, be prevented on this basis. However, it is clear that successful prevention requires a dense network of control, which the editors of this volume regard as a problematic issue in its own right: What about students who are unjustly labeled as dangerous? What are the consequences for their social situation in the school? The ambivalence of the concept of control becomes especially clear in this case. Ralph W. Larkin approaches the problem of controlling this form of violence from the perspective of cultural history and highlights the social development of images of masculinity. The history of masculinity in America is characterized by conflict over the definition of what constitutes masculinity. In Larkin’s view, this conflict is part of the dynamic that underlies violence, including rampage shootings at schools. The media and social discourses are also the focus of the last essay in this section, by Glenn W. Muschert and Massimo Ragnedda, who examine the role of communication processes in school shootings. They argue that these communication processes are of crucial importance for violence control, and that much of what can be observed with regard to school shootings is in fact a mass-mediated phenomenon. They present a model for understanding the types of communication that dominate the discourse around school shootings, and analyze in particular the role of performative scripts behind these incidents which have to be recognized by responding control institutions if they are to be effective.

7.4 The Meso-level: Terrorism Section 3 deals with terrorist violence and the special challenges it poses for the authorities and control mechanisms. Terrorism as a strategy of surprise acts of violence, deliberate violation of taboos, and cruelty always generates a compulsion to act and gives rise to dynamics of escalation, which are difficult for control actors to evade. The authors argue that these relationships of (calculated) uncontrollability must be examined from a historical and sociological perspective and that the latest developments—in particular, the much-discussed advent of a supposedly new kind of terrorism—must be studied with respect to their implications for the question of control. They identify the prerequisites and limits of control over terrorist violence and cite mechanisms of control that have received very little attention to date, from internal limiting discourses to the influence of certain reference groups and constituencies.

Control of Violence

37

Two political killings in twentieth-century Germany—the murders of Walter Rathenau and Hans-Martin Schleyer—form the point of departure of Bernd Weisbrod’s contribution, in which he traces the relationship between the strategic/political aims of terrorist acts and their performative quality. Symbols, rituals, and media coverage form part of the internal logic of these murders and their strategic function and also have an immediate impact on their controllability. In the two cases cited by Weisbrod, communicative battles were waged in the form of public performances in two very different political and social contexts. The two cases show “family resemblances” between terrorist murders both in their “language” and violent practices and in the perpetrators’ voluntaristic and existentialist self-images. Thus, Weisbrod argues, control of terrorist violence can be effective only if it addresses the performative qualities of such violence and its symbolic and emotional aspects. In another historical study, Mate Tokic describes the development of the control strategies of the Australian state to counter the activities of Croatian violent groups in the 1960s and 1970s and examines the conditions under which a policy of relative indifference was transformed into an effective battle against Croatian terrorism. He points out that violence control—especially the control of political violence—is dependent on its political context. Using the example of Croatian separatists in Australia, Tokic shows that the struggle to control terrorist activities was closely intertwined with the struggle for political power and control of the Australian Security Intelligence Organization—a struggle over control of control institutions—and depended on its outcome. Turning to the problem of controlling violence in contemporary societies, Jitka Maleckova critically examines the thesis that terrorist violence today has transformed into a phenomenon distinct from its historical predecessors. Based on available empirical data Maleckova discusses assumptions made by the “new terrorism” school of thought concerning the profile of individual terrorists, the aims and structure of terrorist organizations, the predictability and control of their violent campaigns, and the response they provoke. Her results are mixed. While modern terrorism has certainly developed new features, which have to be analyzed and taken into consideration, it does not necessarily pose a threat that is beyond control. Rather, the available evidence suggests that control mechanisms do have an effect, and are broadly capable of adaptation to new forms of terrorist violence. Friedhelm Neidhardt’s contribution offers a fundamental analysis of the logic of terrorist violence and the conditions and limitations of its controllability. As a phenomenon that is characterized by an out-of-the-ordinary, asymmetrical, clandestine, and non-institutionalizable character, terrorist violence represents a context of action that is difficult to pin down precisely. Provocation and the creation of fear generate compulsions to act and problems of legitimacy for the actors of control, who become part of the independent dynamic of the conflict. All control measures are faced with the dilemma that both by underreacting and overreacting play into the terrorists’ hands. Together with the following two contributions, however, Neidhardt’s contribution points out the significance of reference groups and the

38


bonds between terrorist groups and their supporters, both of which represent influencing factors that may limit their violent campaigns. Effective counter-strategies must take these relationships into account as a factor that may help to contain terrorism. However, Neidhardt holds that terrorism as a strategy of violence is fundamentally independent of preconditions and is, therefore, impossible to “conquer” once and for all. The studies by Stefan Malthaner and by Carolin Goerzig and Khaled al-Hashimi focus on different aspects of the relationships between terrorist groups and their social base, examining the forms of control that result from these relationships. Based on the analysis of two Islamist movements—the Egyptian al-Jamaa alIslamiyya and Hizbullah in Lebanon—Stefan Malthaner examines the dynamics of control emerging from the relationship between militant groups and their social constituencies. He argues that control emerges as a result of orientation toward and interaction with a population valued as the militants’ reference group. The effectiveness of this influence, however, depends on the militants’ perspective and the conflict structure, where Malthaner identifies two opposing, self-reinforcing dynamics: on the one hand, a process of violent escalation leading to tensions with the local population, radicalization, estrangement, and eventually a loss of constraints on violent practices. On the other hand, a pattern of development in which support reinforces the militants’ orientation toward a community, inducing self-restraint in their violent campaign, which in turn reinforces the support relationship. Referring to a later historical phase of the same struggles—the cease-fire initiative and ideological revisions of al-Jamaa al-Islamiyya and al-Jihad in Egypt— Khaled al-Hashimi and Carolin Goerzig analyze the role of internal debate in bringing about and stabilizing the groups’ disengagement from violence. Contrary to conventional counter-terrorism approaches, which argue that terrorist groups should be disintegrated and isolated, al-Hashimi and Goerzig find that it was the groups’ attachment to their followers and to a social base that enabled the debate about the legitimization of violence to emerge and contribute to bringing an end to violence. This inner-Islamic debate now substantially weakens support for alQaeda. It is precisely because it lacks a defined base of support, the authors argue, that the al-Qaeda leadership is able to resist debate and any notion of a truce.

7.5 The Macro-level: Violence in States in Crisis Section 4 deals with changes in state rule and their consequences for violence control strategies. The contributions in this section examine the phenomenon that became the subject of international scientific discourse about two decades ago as “fragile,” “weak,” or “failed” states. However, because the comparatively recent “discovery” of failed states suggests that this could be a new phenomenon, this section is also at pains to draw attention to inconsistencies in the perception of violence in these contexts and, where necessary, to correct one-sided contemporary

Control of Violence

39

interpretations by means of historical comparisons. The contributions on historical and present-day states in crisis stress that crisis situations in states rarely involve a complete loss of control over violence, but rather cause the reconfiguration of strategies, mechanisms, and actors of violence control. In these processes, the lines between legitimate violence control and illegitimate “control violence” become blurred—which once again highlights the ambivalence of control and the epistemological problem that such lines of demarcation are not readily revealed to the analytical gaze of a “neutral” observer. Rather, defining the differences between control and violence depends on subjective perceptions and evaluations according to normative yardsticks—which especially in times of crisis may themselves be called into question. At the beginning of the section, Werner Bergmann and Robert Gerwarth present two historical perspectives on violent processes in the context of rapid social change. Both authors point out that situations of national upheaval involve not only the weakening of the state monopoly on violence, thereby bringing forth opportunities for protagonists of violence, but also call into question the social order as a whole, both territorial and symbolic. Werner Bergmann shows how the underlying objective of anti-Jewish pogroms in nineteenth- and twentieth-century Europe was to eliminate insecurity and restore order. Thus, pogrom violence represents a form of “control violence” which comes into being as a result of the weakness of the state apparatus of repression and the existence of a political vacuum. From the perspective of the actors, this form of violence was justified precisely because of the absence of sanctions by the state as the regulatory force, and this justification occurred long before Jewish people were declared “enemies of the state.” Similarly, the protagonists of violence in Robert Gerwarth’s study on paramilitary violence in Austria and Hungary after the dissolution of the Habsburg empire view themselves as instances of control, and aim to restore by violent means a territorial and symbolic order disrupted by revolution and to prevent further losses of control over society. In this context, the protagonists of violence are young men who were socialized during the war to perceive violence as a “form of expression.” Here, too, violence was able to occur because previously effective mechanisms of control had been undermined by war and revolution, while a new control regime had not yet been able to establish itself. The contributions by Jean-Germain Gros and Andrea Kirschner deal with mechanisms of violence control and different manifestations of loss of control in the so-called “failed states” of the twenty-first century. Jean-Germain Gros points out that “failed states” are not a monolithic phenomenon and goes on to develop a typology of four different phenotypes of failed states, each of which entails a different degree of state loss of control over violence. According to Gros, state failure neither leads automatically to violence nor is it synonymous with losses of control over violence, as control is frequently exercised by non-governmental mechanisms. Nevertheless, Gros holds that the state remains the quintessential supervisory body of violence control. And although he concedes that the international community may contribute in many ways to the failure of states, while its options for “fixing” failed states are limited, he believes that, given

40


the dangers that emanate from such states, restoring state order in these countries remains the paramount responsibility of the international community. The contribution by Andrea Kirschner takes a different approach, focusing on the non-governmental protagonists of violence who view their role as that of authorities controlling violence. Examining vigilante groups in Nigeria, she shows that such protagonists represent a societal answer to a perceived dysfunctionality on the part of the state in controlling violence and maintaining security. Such groups serve as a means of symbolic orientation where the state as a “national idea” is called into question. At the same time, Kirschner holds that the presence of vigilante groups should by no means inevitably be considered a symptom of an actual failure of the state and its loss of control over violence, even though a superficial examination may suggest that this is the case and this assumption is frequently made in the dominant debates on state failure. Instead, she argues that such groups are interlinked with the state by a multitude of practices and complex relationships of competition and cooperation, so that they simultaneously reinforce and weaken different dimensions of statehood. So, from religion, discourse, and theories of self-control, to rampage shootings, suicide bombings, and states torn apart by civil war, this volume covers a broad spectrum of forms of violence and aspects of control. Guided by a common set of categories and a shared perspective, the diverse strands of our three empirical fields converge around the question of control of violence in modern societies, portraying the phenomenon in its multi-layered complexity while still allowing us to trace coalescing lines of argument. The possibilities of comparison between phenomena and mechanisms—both similar and different—reveal the patterns of control and the effects it generates.

References Baacke, D. (1993). Strukturelle und inhaltliche Veränderungen der Jugendphase und Folgerungen für das Gewaltphänomen. In G. Stüwe (Ed.), Jugend und Gewalt. Ist die Gewaltbereitschaft Jugendlicher bereits ein Massenphänomen? (pp. 15–30), Frankfurt a.M.: ISS-Eigenverlag. Bauman, Z. (1989). Modernity and the Holocaust. Cambridge: Polity Press. Beck, U. (1986). Risikogesellschaft. Auf dem Weg in eine andere Moderne. Frankfurt a.M.: Suhrkamp Verlag. Becker, H. S. (1963). Outsiders; Studies in the Sociology of Deviance. New York: Free Press. Becker, P. et al. (2008). Bescheidenes Wissen—unbescheidene Ansprüche. Neurowissenschaft und Gewaltforschung—Ambivalenzen eines neu entstehenden Kontrollregimes. Eine Analyse aus sozialwissenschaftlicher und historischer Sicht. ZiF-Mitteilungen, 4, 5–19. Benjamin, W. (2007). Critique of Violence (Reflections). In B. Lawrence and A. Karim (Eds.), On Violence: A Reader (pp. 268–285). Durham, NC: Duke University Press. Bigo, D. et al. (2007). The Changing Landscape of European Liberty and Security: Mid-Term Report on the Results of the Challenge Project. Research Paper No. 4. Centre for European Policy Studies, Brussels. Black, D. B. (1983). Crime as social control. American Sociological Review, 48(1), 34–45. Blumstein, A. (2003). Youth violence and guns. In W. Heitmeyer and J. Hagan (Eds.), International Handbook of Violence Research (pp. 657–677). Dordrecht: Kluwer.

Control of Violence

41

Brand, U. et al. (2000). Global Governance. Alternative zur neoliberalen Globalisierung? Münster: Westfälisches Dampfboot. Bröckling, U. (2000). Totale Mobilmachung. Menschenführung im Qualitäts- und Selbstmanagement. In U. Bröckling, S. Krasmann, and T. Lemke (Eds.), Gouvernementalität der Gegenwart (pp. 131–167). Frankfurt a.M.: Suhrkamp Verlag. Butterwegge, C. (1997). Jugendgewalt als neue Austragungsform des Generationenkonflikts? In W. R. Vogt (Ed.), Gewalt und Konfliktbearbeitung: Befunde—Konzepte—Handeln (pp. 162–179). Baden-Baden: Nomos. Clapham, C. (2002). The challenge to the state in a globalized world. Development and Change, 33(5), 775–795. Cohen, S. (1985). Visions of Social Control: Crime, Punishment and Classification. Cambridge: Polity. Cohen, S. and Scull, A. (Eds.) (1983). Social Control and the State. Historical and Comparative Essays. Oxford: Martin Robertson. Cooley, C. H. (1922). Human Nature and the Social Order. New York, NY: Scribner. Coolsaet, R. and Van de Voorde, T. (2008). Jihadi terrorism: perception and reality in perspective. In R. Coolsaet (Ed.), Jihadi Terrorism and the Radicalisation Challenge in Europe. London: Ashgate. Daase, C. (1996). Vom Ruinieren der Begriffe. Zur Kritik der Kritischen Friedensforschung. In B. Meyer (Ed.), Eine Welt oder Chaos? (pp. 455–490). Frankfurt a.M.: Suhrkamp Verlag. Deas, M. (1997). Reflections on political violence in Colombia. In D. Apter (Ed.), The Legitimization of Violence (pp. 350–404). New York, NY: New York University Press. Deleuze, G. (1995). Negotiations 1972–1990. New York, NY: Columbia University Press. Dinges, M. (1998). Formenwandel der Gewalt in der Neuzeit. Zur Kritik der Zivilisationstheorie von Norbert Elias. In R. P. Sieferle and H. Breuninger (Eds.), Kulturen der Gewalt. Ritualisierung und Symbolisierung von Gewalt in der Geschichte (pp. 171–194). Frankfurt a.M./New York: Campus. Donajgrodzki, A. P. (Ed.) (1977). Social Control in Nineteenth-Century Britain. London: Croom Helm. Duffield, M. (2001). Global Governance and the New Wars: The Merging of Development and Security. London: Zed Books. Eckert, R. and Willems, H. (2003). Escalation and de-escalation of social conflicts: the road to violence. In W. Heitmeyer and J. Hagan (Eds.), International Handbook of Violence Research (pp. 1181–1199). Dordrecht, Boston, and London: Kluwer. Eichener, V. (1989). Ratio, Emotion und Kognition. Der Modus menschlichen Handelns als abhängige Variable des Gesellschaftsprozesses. Zeitschrift für Soziologie, 18, 346–361. Eisenberg, G. (2002). Die Innenseite der Globalisierung. Über die Ursachen von Wut und Hass. Aus Politik und Zeitgeschichte, 44, 21–28. Eisner, M. (2003). The long-term development of violence: empirical findings and theoretical approaches to interpretation. In W. Heitmeyer and J. Hagan (Eds.), International Handbook of Violence Research (pp. 41–59). Dordrecht: Kluwer. Elias, N. ([1939] 2009). The civilizing process: sociogenetic and psychogenetic investigations, Revised Edition. In E. Dunning, J. Goudsblom, and S. Mennell (Eds.). Oxford: Blackwell. Elias, N. (1996). The Germans: Power Struggles and the Development of Habitus in the Nineteenth and Twentieth Centuries. Cambridge: Polity Press. Foucault, M. (1977). Discipline and Punish. The Birth of the Prison. New York, NY: Random House. Foucault, M. (1991). Governmentality. In C. Gordon and P. Miller (Eds.), The Foucault Effect: Studies in Governmentality (pp. 87–104). Chicago, IL: University of Chicago Press. Foucault, M. (1999). Wie wird Macht ausgeübt? In J. Engelmann (Ed.), Botschaften der Macht. Reader Diskurs und Medien (pp. 187–201). Stuttgart: DVA.

42


Foucault, M. (2004). Sicherheit, Territorium, Bevölkerung: Geschichte der Gouvernementalität I. Frankfurt a.M.: Suhrkamp Verlag. Franz, P. (2000). Wie weit trägt das Konzept “Soziale Kontrolle” bei der Analyse aktueller gesellschaftlicher Entwicklungstrends? In H. Peters (Ed.), Soziale Kontrolle. Zum Problem der Normkonformität in der Gesellschaft (pp. 67–86). Opladen: Leske + Budrich. Gallie, W. B. (1956). Art as an essentially contested concept. The Philosophical Quarterly, 6(23), 97–114. Galtung, J. (1969). Violence, peace, and peace research. Journal of Peace Research, 6(3), 167–191. Garland, D. (2001). The Culture of Control. New York, NY: Oxford University Press. Goffman, E. (1972). Asyle – Über die soziale Situation psychiatrischer Patienten und anderer Insassen. Frankfurt a.M.: Suhrkamp Verlag. Gottfredson, M. R. and Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press. Gramsci, A. (1991). Gefängnishefte. In Herausgegeben von K. Bochmann and W. F. Haug (Eds.), 10 Bände. Hamburg: Argument-Verlag. Heitmeyer, W. (2004). Kontrollverluste. Zur Zukunft der Gewalt. In W. Heitmeyer and H.-G. Soeffner (Eds.), Gewalt (pp. 86–103). Frankfurt a.M.: Suhrkamp. Heitmeyer, W. and Hagan, J. (2003). Violence: the difficulties of a systematic international review. In W. Heitmeyer and J. Hagan (Eds.), International Handbook of Violence Research (pp. 3–11). Dordrecht: Kluwer. Hirschi, T. (1969). Causes of Delinquency. Berkeley: University of California Press. Hirschman, A. O. (1977). The Passions and the Interests: Political Arguments for Capitalism Before Its Triumph. Princeton, NJ: Princeton University Press. Hobbes, T. ([1651] 1970). Leviathan. London: Dent. Hobsbawm, E. (1994). The Age of Extremes: The Short Twentieth Century, 1914–1991. London: Michael Joseph. Hoffman, B. (2006). Terrorismus. Der unerklärte Krieg. Frankfurt a.M.: S. Fischer Verlag. Homans, G.C. (1950). The Human Group. New York, NY: Harcourt. Horwitz, A. V. (1990). The Logic of Social Control. New York, NY: Plenum. Imbusch, P. (2003). The Concept of Violence. In W. Heitmeyer and J. Hagan (Eds.), International Handbook of Violence Research (pp. 13–39). Dordrecht: Kluwer. Imbusch, P. (2008). Kontrolle. Einführung. In P. Imbusch and W. Heitmeyer (Eds.), Integration – Desintegration. Ein Reader zur Ordnungsproblematik moderner Gesellschaften (pp. 463–467). Wiesbaden: VS Verlag. Innes, M. (2001) Control creep. Sociological Research Online, 6(3). http://www.socresonline. org.uk/6/3/innes.html [accessed 01/11/2007] Janowitz, M. (1975). Sociological theory and social control. American Journal of Sociology, 81(1), 82–108. Jensen, G. F. (2003). Social control theories. In R. A. Wright (Ed.), Encyclopedia of Criminology. Chicago, IL: Fitzroy Dearborn. Kaldor, M. (1999). New and Old Wars: Organised Violence in a Global Era. Cambridge: Polity Press. Knöbl, W. (2006). Zivilgesellschaft und staatliches Gewaltmonopol. Zur Verschränkung von Gewalt und Zivilität. Mittelweg, 36(1), 61–84. Leander, A. (2006). Eroding State Authority? Private Military Companies and the Legitimate Use of Force. Rome: Centro Militare di Studi Strategici. Lemert, E. M. (1967). Human Deviance, Social Problems, and Social Control. Englewood Cliffs, NJ: Prentice-Hall. Lemke, T. (2001). Max Weber, Norbert Elias und Michel Foucault über Macht und Subjektivierung. Berliner Journal für Soziologie, 01, 23–46. Lianos, M. (2003). Social control after Foucault. Surveillance and Society, 1(3), 412–430. Liell, C. (2002). Gewalt in modernen Gesellschaften—zwischen Ausblendung und Dramatisierung. Aus Politik und Zeitgeschichte, B 44, 6–13.

Control of Violence

43

Marshall, M. G. and Gurr, T. R. (2005). Peace and Conflict 2005. A Global Survey of Armed Conflicts, Self-Determination Movements, and Democracy. College Park, MD: University of Maryland. Mazower, M. (1998). The Dark Continent: Europe s Twentieth Century. London: Penguin. Mead, G. H. (1959). Mind, Self, and Society – From the Standpoint of a Social Behaviorist. Chicago, IL: Chicago University Press. Milliken, J. (Ed.) (2003). State Failure, Collapse and Reconstruction. Malden/Oxford: Blackwell. Münkler, H. (2002): Neue Kriege. Hamburg: Rowohlt. Neidhardt, F. (1981). Über Zufall, Eigendynamik und Institutionalisierbarkeit absurder Prozesse: Notizen am Beispiel einer terroristischen Gruppe. In H. Alemann and H.-P. Thurn (Eds.), Soziologie in weltbürgerlicher Absicht: Festschrift für René König zum 75. Geburtstag (pp. 243–257). Opladen: Westdeutscher Verlag. Neidhardt, F. (1989). Gewalt und Gegengewalt – Steigt die Bereitschaft zu Gewaltaktionen mit zunehmender staatlicher Kontrolle und Repression? In W. Heitmeyer, K. Möller, and H. Sünker (Eds.), Jugend-Staat-Gewalt, Politische Sozialisation von Jugendlichen, Jugendpolitik und politische Bildung (pp. 233–243). Weinheim/München: Juventa. Neidhardt, F. (1997). Gewalt, Gewaltdiskussion, Gewaltforschung. Bielefelder Universitätsgespräche und Vorträge, 7, 19–28. Newman, K. (2004). Rampage: The Social Roots of School Shootings. Boulder, CO: Perseus Books. Nuscheler, F. (2000). Global governance, Entwicklung und Frieden – Zur Interdependenz globaler Ordnungsstrukturen. In F. Nuscheler (Ed.), Entwicklung und Frieden im 21.Jahrhundert – Zur Wirkungsgeschichte des Brandt-Berichts (pp. 471–507). Bonn: Dietz. Oppenheim, F. E. (1961). Dimensions of Freedom: An Analysis. New York, NY: St. Martin’s Press. Popitz, H. (1992). Phänomene der Macht. Tübingen: Mohr. Risse, T. and Lehmkuhl, U. (2006). Governance in Areas of limited Statehood – New Modes of Governance? SFB 700, SFB Governance Working Paper Series, No 1, DFG Sonderforschungsbereich 700, Berlin, December 2006. Rojek, D. G. (2001). Chinese social control: from shaming and reintegration to “Getting Rich Is Glorious”. In J. Liu, L. Zhang, and S. Messner (Eds.), Crime and Social Control in a Changing China (pp. 89–104). West Port, CT: Greenwood. Rosenbaum, H. J. and Sederberg, P. C. (1974). Vigilantism: an analysis of establishment violence. Comparative Politics, 6(July), 541–570. Ross, E. A. (1896). Social control. The American Journal of Sociology, 1(5), 513–535. Ross, E. A. (1916). Social Control: A Survey on the Foundations of Order. New York, NY: Macmillan. Rotberg, R. I. (Ed.) (2003). State Failure and State Weakness in a Time of Terror. Cambridge, MA/Washington, D.C.: World Peace Foundation and Brookings Institution Press. Schlichte, K. (2000). Wer kontrolliert die Gewalt? Leviathan, 28(2), 161–172. Schlichte, K. (2005). Gibt es überhaupt “Staatszerfall”? Anmerkungen zu einer ausufernden Debatte. Berliner Debatte – Initial, 16, 74–84. Schneckener, U. (2006). Transnationaler Terrorismus. Charakter und Hintergründe des “neuen” Terrorismus. Frankfurt a.M.: Suhrkamp. Schreiber, W. (Ed.) (2002). Das Kriegsgeschehen 2001. Daten und Tendenzen der Krieg und bewaffneten Konflikten. Opladen: Leske + Budrich. Schuppert, G. F. and Zürn, M. (Eds.) (2008). Governance in einer sich wandelnden Welt. Politische Vierteljahresschrift, Sonderheft 41, Wiesbaden. Schwinn, T. (2008). Staatliche Ordnung und moderne Sozialintegration. In P. Imbusch and W. Heitmeyer (Eds.), Integration – Desintegration. Ein Reader zur Ordnungsproblematik moderner Gesellschaften (pp. 469–490). Wiesbaden: VS Verlag. Shrivastava, R. S. (1992). Crime and control in comparative perspective. The case of India. In H.G. Heiland, L. Shelley, and H. Katoh (Eds.), Crime and Control in Comparative Perspectives. Berlin and New York: De Gruyter.

44


Singelnstein, T. and Stolle, P. (2008). Die Sicherheitsgesellschaft. Soziale Kontrolle im 21. Jahrhundert. 2. Aufl. Wiesbaden: VS Verlag. Smith, A. ([1759] 1976). The Theory of Moral Sentiments. Oxford: Clarendon. Smith, A. ([1776] 1991). The Wealth of Nations: An Inquiry Into the Nature and Causes of the Wealth of Nations. London: Random Century Group. Spierenburg, P. (1994). Faces of violence. homicide trends and cultural meanings, Amsterdam 1431–1816. Journal of Social History, 27, 701–716. Spierenburg, P. (2004). Social control and history: an introduction. In: H. Roodenburg and P. Spierenburg (Eds.), Social Control in Europe: 1500–1800 (pp. 1–22). Columbus: Ohio State University Press. Strange, S. (1996). The Retreat of the State: The Diffusion of Power in the World Economy. Cambridge: Cambridge University Press. Sutterlüty, F. (2002). Gewaltkarrieren – Jugendliche im Kreislauf von Gewalt und Missachtung. Frankfurt a.M.: Campus. Thome, H. (2004). Theoretische Ansätze zur Erklärung langfristiger Gewaltkriminalität seit Beginn der Neuzeit. In W. Heitmeyer and H.-G. Soeffner (Eds.), Gewalt (pp. 315–345). Frankfurt a.M.: Suhrkamp. Thome, H. and Birkel, C. (2007). Sozialer Wandel und Gewaltkriminalität. Deutschland, England und Schweden im Vergleich, 1950–2000. Wiesbaden: VS Verlag. Tittle, C. R. (1995). Control Balance: Toward a General Theory of Deviance. Boulder, CO: Westview Press. Tyrell, H. (1980). Gewalt, Zwang und die Institutionalisierung von Herrschaft: Versuch einer Neuinterpretation von Max Webers Herrschaftsbegriff. In R. Pohlmann (Ed.), Person und Institution (pp. 59–92). Würzburg: Könighausen & Neumann. Uvin, P. (1998). Aiding Violence: The Development Enterprise in Rwanda. West Hartford, CT: Kumarian. Vasilache, A. (2006). Hobbes, der Terrorismus und die Angst in der Weltpolitik. WeltTrends. Zeitschrift für internationale Politik und vergleichende Studien, 51, 147–158. Von Trotha, T. (1995). Ordnungsformen der Gewalt oder Aussichten auf das Ende des staatlichen Gewaltmonopols. In B. Nedelmann (Ed.), Politische Institutionen im Wandel, special issue no. 35 of Kölner Zeitschrift für Soziologie und Sozialpsychologie (pp. 129–66). Opladen: Westdeutscher Verlag. Von Trotha, T. (2005). Der Aufstieg des Lokalen. Aus Politik und Zeitgeschichte, 28–29, 32–38. Waldmann, P. (1998): Terrorismus. Provokation der Macht. München: Gerling Akademie Verlag. Waldmann, P. (2003). Terrorismus und Bürgerkriege. Der Staat in Bedrängnis. München: Gerling Akademie Verlag. Wallensteen, P. and Sollennerg, M. (2000). Armed conflict 1989–1999. Journal of Peace Research, 37(4), 635–649. Weber, M. ([1922] 1968): Economy and Society: An Outline of Interpretative Sociology. New York, NY: Bedminster Press. Weller, C. (2003). Perspektiven der Friedenstheorie. INEF-Report 68. Duisburg. Institut für Entwicklung und Frieden. Zartman, I. W. (Ed.) (1995). Collapsed States: The Disintegration and Restoration of Legitimate Authority. Boulder, CO/London: Lynne Rienner. Zinecker, H. (2009). Editorial. In Dies (Ed.), Violence beyond the state – approaches to theory and forms. Behemoth. A Journal on Civilization, 2(2), 1–3.

Part II

Mechanisms and Strategies of Violence Control

An End to Violence Michel Wieviorka

1 The Subject and Violence There can be no discussion of violence today without involving the notions of Subject or subjectivity, in various ways.

1.1 Objectivity and Subjectivity The first thing that has to be stated is that the threat remains of the disarticulation between objective approaches to violence, which may be quantified and which can claim to be universal as they are theoretically acceptable to all, and subjective or relative approaches, which look at what an individual, a group, or a society considers as such at any given time. A legal definition of violence, centered on the state and, in the words of Max Weber, on a legitimate monopoly of force, appears to enable this problem to be set aside and violence simply to be objectivized. In this context, André Lalande’s Vocabulaire technique et critique de la philosophie (1968), backed up by Montesquieu, mentions the “illegitimate or at least illegal use of force.”1 But when the state entrusts private agents with a substantial part of war-making, as is overwhelmingly evident with the US intervention in Iraq (Makki 2004), and when internal security is likewise handed over to the private sector, a trend currently at work worldwide, the state monopoly of legitimate force is called into question and, thus, the possibility of discussing violence objectively, as in Lalande’s definition quoted above. The advent of the age of victims that began in the 1960s considerably strengthens this process of calling into question, and the upsurge of individual identities has considerable “memory” and “victim” dimensions. Many players nowadays are demanding acknowledgment of and, in some cases, reparation for the crimes M. Wieviorka (B) Ecole des Hautes Etudes en Sciences Sociales, Paris, France e-mail: [email protected] 1 “When

we, who live under civil laws, are compelled to make some contract not required by law, we can, thanks to the law, set aside violence” (Montesquieu 1758).

W. Heitmeyer et al. (eds.), Control of Violence, DOI 10.1007/978-1-4419-0383-9_2, C Springer Science+Business Media, LLC 2011

47

48

M. Wieviorka

perpetrated upon their forefathers and, at the same time, appearing in the public arena in connection with the harsh injustices they have suffered or, indeed, may still be suffering to this day. Such movements can be cultural, religious, or ethnic, perhaps national, black, or Indian; they may involve the survivors of a genocide or their descendants, or the parents or children of victims of a dictatorship or totalitarian power. Likewise, in several countries, increasingly diverse and effective mobilization is drawing attention to the violence suffered by women, children, the handicapped, or old people. Such players portray past and present violence not so much from the point of view of the threat to order or calling the state into question, but rather as an experience undergone and its consequences on those undergoing it; they speak of the trauma suffered and its effects over time, for example. Here, violence equals negation of or an attack on an individual’s physical and moral integrity, with implications that may affect succeeding generations. This makes it difficult to develop as a Subject; it invades subjectivity and takes the place of a subjectivization process. From this viewpoint, violence affects individual, personal, and collective existences. The tension between the objectivity and subjectivity of violence is not a purely theoretical problem; it can lead to fierce political debate. In France during the 1980s and 1990s, for example, people wondered whether delinquency and crime were increasing objectively, or whether it was in fact the feeling of insecurity that had increased, without any automatic link with an actual increase in crime, as the left claimed (before gradually moving away from this view of the problem). The harder it is to establish a direct link between acts of violence and their representations, the more the understanding of the one and the other falls into two separate registers which ultimately are almost completely dissociated.

1.2 Classical Approaches When thinking about an end to violence, it is not sufficient just to consider the victims and their subjectivity, however important their point of view may be and however considerable may be their ability to mobilize opinion and the media and to appeal to the state and political leaders. One must also look at the players involved in violence. Now, conventional methods of analysis are scarcely ever concerned with their subjectivity. Some people see violence as crisis behavior, a response to changes in their situation causing one or more players to react, often out of frustration. This approach gains respectability with Alexis de Tocqueville, who explains à propos of the French Revolution that violence was especially marked when the population found its situation improving: “One would say,” he wrote, “that the French found their situation the more unbearable as it improved” (Tocqueville [1856] 1998). Above all, though, it was British and American functionalist or neo-functionalist researchers who were responsible for the rise of this thesis, in the form of the theory of relative frustration, in the 1960s and 1970s. The idea put forward by James Davies, for example, and taken up to a considerable degree by Ted Robert Gurr (e.g., 1980), is in fact

An End to Violence

49

that violence develops when the gap between a group’s expectations and the scope for fulfilling them widens to the point of being intolerable. This type of approach has sometimes produced interesting results. However, in the 1970s various studies revealed its shortcomings and very limited explanatory power. Very differently, a second type of analysis stresses the rational, instrumental nature of violence, including in its collective dimensions—riots and revolution, for example. This may be said to have gained respectability with Thomas Hobbes but really took off in the 1960s, notably through the work of historian, Charles Tilly. For supporters of the thesis of “mobilization of resources,” violence is a resource, a means, which is mobilized by players in order to achieve their ends. Most of the time this idea serves to explain how players excluded from the political field use violence as a way of gaining admission and staying there. Such an idea has the advantage of no longer reducing violence to the notion of reactive crisis behavior; instead, it makes the perpetrator of violence someone who is aware of the issues surrounding the act of violence which, in turn, consequently makes sense. This approach argues that violence should not be separated from the wider conflict in the context of which it may arise, such as industrial action or a farmers’ demonstration. It has considerable explanatory power. Finally, a third type of approach, in fact very wide and diversified, postulates a link between culture and violence. Some writers regard the progress of culture, or rather of civilization, as the opposite of violence, in the tradition of the well-known study by Norbert Elias of the process of civilization, which explains how the modern individual has learned, in court, for example, to control his aggression and check his violent impulses (Elias [1939] 1974, 1975). Other writers stress how some cultures favor violence more than others, possibly through socialization and education—with reference, for example, to the work of Theodor Adorno on anti-semitism (1960). One problem associated with this set of perspectives is that the analysis generally omits political and social mediation and also disregards the historical layer that may separate the time when a personality is shaped and the moment of acting. Conventional approaches to violence should not be forgotten or rejected; they provide a perspective, which may be useful in order to understand a concrete experience of violence. However, they fail to deal with certain dimensions that are nevertheless essential and which the concept of Subject offers a way of comprehending.

1.3 The Subject of Violence Violence may present aspects suggesting a process of loss of meaning: when the player comes to express a meaning that has become lost or impossible and resorts to violence because he is unable to construct the confrontational action that would enable him to assert his social demands or cultural or political expectations, because no political process is available for dealing with them. A lack or loss of meaning does not necessarily lead to a vacuum, a complete lack of meaning and, ultimately, nihilism; such deficiencies often give rise to processes

50

M. Wieviorka

of manufacturing a new meaning of a more or less artificial nature, in other words, detached from reality, resulting in excesses and immoderation. In some experiences, for example, violence is based on an ideology from which it originates and which gives it a substitute meaning, as will be seen below with reference to Italian farleft terrorism. Other cases involve a myth, a discursive construction suggesting the possible integration of elements of meaning, which, in fact, become increasingly contradictory. Here, violence develops when the myth disintegrates and ceases to be viable, whereas the player nevertheless endeavors to keep it alive. In the modern world, though, religion often lends a metapolitical meaning to a violent act which then transcends politics, even if it soon becomes established again at that level. Violence has other aspects, which continue to elude conventional approaches. This is the case when cruelty, gratuitous violence, and violence for violence’s sake are involved; when the player not only destroys another but destroys himself, wipes out his existence by murderous, martyr-style acts. Or when the perpetrator appears to attach no personal meaning to his act, presenting himself as not responsible and claiming simply to have obeyed a lawful authority. This was the line of defense put forward by Eichmann in Jerusalem, as described by Hannah Arendt (1966). The concept of Subject may prove particularly useful for taking account of these different aspects, provided that the definition adopted is not too unimaginative or rudimentary. I, therefore, propose to establish five cases, each corresponding to a type of subjectivity that can be linked to violence (Wieviorka 2005). – The Floating Subject is one who, not managing to become a player, resorts to violence: for instance, the young immigrant from a run-down neighborhood setting fire to cars during the October/November 2005 riots as his only way of expressing, if not specific social demands, then at least his desire to build a life for himself. – The Hyper-Subject compensates for the loss of meaning by excess, to which he gives a new, ideological, mythical, or religious meaning. Here, violence is firmly linked to beliefs; it is the serious commitment of a meaning extending far beyond the bounds of the situation in which it is expressed, and aiming even further still. Islamic martyrism can serve as an illustration of this: the player kills, and in so doing extinguishes his own life, combining tremendous despair with a metapolitical vision that reaches beyond life itself. – The Non-Subject acts violently without in any way involving his subjectivity, at least apparently merely obeying orders, as in Stanley Milgram’s famous experiments (1974). His violence has no meaning from his point of view; it is nothing more than a form of submission to a lawful authority. – The Anti-Subject is that side of the Subject that fails to acknowledge the other person’s right to be a Subject and which can develop only by negating the other person’s humanity. This case corresponds to the dimensions of cruelty or enjoyment of violence for its own sake, as an end in itself. Here, the victim is dehumanized, reified, or animalized and is in every respect the opposite of the Subject. The perpetrator of cruel acts who finds pleasure in violence assumes that position and acts contrary to the humanist dimensions on which the concept of

An End to Violence

51

Subject is normally based—hence the use of the term Anti-Subject. Masochism is a perverted form of this scenario, in which the victim also derives pleasure from his own dehumanization. – The Survivor Subject corresponds to a situation where, before any aggression has actually taken place, an individual may (rightly or wrongly—it matters not) feel that his very existence is threatened, and act violently to ensure his own survival. This typology, briefly outlined here, would certainly deserve to be developed and the proposed terminology is not perhaps the most suitable, but it should be pointed out that until now we have lacked any sociological categories to permit a fuller description of these different cases. It has the advantage of helping us to tackle the most mysterious aspect of violence, which is also the core one: not the frustrations it may reveal, nor the more or less rational calculations made by the person resorting to violence, nor even the culture from which it stems, but the processes of loss of meaning and excess of meaning through which violence may develop, the share of surplus and lack involved, the twisted, corrupted, or sometimes also perverted subjectivity that makes violence possible.

2 Violence and Globalization We can no longer approach the issue of violence today as we would have done only 20 or 30 years ago. The world has changed, considerably, and the processes of globalization are at the heart of these changes. By thinking “globally,” we can approach violence with a fresh or altered perspective.

2.1 The End of the Cold War Let us look at the world as it was in the 1950s or 1960s. Essentially, it was structured by the central conflict between the two superpowers of the day, the United States, on the one hand, and the Soviet Union, on the other hand. The Yalta agreement, signed before World War II was even over, carved the world up into two zones of influence. The Cold War was, thus, a major ideological, economic, and geopolitical confrontation, but it never led either to head-on war or significant unmediated local conflicts. Neither the Korean War nor the Vietnam War pitted the two superpowers directly against each other, nor did they escalate into a much wider world war; they remained localized. Nuclear weapons ensured a degree of prudence between the two blocs and had a deterrent effect; the prospect of their use restrained extremes of action, even in times of high tension, notably the Cuban Missile Crisis in 1962. Warlike violence was, thus, limited throughout the world, as many countries were more or less firmly within the sphere of influence of one or other superpower and everyone knew that a localized war was likely to lead to global conflict. A report by the Human Security Centre in Vancouver, published in October 2005, admittedly forces us to qualify the idea of a world where military violence was

52

M. Wieviorka

lessened by the Cold War. Backed up by figures, this report states that many wars were fought by proxy in the Third World in those days and there was also local violence, in some cases very bloody. One should not, therefore, paint too idyllic a picture of that period. But the Cold War did prevent the escalation, spread, or extension of war, at least in its conventional form. It also had the effect of curbing international terrorism, which was carried on mainly by players claiming to support the Palestinian cause who, as will be seen, never went as far as they do nowadays. The end of the Cold War deprived the world of a way of structuring conflict, which avoided military violence far more than it authorized or facilitated it. After that, new splits appeared, civil wars took on quite a different aspect, and mass outbreaks of new or renewed violence began to occur. Organized crime prospered along with globalization. Whereas the number of conventional armed conflicts between states has decreased by 40% since 1992, according to the Human Security Centre report, and the bloodiest conflicts (those causing more than 1,000 battlefield deaths per year) have declined in number by 80%; and coups d’état or attempted coups have declined to 10 in 2004 compared to 25 in 1963, other forms of violence have increased. “Global” terrorism has struck a number of times, frequently killing and injuring dozens of victims in a single attack. Generally speaking, the percentage of civilian victims compared with military victims has increased considerably. Barbarity has become established in all sorts of parts of the world, including in Europe, where one might think that after Nazism, there would be no more mass crimes of a genocidal nature; the break-up of the former Yugoslavia involved violent “ethnic cleansing,” whereas in the Cold War era that country was in fact considered a factor for international stability. The Great Lakes genocide in Africa left more than one million people dead. And in Iraq today, the war in that country continues with extremely bloody daily acts of violence which could presage a civil war. Armed conflicts now take new forms: asymmetrical wars, for example, or crisis management in a supranational or multilateral context. Military interventions, sometimes by UN-appointed multinational forces, are increasing with the aim, in theory, not of winning in order to impose power, but, rather, of bringing situations of extreme localized violence to an end. The break-up of the former Yugoslavia with violence that lasted almost throughout the 1990s, the horrors of Africa’s Great Lakes, with the 1994 genocide, the violence perpetrated by pro-Indonesian militias following the creation of the independent state of Timor (1999 referendum), the disastrous experience of Somalia (1992–1993), the recent war in Lebanon (summer 2006), and the Darfur crisis are all new configurations of war, in which local confrontations and violence, in some cases charged with nationalist, religious, or ethnic significance, end in joint intervention by armies endeavoring, from outside the theatre of operations, to bring peace and re-energize civilian processes of restoring calm and development. Throughout the Cold War, moreover, nuclear weapons acted as a restraint and even promoted peace. Nuclear weapons have since become a factor, or at least a symbol, of major risks, associated with images of destabilization or regional crisis, notably in the Middle East and Asia, and with considerable problems of proliferation.

An End to Violence

53

Of course, the end of the Cold War does not explain everything, and a more detailed analysis ought, in geopolitical terms, also to cover, notably, the end of colonialism, the decolonization processes, and the ending of dependence for many Latin American societies. But the fall of the Berlin Wall was a turning point. At the time, the Cold War carried the stamp of violence, notably in what were termed the “proxy” wars; its end also meant an ending of these instances of violence. Often it had prevented the intervention of the United Nations (and also of other players, notably NGOs) in a preventive or peacekeeping role. The dawn of a new era brought with it fresh mediations, negotiations, and intervention and, thus, initiated a learning process in negotiated, democratic conflict management. On the other hand, the Cold War kept organized crime at a certain level and held international terrorism in check, because the main players in such violence needed the “sponsorship” of states often themselves within the sphere of influence of the Soviet Union; following its demise, the door was opened to a growth in organized crime and more intense forms of terrorism. The end of the Cold War did not in itself give rise to a fresh period of acts of violence (some of the most spectacular instances of which have been mentioned above), but it did play a large part in some major changes. In the words of the historian Charles Tilly it meant the invention of a new repertoire of violence.2

2.2 The End of the Industrial Age Globalization also meant big changes in the nature of capitalism and its associated forms of domination. The old industrial age, when economic power corresponded more or less directly to social relations located primarily in the factory or workshop, has given way to a phase when production problems seem to be dissociated from problems of economic power. No longer do the bosses hold the central role of dominant players, nor as was once thought do the managers; nowadays that role falls to “global” financial capitalism. Companies’ profits are, therefore, measured by a different yardstick to production, and it is not uncommon for a big group’s share price to rise even as it is announcing mass layoffs and the closure of factories that are still profitable—only not profitable enough. For modern capitalism, the short term prevails over the long term. As Richard Sennett (2007) points out, for example, “In 1960, a company was assessed in terms of predicted profits in three years’ time, whereas in 2000 that timescale has been shortened to three months on average” (437). All the same, conventional forms of work organization have not disappeared. Take the example of the “maquiladora” factories in Mexico, not far from the US border, most of which are dependent on big multinational groups and are part of the

2 Charles

Tilly proposes this concept in The Contentious French (1986), explaining that any population, in a given society, at a given period, has a limited set of collective actions, that is to say, means of acting in concert on the basis of shared interest. That repertoire changes as one moves from one type of society to another.

54

M. Wieviorka

global economy. The forms of worker exploitation that still prevail there often seem to date from another age, so harsh are they and so impenetrable by union action or the most basic protection offered by employment law (Lopez 2007). But there, as everywhere else in the world, globalization means the decline of the workers’ ability to take action, a loss of impact and, even more so, of centrality for the labor movement. Economic power and, thus, the “other side” in the conflict for a would-be labor movement player, is now too far removed from the site of production to allow the existence of a relationship comparable to the one that used to pit the working masses and the employers against one another while at the same time binding them, in the days when Taylorism was at its height. Capital can move at lightning speed, and the identifications that formerly inspired workers to want to become their own bosses and, thus, formed the foundation of counter-offensive protest have waned: how can anyone identify with a job, a type of work, a company if there is no job security, if they know they will have to change their line of work several times in the course of their life, and if the company regards them as eminently “expendable?” How is it possible to devise any long-term action involving a high degree of planning and structure if the work organization is constantly moving around and relocating and if the prevailing individualism and flexibility are depicted as the triumphant opposite of the traditional concept of worker solidarity? The end of the old industrial age did not make the world of work more violent; rather, it led to a loss of fighting spirit, an inability to carry on mass struggles and to link them with counter-projects for society or visions of utopia. Nevertheless, it altered the arena of violence. On the one hand, during the 1970s and 1980s violence took the form of far-left terrorism in several countries just emerging from that old order, notably Italy: students, intellectuals, and also, in some cases, workers, became more radical in order to carry on, by means of armed conflict, a fight which no longer had any meaning or reality in the factories. On the other hand, the decline of the core principle of conflict provided by the labor movement left a vacuum which no other player of similar stature stepped forward to fill, either socially or politically. Throughout the world communism did not collapse solely due to the break-up of the Soviet Union, or the exhaustion of its ideology: its disappearance also owes a lot to the decline of the labor movement which it represented, of which it was sometimes the main representative. And if the social democratic model seems to be on its last legs, or irrelevant, in the old industrial societies, the reason is partly that it continues to postulate strong ties between the party and the unions, in historical contexts where the unions have lost most of their power. Now, the lack of a framework for confrontation between the two sides is always a factor in a decline in social or moral standards and violence. When people’s expectations are not channeled into debate and conflicts between players, they degenerate into cynicism or fatalism, on the one hand, and crisis behavior and violence, on the other hand. One of the reasons for the urban riots in France in October and November 2005 was the lack of any forum, in the poorer districts, for putting the demands of young people—mostly from an immigrant background—into a confrontational setting and dealing with them at the political level. Thirty or 40 years previously, those same districts were “red suburbs” where the Communist Party, as

An End to Violence

55

well as an active, living fabric of associations, effectively channeled social conflict with feedback at political level. Nowadays, the Communist Party has abandoned the field and the old fabric of associations has vanished. The violent rioting, with several hundred vehicles set on fire every night for nearly 3 weeks, was the expression of a keen sense of dereliction and abandonment and also anger, and there was no institutionalized framework for conflict there to give it media coverage. Apart from the experience of the end of the industrial society and the decline of the labor movement, the foregoing remarks lead us to a general hypothesis that can serve as a basis for the analysis: the arena for violence is widening, while the scope for organizing debate and a framework for conflict to deal with social problems is shrinking, lacking, or vanishing. Conversely that arena becomes smaller when the conditions of institutionalized conflict permit a negotiated solution, even in circumstances of great tension between players. Violence is not conflict; rather, it is the opposite. Violence is more likely to flash when a player can find no-one to deal with in his attempts to exert social or political pressure, when no channels of institutional negotiation are available. This proposition should be considered as an analytical tool, not as a hard-and-fast rule—there are situations, experiences, or circumstances where conflict and violence go hand-in-hand. We have linked it to globalization because the more the latter is uncontrolled, purely neoliberal, and knows no borders, the more it undermines the institutions and representative bodies set up to deal with social demands within a framework of conflict. So why not wait for first attempts at establishing courts of law and supranational means of regulating economic life, or perhaps the advent of the alter-globalization movement in the hope that they will, in time, help to redefine the image of globalization and its consequences?

3 An End to Violence: The Victims’ Perspective The analytical bases described above correspond only to certain forms of violence and certain problems and, moreover, in no way permit an approach that could claim to be exhaustive. Our aim here, as throughout this contribution, is more to introduce a type of sociological approach rather than to supply systematic, heavily documented information about a particular topic. Not only does this approach indicate how one may tackle so important an issue as violence, or at least some aspects of it; but also it can extend the analysis of violence by looking at the conditions that may enable us to deal with it. Let us, therefore, return to the first of the points just made, the growing importance of the victims’ point of view.

3.1 Three Registers Democracies are increasingly sensitive to the victims’ viewpoint, and the themes of suffering, trauma, forgiveness, and reconciliation hold a considerable place in the public arena of democratic debate. What does an end to violence mean in a

56

M. Wieviorka

democracy when one is a victim, a descendant of victims, or a survivor? For such individuals and groups, the life-marking experiences of mass killings, genocides, slavery, the slave trade, and other crimes against humanity obviously did not come to a sudden halt on the day when barbarity was ended; they left their marks. An end to violence in fact means dealing with the present-day repercussions of past sufferings. That which is destroyed or altered, in that family of experiences, is not onedimensional; it in fact refers, according to eminently variable conditions, to three separate registers. The first of these is collective identity. Mass destructions not only liquidate human beings, but also, to a greater or lesser degree, a culture, a way of life, a language, a religion—hence the use sometimes of the neologism, “ethnocide.” The destruction of Europe’s Jews by the Nazis and their accomplices, for example, eradicated the Yiddish culture from central Europe and almost wiped out the language. Admittedly the language still exists, but it no longer has the slightest link to living communities, as in the days of the shtetl, the Jewish settlement in central and eastern Europe. It is true, as the work of the historian Jacob Katz showed, that such communities had already been eroded by the modern world and abandoned by many of the people who lived in them even before World War II. But Nazism acted with unprecedented force, practically annihilating that identity, with the result that it would never again contribute anything new, alive, or dynamic to humanity. All that remains is something that has been repressed, with only survivors left to try to keep traces of it alive, at the risk of lapsing into a “lachrymal” past, as it has been termed by the Jewish historian, Salo Baron. That identity fills museums and memories, it has its own memory, but that which gave it meaning has been lost, it no longer corresponds to a constantly evolving history. Here, reparation is impossible; that which has been destroyed cannot be brought back, and irremediably belongs to the past alone. The second register concerns individual participation in modern life. Crimes against humanity do not only affect groups external to modernity; on the contrary, those affected may be directly involved in modernity, or at least in contact with it and likely to be to some extent a party. What is concerned, therefore, is also a person’s ability to exist as an individual and to have access to money, consumer goods, work, housing, health, etc. Being a victim or the descendant of a victim, thus, means not only having been attacked in one’s cultural being and physical integrity; it also means having been treated as a slave whereas, within the same society, other people were free; it means having been deprived of one’s property, one’s rights, the sense of belonging at civil or national level to a larger collective entity than one’s group alone. To carry on the example of Nazism: the German Jews were highly integrated with German society and the German nation, almost assimilated, and when the Nazis told them they had been rejected by society and the nation, many of them failed to understand what they were being told. The great historian and sociologist, Norbert Elias (1991), who fled to the UK in 1935, relates in an autobiographical work how his parents refused to listen when he advised them to flee Germany: “nothing can happen to us, we’ve done nothing wrong, was the gist of what they said to him”. When individual participation in modernity is denied in this way by extreme violence, what is at stake is not only a collective identity, membership of a group, but

An End to Violence

57

an identification, more or less assumed with universal values to which one has been a party, or which have been held up as something for everyone to aspire to, and from which one has been shut out and forcefully expelled. Lastly, there is a third register, which has to do with personal subjectivity, that ability of any human being to be a Subject. Extreme violence annihilates or at any rate severely affects the Subject. It dehumanizes the person, treating him like a thing or an animal, or it may demonize him, attributing evil powers to that individual—in the way that women, in the past, were often called witches. This is why the survivors of a barbarous tragedy sometimes feel they cannot go on living; they have ceased to believe in the humanity of the personal Subject, they have experienced its negation within themselves and they have seen it vanish from their persecutors. How can anyone live after Auschwitz, it has often been asked?

3.2 Dealing with Threefold Destruction If we consider the three registers just described, putting an end to extreme violence means being able to deal with the threefold destruction that has affected first of all a collective identity; second, individuals inasmuch as they participate in modernity; and third, Subjects whose humanity is denied. What can the survivors or descendants of a collective identity defined by destruction do? If all they can put forward is that destruction, and the loss of any way of bringing back to life the collective being that has vanished, then any action they may take, insofar as they are able to express demands, will go no further than calling for acknowledgment of the barbarity that their group has suffered, with material compensation according to the case. If, on the contrary, they are able to advance a positive principle, whatever that may be, the strands of a culture that still has a hope of revival, a view of justice for the society in which they live, a demand for democracy, then the community or group concerned can move forward. Here, an end to violence means creating a “positive” identity, a principle that does not trap people in an identity which is “negative” because it has been destroyed and belongs only to the past. How can one rebuild in the case of the register of individual participation in modernity? Only a full and complete acknowledgment of that which has been prohibited or denied and the resulting wound can provide a satisfactory answer to the victims and those who claim to represent or embody them. The answer here is in the hands of those who hold the power to decide on such an acknowledgment, but who may have ideological or political reasons for not granting it. It may be a matter of protecting persecutors, avoiding re-opening very recent wounds, establishing or maintaining a precarious peace, going along with a consensus that has effected the transition from dictatorship to democracy relatively smoothly. Silence and obliviousness are generally justified by the overriding interests of the community; but they also obviously work in the interests of the persecutors and the guilty parties, to the detriment of the victims. The West German experience for instance, especially from the 1960s on, suggests that a country which decides to undertake the work on

58

M. Wieviorka

itself required by a recent past of extreme acts of violence and mass crimes comes out better than a country which refuses to do so. Debating the past and developing a policy of truth and forgiveness, as South Africa tried to do after apartheid, is the best way of helping the former victims to come back into the fold of the national community. Lastly, can negation be reversed when the victims have been devastated as personal Subjects and deeply dehumanized? If their predominant feeling is that it is no longer possible for them to take back control of their lives and continue living, then the only outcome is a descent into madness or suicide. Among the worst of cases is that where the victim feels, after the event, that his or her own behavior contributed to the negation, their own humanity, and that of others, that they played a part in their own debasement, which may spill over into a kind of inversion, being trapped inside a repugnant image of self which turns into a disgusting character that becomes their public image. Symmetrically it has been found, for example, from knowledge gathered about the Nazi concentration camp experience that in conditions of extreme violence the resources provided by faith or by a previous political commitment increase the likelihood of remaining a Subject despite the dehumanizing environment, as well as making it more likely for the person to rebuild their Subject afterwards. For the three registers of this analysis, the core of the end to violence is the same: it lies in the ability of the group, the individual, or the Subject to move forward. Irrespective of the register, three main attitudes are possible. The first of these involves shutting oneself up in the past, either in the barbarity experienced or in the time that preceded it, which will ultimately then be recalled as a golden age—before the disaster. In Sigmund Freud’s terminology, this attitude is one of “melancholy.” It may lead to demands for reparations, but is much less likely to evolve toward acknowledgment and forgiveness. The second attitude, on the contrary, tries to forget the past, to distance oneself as much as possible from past history, either the period of extreme violence or before that, in an attempt to merge completely into the society or nation in which one lives. In this case, there can be no debate about the past. Finally, the third attitude is to go through a period of “mourning,” again a Freudian concept, and to show that one is able to move forward and live fully in the society and in the nation, while still keeping alive the memory of the earlier experience and its destruction. This third attitude links the past, the present, and the future and is eminently favorable to opening up debate and processes of the “truth and reconciliation” type that have developed throughout the world, in particular in Latin America, following the South African example initiated in 1993. This requires great moral and political strength on the part its promoters. Nelson Mandela talked about this on a number of occasions, for example, with Bill Clinton, who relates their conversation, thus (Clinton 2004: 3): I said, ‘Madiba [Mandela’s colloquial tribal name, which he asked me to use], I know you did a great thing in inviting your jailers to your inauguration, but didn’t you really hate those who imprisoned you?’ He replied, ‘Of course I did, for many years. They took the best years of my life. They abused me physically and mentally. I didn’t get to see my children grow

An End to Violence

59

up. I hated them. Then one day when I was working in the quarry, hammering the rocks, I realized that they had already taken everything from me except my mind and my heart. Those they could not take without my permission. I decided not to give them away.’ Then he looked at me, smiled, and said, ‘And neither should you.’. . . I asked him another question. ‘When you were walking out of prison for the last time, didn’t you feel the hatred rise up in you again?’ ‘Yes,’ he said, ‘for a moment I did. Then I thought to myself, They have had me for twenty-seven years. If I keep hating them, they will still have me. I wanted to be free, and so I let it go.

3.3 Acknowledgment in a “Global” World Crimes against humanity, to confine ourselves to this particular form of violence, have taken place and still take place in arenas that do not necessarily coincide with the framework of modern-day states and nations. One immediate consequence is that the debates they may give rise to, and likewise their legal, political, and institutional processing, cannot, therefore, be restricted to that framework. In a “Westphalian” world, the state provides continuity between the past and the future, and it is within the state that decisions on the granting of rights are made, the processes of political debate take place, steps toward reconciliation or forgiveness are taken, reparations are approved, and so on. This in no way excludes international processes or the institution of courts like the one at Nuremberg to try Nazi criminals after World War II—but such courts exist by virtue of an agreement between states. However, acts of extreme violence, including those that took place in another age, often need to be analyzed and dealt with at “global” level. Looking at the contemporary consequences of the slave trade from the victims’ point of view, for example, means taking into account the historical dimension of the problem—nearly 15 centuries—and considering the role of all sorts of players in different parts of the world, in Africa, Asia, Europe, and the Americas (Pétré-Grenouilleau 2004). Thinking about the major “humanitarian” crises, the ethnic cleansing of the Great Lakes and the Balkans or the killings in Darfur, means bringing into focus processes inherent in Rwanda, former Yugoslavia, or Sudan and also, necessarily, regional and international, geopolitical, and economic dimensions. And if the survivors and descendants of the victims, some of whom have become refugees or exiles, are to emerge from such tragedies and rebuild their lives, that requires the involvement of many players, several of whom are external to the strictly local scene—humanitarian NGOs, international justice, international organizations like the UN or the EU, etc. What these survivors or descendants, the bearers, as Dipesh Chakrabarty (2007) says, of “historical wounds” may demand or hope for may involve the responsibility of several states, for acts for which not all of them are necessarily accountable, states moreover which may no longer exist or whose frontiers have greatly changed since the time of the acts of violence. Today’s players are not those of yesterday, yet accountability for the past is nevertheless likely to be unduly or wrongly laid at the door of one or other of them, so much so that many politicians now feel vexed by the demands for repentance that are made of them. The concept of descendant itself raises problems: how far can

60

M. Wieviorka

one postulate continuity down through the centuries in order to legitimately present oneself as a victim on the grounds of being a descendant of victims? The notion of survivor likewise deserves to be examined: in situations of extreme violence, why should they be representative of all the victims, as if they formed a homogenous group? Such questions make one’s head spin, as Jacques Derrida (1999) indeed showed with reference to forgiveness; how should one respond to the moral demand to forgive the unforgivable? It is neither straightforward nor easy to arrange forgiveness and make it meaningful in a situation whether both victims and persecutors belong to the same nation state and where perpetrators and direct survivors of extreme violence still live. And what should one do if those who can ask forgiveness are not the guilty ones, but simply people in positions of power who personally are guilty of no wrongdoing in connection with the violence? Or if, those who can grant forgiveness is merely more or less distant descendants of the victims? Or if, moreover, the state is not the sole or even the main framework within which these questions should be asked? Was the state that was the Federal Republic of Germany in Cold War days more answerable for Nazism than the German Democratic Republic? Is the state of Israel entitled to represent the victims of Shoah and if so, to what extent? Is it right for heads of state to forgive or to ask forgiveness in place of and on behalf of the victims, of all victims, including those not seeking forgiveness?

4 Dealing with the Violent Player: Subject Policies? Countering violence conventionally means linked policies of repression and prevention, either within the state framework (by mobilizing the police, the courts, schools, etc.), externally (diplomacy, war), or perhaps by combining the two dimensions. The latter is all the more necessary as globalization is blurring the points of reference, and the fight against organized crime and terrorism today, for example, calls for “global” strategies. We believe that our typology of the Subject of violence can make a useful contribution to developing these considerations. – If violence, at least in the dimensions that can be established, corresponds to the Floating Subject, that is to say, the difficulty or the impossibility of converting expectations or demands into action, then the most important thing is to establish or re-establish the conditions that allow conversion to take place. Such a proposition carries on very directly from the remarks made above about the opposition between violence and institutionalized conflict. It in fact means that the best strategy for reducing or preventing violence is to promote the training and development of social or political players responsible for the management—no matter how confrontational—of relations between them: the exact opposite of a breakdown in relations. At the global level, this means more and more players and institutions filling the supranational space all the time. At state level, it means forms of democracy that can resolve the contemporary crisis of political representation and allow the development and recognition of social and cultural players (see the analyses in Wieviorka 2007b).

An End to Violence

61

The same idea applies on smaller scales. Within a company, for instance, the presence of powerful, well-organized unions, while often regarded as a source of problems by management, is in fact also and above all the management’s best bulwark against the risks of a loss and breakdown of social or moral standards; union representatives are in fact a channel for bringing internal problems to light, avoiding the unspoken resentments that fuel a crisis, and negotiating; they also provide some predictability. Refusing even to allow a framework for conflict to exist does not lead to order and peaceful industrial relations; it is more likely to promote crisis behavior, starting with violence. – If violence concerns the Hyper-Subject, with its overabundance of meanings, it calls for symmetrical efforts. Here the problem is not one of a lack of meaning or of a framework for conflict; it has much more to do with the overload that turns a virtual conflict into war and violent breakdowns in relations. What those who can bring any influence to bear on the situation then have to do is to stop the dimensions corresponding to this meaning overload, be they ideological or religious, from over-determining action and preventing any debate or discussion, political or social process, or negotiation. Any intervention on their part is more likely to have some effect if it takes place at a very early stage, before the player has become so wrapped up in their own logic as to permit no concessions and to grant unqualified primacy to the absolute and radicalism. The Floating Subject requires bottom-up strategies, from the absence of conflict and mediation toward the building or strengthening of confrontational relations. The Hyper-Subject requires the opposite, top-down strategies, back down from the metapolitical to the political, from complete breakdown—notably religious— to debate and institutionalized conflict. A significant kind of intervention here may involve efforts to lend weight to those who, within the same ideology, or the same religion, can accept moderation, debate, and conciliation of their sense of identity with universal values of right and reason. This is the case, notably, in western democracies, whenever moderate Islam is respected and recognized there and is also encouraged not to allow Islamism to blight and weaken it. Once they have embraced a mindset of radical purity, violent players are not generally likely to relinquish their beliefs and give up the “all or nothing” that has become their way of thinking. The only way of ending violence in this case is, therefore, by force, repression, and calling in the army and the police. – It is hard to accept the hypothesis of the Non-Subject, for whom violence is meaningless, merely the expression of submission to a lawful authority. Such a hypothesis in fact takes away the violent player’s sense of responsibility, turning him into an automaton, a bureaucrat untroubled by conscience in the service of a machine, someone acting without convictions or passions, who unquestioningly accepts the order or instruction to act. Let us give this hypothesis the benefit of the doubt, though, let us admit that there are some people or situations, which it can explain. The only way to end violence here is to delegitimize the authority involved or, at least, the practices concerned. The admittedly light-hearted

62

M. Wieviorka

ordeals of “ragging” new students, for example, were tolerated in France for a long time, being anchored in a “tradition” itself rooted in certain educational establishments. It took energetic political intervention to delegitimize and ban ragging, and, thus, to make any continuing perpetrators aware of their responsibilities. Much more generally, anything in education, which can raise the sense of individual and collective responsibility, conscience the idea that the individual is accountable for his acts, can only serve to limit the space available for Non-Subject violence. – Cruelty and the violence of the Anti-Subject occur only in very specific conditions. They may, for example, accompany conventional warfare whenever a strong sense of impunity goes hand-in-hand with fear of the enemy, as was the case with the American forces facing the Japanese during the War in the Pacific (notably Dower 1986) or during the Vietnam War (notably in the My Lai massacre where 500 unarmed civilians were brutally killed by an American unit on March 16, 1968), or, to cite a recent example, at Abu Ghraib, in Iraq. Preventing cruelty means putting in place safeguards to prevent players from resorting to pure violence as an end in itself. This applies to the military authorities, in time of war, who must not, in theory and according to the law of war, permit any meanings other than instrumental to occur outside their control and responsibility. It applies to those in charge of armed conflict organizations, who must avoid the dilution of meaning in relation to political ends that the recourse to pure violence implies for their members—unless pure violence is to be used as a means of terrorizing the enemy. And in the case of common cruelty, on the part of criminals, for example, this seems so remote from the problems that can be solved by conventional political responses that it requires different resources to combat it: repression, admittedly, and education, but perhaps also religious, moral, or humanist values, of the kind referred to, for example, by Nelson Mandela at his meeting with Bill Clinton. – The violence of the Survivor Subject, as Jean Bergeret explains, is “dominating and archaic;” it is based on “a primitive fantasy that simply asks the question essential to the individual’s survival: ‘The other person or me?’ ‘Him or me?’ ‘Survive or die?’ ‘Survive at the risk of killing the other person?’” (Bergeret 1995, 46). The problem here is not an end to violence or, as Bergeret says, “controlling violence.” It is a matter of knowing that those who resort to violence do not have the personal resources or the mental models to deal with the situations in which they find themselves. To Bergeret, juvenile violence, the anger, and the hatred felt by France’s “suburban youth,” which come into this category, owe a lot to the shortcomings of adults, who are unable to provide them with suitable identity models. Today, violence is a taboo. Perhaps the last one. This has not always been so, and not that long ago it even had a certain legitimacy, be it extolling revolution, supporting national liberation movements, showing understanding of terrorist groups, or identifying with guerrilla warfare. Rather than analyzing it, intellectuals either supported it or opposed it, according to their political sympathies. The more it is held to be evil personified, the more violence appears meaningless to the point

An End to Violence

63

where, in its most decisive manifestations, it seems to be a product of barbarism or frenzy. The social sciences must not allow themselves to be swayed into accepting the half-baked ideas that reject violence without analyzing it from the point of view of the inhuman, the incomprehensible, and the absurd. Yet, they must also resist the idea of giving meaning to destructive and sometimes self-destructive modes of behavior. They must, therefore, tread carefully and acknowledge that any manifestation of violence, however insignificant, is linked to a meaning, but that link is twisted, perverted, is lost, or artificial. It is all the more difficult to follow that path given that the world in which we operate is no longer “Westphalian.” Violence is a particularly stimulating challenge for the social sciences, as it forces the researcher to do a balancing act and to produce findings ranging from the most personal, private, and subjective to the most general, international, and “global.”

References Adorno, T. (1960). The Authoritarian Personality. New York, NY: Harper. Arendt, H. (1966). Eichmann à Jérusalem. Rapport sur la banalité du mal, Paris: Gallimard. Bergeret, J. (1995). Freud, la violence et la dépression, Paris: PUF. Chakrabarty, D. (2007). Histoire et politique de la reconnaissance. In M. Wieviorka (Ed.), Les Sciences sociales en mutation. Auxerre: Editions Sciences Humaines. Clinton, B. (2004). My Life, London: Hutchinson. Derrida, J. (1999). “Le Siècle et le pardon,” conversation with M. Wieviorka. Numéro 9, Le Monde des débats. Dower, J. (1986). War Without Mercy. Race and Power in the Pacific. New York, NY: Pantheon Books. Elias, N. ([1939] 1974, 1975). Sur le processus de civilisation, vol. 1, La civilisation des mœurs, vol. 2 La dynamique de l’Occident. Paris: Pocket. Elias, N. (1991). Norbert Elias par lui-même. Paris: Fayard (translated from German, 1990). Gurr, T. R. (1980). Handbook of Political Conflict. New York, NY: Free Press. Human Security Centre (2005). Human Security Report. Vancouver, Canada, New York, NY: Oxford University Press. Lalande, A. (1968). Vocabulaire technique et critique de la philosophie. Paris: P.U.F. Lopez, E. (2007). En quête d’identité. Mondialisation, figures de la féminité et conflits sociaux à la frontière Mexique/Etats-Unis. Paris: Ecole des hautes études en sciences sociales. Makki, S. (2004). Militarisation de l’humanitaire, privatisation du militaire. Paris: CIRPES. Milgram, S. (1974). Soumission à l’autorité, Paris: Calmann-Lévy. Montesquieu, C.-L. de S. (1758). L’Esprit des Lois. Amsterdam: La Compagnie. Pétré-Grenouilleau, O. (2004). Les traites négrières. Essai d’histoire globale. Paris: Gallimard. Sennett, R. (2007). Récits au temps de la précarité. In M. Wieviorka (Ed.), Les sciences sociales en mutation (pp. 437–448). Paris: Editions Sciences Humaines. Tilly, C. (1986). The Contentious French. Harvard: World Belknap. De Tocqueville, A. ([1856] 1998). The Old Regime and the Revolutio. Chicago and London: University of Chicago Press. Wieviorka, M. (2005). La Violence. Paris: Hachette-Littératures, coll. Pluriel. Wieviorka, M. (Ed.) (2007b). Le printemps du politique. Paris: Robert Laffont.

Cross-National Homicide Trends in the Latter Decades of the Twentieth Century: Losses and Gains in Institutional Control? Steven F. Messner, Benjamin Pearson-Nelson, Lawrence E. Raffalovich, and Zachary Miner

1 Introduction In an essay published in the Annual Review of Sociology in 1999, Gary LaFree (1999) lamented two limitations in the criminological literature that seriously hindered our understanding of the macro-social determinants of criminal violence in general and homicide in particular. He observed that much of the quantitative research to date had been based on cross-sectional designs.1 While these designs provide useful information about the relationships between levels of theoretically relevant social structural variables and levels of homicide, they are not well-suited for identifying the dynamic processes associated with lethal violence. LaFree also faulted researchers for adopting ahistorical approaches. These approaches tend to treat “social processes as if they were independent of history” (1999, 158). As a result, the interpretations of crime patterns informed by such approaches are typically insensitive to fundamental transformations in the larger macro-social setting. The situation in the discipline has changed appreciably in the 10 years since the publication of LaFree’s essay. Researchers have increasingly employed longitudinal designs based on data not only for single nations, but also for multi-national samples as well. The designs of these multi-national studies—pooled, cross-sectional time-series designs—have two very advantageous features. They typically yield relatively large numbers of observations, thereby increasing the degrees of freedom for multivariate analyses. In addition, it is often possible to implement “fixed-effects” modeling with such data. Fixed-effects models greatly facilitate causal inference by statistically controlling for all time-invariant national characteristics, which lessens

S.F. Messner (B) Department of Sociology, University at Albany, S. U. N. Y., Albany, NY, USA e-mail: [email protected] 1 For reviews of the cross-national research on homicide up through the mid-1980s, see LaFree and

Kick (1986) and Neuman and Berger (1988). Gartner (1995), LaFree (1999), Messner (2003), and Neapolitan (1997) update these earlier reviews with citations to studies conducted in the late-1980s through the 1990s. More recent research is cited below in conjunction with model specification.

W. Heitmeyer et al. (eds.), Control of Violence, DOI 10.1007/978-1-4419-0383-9_3, C Springer Science+Business Media, LLC 2011

65

66

S.F. Messner et al.

the problem of omitted variable bias (Jacobs and Richardson 2008; Neumayer 2003). There has also been a growing interest in understanding current patterns of homicide in a larger historical context. This has been reflected in empirical analyses of trends in homicide for periods that encompass multiple centuries (Eisner 2003, 2008; see also Gurr 1981; Spierenburg 1998), as well as in theoretical efforts to develop explanations of the consequences of large-scale social transformations for changes in levels of lethal violence. Such transformations include broad cultural shifts from collectivism to individualism (Dicristina 2004; Messner et al. 2008; Thome 2007; see also Huang 1995) and the emergence of democratic political institutions (Karstedt 2006; LaFree and Tseloni 2006). The present study continues the exploration of cross-national patterns and trends in levels of homicide guided by both descriptive and explanatory objectives. At a descriptive level, we examine trends in homicide over the period extending from 1950 to 2005—a time span encompassing more than half a century for nations with complete data. We go beyond prior studies that have examined “raw” homicide trends or trends based on moving averages by applying spline regression. Spline regression is a useful smoothing technique to extract the systematic trends in timeseries data and to identify the significant “break-points” in the trends. On the basis of the spline regression analyses, we address the following questions: How prevalent were appreciable, sustained increases in levels of homicide? And among nations characterized by increasing homicide levels, how prevalent were noteworthy reversals in the upward trends, which might be suggestive of homicide cycles?2 Which nations exhibited features of homicide cycles, and what can be learned from the timing of increases and reversals in homicide levels? Our second overarching objective is to assess a possible explanation for changing levels of homicide by drawing upon arguments that were originally advanced by Gary LaFree (1998) to account for the rise and subsequent fall in levels of street crime in the United States in the second half of the twentieth century. Following in the general Durkheimian/Parsonian tradition as applied in criminology (see Messner et al. 2008), LaFree’s general thesis is that trends in crime at the national level reflect the operation of the basic institutions of a society. When institutions lose legitimacy, crime rates increase; when new institutional arrangements take root, crime rates fall. We focus in the present analysis on the institution that traditionally has been assigned primary responsibility for informal social control—the family. The chapter is organized as follows. We begin with a description of the data sources, variables, and measures. We then describe patterns in homicide trends with reference to the results of the spline regression analysis. Finally, we estimate dynamic multivariate models based on a pooled, cross-sectional time-series dataset 2 LaFree

and Drass (2002) examined pooled, cross-sectional time-series data to identify “crime booms” for a large cross-national sample. They report that only about a third (35%) of the nations in their sample satisfy all three of their criteria for a “boom:” positive direction, rapid growth, and sustained change. As explained below, we relax the criterion of “rapid growth” given our focus on “appreciable, sustained increases” in homicide in contrast with “booms.”

Cross-National Homicide Trends in the Latter Decades of the Twentieth Century

67

to assess the “family legitimacy” hypothesis for observed changes in homicide levels during the period under examination.

2 Data Sources, Variables, and Measures As noted, the objective of our spline regression analysis is to describe the systematic trends in homicide rates for a period extending over half a century. The homicide data are taken from the cause of death statistics reported by the World Health Organization (WHO), which is widely regarded to be the most reliable source of information on homicide for cross-national comparison (LaFree and Tseloni 2006, 32). WHO records homicides within the category of “external causes of morbidity and mortality” in its International Classification of Diseases (ICD). A death is classified as a homicide if it results “from injuries by another person with intent to injure or kill” (see http://www.who.int/whosis/mort/download/ftp/documentation.zip). We recorded homicide rates, expressed per 100,000 population, for nations reporting to WHO between 1950 and 2005.3 Our maximum sample contains 43 nations, but as explained below, not all nations are included in the analyses due to erratic data reporting and missing data on other variables. In addition, as is well-known, the nations reporting to WHO tend to be drawn disproportionately from the highly developed western societies (Stamatel 2006; Butchart and Engstrom 2002). Our samples cannot, accordingly, be regarded as representative of the universe of nations during the time period under investigation. These samples do, however, reflect considerable variation in socio-cultural settings and geographic locations. A list of nations is provided in Appendix. Our second, explanatory objective requires the specification of multivariate regression models. Model specification is complicated by the diverse array of theoretical arguments that have been put forth to explain cross-national variation in homicide (contrast, e.g., Antonaccio and Tittle 2007; Butchart and Engstrom 2002; Neumayer 2003; Savolainen 2000) and the ever-present problem of data availability, which is exacerbated when relatively long time periods are considered. Nevertheless, Gartner (1990) provides a useful framework for organizing the various 3 Gary

LaFree generously provided us with WHO homicide data from 1950 to 2000. We updated this data file with observations for additional years extending to 2005 where available. The ICD— currently in its tenth revision (ICD-10)—is periodically updated and categories are often changed in these revisions. This occurred for homicide deaths with the implementation of ICD-10. In ICD10, there is a separate category entitled “Sequelae of assault,” which indicates, “the cause of death, impairment or disability [is] sequelae or ‘late effects,’ which are themselves classified elsewhere. The sequelae include conditions reported as such, or occurring as ‘late effects’ 1 year or more after the originating event.” This category had previously been grouped with the other deaths that were classified as “homicide” (ICD-9 and before). In ICD-10 this category was moved to a separate sub-category under the broader heading of “Sequelae of intentional self-harm, assault and events of undetermined intent.” For our purposes, we have included in the analyses those deaths classified as “Sequalae of assault.” We combined across categories when updating LaFree’s homicide data to encompass more recent years. An account of the development of ICD is available at: http://www.who.int/classifications/icd/en/HistoryOFICD.pdf.

68

S.F. Messner et al.

hypothesized determinants of homicide. She proposes that these determinants can be grouped into four categories representing general contexts: the material, the integrative, the demographic, and the cultural. The material context refers to conditions that produce or reflect economic stress and affluence/deprivation. The integrative context refers to features of social ties that can exert social control. The demographic context refers to the contours of population structure. Finally, the cultural context encompasses norms supportive of violence, sometimes reflected in violent practices that are legitimated (e.g., capital punishment, warfare).4 We selected variables to represent these groupings of socio-cultural determinants, relying on the practices and findings of past research, within the constraints of data availability. Our strategic independent variable, the divorce rate, is a commonly used indicator of family instability and represents the “integrative context.” To maximize sample size and diversity, and to permit analyses over a long period of time, we employ the crude divorce rate, i.e., the number of divorces per 1,000 population. The data sources for the divorce measure, along with much of the additional demographic data described below, are the United Nations Demographic Yearbook and files obtained from the United Nations Statistics Division (2008).5 The divorce rate is a very robust predictor of homicide rates in cross-sectional analyses based on sub-national units for the United States.6 Moreover, Gartner (1990) presents highly suggestive evidence of an effect of divorce on cross-national trends in homicide. She estimates the effect of divorce rates, along with other independent variables, on sex- and age-specific homicide rates using data for a sample of 18 developed nations for the 1950–1980 period. Because the time-varying covariates in her models are only available for 5-year intervals, she “collapses” the homicide data into multiyear brackets, yielding seven time points for sex-specific analyses and five time points for age-specific analyses. She finds that the divorce rate is “by far the most important risk factor” for both males and females in the sex-specific analyses (1990, 100). The divorce rate also is significantly related to homicide rates for older children and adults, but not younger children (see also Hunnicutt and LaFree 2008). Our analyses provide for a more comprehensive assessment of the dynamic effect of divorce rates on total homicide rates by using finer temporal resolution (annual 4 Gartner

does not include a category for the “political context.” Subsequent research has considered the effects of democracy and democratization on homicide rates. See Karstedt (2006) and LaFree and Tseloni (2006). 5 We are very grateful to the UN Demographic Yearbook team at the United Nations Statistics Division for providing us with electronic files containing data on divorce rates, infant births and deaths, population size, sex ratio, and age structure to supplement figures retrieved online from the United Nations Demographic Yearbook. We also appreciate their expert technical assistance concerning the interpretation of these data. The demographic data from the UN are available from the authors upon request. Information on age structure for the United Kingdom for the years prior to 1982 were obtained from the Office for National Statistics, the executive office of the UK Statistics Authority (United Kingdom 2008). 6 See the extensive review of the literature by Parker et al. (1999) for evidence on the effect of the divorce rate on homicide rates in the United States. See Stretesky et al. (2004) for more recent evidence of the divorce effect on homicide rates for US cities.


69

data) for an appreciably larger and more diverse sample, and for a much longer time frame.7 The “material context” has been most commonly represented by indicators of the level of economic development, economic growth, and income inequality. The effects of indicators of economic development and growth are rather mixed, although there is some evidence that both are associated with lower levels of homicide in cross-sectional studies (e.g., Antonaccio and Tittle 2007) as well as in longitudinal studies (Neumayer 2003). We measure the level of economic development in terms of the Gross Domestic Product (GDP) per capita in constant dollars.8 We converted the measure into natural logarithms account for skewed distributions. When expressed as a first difference, this measure reflects short-term economic growth (the annual percent change in GDP/capita). The GDP data are available for download from the Penn World Table website (Heston et al. 2006). Income inequality has been regarded as a robust predictor of national homicide rates on the basis of the cross-sectional literature (Messner 2003; Antonaccio and Tittle 2007), although there is evidence to suggest that its effect may be conditioned by indicators of the generosity of the national welfare system (Pratt and Godsey 2003; Savolainen 2000). The scant evidence based on longitudinal data has been mixed (contrast Neumayer 2003 with Jacobs and Richardson 2008). In any event, the available data on income distribution are not sufficient to allow us to model annual changes in income inequality for an appreciable sample of nations over the time period of interest. We note that Pridemore (2008) has recently argued that poverty, rather than income inequality, is the more important economic determinant of cross-national variation in homicide rates, and he proposes that the infant mortality rate can be regarded as a useful proxy for poverty. In his cross-sectional study of 46 nations circa 2000, he reports a significant positive effect of the infant mortality rate on homicide and no significant effect of income inequality. The validity of the infant mortality rate as a proxy for poverty has yet to be established, but we nevertheless collected data on infant mortality rates for purposes of sensitivity 7 We

recognize that cross-national variation in divorce rates reflects to a large extent the legal and regulatory context. For an extensive discussion of these issues, see the essays in Adams and Trost (2005). A striking example of the impact of legal change in our dataset is an abnormally low number of divorces in Germany in 1978, which reflects the impact of Germany’s Marriage Law Reform Act, passed in 1976. The law established longer “waiting periods” for divorces, which created a 1-year artificial suppression of the divorce rate. We re-estimated the regression analyses reported below with this data point omitted, and the findings for the divorce rates are unchanged. It is of course possible that there are other anomalies in the data that have gone undetected and which bear upon substantive conclusions. We note, however, that the divorce rate is one of the few indicators of family structure that can be used for cross-national, longitudinal analysis, and as noted, there is suggestive evidence that despite measurement error, it may indeed be a useful predictor of homicide trends. Moreover, legal changes pertaining to divorce can themselves be regarded as manifestations of the changing meaning of marriage. Increases in divorce rates concomitant with revisions of the law are thus in some instances indicative of important cultural transformations in the institution of the family. See Cherlin (2004, 852). 8 The specific measure of GDP per capita is the Laspeyeres Index, which is a fixed-base index with 1996 serving as the reference year.

70

S.F. Messner et al.

analyses. The infant mortality rate refers to the number of infant deaths per 1,000 live births. We include three features of population in our primary models: total population size, age structure, and the sex ratio. Total population size is expressed in natural logarithms; the sex ratio is measured as the number of males per 100 females.9 The source for these variables is in most cases the United Nations Demographic Yearbook, supplemented with data from WHO in a few instances.10 The effects of these variables on homicide rates in previous research have been inconsistent (e.g., Jacobs and Richardson 2008; Pridemore 2008; Neumayer 2003), but we include them as controls in deference to conventional practices. An additional population variable that often appears in the cross-national homicide research is urbanism. Once again, the results for this variable have been mixed (e.g., Jacobs and Richardson 2008; Pridemore 2008), but we collected data on the urban population as a percent of the total population to serve as a control (World Bank 2008). Data for the urbanism measure are missing for a fair number of nation/time points. We accordingly estimate models with and without this predictor. The remaining context—the cultural context—has received the least attention in the literature given the difficulties in finding relevant information and devising plausible measures. A few studies have indicated a role for general value orientations, religious values, and values supportive of legitimate violence (for relevant citations, see Messner 2003, 705; see also Antonaccio and Tittle (2007) concerning religion). We do not attempt to capture the effects of cultural factors directly in our regression analyses, but the fixed-effects estimator controls for any time-invariant cultural differences across the nations in the samples.

3 Describing Systematic Trends in Homicide 3.1 The Technique of Spline Regression Modeling Spline regression has been applied effectively for smoothing time-series data in the analysis of trends in homicide rates for US cities (Messner et al. 2005; PearsonNelson 2008). This statistical method helps remove the “noise” that masks the “true” or “systematic” component in the trends. In addition to generating a statistical function that yields predicted values for the smoothed trend, spline regression can be 9 The

recorded sex ratios for Trinidad and Tobago for 1992, 1993, and 1994 diverge widely from that nation’s time series and are implausible. We accordingly omitted these observations. 10 Following the advice of UN personnel, we used estimated data—including population figures— wherever possible, given that these are best for time-series analyses (personal correspondence). When such data are not available, census and other representative survey data have been used instead. Technical details about data sources and estimation procedures used by the UN are provided in the individual data tables of Demographic Yearbook (http://unstats.un.org/ unsd/demographic/products/dyb/dyb2.htm). We used linear interpolation for total population size and age structure when there were gaps in the UN data of less than 4 years. With larger gaps, we used population data from WHO where possible or treated the data as missing.


71

used to identify points along the trend line where the slope changes significantly at an imposed level of significance. This characteristic is particularly useful for studying cyclical patterns in homicide because the presence of a cycle implies a distinct upward turning point and a distinct downward turning point. Spline regression allows us to identify changes in the trend line even with trends in which it is not clear how many knots (or break-points) there may be in the trend (Marsh and Cormier 2002). The mathematical procedure allows for various functional forms, including linear, quadratic, and cubic. The method can thus be applied both when the number of knots is unknown and when the functional form of segments between knots is also unknown. This flexibility is especially useful with data such as the homicide trends for individual countries, since it is impossible to predict a priori the best-fitting functions for the trends or the number of knots in the trend. The problem of estimating the number of knots and the shape of each segment between the knots can be solved by creating a large number of possible knots and trying multiple functional forms along the trend line and then relying on stepwise regression to determine those knots that are statistically significant and that yield the best functional form of the segments (Marsh 1986; Marsh and Cormier 2002). Spline regression models are based on dummy variables that are subject to continuity restrictions so that the regression segments between the series of knots always remain connected. A single equation is allowed to express the entire series of data points in a series (Marsh 1986). This attribute is especially useful for longitudinal data that are not linear. A model for an unknown number of knots with unknown functions can be expressed as (see Marsh 1983, 723–728): y=

3 j=0

β0,j tj +

k 3

βi,j (t − ti )j Di + ε.

i=1 j=0

In this case, y is the dependent variable, t is time, D is the dummy variable, and ε is the usual error term. The dummy variable is equal to zero until the point in the trend line where t is greater than ti (where i = 1, . . . , k), when the dummy variables are equal to one. There are k knots indicated by ti . Stepwise regression is used to select the location of the knots at an imposed level of statistical significance as well as the functional form (linear, quadratic, or cubic) of each of the segments between the knots. A large number of steps are generated by the stepwise regression procedure and variables are added to (or subtracted from) the model at each step to select the knot location and functional form according to the level of significance (see Marsh and Cormier 2002, 53). The present research follows the example of Messner et al. (2005) and PearsonNelson (2008) and uses a significance level of 0.05. The significance level applies both to inclusion (i.e., the significance level to enter the model or SLE) and to deletion (i.e., the significance level to stay in the model or SLS). We note that the number of knots, the timing of break-points, and the form of the function fit for any given trend line is likely to vary depending on the level of significance imposed. As a result, our spline regression analysis is best regarded as an inductive technique for

72

S.F. Messner et al.

discovering potentially suggestive patterns in the data rather than a means of formal hypothesis testing. In our application of spline regression, we are interested in determining which nations exhibited “an appreciable, sustained” increase in homicide during the latter decades of the twentieth century, and among this group of nations, which nations also exhibited a “noteworthy” reversal. There is no theoretical framework to provide formal criteria for these determinations, and thus we make use of the following rules of thumb for exploratory, descriptive purposes. For a nation to qualify as having had an appreciable, sustained increase in homicide rates, there must be at least one spline segment with a positive slope that endures for at least five consecutive years (“sustained”). There must also be an increase in the homicide rate of at least 33% (“appreciable”) when the highest fitted value is compared with the lowest fitted value for a spline segment with a positive slope or joined segments that are rising with no fitted value below the initial “upturn point.” For a nation with an upward homicide trajectory to be classified as one with a notable reversal, there must be an observed break-point at which the positive slope becomes negative (the “downturn point”), followed by a negative slope that persists for at least 3 years. The utility of spline regression techniques for uncovering the underlying patterns of homicide rate trends can be illustrated with a few examples. Below is the homicide trend line for Iceland from 1950 to 2005. The raw homicide rate trend is quite volatile because Iceland has a relatively small population, and the raw rate is highly sensitive to small changes in the number of homicides. Judging the general pattern of the trend line using the raw homicide rates is difficult because of the jagged trend line. However, the application of spline regression techniques indicates that the expected rate is a linear segment that gradually inclines during the second half of the twentieth century. According to our classificatory criteria, Iceland is an example of a country that demonstrates an appreciable increase in the expected homicide trend, but not a cycle, since there is no reversal (Fig. 1). A typical cyclical pattern is illustrated by Denmark. The general pattern of a decline followed by an incline and then another decline could be discerned from the raw data, but the spline regression function clearly highlights three linear segments. Additionally, removing the “noise” locates both the estimated low point (or upturn point) and the estimated peak of the systematic trend line (or downturn point). In this case there is an inflection point in the expected homicide rate at 1959 that marks the end of a downward slope and the beginning of the appreciable increase, which lasts until 1992. Once the trend line of the expected homicide rate reaches 1992, a linear decline follows until the last year of available data (Fig. 2). Norway is another country that experienced a cyclical pattern in the expected homicide rate trend. However, in this case, the pattern incorporates two curved segments, joined by two linear segments. Norway’s expected homicide trend line rises from the first year of available data (1950) and continues to increase until reaching the highest peak in 1986, with a brief interruption in 1984 and 1985. Since 1986 is the highest point in the trend line, and the slight drops in 1984 or 1985 never return

Cross-National Homicide Trends in the Latter Decades of the Twentieth Century Iceland 3.5

Homicide Rate and Expected Rate

3 2.5 2

Homicide Rate Expected Rate

1.5 1 0.5 0 2004 2001 1998 1995 1992 1989 1986 1983 1980 1977 1974 1971 1968 1965 1962 1959 1956 1953 1950 Year

Fig. 1 Spline regression results for homicide rates in Iceland

Denmark 1.6


1.4 1.2 1 Homicide Rate Expected Rate

0.8 0.6 0.4 0.2 0 1998

2001

Fig. 2 Spline regression results for homicide rates in Denmark

1995

1992

1989

1986

1983

1980

1974

1977

1971

1968

1965

1962

1959

1956

1950

1953

Year

73

74

S.F. Messner et al. Norway 1.8


1.6 1.4 1.2 1

Homicide Rate Expected Rate

0.8 0.6 0.4 0.2 0 2004 2001 1998 1995 1992 1989 1986 1983 1980 1977 1974 1971 1968 1965 1962 1959 1956 1953 1950 Year

Fig. 3 Spline regression results for homicide rates in Norway

to the initial levels, 1986 is recorded as the point of reversal into the downward trend, which continues until 2000 (Fig. 3). Venezuela provides an example of a country that experienced an appreciable increase in the expected homicide rate trend, but did not meet the criteria for a downturn. Venezuela’s expected homicide rate increase begins with the first year of available data and continues steadily until 1997, when the rate of increase becomes greater. We identify the upturn date at the beginning of the period—1950—because the initial linear segment is positive, extends over 5 years, and entails an increase of over 33%, even though a sharper incline is initiated in 1997. With respect to a reversal, the 1-year decline that starts in 2003 does not extend long enough to meet the criteria for a cycle, which requires at least 3 years of decline (Fig. 4). A final example is Japan, which did not experience an appreciable, sustained increase in the expected homicide rate trend during the period under investigation. There is a brief rise that starts in 1952 and ends in 1955, but since the expected homicide rate did not increase by one-third during this period, nor did it last for at least 5 years, the criteria for an appreciable increase are not met. After 1955, Japan’s expected homicide trend follows a curving pattern of decrease until the last year of available data (Fig. 5). As noted earlier, our maximum sample of nations with homicide data is 43, but in some cases the available time-series is insufficient for meaningful spline regression modeling. We invoked the following criteria for selecting cases for these analyses. To be included, a nation must have at least 40 years of data with no “big holes,”

Cross-National Homicide Trends in the Latter Decades of the Twentieth Century Venezuela 40


35 30 25 Homicide Rate Expected Rate

20 15 10 5 0 2004 2001 1998 1995 1992 1989 1986 1983 1980 1977 1974 1971 1968 1965 1962 1959 1956 1953 1950 Year

Fig. 4 Spline regression results for homicide rates in Venezuela

Japan


2.5

2

1.5 Homicide Rate Expected Rate 1

0.5

0 2004 2001 1998 1995 1992 1989 1986 1983 1980 1977 1974 1971 1968 1965 1962 1959 1956 1953 1950 Year

Fig. 5 Spline regression results for homicide rates in Japan

75

76

S.F. Messner et al.

i.e., periods with more than three consecutive years of missing data. A nation must also have homicide data at least as recent as the year 2000. The reason for first criterion is obvious: estimating trends and specific features of trends becomes problematic when there are big gaps in the time-series. In addition, we are interested in describing trends that extend approximately to contemporary times and that include the decade of the 1990s, a period of significant changes in homicide rates for many nations (Eisner 2008; LaFree and Drass 2002). See the Appendix for the listing of nations included in the spline regression modeling.

3.2 Results: Cross-National Patterns in Homicide Trends We begin the descriptive analysis by considering the prevalence of appreciable, sustained increases in homicide and the reversal of these trajectories. Applying the criteria outlined above to the results of the spline regression modeling, the data indicate that such increases in homicide were observed in 33 of the 35 nations (94%), i.e., in all nations but Japan (a nation with a decreasing trend) and the Dominican Republic (a nation with an essentially flat predicted trajectory). Moreover, the magnitudes of these increases are noteworthy. To illustrate, we have computed the percentage increase in homicide rates by comparing averaged values for 5-year periods centered on the low points and high points in the modeled trends. The median value is 98%, equivalent to an approximate doubling of the smoothed homicide rate. The spline regression results also indicate that a clear majority of the nations with an upward trajectory also exhibited a significant reversal. A distinct downturn point can be identified in 24 of the 33 nations with increases in homicide trends (72.7%). Accordingly, similar to the patterns that have been observed with data for major US cities, cyclical trends in homicide rates, rather than steadily increasing trends or trends that rise and stabilize, appear to be the norm rather than the exception at the national level when the latter decades of the twentieth century are examined (Joanes 2000; see also Zimring 2007). It is also interesting to consider the possibility that this period was one of increasing homicide rates overall even though homicide levels were cyclical for many nations. In other words, perhaps the “normal” level of homicide was recalibrated at a higher level. We addressed this issue by computing the ratio of the predicted homicide rates for the most recent period averaged over 5 years to the predicted homicide rates for the 5-year period centered on the low point of the cycle. For about a quarter of the nations classified as having had a homicide cycle, this ratio is approximately one or less than one, indicating either a return to the “takeoff” level or a net decrease. The median value, however, is 1.6, indicating that the homicide rates for a typical nation at the end of the period were about 60% higher than they were at the low point for the period. This is perhaps suggestive of upward recalibration in homicide, although it is impossible to know whether some of the “latecomers” with respect to rising homicide levels had yet to reach a downturn point at the last observation point, and whether those nations with negative


77

slopes near the end of the period have continued their downward trajectory in more recent years. In addition to considering the prevalence and magnitude of homicide cycles, it is also highly instructive to examine the timing of the key parameters of these cycles, i.e., the upturn and downturn dates. Previous researchers have called attention to the “synchronization” of rising violent crime rates, including homicide, across many nations (Fukuyama 1999, 31–32). Indeed, on the basis of his examination of European trends in homicide rates in the 1960–1990 period, Eisner (2008) concludes that the high degree of synchronization in these trends implies that some of the key causal factors affecting national homicide rates are likely to be those that extend beyond the boundaries of nation-states. We used the results of the spline regression to identify the estimated dates of any appreciable, sustained upward climb in homicide rates for the respective nations, along with the dates of any noteworthy reversal. The results for the upward climb are presented in Fig. 6, which lists nations by year and decade of the estimated upturn in homicide rates. The specific year of the break-point is indicated in parentheses, although we caution against placing much emphasis on the precise year. These dates are obviously estimates, and we thus concentrate on general patterns. Costa Rica (1960) Portugal (1960) Germany (1950)

Sweden (1961)

Iceland (1950)

Austria (1962)

Ireland (1950)

Bulgaria (1962)

Norway (1950)

Spain (1963)

Venezuela (1950)

United States (1963)

Israel (1954)

Canada (1964)

France (1970)

Netherlands (1956)

Hungary (1965)

Greece (1970)

New Zealand (1956)

Australia (1967)

United Kingdom (1970)

Mauritius (1957)

Italy (1967)

Colombia (1971)

Finland (1958)

Mexico (1969)

Poland (1973)

Trinidad and Tobago (1980)

Denmark (1959)

Switzerland (1969)

Singapore (1976)

Uruguay (1980)

Chile (1997)

l__________________l___________________l___________________l__________________l__________________l 1950s

1960s

1970s

1980s

1990s

Fig. 6 Nations by year and decade of upturn in homicide rates

The figure reveals a concentration of observations in the 1950s and 1960s, although a fair degree of dispersion can be detected as well (the standard deviation of year of upturn is just over 10 years). Of particular interest in the listing is the evidence for geographic/cultural patterns in the timing of the homicide upturn. The Nordic nations and other northern European nations appear to have been in the vanguard for increasing homicide rates. In contrast, the South American nations of Chile and Uruguay, along with the Caribbean nation of Trinidad and Tobago, are clearly “laggards” or “latecomers” in rising homicide levels. To explore more systematically the possibility of geographic/cultural patterns, we have organized the

78

S.F. Messner et al.

nations into the following five geo-cultural groupings suggested by the clustering in Fig. 6: Anglo-America/U.K./Oceania: Canada, United States, United Kingdom, Australia, New Zealand; Central/South America/Caribbean: Chile, Colombia, Costa Rica, Mexico, Trinidad and Tobago, Uruguay, Venezuela; Eastern/Southern Europe: Austria, Bulgaria, France, Greece, Hungary, Italy, Poland, Portugal, Spain, Switzerland; Nordic/Northern Europe: Denmark, Finland, Germany, Iceland, Ireland, Netherlands, Norway, Sweden; Other (Asia/Africa): Israel, Mauritius, Singapore. Using this classification, we conducted a one-way analysis of variance of year of upturn by geo-cultural region. The results, reported in Table 1, reveal that the Nordic/Northern Europe nations do indeed emerge as the vanguard, with a mean year of upturn of about 1954. The “other” category encompassing Asian/African nations follows, although this is essentially a residual category with a heterogeneous mix of nations. The Anglo-American/United Kingdom/Oceania nations are next (mean = 1964) followed by the eastern/southern European nations (mean = 1966). The Central/South American/Caribbean group has the latest average year of upturn (1972), approximately 20 years later than the average for the Nordic/Northern European nations. Although the sample is not a random one, it is useful to apply a test of statistical significance as a rough guide in interpreting whether the observed differences are likely to have been generated by chance. The F-ratio for the one-way ANOVA is 4.344, which is highly significant (p < 0.01). Table 1 Estimated timing of homicide rate upturn by geographic/cultural area Geographic/cultural area

Mean year of upturn

Standard deviation

Number of nations

Nordic/northern Europe Asia/Africa Anglo-America/UK/Oceania Eastern/southern Europe Central/South America/Caribbean

1954.2 1962.3 1964.0 1966.1 1972.4

4.7 11.9 5.2 4.3 15.2

8 3 5 10 7

These geographic/cultural patterns in the timing of the upturn are intriguing, and they are consistent with speculations that at least some of the determinants of rising homicide levels transcended national boundaries and policies in the sense of reflecting social and/or cultural conditions that are likely to be shared to a greater extent by similarly grouped nations. Moreover, the general pattern seems to be consistent with LaFree’s thesis about a prominent role for the institution of the family. Although changes in traditional family arrangements occurred in many nations in the latter half of the twentieth century, the Nordic countries were very much in the vanguard in this institutional domain, as well as in the timing of homicide increases (Amaro 2005; Forsberg 2005; Kiernan 2000; Trost and Levin 2005).


79

Austria (1990) Italy (1991) New Zealand (1991) United States (1991) Colombia (1992) Denmark (1992)

Chile

Poland (1992)

Costa Rica

Finland (1993)

Iceland

Portugal (1984)

Mexico (1993)

Israel

France (1985)

Switzerland (1993)

Mauritius

Singapore (1985)

Bulgaria (1996)

Spain

Norway (1986)

Greece (1997)

Trinidad and Tobago

United Kingdom (1977) Australia (1988)

Hungary (1997)

Venezuela

Germany (1978)

Netherlands (1997)

Canada (1975)

Sweden (1989)

Ireland (2002)

Uruguay

l__________________l___________________l___________________l__________________l__________________l

1970s

1980s

1990s

2000s

No reversal observed

Fig. 7 Nations by year and decade of reversal of increase in homicide rates

Following identical procedures to those in the analysis of year of upturn, we have organized nations by the year of reversal for those nations that exhibited an increase in homicide rates (see Fig. 7). The results reveal an even more striking concentration of observations in a single decade—the 1990s. This difference in concentration/dispersion for year of upturn and year of downturn is reflected in the standard deviations for the respective distributions. The standard deviation of year of upturn is 10.3, whereas the corresponding figure for year of reversal is 6.5. The only geo-cultural pattern that seems readily apparent in Fig. 7 is the representation of Central/South American/Caribbean nations among those classified as “no reversal.” This could of course reflect the fact that these nations are generally “latecomers” in the initiation of homicide increases; hence, any cyclical downturn might occur beyond the time frame under observation. The differing distributions of nations for year of upturn and year of downturn imply that the periodicity in homicide cycles varies across nations, given the greater spread in starting point compared with the concentration of the year of reversal. This is reflected in the univariate statistics for the length of time between year of upturn and year of downturn. The mean number of years for the rising phase of an estimated homicide cycle is 27, with a standard deviation of 10.4 years.

4 Dynamic Modeling of Homicide Rates 4.1 The “Family Legitimacy” Thesis Our second overarching objective is to assess the extent to which the observed changes in cross-national homicide rates can be accounted for by changes in the

80

S.F. Messner et al.

institution of the family. We draw specifically on Gary LaFree’s arguments about the role of declining legitimacy of the family in explaining the rise and then fall of US crime rates over approximately the same period as that under investigation here.11 LaFree hypothesizes that the traditional family was subject to challenges on two major fronts, one reflecting ideological developments and the other transformations in the economy. At the ideological level, the traditional two-parent family came under attack from feminists and other critics of the conventional order as the “bastion of male dominance” (1998, 84). Greater acceptance of sexual freedom and sexual experimentation also stimulated alternative intimate relationships outside of marriage. With respect to the economy, the emergence of an information society placed greater emphasis on knowledge and skills related to the service sector, thereby facilitating and accelerating trends toward increased female labor force participation and the “depopulation” of American homes (see also Cherlin 2004; Fukuyama 1999). The end result of these developments, LaFree hypothesizes, was that “the legitimacy of the traditional family declined enormously” (1998, 85). Moreover, newly emerging family and non-family arrangements were insufficiently developed to assume the functions previously associated with the traditional family. Accordingly, as familial institutional controls eroded, levels of crime climbed. LaFree goes on to speculate that the downturn in US crime rates beginning in the 1990s reflected the stabilization of family structures, as non-traditional family forms became increasingly institutionalized (e.g., blended families, dual-career families, single-parent families, and gay families) (1998, 151). In his words, the capacity for adaptation “is the genius of institutionalization” (151). He points specifically to the leveling off and then decline in divorce rates as suggestive evidence of such adaptation in the realm of the family. The objective of our dynamic modeling is to assess formally LaFree’s “family legitimacy” thesis about the link between divorce rates and homicide rates in a multivariate framework with data for a cross-national sample. We conduct time-series analyses to determine whether the divorce rate, as an indicator of family instability and lack of legitimacy, helps explain some of the changes in homicide rates described above.

4.2 Statistical Procedures Our model specification is determined by the temporal structure of the data. Unit root tests indicate that country-specific homicide rates and divorce rates are difference-stationary processes, so they are measured as annual change (Raffalovich

11 LaFree

focuses on declining legitimacy in two institutional realms in addition to the family: the polity and the economy. He maintains that distrust and collective protests undermined political legitimacy, while inequality and inflation undermined faith in the economy.


81

1994).12 Following Raffalovich (1999), we estimate a fixed-effects variable parameter time-series regression model: βi,k Xk,i,t + εi,t , Hi,t = βi,0 + βi,1 Hi,t−1 + βi,2 Di,t + k=3

where Hi,t is annual change in the homicide rate for country i in year t, Di,t is annual change in the divorce rate for country i, Xk,i,t are annual change in control variables, i,t is a stochastic error term. βi,0 , βi,1 , βi,2 , and the βi,k are parameters to be estimated. Random-effects models are ruled out by the possible correlation of random effects with excluded variables that vary across countries. Instead, we use fixed-effects models to control for unmeasured between-country sources of variability. For all models, the effect (βi,1 ) of the lagged dependent variable (Hi,t−1 ) is allowed to vary among countries. This models a unique dynamic process for each country. The effects of exogenous variables are net of these country-specific dynamics. In addition, we tested whether the effect (βi,2 ) of divorce (Di,t ) varies among geographic regions (it does not). Homicide rate distributions in cross-national samples are typically highly skewed, and as a result, researchers in cross-sectional analyses often convert these rates into natural logarithms. Skewness is not much of a problem with annual change scores. However, the meaning of a given change in the level of homicide might be different for nations with widely divergent levels of homicide. We accordingly estimate regression models with the dependent variable expressed as both annual changes in homicide rates in the original metric and annual changes in logged homicide rates. The change scores for logged homicide rates reflect proportional change. Models are estimated with the iterated generalized least-squares procedure in Eviews-6 (QMS 2007:495–506). We use robust standard errors throughout, correcting for both heteroscedasticity and cross-sectional interdependence. Examination of residual correlograms indicates no residual autocorrelation for any country in either level or log specifications of changes in homicide rates. The variable parameter time-series regression models can be estimated with an unbalanced design, and thus nations do not need to have complete data to be 12 Differencing

the variables (measuring them as annual change) removes all trend components. Therefore, we are analyzing short-run relationships between independent and dependent variables. This is necessary to mitigate the problem of spurious regression: regression with differencestationary variables will find “significant” relationships even when the variables are in fact independent. An exception is if the variables are “co-integrated” (i.e., share the same randomtrend component). We conducted co-integration tests. The results indicate that divorce rates and homicide rates may be co-integrated, suggesting that divorce rates increase homicide rates in the long-run as well as in the short-run (the latter results are reported below). The implications (if any) of this tentative finding are not immediately apparent to us, but the issue warrants further attention in the future. We thank Christoph Birkel for bringing this to our attention.

82

S.F. Messner et al.

included in the analyses. We exclude those nations that do not have observations for at least eight time points on the dependent variable (homicide rates) and the independent variable of primary interest (divorce rates) because such cases do not contribute much information to the modeling of change. The Appendix identifies the sample of nations included in the dynamic regression analyses.

4.3 Results The results of the regression modeling of changes in homicide rates in the original metric are reported in Table 2. Model I includes controls for changes in the level of economic resources (Ln GDP/capita) and three features of demographic structure: percentage of the population in the high risk 15–24 age group, sex ratio, and population size. Model II adds the annual change of infant mortality rates as a possible proxy for the growth or decline in levels of poverty. Model III substitutes changes in the relative size of the urban population for the infant mortality rate. Finally, Model IV is based on the full set of control variables. The finding of greatest theoretical interest in Table 2 is the significant positive effect for the divorce rate. Consistent with the family legitimacy thesis and Gartner’s results for a smaller sample of nations and a shorter time period (1990), increases in the divorce rate are associated with rising homicide rates, and thus decreases in the

Table 2 Regressions of changes in homicide rates on changes in divorce rates and control variables Independent variables

Model I (N = 1285)

Model II (N = 1282)

Model III (N = 1132)

Model IV (N = 1129)

Constant

Infant mortality

–0.004 (0.007) 0.113∗∗ (0.034) –0.665∗∗ (0.206) 0.028 (0.017) 0.019 (0.031) 1.517∗∗∗ (0.393) –

0.008 (0.009) 0.108∗∗ (0.034) –0.724∗∗∗ (0.211) 0.021 (0.016) 0.018 (0.030) 1.364∗∗∗ (0.367) –

Urban (%)

–

–0.009 (0.009) 0.113∗∗ (0.034) –0.744∗∗∗ (0.207) 0.025 (0.017) 0.018 (0.031) 1.462∗∗∗ (0.389) −0.009 (0.004) –

Log likelihood

−2198

−2163

0.007 (0.009) 0.109∗∗ (0.035) –0.695∗∗ (0.212) 0.023 (0.016) 0.018 (0.030) 1.397∗∗∗ (0.368) 0.005 (0.006) 0.024 (0.019) −1909

Divorce rate GDP/capita (Ln) Age 15–24 (%) Males/females Population (Ln)

∗∗ p

0.023 (0.019) −1941

< 0.01 ∗∗∗ p

Control of Violence: Historical and International Perspectives on Violence in Modern Societies

Violence in Europe: Historical and Contemporary Perspectives

Violence in Europe: historical and contemporary perspectives

Children's Perspectives on Domestic Violence

Multidisciplinary Perspectives on Family Violence

Cultures of Violence: Interpersonal Violence in Historical Perspective

On violence

On violence

On Violence

Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History

International Violence Against Women

International Handbook of Violence Research

Islam and Violence in the Modern Era

Islam and Violence in the Modern Era

Blood and Violence in Early Modern France

Islam and Violence in the Modern Era

Islam and Violence in the Modern Era

Violence in the Home: Multidisciplinary Perspectives (Psychology)

Violence in the Home: Multidisciplinary Perspectives (Psychology)

Religious Violence Between Christians and Jews: Medieval Roots, Modern Perspectives

Cultural Shaping of Violence: International Perspectives (Comparative Cultural Studies)

Fighting for Girls: New Perspectives on Gender and Violence

Violence and Non-Violence in Africa (Routledge Studies in International Relations and Global Politic)

Aggression and Violence in Adolescence

Lionel Robbins (Historical Perspectives on Modern Economics)

Random Violence

Intimate Violence

Punishing Violence

Violence and Phenomenology

Violence, Culture And Censure

Control of Violence: Historical and International Perspectives on Violence in Modern Societies

Violence in Europe: Historical and Contemporary Perspectives

Violence in Europe: historical and contemporary perspectives

Children's Perspectives on Domestic Violence

Multidisciplinary Perspectives on Family Violence

Cultures of Violence: Interpersonal Violence in Historical Perspective

On violence

On violence

On Violence

Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History

International Violence Against Women

International Handbook of Violence Research

Islam and Violence in the Modern Era

Islam and Violence in the Modern Era

Blood and Violence in Early Modern France

Islam and Violence in the Modern Era

Islam and Violence in the Modern Era

Violence in the Home: Multidisciplinary Perspectives (Psychology)

Violence in the Home: Multidisciplinary Perspectives (Psychology)

Religious Violence Between Christians and Jews: Medieval Roots, Modern Perspectives

Cultural Shaping of Violence: International Perspectives (Comparative Cultural Studies)

Fighting for Girls: New Perspectives on Gender and Violence

Violence and Non-Violence in Africa (Routledge Studies in International Relations and Global Politic)

Aggression and Violence in Adolescence

Lionel Robbins (Historical Perspectives on Modern Economics)

Random Violence

Intimate Violence

Punishing Violence

Violence and Phenomenology

Violence, Culture And Censure

Recommend Documents