International Political Economy Series

International Political Economy Series General Editor: Timothy M. Shaw, Professor of Commonwealth Governance and Development, and Director of the Institute of Commonwealth Studies, School of Advanced Study, University of London Titles include: Pradeep Agrawal, Subir V. Gokarn, Veena Mishra, Kirit S. Parikh and Kunal Sen POLICY REGIMES AND INDUSTRIAL COMPETITIVENESS A Comparative Study of East Asia and India Roderic Alley THE UNITED NATIONS IN SOUTHEAST ASIA AND THE SOUTH PACIFIC Dick Beason and Jason James THE POLITICAL ECONOMY OF JAPANESE FINANCIAL MARKETS Myths versus Reality Mark Beeson COMPETING CAPITALISMS Australia, Japan and Economic Competition in Asia-Pacific Deborah Bräutigam CHINESE AID AND AFRICAN DEVELOPMENT Exporting Green Revolution Kenneth D. Bush THE INTRA-GROUP DIMENSIONS OF ETHNIC CONFLICT IN SRI LANKA Learning to Read between the Lines Steve Chan, Cal Clark and Danny Lam (editors) BEYOND THE DEVELOPMENTAL STATE East Asia’s Political Economies Reconsidered Abdul Rahman Embong STATE-LED MODERNIZATION AND THE MIDDLE CLASS IN MALAYSIA Dong-Sook Shin Gills RURAL WOMEN AND TRIPLE EXPLOITATION IN KOREAN DEVELOPMENT Jeffrey Henderson (editor) r INDUSTRIAL TRANSFORMATION IN EASTERN EUROPE IN THE LIGHT OF THE EAST ASIAN EXPERIENCE Takashi Inoguchi GLOBAL CHANGE A Japanese Perspective Dominic Kelly JAPAN AND THE RECONSTRUCTION OF EAST ASIA

L. H. M. Ling POSTCOLONIAL INTERNATIONAL RELATIONS Conquest and Desire between Asia and the West Pierre P. Lizée PEACE, POWER AND RESISTANCE IN CAMBODIA Global Governance and the Failure of International Conflict Resolution S. Javed Maswood JAPAN IN CRISIS Ananya Mukherjee Reed PERSPECTIVES ON THE INDIAN CORPORATE ECONOMY Exploring the Paradox of Profits CORPORATE CAPITALISM IN CONTEMPORARY SOUTH ASIA (editor) r Conventional Wisdoms and South Asian Realities Cecilia Ng POSITIONING WOMEN IN MALAYSIA Class and Gender in an Industrializing State Ian Scott (editor) r INSTITUTIONAL CHANGE AND THE POLITICAL TRANSITION IN HONG KONG Mark Turner (editor) r CENTRAL–LOCAL RELATIONS IN ASIA-PACIFIC Convergence and Divergence? Fei-Ling Wang INSTITUTIONS AND INSTITUTIONAL CHANGE IN CHINA Premodernity and Modernization

International Political Economy Series Series Standing Order ISBN 978-0-333-71708-0 hardcover Series Standing Order ISBN 978-0-333-71110-1 paperback (outside North America only) You can receive future titles in this series as they are published by placing a standing order. Please contact your bookseller or, in case of difficulty, write to us at the address below with your name and address, the title of the series and the ISBN quoted above. Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21 6XS, England

The Intra-Group Dimensions of Ethnic Conflict in Sri Lanka Learning to Read between the Lines Kenneth D. Bush Assistant Professor of Conflict Studies St Paul University Ottawa Canada

© Kenneth D. Bush 2003 Softcover reprint of the hardcover 1st edition 2003 978-0-333-71456-0 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The author has asserted his right to be identified as the author of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2003 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N. Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 978-1-349-40340-0 DOI 10.1057/9780230597822

ISBN 978-0-230-59782-2 (eBook)

This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Bush, Kenneth D., 1961– The intra-group dimensions of ethnic conflict in Sri Lanka : learning to read between the lines / Kenneth D. Bush. p. cm. – (International political economy series) 1. Sri Lanka–Politics and government–1978- 2. Tamil (Indic people)–Sri Lanka–Politics and government. 3. Sri Lanka–Ethnic relations. I. Title. II. Series. DS489.84.B87 2003 954.9303’2–dc21 2003050454 10 12

9 11

8 10

7 09

6 08

5 07

4 06

3 05

2 04

1 03

For Brennan and For Thiruvan, Kabir, Robinston, Nawshaat, Dulekshan, Shiphra, Pravin, Janavee, Ajanthan, Karoon, Shantini, Iphran, Vijayakanth, Shivakumari, Zhawulhuk, Pamila, and Munaz and all the other children who have been touched by militarized violence in Sri Lanka That they might come to know the full blue of this chameleonizing Batticaloa sky and become innocent again quickened into verb pure verb

This page intentionally left blank

Contents List of Tables

ix

List of Figures

x

List of Abbreviations

xi

Acknowledgements

xiii

Preface

xvii

Map of Sri Lanka Part I

xx

Introduction

1 3 3 6 7

1 Beyond Billiard Ball Analysis Identity and security Rethinking identity Linking security and identity 2 Learning to Read between the Lines Patterns of violence Methodology

10 11 12

3 An Overview of Sri Lankan Politics Introduction to Sri Lanka Axes of identity On powerful neighbors

29 29 45 70

Part II

73

Critical Junctures

4 Critical Juncture I: 1948 Independence and the Disenfranchisement of the Plantation Tamils The context The episode Model I Model II

75

5 Critical Juncture II: 1956 Election and the Premiership of S.W.R.D. Bandaranaike The context The episode Model I Model II

84

vii

75 76 77 81

84 88 90 90

viii Contents

6

Critical Juncture III: 1971 JVP Insurrection and 1987 JVP Resurgence 1971 JVP insurrection 1987 JVP resurgence

100 103

7

Critical Juncture IV: 1983 Riots The context The episode Model I Model II

118 118 123 125 127

8

Critical Juncture V: 1987 Indo-Sri Lanka Agreement The context The immediate pre-agreement political–military landscape The episode The intra-group dimension of an inter-group alliance Model I Model II

135

9

Critical Juncture VI: February 2002 Ceasefire Agreement (CFA) The context The episode Model I Model II Conclusion

Part III 10

Conclusion

Fitting the Pieces Together Summary of critical junctures Accounting for differences in levels and patterns of violent conflict The theoretical contribution of this study Conclusion

99

136 141 144 146 147 152 156 157 161 162 163 175 177 179 179 183 194 201

Notes

204

Bibliography

245

Index

261

List of Tables 3.1 5.1

Demographic composition of the Northern and Eastern Provinces (1981) Results of the 1956 general elections

ix

40 88

List of Figures 2.1

Mediating factors of ethnic conflict in intra-group and inter-group arenas 4.1 Elite-non-elite conflict and the disenfranchisement of plantation tamils 4.2 Ethnicization of elite conflict and the disenfranchisement of plantation tamils 8.1 Provocation–retaliation–consolidation dynamic 8.2 Sub-group position on the Indo-Sri Lanka agreement: inter-group alliances and intra-group conflicts 10.1 Summary of findings at the inter- and intra-group levels in Sri Lanka

x

19 82 83 140 146 180

List of Abbreviations AWB B–C Pact CFA CJC CP CPA CRD CTC CWC DDC DJV DUNF EBP ENDLF EPDP EPRLF EROS FEER FP GSL IPKF PIRA IPLO IRB JSS JVP LSSP LTTE MEP MIRJE MOU NGO NLSSP NMAT NSSP

Afikaner Weerstandsbeweging Bandaranaike–Chelvanayakam Pact Cease Fire Agreement Criminal Justice Commission Communist Party Centre for Policy Alternatives Committee for Rational Development Ceylon Tamil Congress Ceylon Workers’ Congress District Development Council Deshapremi Janatha Viyaparaya (Patriotic People’s Organization) Democratic United National Front Eksath Bhikkhu Peramuna (United Front of Monks) Eelam National Democratic Liberation Front Eelam People’s Democratic Party Eelam People’s Revolutionary Liberation Front Eelam Research Organization of Students Far Eastern Economic Review Federal Party Government of Sri Lanka Indian Peace Keeping Force Provisional Irish Republican Army Irish People’s Liberation Organization Immigration and Refugee Board, Government of Canada Jathika Sevaka Sangamaya Jathika Vimukti Peramuna (People’s Liberation Front) Lanka Sama Samaja Party Liberation Tigers of Tamil Eelam Mahajana Eksath Peramuna Movement for Inter-Racial Justice and Harmony Memorandum of Understanding Non-Government Organization Nova Lanka Sama Samaja Party – Trotskyist National Movement Against Terrorism Nava Sama Samaja Party xi

xii List of Abbreviations

PA PLOTE PTA RPG SLA SLFP SLMC SLMM SLMP STF TC TELO TNA TULF UCPF UNP USCR UTHR

People’s Alliance People’s Liberation Organization of Tamil Eelam Prevention of Terrorism Act Rocket-Propelled Grenade Sri Lankan Army Sri Lanka Freedom Party Sri Lanka Muslim Congress Sri Lanka Monitoring Mission Sri Lanka Muslim Party Special Task Force Tamil Congress Tamil Eelam Liberation Organization Tamil National Army Tamil United Liberation Front Up-Country People’s Front United National Party United States Committee for Refugees University Teachers for Human Rights

Acknowledgements The story of the writing of a book is always more interesting that the book that finally gets written. My first extended visit to Sri Lanka started in October 1980 – three years before the riots which would set the island ablaze. There are literally hundreds of people who helped me – and who continue to help me – along the path of understanding. The best part about completing this book is that I am finally able to thank publicly those who encouraged and supported me at various points in the process. I don’t know where I would be without them, but I do know that this book would never have been written. Throughout my academic life, my family has provided inspiration and support. From her kitchen table, my mother, Carrol Robb, has fed far more than 5,000 people with far less than five loaves of bread and two fish. From her and my step-dad, Jim, I have learned that family, in its truest form, excludes no one. It has been an indescribable joy to watch brothers and sisters become mothers and fathers. Every one of them contributed more than they know to enabling me to complete this book: Paul Bush and his wife, Sue Martin; Pamela Bush and her partner, Marc Pelletier; Kevin Bush and his wife, Barbara Bush; Christina Robinson; and Todd Baseden. My nieces and nephews (Katie, Patrick, and Ally Bush, Benjamin Prentice, Andre and Luc Pelletier, Jordan and Natasha Bush, and Taylor Robinson) always put things back into perspective when inhumanity seemed to make despair the only possible human response. On the Goodman side, Eric provided much appreciated technical expertise and good paternal advice. Connie Huggett maintained one of my many homes while I was recovering academic. Likewise, Sue and Jim Chalmers provided continuous support and encouragement. Tracey Goodman has sustained me from India and Italy to Ithaca and Geneva. From her, and from our son Brennan, I’ve learned more than from anyone within the walls of a university. Over the years, many people have nurtured a supportive environment within which to develop ideas and beliefs, including Milton Esman, Nalini Devdas, Fen Hampson, Bruce Matthews, Carlton Hughes, Taiaike Alfred, Tim Shaw, Erin Baines, and David Power. In Washington, the engaging conversation and unequalled hospitality of Maurice Rabot and Mary Doyle easily eclipsed American Political xiii

xiv Acknowledgements

Science Association Meetings – not surprising given the Rabot family tradition I first encountered over 20 years ago. Within this group of people, I am most indebted to John Rabot, who is teaching me as much now as he did when he was alive. In the writing of this book in its thesis form, a place of honour goes to Norman Uphoff. His razor-sharp pen pushed me to refine ideas and words that were embarrassingly blunt. His ‘adversarial’ reading of drafts was unstintingly collegial and constructive. In his dual role of mentor and tormentor, Norm provided the intellectual prodding that pushed me much further than I would otherwise have gone. I wish that every graduate student could find such an advisor and role model. Throughout the years, I have enjoyed the hospitality and assistance of a wide range of people and organizations in Sri Lanka: Anthea Mulakala (DFID); Peter Kuperus (Royal Netherlands Embassy); Ruth Archibald (Canadian High Commission); Colin Glennie (UNICEF); Mariana Todorova (World Bank); Stina Karltun (SIDA, Sri Lanka); Reinhardt Bolz (GTZ); Sunil Bastian (International Centre for Ethnic Studies), Father Harry Miller (Batticaloa Peace Committee); Father Paul Casparsz (Satyodaya); F.Paul Satkunanayagam (Butterfly Peace Garden); Steve Hollingworth (CARE International); Phil Esmonde (OXFAM UK); Mark McKenna (Asia Foundation); Norbert Ropers (Berghof Foundation for Conflict Studies, Sri Lanka); Pauline Taylor-McKeown (Save the Children UK); Georg Frerks (Clingendael Institute); Jeevan Thiagarajah (Consortium of Humanitarian Agencies); William Knox and colleagues (Peace and Community Action); Dr P. Saravanamutu (Centre for Policy Alternatives); Dr W. Abeydeera (Sri Lanka-Canada Development Fund/ Centre for Development Faciliation); and Don Brownell (Canadian Federation of Municipalities). A special thanks goes to the Renuka Hotel in Colombo 3. And most of all, I thank Paul Hogan and the Butterfly Peace Garden of Batticaloa who have taught me that the more I learn, the less I know – and as importantly, that it is the space between the words, not the words themselves, that is important. This book benefited from research commissioned by a host of organizations, including: OECD DAC Working Group on Conflict, Peace, and Development Cooperation; the World Bank; UNICEF; the Canadian Department of Foreign Affairs and International Trade: the Canadian International Development Agency; the Swedish International Development Agency; the British Department for International Development; and South Asia Partnership. While writing this book, I tried to juggle a daunting number of activities and commitments. In the early stage of this process, Sid Tarrow at

Acknowledgements xv

Cornell University cautioned me (in that avuncular way of his): ‘Bush, you can’t dance at all the weddings and cry at all the funerals.’ He was right. But he was right for the wrong reason. In the course of my research, I realized that there are far more funerals than weddings in the Sri Lankas, Bosnias, and Rwandas of this world. Every day spent in the luxury of writing is a day that sees the massacred buried and the detained tortured. Andre Dumas sums it up most urgently: ‘There comes a point when words are useless commentary. Immodest babbling beside the hard reality of suffering.’ These observations by Tarrow and by Dumas present us with the challenge: the need both to understand and to change those structures and processes which degrade our common humanity. I reiterate that I, alone, am responsible for the statements and assessments that follow.

This page intentionally left blank

Preface The objective of this book is two-fold. First, it seeks to refine and advance our understanding of the dynamics underpinning ‘ethnicized’ conflict (both in its violent and non-violent manifestations) through the development of an analytical lens which enables us to understand how the ebb and flow of relations within ethnic groups affects relations between group, for good or for ill. Secondly, this book seeks to deepen our understanding of the nature and evolution of conflict in Sri Lanka. The use of a case study methodology is intended to serve as an illustration of the way such a multi-level lens (linking intra-group and intergroup structures and processes) might be applied to other cases of ethnic conflict. Two sets of inter-locked puzzles compelled the writing of this book. One influenced the development of the methodology, the other influenced the selection of Sri Lanka as a case study. The first puzzle was the observation that a pattern of gradual improvement in intergroup relations in conflict-prone regions around the world is often shattered by unexpected eruptions of violence that are inexplicable if our analysis is limited to inter-group inter-actions. The second was the realization that even in the midst of protracted militarized conflict in Sri Lanka, there exists a systematic network of communication and cooperation across the principal battle lines. This is evident in the formal and informal negotiations within and between governmentcontrolled and rebel-controlled areas. Consequently, we have seen such incongruous events as: island-wide National Immunization Days (Bush 2001; Hull 1999); island-wide O-level and A-level exams (requiring exam papers to cross the battle lines and be written in examination centers in rebel-held areas);1 the return of electricity in the late 1990s to rebel-controlled areas from government-controlled areas in the East;2 and the unique situation in which Government-appointed officials live and function in rebel-controlled territory to provide humanitarian services to war-affected populations and to maintain essential services and administrative structure – a situation described by one colleague as Casablanca-esque, because you are never quite sure whose side anyone is on. Perhaps the most bizarre episode within this web of inter–intra-group relations was the 1991 provision of weapons and material by the President of Sri Lanka to anti-government rebels which xvii

xviii Preface

ultimately lead to impeachment proceedings that were extinguished only by adroit constitutional manoeuvring and strong-arm politics, albeit at a considerable political cost.3 The search for understanding and explanation requires us to dig through the layers of rubble, human devastation, competing explanations, accusations and counter-accusations. Eventually, we find ourselves in the thick and murky world of intra-group politics where ruptures – or rapprochement – at the inter-group level of relations are often compelled or constrained by internal or intra-group structures and processes. And even when international forces affect the dynamics of peace and conflict between identity groups, we see that they are frequently filtered through the intra-group arena. When we move from the clean world of theory to the dirty world of identity politics, the implication of this study is this: any attempt to ‘deal with’ violent ethnicized conflict which does not consider the internal politics of groups will, at best, have temporary or minimal impact. At worst, it will have a negative impact. This is especially salient in the shadow of the final, blood-soaked, decade of the last century, when the world watched the slaughter of 200,000–250,000 people (largely civilians) in the Balkans; and as horrifically, the butchering of 600,000–800,000 children, women and men in Rwanda over a period of 16 weeks – illustrating that the panga could be even more efficient than the gas chambers of Nazi Germany. It may be appropriate at this stage to make clear what this book is not. It is not a political history of post-colonial Sri Lanka. While such a volume would be welcome and necessary, it is beyond the scope of the current work. This study attempts to capture the essential dynamics of intersecting conflicts rather than to ‘tell the whole story.’ Neither is this book an anthropological study of ethnicized violence in Sri Lanka – although, it certainly draws on the excellent work of those scholars who have examined the deep cultural dimensions of conflict over time such as Bruce Matthews, Jonathan Spencer, Patricia Lawrence, Steven Kemper, and K.N.O. Dharmadasa. In this book, anthropological details become salient to the discussion when they are manipulated instrumentally to serve particularistic interests, that is, when they become political resources in the hands of ethnic mobilizers. This book seeks to understand an under-explored dimension of ethnicized violence. It develops a methodological argument which has practical implications. It has a particular interest in identifying and analysing ‘critical junctures’ – those turning points at which relations within and between groups have the potential to be altered fundamen-

Preface xix

tally, and to thereby affect the dynamics of conflict or peace, e.g. events which polarize, fragment, consolidate, or reconfigure intergroup or intra-group relations. Far from being a comprehensive history, this book’s focus on critical junctures is an effort to understand those moments which indelibly affected the trajectory of political events in Sri Lanka. If this sheds light on the dynamics of peace and conflict in Sri Lanka, then we can take the next logical step and test this approach in other cases. From a practical perspective, it compels us to develop and employ conflict management strategies which explicitly link interventions at an inter-group arena with those at the intragroup level.

xx

Source: Courtesy of the General Libraries, The University of Texas at Austin.

Part I Introduction

1 Beyond Billiard Ball Analysis

The purpose of the current chapter is to sketch out briefly some of the thinking and issues surrounding the linkages between identity, security and violent conflict as reflected in the dominant literature on ethnic or nationalist conflict. Chapter 2 builds on this overview by setting it in the Sri Lankan context.

Identity and Security1 Bosnia. Rwanda. Sri Lanka. Northern Ireland. At he dawn of the 21st century, places have become metaphors for violent ‘communal’ or ‘identity-based’ conflict. As Donald Horowitz has observed, such conflicts have ‘fought and bled and burned [their] way into public and scholarly consciousness.’ Yet, at the same time, we find ourselves illequipped intellectually and institutionally to face the complex challenges posed by identity-based violence in a new, and very different, post-Cold War world. While the collapse of the Communist Block and galloping globalization have fundamentally altered the structure of geopolitics, we find ourselves in a situation in which the conceptual frameworks, theoretical tools, institutional response mechanisms, and menu of policy prescriptions are all indelibly infused with a Cold War political logic. Indeed, in the wake of the attacks of 11 September 2001 – or more specifically, the Anglo-American military response of 7 October 2001 – very nuanced and complex conflicts around the world have been strained through a simplistic lens to be reframed as one-dimensional confrontations between ‘terrorists’ and antiterrorists, not unlike the dichotomization that was sustained in the earlier era of bi-polarity. 3

4 Introduction

The search for effective responses to identity-based conflicts requires more than the current mainstream (read ‘realist’) efforts to fit them into the existing approaches to security. Such conflicts push us to rethink not only the way we deal with security challenges, but the way we define and understand security issues. Correspondingly, identitybased conflicts push us to rethink our understanding of collective identity – its formation, mobilization, politicization, and most importantly, its connection to violent conflict. Through the process of ‘rethinking’ violent conflict and identity, this book argues that there is a need to both build on, and go beyond, well-travelled realist paths. While a sophisticated version of realism offers some insights into the dynamics of identity-based conflict, there is a need to go further in our thinking about, and responses to, such conflict because even in this version, identity is still cast as being fixed, coherent, and selfcontained. The development of an understanding of group identity (self-defined and other-defined) requires a thorough interrogation of the specific linkages between the individual and the larger group with which he or she identifies. This book argues, and hopefully demonstrates, that a relational, historical, and dynamic understanding of identity is crucial not only for coming to terms with the connections between security and (ethnic) identity, but also for constructing effective strategies for the management or resolution of conflict.2 When this argument is pushed to its limit, it holds that complete security – that is, the complete absence of threat or danger – even if it was possible, would destroy a necessary precondition for politicized group identity. If the absence of threat challenges a group’s sense of identity, then, as we see in current communal conflicts, the construction of threat may also consolidate identity. This is the crucial link between identity-based conflict, security, and insecurity. Despite the years that have passed since the fall of the Berlin Wall, we still have not grasped the full implications of this linkage. We still bear the Cold War like an incubus. There have been some calls for an integrated multidimensional approach to security,3 yet so far, the response by scholars and practitioners has been lacking. The politics of funding, disciplinary rigidity, vested interests, and parochial academic feudalism, all contribute to explaining why a concerted and integrated approach has not yet been developed.4 The dominant tendency in current approaches to the study of ethnicized conflict is to fit (or force) this phenomenon into the pre-existing (albeit somewhat modified) realist epistemological frameworks. There is

Beyond Billiard Ball Analysis 5

an overwhelming focus on inter-communal relations or what can be called ‘billiard ball’ models of communal conflict. Communal groups are represented as the functional equivalent of states: unitary, powerseeking, (though interestingly, not necessarily ‘rational’) actors in a Hobbesian world. In other words, communal groups are viewed as being analogous to the state epistemologically and ontologically. Like states, such groups are seen to constitute stable and unified entities, and to act as coherent and separate totalities. Thus, the billiard-ball model, which is based on relations among separate states as unified entities, now includes inter-ethnic group relations, each of which constitutes a unified and separate totality, that is, self-contained and self-propelling entities. In effect, realism simply adds the notion of ethnic identity to its basic assumption that the position of a collectivity, whether it be a state or a group, in an anarchical system is the primary causal variable in the area of security. The present state of theory and practice in Conflict Studies underscores the need to link the logics of security and identity as a precondition to the effective management of identity-based violent conflict. As discussed further below, this is not an easy task, given the fact that the extension of the analytical boundaries of security beyond the historically privileged areas of state and military requires both a substantial critique of the realist articulation of international relations theory and a more critical understanding of identity.5 It is important to note that although a critique of realism enables one to break out of the narrow logic of security fixated on an inter-state level of conflict – and thereby makes it possible to pair the concept of security with a notion of identity – such a critique still leaves unanswered the question of what is meant by identity. That is to say, a simple integration of the notion of identity into the logic of security remains partial or confused unless and until the question of identity is posed as a proper object of theoretical and historical inquiry – as is attempted in this book through the development and application of a Two-Level Critical Junctures Approach to ethnic conflict analysis. At a broader theoretical level of analysis, realist accounts of ethnic conflict share two particular traits. First, they view the rise of ethnic conflict as a by-product of anarchy. While this illustrates that the realist analytical framework is capable of finding a place for identity within their conception of security, it does not distract the singlepointed focus of realists on the Third Image – the anarchic international system. The second trait follows from the first: although political events may push realists to address the phenomenon of politicized

6 Introduction

identity (ethnic identity), their lack of a theory of nationalism allows them only to cast ethnicity in broad causal terms as a spur to violent ethnic conflict. The linkages between cause and effect are left undeveloped, and consequently the adjective ‘ethnic’ is often brandished crudely as an explanation, rather than a description, of a conflict. One thing has become clear from even a cursory examination of cases of ethncized violence : ‘identity’ does not mobilize individuals (as primordialist proponents would argue), rather individuals mobilize identity – from among a menu of possible identities within limits set by the broader opportunity structure and the specific political, economic and social context within which identity is articulated. Typically, such mobilizers are part of a larger organization, or they may be seeking to create or reinforce a larger organization.

Rethinking identity Just as identity-based conflicts force us to rethink our understanding and approach to security, so do they force us to rethink our understanding of collective identity – its formation, mobilization, politicization, and, most importantly, its connection to violent conflict. While the conflicts noted above – Bosnia, Sri Lanka, Rwanda, and Northern Ireland – are all instances of communal violence, the salient axis of confrontation differs between cases. They have been cast variously as ‘ethnic,’ ‘tribal,’ and ‘sectarian’ conflicts. While it is important and useful to recognize the variety of communal conflicts, this book takes as a starting point the understanding that the different axes of conflict evident across the cases noted above, are also evident and significant within cases of communal conflicts. Communal groups may be internally divided into sub-groups along a range of differentiating axes of identity, such as gender, regional affinity, religion, sect, dialect, caste, political affiliation or ideology, and socio-economic status. The pattern of communal conflict is not simply determined by the interaction of communal groups like the action–reaction dynamic of billiard balls, it is also affected by the constellation of shifting factors within each communal group. As discussed further in the next chapter, there is no essentialist quality to ‘ethnic’ identity. It is, by definition, contextual, in so far as membership in an ethnic community is premised upon a shared belief in a common ancestry, which becomes a discursive mechanism for binding people together in a (ethnic) community, allowing them to overlook their internal differences and conflicts, while perceiving


themselves as ontologically different from other identity groups. When ethnic identity comes to fore, all other dimensions of identity do not disappear. Rather they are subordinated to a cultural unity defined in terms of ethnicity. In this way, ethnicity functions as a unifying entity. ‘A certain degree of cultural homogenization, or at least the existence of an important cultural text with which all members of a group identify themselves, is indispensable for the formation of an ethnic group.’ Thus, we can define an ethnic identity as being constructed both historically and symbolically around a particular cultural tradition. It is also clear that ethnic identity has two inter-related, constitutive, components. First, the idea that one belongs to a group which is distinct, separate, and unique from all other groups. And second, the presence of an outsider group which acts as a whet stone that sharpens the insider group’s sense of its separate self. Ethnic identities are always relational. They are constructed not in isolation but through interaction with other groups. For this reason, the creation of cultural boundaries, as Sri Lanka illustrates, has been central to the construction and maintenance of politicized ethnic identities – a phenomenon equally evident in the Balkans, Northern Ireland, and Rwanda. Just as every identity conflict is not an ethnic conflict, neither is every contemporary conflict an identity conflict. This caveat is especially noteworthy in a post-Cold War era in which some analysts are tempted to see identity conflicts around every corner – underpinning, for example, everything from vulgar and unsophisticated assertions of ‘clashes of civilizations’ to patently racist and xenophobic policies of the United States and Britain’s ‘War on Terror.’ While the mobilizational utility of identity has made it a conspicuous feature in contemporary politics, it is not necessarily the principal source of conflict. Non-identity based conflicts may include those precipitated by environmental scarcity (Homer-Dixon 1991), hegemonic competition, ‘development wars’ (Miller 1992), resource or territorial disputes, and the straightforward greed and self-aggrandizement that characterize kleptocracies around the world (Berdal and Malone 2000; Musha and Fayemi 2000). Any examination of linkages between identity and conflict must begin with the obvious recognition that the sources and manifestations of conflict are both complex and variable.

Linking security and identity An assessment of current mainstream approaches to the analysis of identity-based conflicts must inevitably consider the realist character-

8 Introduction

ization of the rise of nationalism and the collapse of the Soviet Union and Yugoslavia. The starting-point for such analyses is that ‘the insecurity and conflict [was] caused by the situation of anarchy, in which states and other social groups must look to their own devices for protection against possible depredations by others.’6 Accordingly, Barry Posen accounts for aggressive behavior by ethnic groups with the tautological assertion that the collapse of central authority in such places as the Soviet Union and Yugoslavia precipitates an authority crisis. Under these conditions, realists assert that ‘ethnicity comes to the fore as a logical basis for forming self-protection groups in the face of a general security dilemma.’7 However, given the heterogeneity of groups, the multiplicity of the axes of identity that might be politicized, and the variability of responses in different cases at different historical junctures, such realist assertions hold up neither as an axiom of social behavior, nor as a useful working assumption for the examination of the linkages between identity and security. In fact, by focusing solely on the violent dimensions of inter-ethnic relations, one misses the opportunity to examine when and why ethnicity does not become a chauvinistic and hegemonic political logic infusing all forms of interaction. Even in the midst of classic cases of protracted ‘ethnic’ conflict, there are many instances of individual and collective behaviour which ‘violate’ realist assertions. For example, based on years of studying conflict and discord in the Middle East, Edward Azar concludes: ‘Conflictual and cooperative events flow together even in the most severe of intense conflicts. Cooperative events are sometimes far more numerous than conflictual ones even in the midst of intense social conflict situations. However, conflictual events are clearly more absorbing and have more impact on determining the consequent actions of groups and nations [and one might add, on determining the outsider’s impression of the conflict]’.8 While ethnicity may come ‘to the fore as a logical basis for forming self-protection groups in the face of a general security dilemma’ as realist propose, this is by no means the only, or necessarily the most important, social dynamic at work. The Sri Lanka case (and others noted briefly throughout this book) point to the importance of selfinterest or particularistic interests which may be affected only marginally by inter-ethnic barriers. In other words, just as there are non-ethnic bases for individual self-definition, there are non-ethnic bases for social interaction. To focus narrowly of ethnic categories may serve to misrepresent the dynamics of conflict and obscure potential avenues for nurturing peace. More importantly, such a narrow analytic


lens may serve to exacerbate violent conflict by legitimizing politicized borders of identity. The particular self-definition of a group (within the constellation of surrounding groups) confers legitimacy, credibility, and status upon those who have done the defining, implicitly and automatically. It is universal that all ethnic mobilizers seek to legitimate their particular representation of the group as continuous, unbroken, and primordial. But the acceptance of these claims of primordialism at face value disregards the inevitability of change in culture and identity over time and in different settings. More importantly, the acceptance of a fundamentally political representation of the past endorses the political objectives its sponsors seek to achieve in the present. Discursive constructs have tended to be accepted as ‘empirical facts’ that need no problematization. Thus, in effect, the a priori acceptance of the ‘legitimacy’ of ethnic boundaries in violent conflicts by realist analysts, serves to entrench the ethnic chauvinists who politicize and sustain such divides. Part and parcel of this process is the marginalization of moderates, and non-ethnically organized institutions. Such analytical lenses filter the policy process, and ultimately subsidize ethnic violence. This book is structured as follows. The next chapter introduces the theoretical approach adopted in the study. It is followed by a chapter introducing the actors and politics in Sri Lanka. The following chapters apply the theoretical discussion to six critical junctures. The final chapter considers the implications of this study for (1) addressing conflict in Sri Lanka, and (2) the theory and practice of conflict resolution and management. It wrestles with the question: where do we go from here to develop more satisfactory analyses of ethnic conflict and more promising means of dealing with a phenomenon which appears at present to be escalating worldwide, mocking the optimistic predictions of an earlier generation of social scientists who anticipated that modernization would create a harmonious ‘melting pot.’ However, as the expectations for a ‘melting pot’ give way to the reality of a ‘boiling pot,’ the onus falls back onto academics, inside and outside universities, to come to terms theoretically and practically with this increasingly violent phenomenon.

2 Learning to Read between the Lines

This book uses the term ‘ethnicity’ broadly to refer to the ‘subjective, symbolic, or emblematic use by a group of people of any aspect of culture in order to create internal cohesion and differentiate themselves from other groups’ (Brass 1991: 19). This understanding of ethnicity recognizes that the internal and external functions of group identity are cohesion and differentiation. Central to this sense of identity is a shared belief in common descent, birth or kinship which may be (but usually is not) based on biological fact. The attempt to understand the role of sub-groups in ethnic conflict through a review of the literature proves unsatisfying. It reveals an overwhelming focus on inter-group relations, treating ethnic groups as homogenous entities, employing what is described in the introduction as a ‘billiard ball’ model of ethnic conflict – solid masses moving at different speeds and on different trajectories, sometimes colliding. Conspicuously absent is a recognition of the need to ‘read between the lines’ – or more accurately ‘within the lines’ – of ethnic groups in conflict, that is, to consider the internal structure and dynamics of groups and to assess the ways in which these affect inter-ethnic group relations in conflict situations. Although the inter-group level of analysis is useful in illuminating some features of inter-ethnic relations, it is, for example, unable to explain when steadily improving inter-ethnic relations are ‘suddenly’ shattered by violent outbursts.1 Often the outbreak of inter-ethnic violence is preceded not by a deterioration of inter-group relations, but by changes in intra-group relations. Thus, sub-groups are also important units of analysis. Although a chronological sequence of events does not necessarily indicate a causal relationship, it appears that under certain conditions intra-ethnic group relations exercise an overriding influence on inter-ethnic group rela10

Learning to Read between the Lines 11

tions – often in quite unexpected ways. The reverse can also be true: under certain conditions inter-ethnic relations exercise an overriding influence on intra-ethnic relations.2 This leads to the unsatisfactory proposition that inter-ethnic group relations may both condition and be conditioned by intra-group dynamics. Further, intra-group dynamics may have a variable impact; sometimes exacerbating and sometime dampening inter-ethnic conflict. While a reading of the literature on ethnic conflict may prompt the observation that the intra-group dimension of inter-ethnic conflict is both interesting and neglected, it remains to be demonstrated exactly why and how the intra-group dimension is related to inter-group conflict dynamics, and vice versa. Concomitantly, it needs to be demonstrated how these linkages relate to policy alternatives and conflict management. This requires more than the examination of the correlation of inter-group and intra-group conflict by charting the coincidence of the two, since such an approach would be unable to consider those instances when changes in the intra-group arena are not coincident with a change at the inter-group level. Rather, it is necessary to examine the relationship between these levels of interaction by asking when and why changes at one level have – or do not have – an impact on the behavior of actors at another level under conditions of violent ethnic conflict. This study does not venture a full-blown theory of ethnic politics. It does not make assertions of universal, invariant causal relationships between independent and dependent variables. Rather, it focuses attention on what are judged to be crucial mediating factors in conflictual ethnic politics without imposing a priori assumptions of causality among them.3 While each of these mediating factors can be analytically separate and discrete, in practice they interact and affect each other. Thus, for example, the mobilization of group identity and the mobilization and extraction of resources may be mutually reinforcing in certain cases. Rather than try to disentangle them artificially, it is better to recognize the possibility of interdependence and to draw insights from case studies into when, why and how they interact. Taken together mediating structures represent a series of prisms which focus or diffuse both inter-ethnic and intra-ethnic relations to make them more or less conflictual, indeed, more or less violent.

Patterns of violence Often, ‘empirical’ studies of ethnic conflict count up instances of violence within a case and point to the final total as some kind of

12 Introduction

indicator or descriptor of ethnic violence (Gurr 1993; Carment 1993). The large-N studies of Ted Robert Gurr epitomize this approach (Gurr and Harff 1994, esp. 160–6). However, when one unpacks the cases with which one is most familiar, we usually find the conflation of different types of violence and different types of victims. In the context of the current study, particularly noteworthy is the failure to distinguish intergroup violence from intra-group violence. This is particularly important in the Sri Lankan case where violence within the majority (Sinhalese) community vastly surpassed inter-group violence during the People’s Liberation Front (JVP) resurgence from 1987 to 1990 resulting an estimated 40,000 deaths in a little over two years. Here, the implications of conflating the two arenas of violence are significant, because it risks representing the unprecedented violence in the Sinhalese intra-group arena as an escalation of inter-group conflict. Not infrequently, the media deals with such complexity by simply ignoring it. While it is important to maintain the distinction between arenas of violence, it is equally important to recognize the inter-connections. To cast the JVP resurgence exclusively as an intra-group phenomenon, strips it of the context which gives it broader meaning in the conflict and neglects its inter-group dimensions. The violence of this episode was different from that in the inter-group arena, but it is not separate from it.4 Although the necessary conditions were present for an increase in Sinhalese intra-group violence (such as high youth unemployment and disillusionment, waning legitimacy of the ruling regime, pervasive sense of injustice, and so on), the trigger for this episode was unquestionably rooted in the inter-group conflict – the signing of the Indo-Sri Lanka Agreement to settle the Sinhalese–Tamil conflict. Moreover, Sinhalese sub-groups were mobilized and animated by competing responses to the ‘Tamil Question.’ To overlook these inter-group dimensions would neglect a critical component of that episode in the conflict. The fact that this episode in the Sri Lanka conflict is dominated by intra-group Sinhalese violence does not detract from the ‘ethnic’ quality of the struggle. In keeping with Brass’s definition of ‘ethnicity,’ the period was infused with sub-groups’ use and manipulation of symbols and emblems of Sinhalese culture ‘in order to create internal cohesion’ and to differentiate themselves from each other in the intragroup and inter-group arenas.

Methodology A key influence on the development of the methodology developed in this book is the classic work by Graham Alison (1969, 1971), despite


the obvious difference in the central objects of analysis – the Cuban Missile Crisis on the one hand, and conflict in Sri Lanka on the other. Although 30 years have passed since the publication of that article and book, they still have much to offer conceptually and analytically by demonstrating that the application of different conceptual lenses (‘models’) to the same event may generate very different explanations and responses. However, there is a need to push this analysis still further. The application of different analytical lenses may not just highlight different facets of the same reality; it may highlight parallel but separate realities which intersect in different ways, at different times, with different results. Analyses or responses to instances of identity-based conflict must consider the possibility that different groups may rally different facts to describe their realities and to justify their behaviours. In essence, the current study is an attempt to understand how the intersection of inter-group realities and intra-group realities ‘play out’ in the arena of ethnic competition and conflict in Sri Lanka – an issue that is no less salient in Northern Ireland, Israel/Palestine, or Afghanistan. It is necessary to briefly explain the terminology and categories that will be employed. ‘Ethnic group’ is used to designate a collectivity which defines itself as distinct from other collectivities according to broadly accepted cultural markers of difference such as language, religion, sense of share origin, real or fictive kinship or descent, etc. (Tamils, Sinhalese, Muslim, Burgher). Ethnic groups however are rarely, if ever, homogenous. Within ethnic groups there are identifiable subgroups which distinguish themselves from each other within the boundaries of that ethnic group. Differentiation may be based on a variety of markers (such as dialect, caste, political affiliations, geography, class, gender). Thus, for example within the Tamil ‘ethnic group’ one may identify Jaffna Tamils, Plantation Tamils, Colombo Tamils, South Indian Tamils, Hindu Tamils, Christian Tamils, and caste-based grouped groupings such as the Vellalas, the Karaiyars, and the Pariah. These social, political, and economic groupings within ethnic groups will be identified as ‘sub-groups.’ While members of a ‘sub-groups’ share a common sense of identity, sub-groups themselves may possess different degrees of organization. This affects the efficacy of any attempt to embark on collective forms of action. Like Tilly (1978), my use of the term ‘organization’ does not necessarily imply ‘formal organization.’ The more extensive the common identity and internal networks of the group, the more organized it is. ‘Mobilization processes do not have to start from scratch. They can build on pre-existing networks of informal relations as well as on

14 Introduction

pre-existing networks of formal organizations, political and otherwise’ (Kriesi 1988: 362). Accordingly, the term ‘ethnic organization’ is used to identify those organizations which attempt to mobilize individuals along axes of ethnic identity, e.g., in Sri Lanka, Buddhist Patriotic Organizations, the National Movement Against Terrorism (NMAT), the Liberation Tigers of Tamil Eelam (LTTE), and ethnically based political parties. In other cases such as Northern Ireland, this would include the Provisional IRA, the Protestant Orange Order. In apartheid South Africa, it includes the National Party, the Nederduitse Hervormde Kerk and the neo-nazi the Afrikaner Weerstandsbeweging (AWB). The basic analytical distinction put forward in this book is that between the ‘intra-group’ or ‘intra-ethnic arena’ and the ‘inter-group’ or ‘inter-ethnic arena.’ The intra-ethnic arena refers to the interaction between sub-groups within the same ethnic group whereas the interethnic arena refers to the interaction between sub-groups across ethnic group boundaries. Thus, the confrontations between Tamil paramilitary organizations throughout the 1980s and 1990s5 and the feuding within the Republican Irish People’s Liberation Organization (IPLO) in Belfast in August 1992, are both examples of sub-group activity in the intra-group arena. While the 1977 post-election violence in Sri Lanka was initially intra-ethnic in scope (between the in-coming and outgoing Sinhalese-dominated political parties), it also spilled across ethnic group boundaries into the inter-ethnic arena. Although this basic distinction is maintained throughout this book, it is clear that sub-groups engage in activities within the inter-group arena in order to pursue intra-group goals, just as they engage in activities in the intra-group arena in order to pursue inter-group goals. The terms ‘inter-group’ and ‘intra-group’ presuppose that there exist primary and secondary social boundaries according to which individuals define themselves, and distinguish themselves from others. Thus, in Sri Lanka, the principal inter-group boundary separates Tamils from Sinhalese6 while numerous intra-group differences further divide these primary groups into sub-groups along lines of regional origin, class, caste, political allegiance, political ideology, and so on. While this is a parsimonious way of organizing a study of ethnic groups in conflict, it may also be criticized for being too mechanistic or simplistic because in practice inter-group and intra-group boundaries are usually more fluid than the terminology implies. Identities are complicated, as seen, for example, in instances of ‘inter-communal’ or ‘inter-ethnic’ solidarity or cooperation7 and in multiple membership and allegiances within the intra-group arena – even during protracted conflict.


Although inter-group and intra-group boundaries may be permeable or ‘fuzzy’ at the edges, they nonetheless represent a core set of attributes which define and animate a sub-group at specific points in time. Without criteria to organize groups and sub-groups, there is no way to identify when and why boundaries are more or less fluid and how this may affect patterns of conflict. Cross-cutting affiliations do not necessarily challenge the existence or particular location of a group boundary. Indeed, it is only by identifying boundaries that cross-cutting affiliations can be identified. They represent a phenomenon which needs to be fully examined and explained because they suggest possible sites and opportunities for constructing mechanisms of conflict reduction. A two-level critical junctures approach to ethnic conflict analysis An examination of conflict in Sri Lanka might be undertaken in many different ways. In this study, the analytical chapters are organized according to ‘episodes’ which mark ‘critical junctures’ in the conflicts. A critical juncture is a turning point in which relations within and between groups are altered fundamentally. This is evident, for example, in the polarization, fragmentation, consolidation, and reconfiguration of inter-group or intra-group relations. The disaggregation of the cases into episodes allows for a better appreciation of the evolution of a conflict. It also allows different types of conflict (linguistic, caste, territorial, and so on) to be compared within the same study. There are many kinds of data that can be used to construct indicators of increased or decreased conflict: deaths due to political violence, number and intensity of ‘disturbances’ such as those reported in detailed chronologies of violence;8 human rights violations as reported by national and international organizations; political assassinations; demonstrations; strikes; protests with or without violence; riots; explosions; and property damage. The list could continue at some length.9 Any or all of these can provide the basis for an empirically grounded examination of patterns of conflict. This study does not examine all possible instances of conflict as is done in some studies employing events analysis. It does develop some novel techniques for assessing qualitatively and quantitatively the numerous increases and decreases in the level of conflict in each case. Although critical junctures are usually coincident with changes in the empirical pattern of conflict (i.e. an increase or decrease in violent incidents), this need not be so. An advantage of a critical junctures approach to the study of ethnic conflict is that it enables the disarticulation of dauntingly ‘big struc-

16 Introduction

tures’ and ‘large processes.’ Consequently, it is less difficult to identify how and why events at a micro level have a larger impact beyond immediate circumstances. Following the advice of Tilly (1984), the approach is both ‘concrete’ and ‘historical;’ that is, it is ‘concrete’ in terms of having ‘real times, places and peoples as … referents and in testing the coherence of the postulated structures and processes against the experiences of real times, places and peoples’ (ibid. 14). And, it is ‘historical’ in terms of ‘limiting [its] scope bounded by the playing out of certain well-defined processes, and in recognizing from the outset that time matters – that when things happen in a sequences affects how they happen, that every given structure or process constitutes a series of choice points. Outcomes in a given point in time constrain possible outcomes in later points in time’ (Tilly 1984: 14). A potential disadvantage of the critical juncture approach is that it often involves detailed, ‘thickly described’ analyses of events (Geertz 1973). As a result, the study risks becoming bulky and intimidating for the reader. To alleviate this danger, the chapters analysing critical junctures are prefaced by a discussion of the context within which an event occurred. In addition, they are followed by a summary of findings in a table format. Having identified critical junctures in each of the conflicts, the associated episodes are examined through two separate sets of ‘analytical lenses:’ Model I examines the inter-group level of interaction, and Model II examines the intra-group level of interaction. The term ‘model’ is used following Allison (1969, 1971): Analysts think about problems of … policy in terms of largely implicit conceptual models that have significant consequences for the content of their thought. … Clusters of … assumptions constitute basic frames of reference or conceptual models in terms of which analysts both ask and answer the questions: What happened? Why did the event happen? What will happen? … In arguing that explanations proceed in terms of implicit conceptual models, this essay makes no claim that … policy analysts have developed any satisfactory, empirically tested theory (Allison 1969: 689–90).10 By applying the two analytical lenses concurrently, the utility (and limitations) of each can be evaluated. It is then possible to establish the relative impact and importance of inter-group (Model I) and intragroup (Model II) factors at the critical junctures of a conflict. It is clear from the case of Sri Lanka that an exclusive focus on inter-group


factors reveals and explains only part of the ‘conflict story.’ Although there may be episodes in ethnic conflicts which are essentially intergroup in structure (i.e. with a minimal intra-group dimension), none were found in this study. In every episode, intra-group factors exercised a ‘notable,’ ‘equal,’ or ‘dominant’ influence vis-à-vis inter-group factors.11 This does not reduce the need to study the inter-group dimensions of ethnic conflict. It does, however, underscore the importance of bringing intra-group factors more centrally into ethnic conflict analysis. While an inter-group analytical lens may reveal important features of a conflict, there is a danger that it may generate only a partial understanding of the dynamics of ethnic conflict, or worse, that it may misrepresent the underpinning dynamics of conflict. An exclusive Model II focus is no less susceptible to the misrepresentation of a conflict. However, if applied conscientiously, it may be able to explain when and why inter-group polarization, conflict escalation and intransigence are a function of intra-group, rather than inter-group, processes. In the final analysis, the issue should not be framed as a choice between Model I or Model II. It is a matter of integrating both into a single analytical framework. The implications of thinking in ‘either–or’ terms rather than ‘both–and’ terms are addressed by Uphoff: ‘Analytically, choices can be reduced to simple ‘either–or’ alternatives, but these are usually creations of the mind. Where more expansive possibilities framed in ‘both–and’ terms are conceived and tenable, we will derive greater benefits from existing resources, resources and material’ (1992: 25). The mediating structures and processes of ethnic conflict This study proposes that there is a set of mediating structures and processes which conditions inter- and intra-group relations to make them more or less conflictual:12 (i) the mobilization and politicization of group identity; (ii) the mobilization of and competition for resources within and between groups; (iii) the changing economic, social, and political context (opportunity structure); (iv) the variable role of state actors. While each of these factors has been elaborated and applied separately in previous studies, this study integrates them into a single analytical framework and demonstrates their utility for developing an under-

18 Introduction

standing of ethnic conflict which explicitly incorporates the interaction between intra-ethnic and inter-ethnic political arenas. Figure 2.1 illustrates the way in which the four sets of mediating factors are thought to affect behavior in both the intra-group and inter-group arenas. The arrows connecting each of the sets of mediating factors illustrate their interaction. The revolving circles represent the dynamic of the specific mediating factors among sub-groups in the intra-group arena, a mediating dynamic which continues into the inter-group arena through the broken line (representing the permeability of the boundary between arenas). This broken line indicates the interface through which internally driven behavior of sub-groups crosses into the inter-group arena. Thus, for example, while identity mobilization takes place within Intra-group Arena A (between Subgroup A.1, Sub-group A.2, and so on) and within Intra-group Arena B (between Sub-group B.1, Sub-group B.2, and so on), the Inter-group Arena also provides space for sub-groups to engage in identity mobilization. If fear of domination by the opposing ethnic group is an important facet of intra-group identity or a necessary component for intra-group solidarity, then an attack by Sub-group A.1 across the interethnic divide on any Sub-group in Intra-group Arena B in order to elicit a backlash would represent the extension of the process of identity mobilization into the Inter-group Arena. Similarly, the other mediating factors have potential dual impact across arenas, that is, both within and between ethnic groups. By developing and applying an analytical framework based on a consistent set of mediating factors, it is possible to understand better the political salience of the inter-group/intra-group nexus both within a case and across a range of cases. By focusing on the mediating structures and processes of ethnic group relations rather than on the ebb and flow of violence, it allows one to examine instances in which intra-ethnic activities alleviate conflict as well as exacerbates it. The rest of this chapter will introduce and discuss each of the mediating factors of ethnic conflict in the intra-group and inter-group arenas. The mobilization and politicization of group identity The recognition that ethnic groups are internally heterogeneous is a necessary but not sufficient starting point for an examination of the connections between inter-group and intra-group behavior under conditions of conflict. Not only are ethnic groups heterogeneous but, as will be discussed in greater detail below, individual and group selfidentification is ‘situational,’ ‘contextual,’ ‘mutable,’ ‘non-fixed,’ and

Figure 2.1

Mediating factors of ethnic conflict in the intra-group and inter-group arenas

20 Introduction

‘fluid’ (Horowitz 1985: 7; Manor 1988: 12; Nagel 1986: 95; Nash 1986: 6). As James Manor aptly notes: ‘… the crucial point is that most people … appear to be able to shift their attentions and passions from one identity to another, frequently and with great fluidity. Some of these multiple identities can exist along side one another in one person’s mind, though one of them usually tends to predominate at one particular time’ (Manor 1988: 12). Although Manor’s example is drawn from an Indian case, the phenomenon he describes is evident in all cases of conflict in which boundaries of identity constitute shifting battle lines. To argue that identities are mobilized and politicized is to wade into an on-going debate over the nature of ethnic group identities. A thorough assessment of this debate would require more space than is appropriate here. It is sufficient to note that the dominant point of reference is the disagreement between those who argue that ethnic identities are primordial and those who prefer to view them as instrumental for other interests or forces (McKay 1982; Scott 1990). Each position offers an opposite explanation for the ubiquity of ethnic conflict: the first argues that it results from the clash of rigid and unchanging group identities, while the latter presents it as a consequence of the plasticity and manipulation of collective identities. This book employs a conception of politicized group identity which attempts to steer between static primordial explanations and overly fluid instrumentalist explanations. Group identities are viewed as being neither primordial nor instrumental,13 but contingent and contextual. Horowitz sums it up most succinctly: Group boundaries are made of neither stone nor putty. They are malleable within limits. The mutability of boundaries does not mean that ethnic affiliations are merely ‘strategic,’ that they can be called forth whenever it is convenient to do so in the quest for competitive advantage or can be willed into being in the service of economic interest … What is necessary therefore is a sense of the mutability of group boundaries and yet their dependence on antecedent affinities that are not easily manipulated. To overemphasize one is to mistake the bases of conflict. To overemphasize the other is to miss opportunities for policy innovation.14 In other words, although the political salience of boundaries of identity is not hard and fast, the boundaries themselves cannot be created out of thin air.15


‘Contingency’ is used to refer to a particular type of group selfidentification in which members define themselves along a range of different and sometime competing axes of self-identification. Under different conditions, different traits will be the primary point of reference guiding behaviour. Far from being fluid, self-identification is affected by the particular context of expression as well by as particular understandings of past articulations of identity. While cross-cutting cleavages have the potential to divide groups, they may also reconfigure individual affiliations and thereby create new groups, bound together by new and different bases of group loyalty. For the concept of ‘contingency’ to be analytically relevant, it is important that multiple axes of identification not merely exist, but that they be politically salient in group interaction. All groups are potentially contingent, since there is a variety of axes along which individuals can group themselves. However, the particular expression of group identity at any point in time is a function of the conjunction and constellation of factors, including state structures and societal processes as well as the activities of ethnic mobilizers. Importantly, in many cases there appears to be a mutually reinforcing relationship between the mobilization of identity and the mobilization of resources for ethnic group collective action and the role of the state actors. The mobilization of identity is a process whereby particular axes of identity within heterogeneous groups become more or less politically salient, thereby affecting both intra-group boundaries as well as intergroup boundaries. It is a process which highlights or even inserts markers of difference between and within groups while obscuring possible markers of similarity. It appears that the mobilization of one group may itself stimulate countermobilization by other groups. While this process is largely catalyzed and articulated by mobilizers, it is channelled through existing state and social structures, processes, networks and institutions. The mobilization of identity merits attention because it sheds lights on the drawing of the dividing lines-cum-battle lines; it is a critical component in the construction and maintenance of a sub-group’s claim to legitimacy which consequently affects both the efficacy of group boundary maintenance and the mobilization of resources. The point here is two-fold: first, identities are historically constructed, and second, they are always relational and multiple. Because the individual-self may be a member of many different groups, it is essentially plural, and is therefore continually open to redescription.

22 Introduction

‘Ethnic’ identity, can only function as a mechanism of identification within a specific context which privileges the ‘subjective, symbolic, or emblematic use … of any aspect of culture’ in the pursuit of cohesion and the differentiation (Brass 1991: 19). Thus, ethnic conflicts are distinguishable from other types of identity conflict which are characterized by the mobilization of non-cultural markers of difference – such as class, ideology, gender, or kinship – in the pursuit of group cohesion and differentiation. By maintaining the distinction between ethnicity and other bases of group identification, we can examine the linkages between them. As Rothschild points out: ‘the contemporary rise of ethnic individual identification and the trend towards ethnic group politicization do not mean that all other orientations, affiliations, differentiations, segmentations, or conflicts have been neutralized or eliminated. They interact with ethnicity – within individuals, within states and across states.’16 Particularly intriguing in the context of the current research project are those occasions in which a shift in the politically salient axes of identity is reflected in a corresponding shift in the axes of conflict.17 For example, recent anthropological studies have revealed that the present dynamics of conflict in Sri Lanka, and the ethnic identities upon which it is based, are ‘radically different from earlier conflicts and identities’ (Nissan and Stirrat 1990). Pre-colonial states in Sri Lanka were founded on kinship and political structures which were indifferent to the cultural and linguistic composition of the society. More significantly, it has been suggested that riots during the colonial period were markedly different to those of the post-colonial period in that they were usually aligned along religious lines rather than ethno-linguistic lines – for example: Sinhalese Buddhist attacking Sinhalese Catholic; Tamil Hindu attacking Tamil Catholic; Buddhist, Hindu or Catholic attacking Muslim; and Muslims attacking all back in return (Nissan and Stirrat 1990; Spencer 1990: 5). This stands in stark contrast to the dividing lines today where ethno-linguistic antagonisms reflect a wide range of interactions, including the relations between Tamil and Sinhalese Christians (clergy and lay people).18 Other cases reveal similar reconfigurations of the boundaries of group identity and conflict. For example, in Northern Ireland during the Industrial Revolution, there was a consolidation of a working class identity across the sectarian divide – abetted by working class leaders and trade unionists. Fearing this challenge to its class and economic position, the Tory elite actively fostered sectarian antagonisms among its opponents as a means of preserving its prerogatives (Lebow 1974:


209). As a result, a class-based form of social organization gave way to the sectarian forms evident today. However, one should be careful not to overdraw this facet of Northern Irish inter-group relations. Non-class axes of group identity were also salient then, just as class-based axes are still salient now. Such examples illustrate that different sub-groups and facets of collective identity may become more or less salient over time under different conditions. While it is useful to distinguish between inter-group and intra-group levels of interaction, it is essential to bear in mind that this separation is itself contingent upon socially constructed boundaries that do not necessarily define all instances of social interaction (see for example, Darby 1990a). In other words, it is necessary to examine not just how such boundaries condition inter-ethnic relations, but how such boundaries come to be constructed and maintained. By understanding how identities are internally mobilized, one gains insight into how a particular identity may constitute the basis for intransigence and violence between groups. Conversely, the examination of intra-ethnic identity mobilization may shed light on the ways in which shared facets of identity and cross-cutting cleavages may provide the basis for peaceful coexistence as well as conflict management and resolution. The mobilization of resources Ethnic identity becomes politically salient and potentially volatile not merely when individuals come to view themselves collectively as a group with shared experiences, interests and ideas. In order to act collectively, a group also needs a variety of resources.19 While a collectivity may come to define itself as ‘ethnically’ distinct, its political efficacy is heavily dependent on its ability to mobilize and extract resources (McCarthy and Zald 1977).20 Although resource mobilization is a necessary rather than sufficient requirement for ethnically-based collective action, by focussing on resources and their use, ‘we come closer to the proximate causes of the effects we wish to understand’ (Uphoff 1990: 321). As with identity mobilization, the process of resource mobilization is not an autonomous one; it is catalyzed by individuals and a wide range of organizations,21 associations and institutions.22 These may, or may not, t have been founded to engage in activities related to ethnic identity. Organizations or institutions created for one purpose (‘ethnically neutral’) can under certain conditions become ‘ethnicized’ and come to play a role in directing ethnic relations through the mobilization of both collective identity and resources.

24 Introduction

As Charles Tilly points out, ‘entire ethnic groups almost never mobilize or act collectively, but ethnic groups serve as bases for mobilization and collective action when the actions of outsiders either threaten to exclude them from shared and collectively controlled opportunities or open up new niches of collective competition’ (Tilly 1991: 5). He continues, ‘ethnic entrepreneurs (who are often professional brokers such as intellectuals and politicians) play exceptional parts in such mobilization.’ What Tilly does not develop is the ways in which internal heterogeneity and intra-ethnic competition may play a role in the process of resource mobilization. Nor is it clear why one axis of identity out of many possible axes comes to constitute a dividing line. To develop an understanding of the nexus between the intra-ethnic and inter-ethnic arenas, it is necessary to examine sub-group resource mobilization in both the intra-group arena and the inter-group arena. Not only do sub-groups compete with each other for resources from both arenas, but they attempt to utilize those resources within both arenas as well. This may result in unexpected coalitions such as the well-founded charges in 1991 that the late Sri Lankan President Premadasa had been responsible for supplying weapons and material to his arch-rivals, the LTTE, in an effort to apply indirect pressure on mutual enemies and thereby defuse the escalating challenges from within his ethnic constituency.23 This is a crucial example of a Tamil sub-group extracting resources from the other side of the inter-ethnic divide for use within the intra-group arena.24 Other cases also illustrate similar instances of ‘cross-border shopping.’ In Northern Ireland, Loyalist and Republican paramilitaries have cooperated with each other in the assassination of security forces and prison officials as well as individuals within their respective organizations (Feldman 1991; Maguire 1988);25 just as they have cooperated in dividing up territory for racketeering purposes, including the running of Loyalist extortion rackets in Republican territory (Anderson 1994; Clare 1989, 1990; Darby 1986; Bruce 1992: 192–8, 245–8; Shivers and Bowman 1985: 21; Irish News, 3 August 1991: 1). The case of Sri Lanka illustrates the importance of considering the international dimension of intra- and inter-group resource mobilization. The role of India26 has been variable, but continuous. This has included: provision of arms and training bases to Sri Lankan Tamil paramilitaries from the early 1980s to the assassination of Rajiv Gandhi; sanctuary for Tamils displaced by violence; and the failed Indo-Sri Lanka Agreement – an effort which deployed 60,000–80,000 Indian troops on the island from 1987 to 1990 to impose a peace


settlement in Sri Lanka (de Silva 1993; Gunaratna 1993; HellmanRajanayagam 1989). Beyond the Indian subcontinent, most, if not all, Tamil paramilitaries have raised voluntary and non-voluntary contributions from Tamils of the diaspora in Europe and North America. International lending institutions, development Non-Governmental Organizations (NGOs), and international human rights monitoring agencies are all part of the international network of actors which have exercised varying degrees of influence on events in Sri Lanka (Bush 1999, 2001). Resource mobilization has been an important point of reference in the social movements literature (Tarrow 1989). However, empirical points of reference of this literature have tended to be industrialized countries of the West, and ethnic dimensions have been given scant attention. There is a need to broaden the empirical scaffolding of resource mobilization analysis developed in the social movement literature and to apply it beyond the Western arena to cases of contention where ethnic identity constitutes a central basis for conflict. Opportunity structures In this book, the concept of political opportunity structures follows that of Sidney Tarrow: Political opportunity structures provide groups with resources that increase the effectiveness of their protest – for example, a sympathetic press, or political parties seeking electoral advantage, or ‘conscience constituents’ who bring outside resources and legitimacy to the movement. It encourages unrepresented groups to protest in the belief that the costs of insurgency have been lowered – as when a sympathetic political party comes to power and makes it clear that it will not support repression. Political opportunities cannot make the poor conscious of grievances of which they were formerly unaware, but it can help them to detect where and how the system is most vulnerable, enabling them to overcome their habitual disunity and lack of information (1989: 36). The following facets are particularly important: inter-elite group conflict; presence or absence of influential allies; stability or instability of political alignments including electoral balance; the degree of openness or closedness of the formal political system; and the presence or absence of repression.

26 Introduction

The strategies and tactics of the mobilizers of ethnic sub-groups are drawn from a set of choices conditioned by the political opportunity structure (including the expected response of state actors as well as other ethnic groups). In other words, changing political opportunity structures affect the strategic and tactical interaction between and within ethnic groups. Central variables in an examination of the effect of changes in the opportunity structure are international actors, along with domestic state actors. While opportunity structure is an important mediating factor that conditions the context in which ethnic actors act, its influence is seldom uniformly experienced by all actors. The opening of certain opportunities for some actors closes avenues for others. The differences in sub-group abilities to respond to and affect change in the opportunity structure are especially important in the intra-ethnic arena because they alter capacities to act both in that arena and the interethnic arena. It is important to bear in mind the inherent risk of tautology in the application of the concept of political opportunity. One needs to explain those instances when movements or activities are launched despite inhospitable opportunity structures as well as instances in which groups fail to act despite conducive opportunity structures. For this reason, it is necessary to frame ‘opportunity structure’ as a contextual, rather than independent, variable. This acknowledges that while opportunities are important, we need to consider the impact of the other mediating factors identified here. State actors Research on state actors has generated a growing collection of case studies detailing the various roles they may play in the structuring of ethnic relations, including the creation of conditions conducive to intergroup conflict and accommodation.27 An important contribution of these studies is that they challenge the tendency to conceptualize ethnic groups as given, self-contained entities. These studies attempt to illuminate the state structures and processes which condition the construction and articulation of identities and inter-group relations. By casting ethnic groups and identities as dependent variables, these studies explore the underpinning dynamics of inter-group and intragroup relations. While these studies have made an important contribution to the understanding of ethnic phenomena, they are limited by their topdown theoretical orientation. In this sense they share a common short-


coming with the more general studies of ‘the state:’ the relative invisibility of societal structures and processes in the explanation of conflict and cooperation. This attributes too much agency to state actors and not enough to societal forces. In extreme cases, societal structures and processes are treated as little more than putty in the hands of an allpowerful state. While actions by state actors such as their differential allocation of resources may help to explain the political rise of certain social groups (for example, Laitin 1985, 1986; and Warnapala and Woodsworth 1987; Herring 1998), this factor is less helpful in explaining the tenacity and influence of those social groups which did not receive similar treatment. Heavily state-centric explanations may have been tactically acceptable in order to reinsert the state back into an academic agenda weighted otherwise towards societal reductionism.28 The pressing task now at hand is to restructure that artificially dichotomized debate of state or society into a more subtle and sophisticated understanding framed in terms of state and societal influences on political behavior. This requires an understanding of the degree to which, and the conditions under which, state actors are actually independent of ethnic actors. That is, there is a need to establish the degree of state autonomy (the ability of state actors to formulate goals not simply reflective of particularistic group interests) and of state capacity (the ability to implement them). In Sri Lanka (and in other cases) state actors appear frequently to be ‘semi-autonomous’ since efforts (particularly conciliatory or accommodative efforts) are often subverted by the negative response from one or another segment of society. In Sri Lanka, the association of state actors with vigilante groups and the orchestration of anti-Tamil riots illustrate that portions of the structures of state may be ‘infiltrated’ as well as pressured by societal forces (see Chapter 7). The dynamics and significance of ‘semi-autonomous’ state actors are empirical as much as theoretical issues. If state actors are semi-autonomous, then by definition they are also semi-dependent. The shifting lines of autonomy, capacity and responsiveness are important points of reference in this study. By systematically examining relations within and between ethnic groups, the variability of state autonomy, capacity and responsiveness will be more clearly understood.29 While actions of state actors may precipitate new forms of ethnic mobilization, these do not determine the outcome. The internal solidarity achieved by ethnic groups depends also on the other mediating factors introduced here,

28 Introduction

namely the mobilization of identity and resources, and the opportunity structure. Having sketched out the theoretical framework in this chapter, the next chapter provides a general overview of politics in Sri Lanka – a prerequisite to an examination of specific critical junctures in the ethnicized politics of Sri Lanka.

3 An Overview of Sri Lanka

Introduction to Sri Lanka In 1948, when Sri Lanka (then Ceylon)1 gained political independence from Britain, the knowledgeable observer would have felt confident in predicting that the country would quickly adjust to its new sovereign and independent status. Unlike India, Sri Lanka attained independence rather untraumatically. Practically speaking, Sri Lanka’s independence followed on the coat-tails of the Indian struggle. The British left amicably and were perceived as a trustworthy ally by both Sinhalese and Tamil political elites. Sri Lanka enjoyed a high literacy rate and a politically sophisticated electorate which had exercised universal adult franchise since the early 1930s. In the early 1980s, it was still common for Sri Lanka to be labelled a ‘model Third World democracy’ with a robust parliamentary system which regularly voted the incumbent party out of office. Although there were debates between ‘welfareists’ and ‘economic liberals’ over the role of the state in the economy (Moore 1990), the overall quality of life was high. Indeed, Sri Lanka was an envied anomaly because of its ability to maintain such a relatively high standard of living despite its low GNP per capita.2 However, since 1983, the contemporary history of Sri Lanka has been steeped in militarized violence. An estimated 66,000 people have disappeared or been killed in inter-group violence associated with the efforts of Tamil paramilitaries (most notably, the Liberation Tigers of Tamil Eelam) to create a separate state in the North and East of the island. Assassins and suicide bombers have murdered politicians of all political stripes including Ranasinghe Premadasa, the President of Sri Lanka and Rajiv Gandhi the Prime Minister of India. And in 2000, President Chandrika Kumaratunge narrowly escaped being killed by a 29

30 Introduction

suicide bomber in Colombo, but was blinded in one eye. Moderate Tamil politicians have been murdered, throughout this period: including a number of mayors of Jaffna; A. Amirthalingam, leader of the TULF (Tamil United Liberation Front) in 1989 and Neelan Tiruchelvum in July 1999, a sitting MP and internationally recognized, Harvardtrained, lawyer and scholar. More than 1.7 million people (or one-tenth of the total population) have been displaced from their homes at one time or another.3 An additional 200,000 persons are estimated to have sought refuge internationally. The share of internally displaced households ranges from 40 per cent in Ampara District to 95 per cent in Mannar. More than one third of all homes have been damaged or destroyed in the NorthEast Province (GSL 2001). From 1983 to late 2001 (except for a fivemonth gap between January and June 1989) Sri Lanka has been governed under emergency regulations.4 The ceremonial guard and simple police force of 1948 (backed up by a defence pact with Britain) has swelled to an armed force of approximately 235,000 men and women (Kelegama 1999) – however, this figure does not include the tens of thousands of crudely armed village militias (or ‘Home Guards’) nor the pro-government Tamil Paramilitary Organizations. Further, any calculation of island-wide armed stakeholders must include the LTTE, estimated to be 13,200 (Jane’s 2000) and the estimated 30,000–40,000 army deserters who may, or may not, have kept their weapons.5 Child soldiers are used systematically in war fighting, particularly, but not exclusively by Tamil Paramilitaries.6 Some have estimated that the country that had rightly prided itself for high literacy rates and educational standards now has up to 50 per cent of pupils leaving school in some LTTE-controlled areas to join the rebels (BBC 1998). A conservatively estimated 25,000 antipersonnel land mines ‘are scattered widely [and] sometimes indiscriminately’ in the North and East instilling the terror of militarized violence step-tostep as well as day-to-day.7 Army recruitment and compensation have become the primary source of resources transferred into the economy of the rural poor in the Sinhalese-majority regions of the South. As a percentage of the World Bank-defined rural poverty line, ‘military remittances’ increased from 5 per cent to 32 per cent between 1985 and 19978 – surpassing foreign remittances as the principal means of economic sustenance. In a perverse way, military employment has functionally become a youth employment programme cum poverty alleviation scheme. Related to the size of the armed forces, is the bloated size of the military budget.

An Overview of Sri Lanka 31

In a recent and rigorous study on the costs of the war in Sri Lanka, it has been estimated that from 1982 to 1996 defence expenditure as a percentage of total government expenditure has increased from 3.1 to 21.6 per cent. As a percentage of the GDP, the defence budget has increased from 1.1 to 6.0 during the same period of time.9 Thus, military expenditures dwarf combined expenditure on education and health, the consequences of which affect availability and access to basic social services by all Sri Lankans inside and outside of the war zones. So despite an economy growing at 5.5 per cent in the late 1990s (Dhume 2000), Sri Lanka has the second highest level of military expenditure as a percentage of GDP of all Asian countries (6 per cent – NGO Forum 1998; Kelegama 1999; and Arunatilake et al. 2000). In addition to the acute economic dislocation which would accompany any lasting cessation of militarized conflict (from the household to the national level), the sociological and anthropological impact would be profound. An entire generation of children are growing up without fathers, brothers, and uncles as the number of men in the armed forces (as well as lower number of women) has ballooned. The active or ‘hot’ war zones of Sri Lanka are located in the North and the East of the island. Within these areas, there is a complex and fluctuating system of territorial demarcation and control between various armed groups – accordingly, language expands to accommodate and describe these geopolitical realities: cleared and uncleared areas, newly cleared areas, gazetted areas, grey areas, and border areas. Until the LTTE stepped up its terrorist campaign of bombing in the largely Sinhalese South of the island in the early 1990s, Colombo was relatively untouched by the war raging at times only 200 kilometers away. The massive LTTE bombing of the World Trade Centre and the adjacent Galadari and Hilton hotels (killing 86 people and injuring 1400) in early 1996 and its bombing of the Bank of Ceylon in October 1997, brought the violence to the heart of Colombo Central Business District. Continuing campaigns of LTTE suicide bombers have provoked road blocks, security checks, and continued mass arrests of Sri Lankan citizens of Tamil origin – reinforcing the presence of warinduced insecurity, even while the military fighting is limited to the North and South.10 The dynamic of warfare between the state and the LTTE is one in which civilians are the principal victims as both direct and indirect targets of violence and intimidation. However, it is also a war which has seen direct and massive confrontations between government and LTTE troops in ‘conventional’ military battles – indeed in

32 Introduction

late 1998, one battle saw an estimated 2000 troops killed in 36 hours. Prisoners are seldom taken. The laws of war are rarely applied. No one on the island has been untouched by the militarized violence over the last 20 years. Although these figures convey a sense of the escalating militarization of the economy, society, and polity in Sri Lanka, they do not distinguish the relative impact of inter-group and intra-group violence on the overall pattern. In particular, they obscure the significant contribution of intra-group conflict to the worsening of violence in the late 1980s. From 1987 to 1989, in the Sinhalese south of the island, an extremist Sinhalese group called the JVP ((Janatha Vimukthi Peramuna – People’s Liberation Front) launched a sustained and bloody attempt to overthrow the Sri Lankan State. At the height of this challenge it was reported that 500–700 people were being killed each week in intragroup violence (see the discussion of the ‘JVP’ below and Chapter Five).11 It is estimated that 40,000 people were killed or disappeared during this period. JVP violence surpassed Sinhalese-Tamil violence in sheer viciousness and volume, i.e. intra-group violence eclipsed intergroup violence. But a ruthless counter-attack by state and quasi-state12 forces in the fall of 1989 annihilated the JVP leadership as well as thousands of supporters and innocent victims, thus extinguishing this particular source of violence for the time being. However, as military attention was being redirected to the Tamil areas of the north and to the ethnically mixed areas of the east, an assassination of the president again thrust Sinhalese intra-group politics to the fore. Political uncertainty created conditions in which Sinhalese sub-groups sought to capitalize on emerging political opportunities while protecting themselves from potential losses and challenges. This on-going struggle and its outcome will inexorably affect Sinhalese–Tamil relations since different groups are more or less accommodating of Tamil grievances, and Sinhalese intra-group competition often includes mobilization of support based on anti-Tamil sentiment. Intra-group violence is also evident within the Tamil community in the form of Tamil paramilitary feuding and, more significantly, the mass intimidation and brutalization of the civilian population by Tamil paramilitaries (Hoole 2001; Hoole et al. 1992; Bush 1993;) Both Sinhalese and Tamil intra-group violence have exacted a considerable toll. Within this context of inter- and intra-group violence, Sri Lankan citizens from all communities have been subjected – to varying degrees – to individual and mass arrests and incommunicado detention, abduction, torture and frequently murder by state, quasi-state, and anti-state actors.


Two points need to be stressed. First, while the JVP insurrection, Tamil paramilitary feuding, and the intimidation of civilians may be accurately identified as intra-group phenomena in terms of the perpetrators and victims of violence, they have crucial linkages to the inter-group level of interaction. For example, Tamil paramilitary feuding has often relied on weapons and materiel supplied by the Sri Lankan State, and the JVP insurrection was precipitated and sustained by the criticism of the Sri Lankan government’s approach to ‘the Tamil Question.’ Although inter- and intra-group violence are distinct, they are nonetheless inter-related. Second, although JVP violence was directed against the state and its supporters – that is, against members of the same ethnic group as the perpetrators13 – its underlying motivation and objectives were nonetheless ‘ethnically based,’ following Brass’s (1991) definition of ethnicity specified in Chapter One. That is, it was a struggle in which the subjective, symbolic and emblematic aspects of culture were employed to create internal cohesion and to differentiate sub-groups from each other in both the inter-group and intra-group arenas. It was a battle between sub-groups exclusively within the same intra-group arena over the definition of ethnic group identity and the right to be cast as the legitimate representative of that group. Academic explanations of the past two decades of ethnic violence in Sri Lanka are curious and disappointing, not only because of their relative neglect of the intra-group dimensions of the conflict but also because they tend not to ask the question of why or how a ‘model Third World democracy’ was transformed into ‘the Beirut of South Asia.’14 Rather than seeking to explain the disintegration of Sri Lanka as a transformation, there has been a tendency to project the Tamil–Sinhalese divide of today anachronistically onto a misreading of the past. Thus, the dominant research focus of those addressing the ethnic conflict has sought to explain the continuity rather than the upsurge of violence.15 For example, Gananath Obeyesekere writes that prior to the 16th century: there were historically two major opposed ethnic identities, Sinhalese and Tamil. The historical conflicts between Sinhalese and South Indian invaders reinforced and stabilized the SinhalaBuddhist identity … these people could be mobilized by their rulers to fight foreign invaders … the myths became, on occasion, rallying points for Sinhalese Buddhism (Obeyesekere 1979: 282–3).

34 Introduction

This interpretation is not that different from the much-cited ‘Cleghorn minute’ of 1799: Two different nations, from a very ancient period, have divided between them possession of the island: First the Sinhalese inhabiting the interior of the country, in its southern and western parts, from the river Wallouve [sic] to that of Chilaw, and secondly, the Malabars [i.e., Tamils or South Indians], who possess the northern and eastern districts” (cited in Kemper 1992: 144). This passage represents a particular colonial interpretation of the past which has conditioned the analysis of conflict dynamics and the definition of the groups in conflict, casting the conflict as the bipolar interaction of relatively homogenous groups. However, more critical scholarship has challenged this interpretation: ‘pre-colonial and most of colonial history does not conform to the model of two opposed nations imposed upon it by present day rhetoricians’ (Nissan and Stirrat 1990: 24).16 Not only did Tamils and Sinhalese live together peaceably for most of two millennia, but there was considerable social, political and economic commonality between them. While there were dynastic wars, the Sinhalese–Tamil communal violence dates from after Independence. ‘This is not to say that there were no differences between groups of people living in the island: the point is simply that differences of language, custom and religion were made into something new by the devices of the modern state … .’ (Nissan and Stirrat 1990: 24). Even today, in the midst of severe violence, there exist pockets and processes of inter-group cooperation and peaceful coexistence. The violence and group boundaries are not continuous, undifferentiated, or impenetrable. Sri Lanka exhibits a pattern identical to that identified above by Edward Azar in his studies of the Middle East: ‘Conflictual and cooperative events flow together even in the most severe of intense conflicts. Cooperative events are sometimes far more numerous than conflictual ones even in the midst of intense social conflict situations. However, conflictual events are clearly more absorbing and have more impact on determining the consequent actions of groups and nations [and, one might add, on determining the outsider’s impression of the conflict].’ Turning his attention to prescriptive concerns, Azar adds: ‘Cooperative events are not sufficient to abate protracted social conflicts. Tension reduction measures may make the conflict more bearable in the short term, but conflict resolution involves a far more complex process than mere conflict management’ (Azar 1986: 30–31).


The multiplicity of competing histories in Sri Lanka and their utility for politicizing identities and mobilizing populations will be addressed in greater detail below and in subsequent chapters. The point emphasized here is that the casting of present antagonisms as the continuous trajectory of the past obscures the examination of both the patterns of conflict and the sources of group identity. The conflict is neither as dichotomous nor as homogeneous as often suggested. This chapter begins its examination of the conflict by considering the political salience of the heterogeneity of the groups in conflict. Once the boundaries and significance of inter-group and intra-group relations have been drawn, the next chapter will examine the impact of these divisions on the course of the conflict. Who counts? Most studies of Sri Lanka begin by categorizing the groups according to the demographic information contained in the most recent full census (1981). Accordingly, groups are organized and defined according to the following categories: Sinhalese (74 per cent): predominantly Buddhist and Sinhalaspeaking. Tamils (18.2 per cent): generally divided into: Sri Lankan Tamils (12.6 per cent): predominantly Hindu and Tamil-speaking; Up Country Tamils17 (5.6 per cent): predominantly Hindu and Tamil-speaking tea estate workers in the Central Highlands. Muslims/Moors (7.5 per cent): Muslims, mostly Tamil-speaking. Others (0.04 per cent): including the Burghers, the Eurasian descendants of Dutch colonialists. While these categories provide a first cut at understanding one level of social interaction, it is necessary to ‘read between the lines’ to appreciate the impact of the internal or sub-group dimension of inter-group relations. None of these categories is internally homogeneous. There are important sub-groups and competing axes of identity which affect intra- and thus inter-group behavior. Thus, the developing an understanding of conflict in Sri Lanka must start with a recognition that the major groups are internally divided into politically salient subgroups along a range of differentiating axes such as political party affiliation, regional identification, religion, caste, and class. This approach stands in contrast to the representation of the conflict as the stark and violent

36 Introduction

confrontation between ‘the Tamils’ and ‘the Sinhalese.’ As outlined in Chapter One, to make sense of the patterns of ethnic violence it is necessary to delve more deeply than the interaction between warring ethnic groups; it is essential that the structure and dynamics within the groups in conflict must be given a central position in analysis. Subgroup identities also need to be put into a regional context. The fact that the Sinhala language and culture have no center outside of Sri Lanka has been used by political entrepreneurs to create both a siege mentality and a regional minority complex among the Sinhalese. Although the Sinhalese are a majority within Sri Lanka, many nonetheless perceive themselves to be a minority in relation to the 50 million Tamils of Southern India. Thus, the categorization of groups into minority Tamils and majority Sinhalese masks a phenomenon which could be termed a ‘double minority complex.’ This exercises considerable influence on the perceptions and behavior of both the Sinhalese and Tamil parties to the conflict. A senior diplomat in Sri Lanka’s capital, Colombo, sums it up well: ‘There are three minorities in this country: the Tamils, the Muslims, and the Sinhalese.’18 To the extent that this condition affects behavior within the Sinhalese community, the demographically defined minority and majority categories within Sri Lanka are inverted. Residential patterns Inter-group patterns It is useful to highlight those politically relevant spatial features which are not immediately evident in the first cut at demographic analysis. The standard demographic data offered above must be disaggregated by region and administrative district in order to develop a clearer picture of social and residential patterns. Unfortunately, this examination is limited to 1981 census data which do not reflect demographic changes since then, such as the massive displacement of people due to intra- and inter-group violence. Since the mid-1980s, the number of Internally Displace Persons (IDPs) has fluctuated between 500,000 to 1,000,000. As noted above, ‘the’ war is actually the cumulative violence undertaken by a variety of groups in different arenas, including the intimidation of civilians by Tamil paramilitary forces; the LTTE’s19 version of ‘ethnic cleansing’ against Muslims on the East Coast and in Mannar since 1990; the ‘excesses’ of the Sri Lankan Security forces; and in 1987–89, the ‘excesses’ of the Indian Peacekeeping Force in Sri Lanka under the auspices of the India–Sri Lanka Agreement (see Chapter 8).


The spatial concentration of Sinhalese and Tamil ethnic groups – with the important exceptions of Trincomalee and Ampare districts20 – helps to create conditions conducive to the development of a territorially based group consciousness. This has been further facilitated by the economic isolation of the Northern and Eastern Tamil areas and the destruction of normal life by the activities of military and paramilitary forces.21 A territorially based Jaffna Tamil group identity has been mobilized and harnessed by political entrepreneurs in a call for the creation of a separate Tamil state of Tamil Eelam. It should be noted, however, that the territorial scope of the proposed Tamil Eelam corresponds neither to the 1981 demographic composition of a number of the districts it would incorporate, nor to the historical geographic boundaries of the pre-colonial Tamil kingdom of Jaffna (see below). For example, despite the LTTE claim to incorporate in its separate state the entire eastern coast of Sri Lanka, the demographic composition of the administrative districts in this region is precariously balanced among Muslim, Tamil and Sinhalese populations – with the exception of Batticaloa District. Moreover, Tamil Eelam incorporates territory in which ‘Sri Lankan Tamils’ constitute a minority of a district’s population. According to the LTTE maps22 the proposed Tamil Eelam claims more than threequarters of the island’s coastal areas, including the Puttalam District (Sri Lankan Tamil proportion of district population: 6.7 per cent), the Northern Province, and the entire Eastern Province (Sri Lankan Tamil portion of the districts’ populations: Trincomalee, 33.8 per cent; Batticaloa, 70.8 per cent; Ampare, 20.1 per cent). These territorial claims appear to encompass an area over twice as large as the territory controlled by the Jaffna Kingdom when the Portuguese arrived in 1505 (cf. Mendis 1957: 50). That this demographic distribution and historical boundaries do not correspond with territorial claims of the separatists does not invalidate a territorially based Tamil identity, but it does draw attention to the way in which this has been harnessed and mobilized by political entrepreneurs. The spatial concentration of Tamil residents has been used by Sinhalese chauvinists to justify acts of communal violence when they identify all Tamils with the small number of violent Tamil militants. This collectivization and attribution of guilt underpinned the government’s military offensives and embargoes of the entire Northern Province from late 1986 onwards. The fact that Tamil paramilitary organizations are based in the north and east within the Tamil civilian population has led to the blurring of the distinctions between combat-

38 Introduction

ants and civilians in the minds of Sinhalese chauvinists and in military practice. Even those Sinhalese extremists who claim to distinguish Tamil civilians from Tamil Tigers have thus far failed to translate this into realistic policy prescriptions.23 Spencer has noted a parallel process among Tamil chauvinists who argue that ‘all Sinhalese are inherently violent and potential murderers like the relatively small number who were actually involved in the 1983 riots’ (1990a: 298). This rationale is apparent in LTTE attacks on defenseless Sinhalese and Muslim civilians – though this should not obscure the tactical and strategic considerations of such acts. All this has contributed to the development of territorial attachments by sub-groups within the Sinhalese and Tamil communities. The extreme manifestations of ethnic territoriality among Jaffna Tamils is the call by some for a separate state of Tamil Eelam. Among some Sinhalese, ethnic territoriality is expressed as the claim to absolute control over the entire island through a unitary state which fuses and protects Sinhala language, Sinhalese–Buddhism and Sinhalese nationalism, refusing to accord equal status to non-Sinhalese citizens. Yet, ethnic territoriality need not be expressed in such extreme terms; calls for a genuine devolution of power and the creation of a federal system of government are moderate variations of this territorial vision of group identity and rights. While there is a spatial dimension to intergroup relations, there are further geographical distinctions to be drawn within Sinhalese and Tamil communities. Sinhalese intra-group patterns The Sinhalese are divided culturally and geographically into the upcountry (Kandyan) Sinhalese and the low-country Sinhalese24 (roughly 40 per cent and 60 per cent of the Sinhalese population, respectively). Differences between these sub-groups are based on the historical fact that the Kandyan Kingdom successfully resisted colonial rule until 1815 when it was conquered by the British. The low-country Sinhalese, particularly in the coastal areas, were exposed to the various influences of the colonial powers (Roman Catholicism, Protestantism, Roman Dutch law, English medium-education) from about 1500 on, while the Kandyan Sinhalese continued to resist colonial control even after 1815 – staging two rebellions against the British in 1817/18 and 1848. The Kandyan Sinhalese maintained their separateness through limited intermarriage with low-country Sinhalese and the maintenance of their distinctive social organization and legal system (Wilson 1974: 38–43). Intra-group relations have fluctuated over time. In 1927, a number of


Kandyan Sinhalese political organizations reported to a constitutional commission that ‘[the Kandyan Sinhalese] never ceased to regard the low country Sinhalese as foreigners’ and that a federated Sri Lanka, in which the Kandyan Sinhalese provinces would constitute one of three units, would be the most appropriate system of government (quoted in Wilson 1974: 43). While differences continue to exist, a political accommodation has been worked out between up-country and lowcountry Sinhalese, and Kandyan Sinhalese are conspicuous on the political landscape. The political salience of this intra-group division is discussed in the following chapter. As discussed below, caste differences correspond to the regional distinctions between these two Sinhalese sub-groups and exercise an even greater impact on the separation of these sub-groups. Tamil intra-group patterns Examination of the intra-group divisions of the Tamil community is facilitated by a geographical correspondence to major social, political, and economic differences between groups. Accordingly, it is possible to identify the following principal sub-groups: Wet Zone Tamils, Jaffna or Northern Tamils, East Coast Tamils, and Up Country Tamils. Not surprisingly, within each of these groups there are further subdivisions and stratification (for example, caste, economic and social class, and political allegiance). While each of these subsequent axes of differentiation is politically salient, the geographical designation is accorded primacy in this thesis because in Sri Lanka, where one lives is central to the understanding of who one is. As Jonathan Spencer puts it: ‘identity is a matter of belonging to place as much as having a history’ (1990: 10). Members within these Tamil sub-groups recognize a status hierarchy which conditions intra-group relations: Colombo Tamils and Jaffna Tamils place themselves on top, East Coast Tamils are located in the middle, and the Up Country Tamils are cast to the bottom.25 This division is further underscored by caste differences (see below) and the distrust of many East Coast Tamils of the economic domination by Jaffna Tamils (Obeyesekere 1984). The impact of the cultural differences and natural antipathies between the East Coast Tamils and Jaffna Tamils – that once may have encouraged each to pursue a separate political agenda – have been reduced by the oppressive military actions of the government which treated all Tamils as ‘Tigers,’ sympathizers, or potential supporters. This common experience as victims has helped to push the two groups closer together. This appears to have encouraged

40 Introduction

some East Coast Tamils to be more receptive to the most powerful Tamil paramilitary, the LTTE, despite the Jaffna Tamil complexion of the organization,26 though as noted below in the discussion of paramilitary and regional allegiances, there are limits to this intra-group solidarity. The sensitive demographic balance on the East Coast is a crucial difference between the Northern and Eastern Provinces, a difference which has been reflected in the dynamics of violence and the displacement of persons. While the Northern Tamils clearly constitute the majority group in the Northern Province, there is no demographic advantage of Tamils in the Eastern Province where Tamil-speaking Muslims find themselves pivotally placed in a precarious ethnic balance (see Table 3.1). There, the Muslims constitute a group able to substantially help or hinder either Tamil or Sinhalese efforts to capture political and military Table 3.1 (1981)

Demographic composition of the Northern and Eastern Provinces

Eastern Province District Trincomalee Batticaloa Ampare Total % of Eastern Province Population

Sinhalese

SL Tamil

P. Tamil

Muslim

Other

33.6 3.2 37.6 24.92

33.8 70.8 20.1 40.90

5.6 1.2 0.4 1.23

29.0 24.0 41.6 32.28

1.0 0.8 0.3 0.66

Sinhalese

SL Tamil

P. Tamil

Muslim

Other

0.6 5.1 16.8 8.1 2.98

95.3 76.0 56.9 50.6 86.35

2.4 13.9 19.4 13.2 5.71

1.6 4.9 6.9 26.6 4.73

Sinhalese

SL Tamil

P. Tamil

Muslim

Other

5.6

7.1

0.7

Northern Province District Jaffna Mullaitivu Vavuniya Mannar Total % of Northern Province Population

0.1 0.1 .02 1.5 0.22

National

74.0 Source: Manor 1984 and CRD 1984: 21.

12.6


control. The Muslim vote is able to tilt the balance in at least 15 electorates in the country – a fact Muslims have regularly used to their advantage through strategic voting (de Silva 1986: 233). Consequently, they have been variously wooed and attacked by militaries and paramilitaries on both sides. Until recently, relations between East Coast Tamils and Muslims fluctuated between cooperation and conflict depending on the political context. Although in the past, alliances have been struck between the Muslims and Tamil politicians claiming to represent ‘the Tamil-speaking people of Sri Lanka,’ these have been tactical arrangements on the part of the Muslims who ‘have no great emotional commitment to the Tamil language’ (de Silva 1988: 203). Despite a common language, the two groups are decisively separated on the basis of religion (de Silva 1986: 228; de Silva 1988). As important, Muslim suspicions of Tamil political leadership are maintained by popular memories of the 1915 anti-Muslim riots which were justified at the time by the Tamil leader, Sir Ponnambalam Ramanathan. Surprisingly, given past history, there has been sufficient common ground to have encouraged a few of Muslims to join the LTTE (Ali 1993; UTHR 1991: 53; Gunaratna 1987: 21; personal interviews). This does not mean that the East Coast Muslim and Tamil political objectives are one and the same; but it does illustrate the variable and nuanced relationship between the two groups. The overall pattern of relations has been one in which Muslims have supported the non-Tamil political parties.27 Ali (1993: 6) observes: At times of momentous decisions in the national legislature, such as when the Ceylon Citizenship Bill, which disenfranchised hundreds of thousands of Plantation Tamil workers in the island; the Sinhala Only Act, which deprived the Tamil language of equal status in the country; and the University Entrance Examination Standardization Scheme, which removed the competitive edge hitherto enjoyed by the Tamil Colleges in the North; were introduced, the Muslim leaders rendered total support to the government of the time. In return for this support, the government of the day bestowed upon the Muslim community certain favours and privileges which, in the eyes of the Tamil politicians, were deliberate attempts by the state to oppress the Tamil community. Since June 1990,28 the two communities have been violently separated as a result of the LTTE campaign to drive Muslims out of the North and East by terror. The motives underpinning this LTTE version

42 Introduction

of ‘ethnic cleansing’ are more fully examined in following chapters. In October 1990, the LTTE banished all Muslims living in the northern districts of Mannar, Mullaitivu, Kilinochchi and Jaffna with the order ‘leave or be killed.’ Some villages were given as little as two hours to vacate their homes (USCR 1991: 25–28). This process was repeated on the East Coast. The consequence of this strategic shift in relations between the LTTE and Muslims has been the massive exodus of Muslims from the East Coast. It is estimated that approximately 60,000 ‘evicted’ Muslims from these northern districts fled south seeking refuge. In July 1991, it was estimated that 40,000 Muslims were living in some 60 camps in Puttalam District. The Muslim Refugee Rehabilitation Organization estimated that at that time there were more than 100,000 displaced Muslims in Sri Lanka (Sri Lanka Monitor, July 1991; USCR 1991: 25–28). By September 1992, the Island International (p. 4) reported that some 150,000 Muslims had been displaced from the North. Even as a formal ceasefire agreement was signed in February 2002, and peace talks commenced shortly thereafter, the central concern of Muslim leaders in the North and North East was that their human rights would be bartered away by the government in negotiations with the LTTE for a post-conflict settlement. Or, more diplomatically, that there will be insufficient safeguards to ensure the protection and promotion of their human rights. Both Muslims and Tamils pointed to the apparent unwillingness of government forces to respond to the increase in extortion, kidnapping, and theft in the North and North East during the period of parallel unilateral ceasefires (1 December–2 February), affecting both Tamil and Muslim communities. 29 The Sri Lankan government has always been aware that a coalition of Muslims and Tamils on the East Coast could effectively challenge Colombo’s authority. Thus, the government has employed a twopronged strategy to lessen this possibility. First, it has attempted to alter the demographic composition of the East Coast by increasing Sinhalese settlement in, and control over, areas claimed by separatists to have been traditionally inhabited by Tamils. This so-called ‘West Bank scheme’ envisions the resettlement of thousands of armed Sinhalese ‘settlers’ trained in self-defence on government-owned land in this area.30 By 1985, over 50,000 Sinhalese had already been resettled on traditional Tamil land (including ex-convicts, retired military personnel and families displaced by the massive Mahavelli water project). Unable to affect the government’s colonization policy, some Tamil paramilitaries responded with terror and violence against these


settlements – including the brutal massacre of whole communities.31 The LTTE has also used massacres as punitive measures. Thus, for example on September 18, apparently in retaliation for the deaths of civilians killed in air force strikes on Puthukudiyiruppu three days earlier, suspected LTTE members hacked to death some 48 Sinhalese villagers and shot six others in attacks on three villages in eastern Sri Lanka. The killings sparked large anti-LTTE demonstrations in Colombo (Human Rights Watch 2000). In the East, the unhealed wounds of the Kattankudi Mosque Massacres on 8 March 1990 – in which 103 people were murdered – sustain deep levels of distrust between the Muslim and Tamil communities in the District. Massacres in Eravur (121 killed) as well as in border villages are very much a point of reference in the Muslim community’s relations with the LTTE.32 The second part of Colombo’s strategy has been to encourage the separate articulation of Muslim and Tamil interests and identities – a move which elicited the full participation of Muslim sub-group mobilizers. Education policy was conveniently pressed into service to allow Muslim children to study in ‘Muslim schools’ (up until 1974), which helped to emphasize religious difference while de-emphasizing the shared Tamil language. The establishment and expansion of these schools, it must be emphasized, contradicted the principle of nonsectarian state education which has been the declared policy of all governments since 1960 (de Silva 1988: 211). The Sri Lankan government has created and armed a Muslim civilian militia (‘the Muslim Home Guard’) ostensibly for community protection. The Home Guards are village militias armed by the Government of Sri Lanka ostensibly to protect communities in border areas from LTTE attack. This was the brainchild of the late Lalith Athulatmudali when he was the Minister of International Security in the mid-1980s which aggravated the militarization of the conflict. There have been instances where the Home Guards fled from LTTE attacks on their villages, but returned from the jungles to massacre neighbouring Tamil villages. According to the Human Rights Watch World Report 2000: ‘The security forces’ use of home guards and armed ex-militant Tamil groups as auxiliary units to aid in military operations continued to draw criticism from human rights defenders. These groups have engaged in illegal detention, murder, abduction, extortion, assault, torture, forced conscription, and forced eviction.’33 As far as selfprotection is concerned, the Muslim Home Guards appear most often to be involved in retaliatory massacres of Tamil children, women and men following LTTE atrocities than in the guarding of Muslim

44 Introduction

communities (e.g. Pieris and Marecek 1992; Abeyesekere 1992). The consequence of this government policy has been the provision of the material means to pit Muslim groups against Tamil groups militarily. The victims, by and large, are innocent civilians caught in the middle. This is graphically evident in the use of massacres by all combatant groups in Sri Lanka. Government policies have served to harden and antagonize relations between East Coast Tamil and Muslim communities. The provision of resources particularly in the form of arms and separate education contributed to the mobilization of sub-group identity on the basis of religion while discouraging possible unification on the basis of a common language. Schisms have also appeared within the Muslim community. The Sri Lanka Muslim Congress (SLMC), the first explicitly Muslim political party, was formed in the early 1990s to represent Muslim interests. However, it has been bogged down in a community-wide debate over who or what constitutes the ‘Muslim voice.’ Through out 2002, antiCeasefire elements within parliament – especially from within the President’s Office – actively and specifically played on the insecurities of Muslims in the East and stoked intra-Muslim tensions in an effort to stymie the peace process. (See Chapter 9). The tactical alliances between leaders of different Muslim factions and the Government and LTTE galvanize coalitions of convenience designed ultimately to win control within respective communities. Muslims in Sinhalese-speaking areas of the country are reported to be losing their ability to speak Tamil to their co-religionists on the East Coast, thus suggesting the beginnings of a linguistic as well as a regional divide within the Muslim community (Ali 1992: 6). The Up Country Tamils (concentrated in the Hill Country, especially Nuwara Eliya District)34 constitute another significant Tamil sub-group. The social distance between this and the other Tamil sub-groups is illustrated by a project initiated in the 1970s by the late Bishop Leo Nanayakkara to educate Jaffna Tamils about the poverty-stricken plight of Up Country Tamils. He was struck by two observations: 1.) the Jaffna Tamils’ lack of awareness concerning the conditions on the tea plantations; and 2.) the low regard in which Up Country Tamils were held by Jaffna Tamils.35 This latter point is frequently reflected in the Jaffna Tamil use of ‘low Tamil’ when addressing Up Country Tamils. A more telling indication of the differences between these sub-groups is the tendency of Jaffna Tamils and East Coast Tamils to date the beginning of the conflict from either 1956 when the Sinhala-only language policies were implemented or the early 1970s when the edu-


cation policies and a republican constitution were introduced. Not one Jaffna Tamil or East Coast Tamil interviewed in the course of field work mentioned the historical referent which dominated the political landscape of the Up Country Tamils: the three citizenship acts of 1948/49 which deprived them of the vote they had exercised under the British. This suggests the insularity and social differences between the Tamil sub-groups as well as their very different experiences of the conflict.

Axes of identity In addition to the significant geographical elements to sub-group selfidentification in the inter-group and intra-group arenas, there are other salient axes of identity. Myth, mobilization, and identity The early history of Sri Lanka is shrouded in myth. Sinhalese Buddhist texts – the Dipavamsa (4th century A.D.), the Mahavamsa (late 5th century), and the Thupavamsa (13th century) – chronicle a history of kings and gods, myths and magic (Smith 1978; Kemper 1991; Tambiah 1992). It is a mythic history in which the Buddha visits Sri Lanka and establishes the island as the repository of his teaching, thus ideologically fusing Sinhalese religion (Buddhism), nationalism and ethnic identity. According to tradition, the island was first settled by a north Indian ‘prince,’ Vijaya, and his followers, supposedly descended from a lion (‘Sinha’).36 Despite the fantastic mythological details and lack of independent evidence, the story has not diminished in its ‘potential as a vehicle for moral feeling and ideological assertion’ (Kemper 1991: 111). The transformation of the Vijaya story from myth into history gained momentum during the Buddhist revival of the late 19th century when Buddhist monks and laity mobilized against the British in an effort to re-assert the social, political and economic control they had exercised in the pre-Colonial period – particularly in the realms of land tenure and education. By the early 20th century, politicized Buddhism had become a solid basis from which to resist British rule. Although the mobilization was initiated in reaction to British control, it was harnessed in the political and economic competition with minorities (Tamil, Muslim, Burgher, Chettiar traders), some of whom had achieved considerable professional and material success under colonial rule. A pro-Buddhist group identity was fostered through the

46 Introduction

cultivation of chauvinistic sentiments toward non-Buddhists and nonSinhalese. For example, Anagarika Dharmapala, a literary leader of the Buddhist revival, referred to minority communities as ‘infidels of degraded race’ and propagated the myth of the Sinhalese as members of a pure, racially superior, Aryan race. This was employed by some Sinhalese to apply a biological gloss to their territorial claims over the entire island. Colonial historiography, including the early work of the distinguished South Asianist Max Muller, reinforced such racialist claims, and the Tamils were similarly miscast as members of the ‘Dravidian race.’37 The political assertion of Buddhist identity illustrates the instrumental mobilization of a reconstructed past rather than a return to the some absolute Buddhist ideal. What becomes evident is the reactive impetus underpinning the Buddhist mobilization, i.e. much (but by no means all) of the momentum for the ‘Buddhist revival’ was generated as a response to challenges posed by the British. Because British colonial employment and education policies were premised on religious identities, religion became the principal currency of the political marketplace. As a result, even employment, education, and economic grievances came to be expressed in the colonially defined terms of religion. Instead of mobilizing multi-religion coalitions to challenge colonial policies on education and so on, groups began to mobilize on the basis of religion in order to extract concessions for particularistic groups, often at the expense of other groups. While the process by which group identity is fostered in reaction to out-group challenges continues in the post-colonial period, the boundaries of group self-definition have changed from religion to ethnicity. The reactive dynamic of group formation is also reflected within the Tamil community in what Pfaffenberger (1990) has termed the political construction of defensive nationalism. Not only is an overarching conception of ‘Tamil-ness’ developed in counter-position to ‘Sinhaleseness,’ but it is also evident at the intra-group level where, for example, the definition, articulation and politicization of Up Country Tamil identity is as much a reaction to the militant nationalism of the Northern Tamils as it is a reaction to the Sinhalese (personal interviews, Kandy, May–July 1992). The political manipulation of Sinhalese myths and symbols has become a common means of mobilizing support and legitimating actions of competing groups within the Sinhalese community. For example, the fusion of political and religious ideology is clearly illustrated in a ceremony presided over by Buddhist clergy in which two of


the most powerful ministers of the Jayewardene Cabinet pledged Buddhist oaths ‘to save Sri Lanka and its people from terrorism’ and laid wreaths at the statue of King Dutta Gamani, a religio-mythic hero of the Sinhalese chronicles heralded for expelling Tamils from Sri Lanka (Far Eastern Economic Review, 21 February 1985: 37). Similarly, the efforts of former president J.R. Jayewardene to update the Mahavamsa was intended to locate himself and his political party (the United National Party) among the heros, kings and gods of the epics and thereby establish his Sinhalese Buddhist nationalist credentials for posterity (Kemper 1991). A fundamentalist acceptance of this mythic version of history is prevalent in the both mainstream media and staterun education system (Perera 1993; Siriwardene 1984). Separatist elements within the Jaffna Tamil community similarly rally historicist arguments in support of their claims (e.g. Ponnambalam 1983). It is argued that because an independent Tamil Kingdom existed in Jaffna when the first colonial power (Portugal) arrived in 1505, the Northern Tamils should have received their independence when Sri Lanka was granted statehood in 1948. Because the island remained undivided at independence, Ponnambalam argues that the Jaffna Tamils have been dominated by the Sinhalese colonial state. It is important to add that there are competing voices which have argued not for separation but for increased Tamil autonomy within the Sri Lankan state. Yet calls for devolution have not received a serious consideration, and a federal system of government has been vigorously resisted by Sinhalese politicians who fear the loss of control might be the first step towards separation. Although some Jaffna Tamils employ historicist arguments, there is no Tamil equivalent to the Buddhist Chronicles which can be rallied for the maintenance and politicization of group identity. HellmannRajanayagam argues that ‘one reason why Tamil nationalist historiography is less strident than Sinhalese nationalist historiography is that Tamil identity can be sought in a much wider range of cultural resources, Indian as well as Sri Lankan’ (see Spencer 1990: 7). This argument suggests that the form or shape of group identity is related to the availability, mobilization and perhaps manipulation of ‘cultural resources.’ A catalyzing element of conflict escalation in Sri Lanka – and every other case of ethnicized violence – is: (1) the competing representations of the past, (2) understandings of the present, and (3) the linkages between the two. Under conditions of ethnic tension any characterization of events may viewed as ‘political’ by virtue of the

48 Introduction

context in which it is voiced, and heard. The language of representation, as well as the representation itself, may be accused of containing political ‘bias.’ Rather than obscuring an understanding of the conflict, the multiplicity of competing and often antithetical histories offers insights by pointing to the structures and processes by which individuals and groups are mobilized. An examination of the nuances of this process is developed in subsequent chapters. Caste It is paradoxical that despite the importance of caste in structuring social behavior, it is extremely difficult to obtain useful statistical or qualitative data. Caste is rarely discussed openly in conversation but its impact is evident at all levels and in all forms of social interaction (such as politics, education, employment).38 While caste distinctions differ between the Sinhalese and Tamil communities, ‘in both cases, caste is used to define a corporate group formed through familial descent and [endogamous] marriage’ (Matthews n.d.(a): 1). Caste is a factor contributing to the intransigence of both the Sinhalese extremist JVP and the Tamil paramilitaries which are disproportionately low caste. Calls for the dismantling of the caste system in the ‘revolutionary programs’ of both Tamil and Sinhalese militant organizations (albeit often in expediently elliptical terms) mean that failure to achieve militant ethnic sub-group goals entails losses in the intra-group as well as the inter-group arena. Thus, some of the warriors are fighting a two-front war, against casteism in the intragroup arena and against the ‘opposing’ ethnic group in the intergroup arena. Sinhalese Statistics have not been collected on caste since the 19th century, but it is estimated that the current Sinhalese caste composition is (in ranked descending hierarchical order): Goyigama (‘wet rice farmers’) 60 per cent; Karava (‘fishermen’) 10 per cent; Salagama (‘cinnamon peelers’) 7 per cent; Batgama (‘former Palaquin Carriers’) and Wahumpura (‘jaggery or raw sugar makers’) 11 per cent; Durava (‘toddy tappers’) 4 per cent; others 8 per cent.39 As noted above, a further status distinction may be identified between up-country and the lowcountry Sinhalese. Caste is a feature in the mobilization of support by all major Sinhalese actors in Sri Lanka: political parties;40 the JVP;41 the Sangha (Buddhist clergy).42 Caste cross-cuts both Buddhist and Christian communities.43


Historically, there have been tensions between the Goyigama and the castes immediately below it. As the Karava, Salagama, and Durava began to convert to Roman Catholicism in the 16th century, they also broadened their employment opportunities and gained economic independence from the aristocratic branches of the Goyigama. Also the Karava, Salagama, and Durava tended to live in the coastal areas, whereas the Goyigama were more often inland. Such geographical separation allowed the Karava, Salagama, and Durava to assume the foremost social positions in coastal areas. The Karava, Salagama, and Durava did not accept their status as being ‘inferior’ to ‘farmers’ – only to ruling castes, i.e. nobles. The idea of having ‘noble’ farmers did not gain currency among the coastal castes. The incongruence between their economic power and subordinate social position soon became apparent. This has underpinned a ‘marked escalation in [inter-caste] conflict’ (Roberts 1982: 133) which sometimes spills into the parliamentary and extra-parliamentary political arena. The term ‘upwardly mobile castes’ has been used to identify those who are low in the traditional caste hierarchy, but who attempt to compensate through economic advancement.44 This suggests a tension between two competing forms of social stratification: caste and economic class. Caste advancement is far slower than economic advancement – though some members of lower middle castes have been known to appropriate the symbolic behavior of the higher castes, for example, the adoption of higher caste dietary customs, marriage and birth rituals, and Goyigama names. In this way, individuals have attempted to transcend their original caste ranking. This is more acceptable in Sri Lanka than in India if the claimed change in status is further buttressed by educational and income advantages. Whereas caste advancement is an intra-group process guided by an intra-group social stratification system, economic stratification is a ranking system which cross-cuts inter-ethnic groupings. While casteadvancing activities may only be undertaken within the intra-group arena, economically advancing activities may be fruitfully undertaken in the inter-group arena, for example, through attacks on Tamils or through the development of mutually beneficial joint business enterprises. A strategy of attacking low caste and economically depressed groups may be further fuelled by the representation of the Tamils as economically advantaged as a group (excluding the Up Country Tamils). This point of intersection between the inter- and intra-group arenas in Sri Lanka is examined in subsequent chapters, but it is useful to note that one of the distinctive features of the 1983 riots is that for

50 Introduction

the first time, violence was directed against the ‘Tamils as a group.’ Previously, middle and upper class Tamils had been relatively insulated from mob violence (though occasionally their shops would be burned down if they were located in a lower class neighborhood). Thus, the year 1983 marks a turning point in the delineation of the axes of conflict. Any protective effect of class was eliminated when the Tamil community came to be externally defined wholly in terms of ethnicity regardless of class or any other forms of differentiation. The result was the functional homogenization of ‘the other’ by Sinhalese mobs and, increasingly, by Sinhalese mobilizers. Low caste was effectively used in the mobilization activities of the late Cyril Mathew, a United National Party (UNP) cabinet minister and representative for the interests of his Wahumpura caste, particularly in the economically depressed constituency of Borella (Colombo 8). It was from these areas that government lorries transported Sinhalese thugs to murder and loot throughout the city. The economically depressed Sinhalese lower castes constituted a resource that was usefully mobilized by Mathew and other UNP politicians. (Tambiah 1996: 82–100; Hyndman 1988; Hoole et al. 1992; Tambiah 1986). Tamil Caste is as significant in the Tamil intra-group arena as it in the Sinhalese intra-group arena. In the late 1960s there was considerable low-caste agitation against the dominance and discrimination by the high-ranking Vellalars, who are equivalent to Sinhalese Goyigamas (Shanmugathasan 1993; Pfaffenberger 1990). Caste differences underlie and maintain the separate regional identities of the Jaffna Tamils, the East Coast Tamils and the Up Country Tamils. It is also reflected in the composition and behavior of the Tamil paramilitaries. It is important to note that the Jaffna Tamils are a much larger community with a distinctly different caste structure than the East Coast Tamils. The Jaffna Tamils’ matrilineal structures are more closely related to the system in Kerala, whereas the East Coast Tamil and Up Country Tamil caste systems parallel that of Tamil Nadu. This regional caste matrix is rarely considered in the examination of the Sri Lankan conflict. Consequently, a prevalent assumption has been that there is ‘natural’ affinity between the Tamils in Sri Lanka and the 50 million Tamils of Southern India which served to sustain a continuous stream of support from South India to Jaffna. However, even before the 1991 assassination of Rajiv Gandhi – which turned many South Indian Tamils against the Sri Lankan Tamil paramilitaries and increased pres-


sure within India to repatriate the more than 150,000 Tamil refugees in Tamil Nadu – this relationship between South Indian and Sri Lankan Tamils was variable. The level and nature of support by South Indian public and the State Government of Tamil Nadu fluctuated over time, from the high levels following the anti-Tamil riots of 1983 to its low during the Kumaratunge Regime (1994–2001). Of the many Jaffna Tamil castes, the most important ones are, in ranked hierarchical order: Priest: Brahmins (relatively few)45 ‘Clean Castes’: Vellalar (‘farmer’) 50 per cent;46 Koviya (ritual ‘cooks’ to the Vellalar) r 7 per cent; Karaiyar or Karava (‘fishermen’) 10 per cent ‘Service Castes’: Taccar (‘carpenters’); Kaikular (‘weaver’); Vannar (‘washermen’) around 15 per cent ‘Untouchables’: Pallar 9 per cent; Nalavar 9 per cent Despite paramilitary claims to the contrary, caste is a factor conditioning paramilitary politics. It is popularly held that ‘some if not all the militant groups had been led by youth of the Vellalar caste or Karaiyar caste: thus the LTTE are said to be led by the Karaiyars of Valvettithurai, and that the Tamil Eelam Liberation Organization (TELO), wiped out by the former, was a Vellalah organization’ (Tambiah 1993: 11). In other words, despite the multi-caste mobilization and composition of paramilitary organizations, radical politics coexist with social orthodoxy in matters of caste. There have instances where caste hierarchy has been in conflict with the paramilitary activities and the vicissitudes of life in the war zone of the North and East. For example, there are cases where the paramilitaries have not challenged the traditional restrictions and attitudes on the access of lower castes to wells under high-caste control or ownership (ibid.: 11). The following incidents recounted by Tambiah illustrate that both military and paramilitary activities in the North and East are undertaken within the framework of caste which both affects and is affected by those activities. In an ironic twist, certain actions taken by the occupying armies – Sri Lankan and Indian – have actually loosened up caste avoidances and prejudices. For example, when the two armies have frequently declared a curfew, they have ordered civilians to gather in specified Hindu temples allegedly for their collective protection. This means

52 Introduction

that all castes will have to enter and occupy the temple, and frequently do so. But there are sticky episodes. On one occasion a temple nominated for occupation was the Karaveddy Pillayar Temple, whose annual festival was to take place in a month’s time. The trustees of the temple closed it giving admission to no one. Young men of the area who could not take refuge were arrested. A more subtle interplay of caste aversion was witnessed in the Vellani Pillayar Temple, which was made available for the occupation of refugees. Apparently the Pallars (of low caste) circulated the joke that while ‘our boys are inside the temple, their (high caste) boys are outside’. This inversion of traditional inclusion–exclusion was brought about by the high caste refugees keeping away from their social inferiors (Tambiah 1993: 11). Until recently, the Vellalars maintained strict control of Jaffna society. More broadly, they have assumed the role of defining Tamil civilization and what it means to be Tamil in Sri Lanka. This privileged social position is reinforced economically; the Vellalar are the major landowners in the 436 square miles of the Jaffna peninsula and they have traditionally monopolized both cultural leadership and the professional and political opportunities (Matthews n.d. (a): 5). However, this Vellalar dominance is increasingly challenged by the non-Vellalar castes (Pfaffenberger 1990), particular as Northern Tamils flee the violence of both the Sri Lankan security forces and the LTTE. The material basis of Vellalar control is further reduced by the LTTE practice of forcing all fleeing civilians to surrender their property on departure, as well as pay an exorbitant ‘exit tax.’47 As Hellman-Rajanayagam points out: ‘the conservative Jaffna Tamils will not gladly suffer mere teenagers with questionable political ideas and of questionable caste lording it over them’ (1989). It is thus unclear whether the LTTE will be able to gain widespread Vellalar support when one of its stated objectives is the dismantling the caste system. This has also been a factor dissuading Tamil political parties from incorporating the dismantling of the caste system into their political platform. In electoral politics, it was recognized that to do so would alienate the socially conservative high caste vote which is the majority of the Tamil community (Shanmugathasan 1993: 22). LTTE activities on the East Coast are undertaken within a very different caste context. There, the Vellalar are a small and relatively powerless minority. The dominant caste is the Mukkuvar (related to the Jaffna Karava). This has implications for the LTTE for it suggests that the


support of non-Vellalar castes on the East Coast is necessary to off-set the Vellalar dominance in the North. However, the variation in the caste systems and, importantly, the regional attachments of many East Coast Tamils dilute some LTTE efforts in the East. This is not to say that the LTTE is not supported in the East, but the extent of support is difficult to gauge and any support is as likely to be based on the need for protection from government forces as well as on support for the LTTE. According to Ivan, the ‘struggle’ of the Northern Tamil militants is ‘not only a struggle for national freedom, but also a struggle against Vellalar domination. … The struggle of the militant Tamil youth of the north is not purely an ethnic struggle, a class struggle, or a caste struggle. It is really a peculiar combination of all these three factors’ (Ivan 1993: 27). Caste, thus, exercises an influence in both the intra-group and inter-group arenas. Because of the relative invisibility of Up-Country Tamils in mainstream political discussions, it is perhaps useful to underscore the importance of bringing them into the political picture. In the mid to late 1980s there were tentative efforts by Northern Tamil militant groups to enlist Up-Country Tamil youths into the broader Eelam struggle. Indeed, P. Chandresekeran, a significant political activist in the plantation sector, contested the 1989 elections under the People’s Liberation Organization of Tamil Eelam (PLOTE) banner. Such contacts have been relatively limited and results have been minimal. Nonetheless, Government fear of such infiltration has led to mass round-ups arrests of Up-Country Tamils by Police in the Hill Country. The hitherto failure of Northern militant efforts to expand the Eelam struggle into the Plantation areas suggests the fundamental social tension between the Northern Tamil and Plantation Tamil political projects. Ironically, the more severe the crackdown on the alleged political activities of the Up-Country Tamils, the more likely it may be that they will respond to Northern militant overtures, or fashion their own physical force response. Indeed, despite the apparent incongruity of the arrangement, a number of Up-Country Tamils teamed up with the Marxist Sinhalese JVP during the terror of the 1987–89 Insurrection. Clearly, political convenience may go a considerable distance towards blending the ostensibly incompatible. Changes within the Up-Country Tamil community which are altering the politics of identity and confrontation include the increases in literacy, education, and political awareness. These changes are reflected in the shifting political demands of Up-Country Tamils from issues of

54 Introduction

citizenship and labour to issues of political representation and governance. The emergence in the early 1990s of a Up-Country Tamil political movement called the Up-Country Peoples’ Front (UCPF), led by P. Chandrasekeran, is a development that presages the political mobilization that is under way. The significance of the UCPF lies less in the organization per se, than in the internal social changes noted above that have made this development possible. As one researcher noted, ‘in the course of my research, I have met ten Chandrasekerans.’ One community worker predicted that it will be the next generation of Up-Country Tamil youths – those who are educated and unemployed – that will be the basis of the politicization of the Up-Country Tamils interests. And indeed, in the mass violence that spilled through the up-country in October 2000, Chandrasekeran was again arrested for inciting mobs against government officials. For Up-Country Tamils – no less than for Northern and Eastern Tamils, or for unemployed Sinhalese youth in the South – the likelihood of using ‘extraparliamentary’ or ‘extra-legal’ means of pursuing their political interests is directly related to the legitimacy and efficacy of existing legal and political channels. This does not bode well for the future of the Up-Country Tamils at a time of increasing mobilization, politicization, and confrontation. Political parties Political parties in Sri Lanka are central actors in the mobilization of the identities and resources which sustain sub-groups. They occupy an interstitial position between state and society and are therefore uniquely placed to cultivate and exploit resources from both arenas. They both reflect and shape (within limits) the sentiments and interests of sub-groups within their ethnic-cum-electoral constituencies. Political parties and particular politicians have been central actors in generating violence within and between groups. It is widely believed that a UNP Minister was active in inciting and coordinating the 1983 anti-Tamil riots (see Chapter 7). The United National Party (UNP) government of the late President Premadasa was implicated in the organizing of a vigilante force called the ‘Black Cats’ which was allegedly responsible for the murder of hundreds, perhaps thousands, of political opponents including members of the opposition party, the Sri Lanka Freedom Party (SLFP).48 The UNP support for, and the SLFP rejection of, the 1987 India–Sri Lanka Accord was a fundamental political disagreement between the two principal Sinhalese parties with direct violent consequences in the inter-group arena (such as the intro-


duction of Indian troops into the battle against Tamil paramilitaries in the north and east) and in the intra-group arena (violent attacks against the UNP government by Sinhalese sub-groups). That elections are still a regular feature of the Sri Lankan political landscape (in those areas under central government control) further highlights the relevance and potential influence of political parties. On the other side of the ethnic divide, the major Tamil political party since 1976, the TULF, has oscillated between a constitutional program seeking regional autonomy and a revolutionary program seeking the creation of an independent state of Tamil Eelam. Many of the active Tamil paramilitary soldiers today were politicized and radicalized by the Eelamist political rhetoric of the 1970s but were subsequently disappointed with the lack of results. Consequently many adopted violence as the means of achieving their political aspirations.49 Since 1972, Sri Lanka has officially been proclaimed (in its constitution) as a socialist democracy. Although significant portions of the island in the North and East are beyond the control of the central government. Thus, areas are, or have been, under the direct control of the Sri Lankan Military, as well as the LTTE. As Matthews puts it: Sri Lanka ‘in some sense sustains a political democracy. It can be argued that it is a quasi-democracy or a democracy under seige but nonetheless a system of multiparty elections remains in place at four levels (presidential, parliamentary, provincial and local)’ (Matthews 1992: 215). The political system in Sri Lanka is characterized by the regular splintering of political parties and interest groups.50 This is particularly evident within the Sinhalese community which is riddled with a shifting constellation of organized interests. The major opposition party, the Sri Lanka Freedom Party (SLFP) was formed in 1951 by S.W.R.D. Bandaranaike when he abandoned the ruling United National Party. Although the two parties may be distinguished on the basis of ideological predisposition (i.e. the left-leaning SLFP versus and the rightleaning UNP), the principal cause for Bandaranaike’s departure was not ideological; rather, it was his disgruntlement with being overlooked for the leadership of the party. The politics of personality and patronage were underscored as a feature of political life in Sri Lanka when the first prime ministership passed from father to son, from D.S. to Dudley Senanayake in 1952.51 S.W.R.D. Bandaranaike’s sweep to power in 1956 marked the establishment of the first non-UNP government. The success of the SLFP lay in its ability to mobilize popular support by politicizing and harnessing religion and language, that is, Buddhism and Sinhala. Bandaranaike was a crucial

56 Introduction

catalyst in both the politicization of ethnicity and the ethnicization of politics. The shrewdness and expedience of his political program are suggested by the biographical ‘facts’ of S.W.R.D. Bandaranaike: born Solomon West Ridgeway Dias; English educated; raised Anglican; a poor Sinhala speaker and a late convert to Buddhism. As is often the case, the sectarianism unleashed by Bandaranaike boomeranged in September 1959 when, as Prime Minister, he was assassinated by a Buddhist monk with connections to the same militant Buddhist sects which he had cultivated and relied upon in his successful bid for the prime ministership.52 In summary, the political life and death of S.W.R.D. Bandaranaike highlight some of the features of conflict in Sri Lanka: volatility, political manipulation of cultural symbols, identity mobilization, and the politicization of language and religion. The strategy of ‘playing the ethnic card’ quickly passed to other Sinhalese sub-groups as illustrated below in the successful intra-group challenge to the Bandaranaike–Chelvanayakam Pact (‘B–C Pact’) in 1957. Indeed, once ethnic identity was successfully mobilized by S.W.R.D. Bandaranaike, others sought to do the same in order to compete within the Sinhalese political arena. Once again, the intra-group dimensions of inter-group relations become evident. The fact that some form of electoral politics has operated in Sri Lanka since the early 1930s has had an impact on the patterns of conflict and social relations more generally (at both inter-and intragroup levels). In the past, the need to rally votes has constrained the political latitude of politicians. Thus, for example, in 1957 Prime Minister S.W.R.D. Bandaranaike was forced by Sinhalese intra-group pressure to renege on the B–C Pact, a major agreement struck with Tamil leaders which would have increased the autonomy of Tamil regions and, in retrospect, might have helped to pre-empt the subsequent spiral into ethnic violence.53 It is neither coincidental, nor a footnote in history, that a central figure in the mobilization of opposition to the B–C Pact was J.R. Jayewardene, then an opposition party MP and later prime minister and president from 1977 to 1988. His famous October 1957 march to Kandy to invoke the blessings of the devales (gods) in his anti-Pact campaign was calculated to win the support of segments within the Sinhalese community by first pandering to their fears of ‘encroaching’ Tamil power (especially economic and political) and then presenting the agreement as a contribution to such an encroachment. Especially noteworthy in Jayewardene’s attack of Bandaranaike is that he harnessed the Sinhalese chauvinist sentiment which Bandaranaike himself had so effectively mobilized in his successful bid for the prime ministership in 1956. The impact of past mobilization


of identities on conditioning the rules of the political game is considered in the final Chapter of this thesis. Following the assassination of S.W.R.D. Bandaranaike, the leadership of the SLFP was taken up by his widow Sirimavo Bandaranaike, who twice held the position of Prime Minister (1959–65 and 1970–77). The SLFP suffers from seemingly endless internal squabbling, factionalization, and feuding. Over the last few years, the main internal divide has been between the ‘Anura’ and the ‘Chandrika’ factions, named because for their associations respectively with the son and daughter of Mrs Bandaranaike. Each has jostled for position in the battle for the SLFP leadership held by their aging mother.54 In 1994, leadership of the Party (under the banner of the Peoples’ Alliance) was assumed by the daughter, Chandrika Bandaranaike Kumaratunge, while the brother decamped to the UNP to pursue his aspirations from the other side of the political divide. The SLFP vehemently opposed the 1987 India– Sri Lanka Peace Accord and during the 1987–90 period was suspected of ‘consorting with the JVP [Sinhalese extremists] in an effort to topple the UNP government.’55 The UNP and SLFP alternately in forming the Sri Lankan government until the landslide victory of the UNP under J.R. Jayewardene in 1977, when it won 85 per cent of the available seats. The massive UNP majority in parliament allowed Jayewardene to effect a number of farreaching political manoeuvres: he reduced the power of parliament and concentrated control in the presidency (Matthews 1992; Wilson 1980; Manor 1984);56 he stripped the SLFP leader of her civic rights and ‘expelled’ the only Tamil party from parliament in one of 17 constitutional amendments since 1978; and extended the life of the sitting parliament through a dubious referendum in December 1982, thus avoiding general elections for 11 years, i.e. 1977 until late 1988 – when Ranasinghe Premadasa, Jayewardene’s Prime Minister, assumed leadership of the UNP and narrowly won the presidential election. The UNP, like the SLFP, is racked by internal factions. The battle for leadership of the UNP following Jayewardene’s announcement not to seek re-election in the presidential elections of 1988 set in motion an internal power struggle which is still reverberating within the party. For example, the 1991 impeachment procedings against President Premadasa were launched by dissenting UNP factions allied with an array of non-UNP political parties. As Matthews accurately points out: The crisis was primarily the result of an internal UNP revolt over the style of presidential leadership. At the heart of the matter is con-

58 Introduction

frontation between the old elite of powerful families (walawe deshpalakas) and that of a new order which claims to transcend class distinctions and tradition. President Premadasa avers that this ‘class struggle’ is what caused the attempted coup. This incident was representative of a genuine dislike by certain key elements in the UNP by the way in which power had come to be exercised by the incumbent president. On the issue of ‘class struggle’ within the UNP leadership, it can be argued that a majority of UNP parliamentarians selected by President Premadasa in 1989 were from more commonplace or less-than-privileged backgrounds than those MPs from former UNP governments … with the result that many MPs and even cabinet ministers are less well-educated, less sophisticated and less ‘Westernized’ than their predecessors in previous parliaments (Matthews 1992: 219).57 The impeachment procedings would not have been possible without the support of dissenting members within the UNP. This included the backing of three UNP ministers (the late Lalith Athulathmudali, G.M. Premachandra, and Chandra Gankanda) and 44 other UNP members of Parliament. The politics behind the impeachment efforts reveal dissenting factions within both the UNP and the SLFP, e.g. members within the SLFP either refused to endorse the proposal (such as former Speaker Stanley Tillekeratne) or were not approached due to uncertainty concerning their disposition.58 Although the impeachment challenge was adroitly suppressed by Premadasa, one by-product of the incident was the creation of new political party, the Democratic United National Front (DUNF), by the parliamentary coup leaders, both of whom were former UNP cabinet Ministers and jilted presidential candidates (the late Lalith Athulathmudali, former Minister of National Security under UNP President Jayewardene and the late Gamini Dissanayake, former Minister of Lands, Land Development and Mahaweli Development). The dispute over control of the DUNF was ostensibly settled by agreement on a system where the leadership would alternate annually between Athulathmudali and Dissanayake. The assassination of Athulathmudali a few weeks after this arrangement was agreed upon set in motion the usual set of rumors and allegations connecting the internal DUNF struggle to the assassination. Within months of the assassination, Dissanayake was negotiating with the UNP about joining the ranks in order to form a ‘National Government to solve the National Issue’ when he was also assassinated. In early 2000, gunmen in police uniforms


killed the chief suspect in the killing Athulathmudali.59 Both murders remain unsolved. Aside from the politics of personality that catalyses the splintering of political groups, the late Dissanayake and the late Athulathmudali represent the Anglicized elite of the Jayewardene Old Guard.60 Initially, they represented a reaction against the late President Premadasa whose low caste and shrewd mobilization of the poor and marginalized (often through crude economic incentives) enabled him to build a strong supportive constituency.61 Premadasa’s populism and departure from the economically liberal UNP policies alienated members of the business community. If the UNP is seen to be abandoning its business-oriented platform and if the nationalization legacy of the SLFP weakens its credibility among the business community, then there is an important and powerful segment of society to be mobilized. It is this segment which the DUNF hoped – but ultimately failed – to capture (personal interviews with Athulathmudali, Colombo, May and June 1992). However, Premadasa’s introduction of populism into the mainstream political arena, has DUNF and the SLFP/ PA to ‘play the populist card’ in the broader mobilization of support as well. The Anglicized and elite character of the DUNF leadership, however, made it difficult to credibly or effectively employ populist tactics in the mobilization process. Interestingly, the election of Ranil Wickramasinghe as prime minister in 2001, appears to have been more successful in consolidating the support of both the business community and the large segment of poor and marginalized Sri Lankans with the help of bold symbolic acts within the first few months of power: the maintenance of parallel unilateral ceasefires leading to the joint ceasefire agreement signed on 22 February 2002; improved humanitarian cooperation such as easier movement of goods and people between cleared and uncleared areas; the opening of large sections of formerly blocked roads (most notably the A9 to Jaffna officially opened on 7 April 2002, and the main thoroughfares of Colombo opened immediately after his election); and his trip to Jaffna – the first by the Sri Lankan head of state since 1983.62 Until 1981, the political concerns of Tamils in the North and East were articulated principally through the Tamil United Liberation Front, a political party formed in 1976. The next year, the TULF won the parliamentary seat in every constituency in which there was a Tamil majority. Although distinct from the militant groups engaged in military activities, it shared the same objective; TULF MPs were elected to the Sri Lankan parliament on a political platform that favored the creation of a ‘secular, socialist state of Tamil Eelam.’ Significantly, this

60 Introduction

position was later moderated to one which would accept ‘any viable alternative.’63 TULF was marginalized in Sri Lankan politics by its expulsion from parliament in August 1983, when through the use of an extraordinary parliamentary procedure, the UNP-dominated government approved a motion of non-confidence in the TULF leader of the opposition, A. Amirthalingam (see Chapter 8: 1987 Indo-Sri Lanka Agreement). This was followed by the Sixth Amendment to the Constitution which required all MPs to take an oath to the ‘unitary constitution’ of Sri Lanka. TULF members of parliament were subsequently expelled when they refused to take an oath of loyalty under which they would have had to renounce even peaceful efforts towards separatism in the north (the Sixth Amendment). This expulsion alienated the Tamil community from the parliamentary process and consequently, political grievances came to be increasingly expressed through the actions of militant Tamil groups both in the inter-group arena (e.g. attacks on Sinhalese military personnel, Sinhalese settlers, and random Sinhalese victims of terrorist attacks) and in the intra-group arena (internecine feuding, the murder of dissenters, the assassination of Tamils working for the Sri Lankan government such as policemen, and assassination of moderate Tamil politicians including the mayor of Jaffna in 1975, the TULF leader Amirthalingam in 1989, and Dr. Neelan Tiruchelvum in 1999).64 The military and paramilitary allegiances The government security forces The escalation of inter-group ethnic conflict in Sri Lanka has simultaneously increased the size and narrowed the ethnic composition of the government security forces. The ethnicization of security forces has been facilitated by the decreased supply of Tamil recruits.65 The growing animosity of Tamil civilians who have suffered the brunt of military offensives and military abuses has all but eliminated the number of willing Tamil recruits into the Sri Lankan security forces; but it is also the result of Tamil paramilitary assassinations and intimidation of Tamil security personnel from 1977 onward.66 Between 1977 and June 1983, out of the 37 soldiers or policemen attacked by Tamil paramilitary organizations, 17 were Tamil (Meyer 1984: 143). By 1987 only 600 (2.4 per cent) of the 25,000 members of the national police force were Tamil (The Island, 18 August 1987). A Brigadier General claimed that ‘the Army would love to have Tamil soldiers,’ but members of pro-government Tamil paramilitaries have responded that they would not join until the LTTE is defeated


and elections have been held in the north and east of the island (personal interviews, Sri Lanka, May–July 1992). As noted in Chapter 2, although the non-LTTE Tamil paramilitaries will not join the Sri Lankan military, they have worked side-by-side in the war zones of the north and east since 1987. This is not really an alliance, however, and most military officers hint at friction between the regular forces and the ill-disciplined pro-government Tamil paramilitary ‘cowboys.’ On the east coast in Batticaloa, residents have reported human rights abuses committed by a squad of pro-government paramilitary soldiers from TELO under the leadership of a Sri Lankan Army officer. This is an example of how extremist elements within the government military may exploit intra-group tensions to conduct ‘inter’-group violence. The reverse side of the same process is evident in episodes when the Sri Lankan military rounds up a group of Tamil villagers and has masked TELO members identify those who allegedly are collaborating with the LTTE. This is an example of intragroup rivalries employing inter-group resources. In 1962 and 1966, factions within the army and police attempted two coups d’état (Horowitz 1980; Nyrop 1985: 446–9). The question of the loyalty of the military became a public issue when Rajiv Gandhi was assaulted by a member of Sri Lankan naval honor guards following the signing of the India–Sri Lanka Agreement on 19 July 1987. There is a variety of cliques, power blocks, and professional jealousies within the military. Although the military has not ventured into the formal arena of politics, there is considerable disgruntlement among officers at the highest levels for a number of reasons. Members of the military feel that a lack of political will by politicians is hampering the war effort, while at the same time, political incursions into the military domain are increasing. The resources allocated to the military are viewed by many officers as inadequate for the task they face. And finally, there is dissatisfaction at having to work alongside ill-disciplined progovernment Tamil paramilitaries and at having been subordinated to the Indian Peacekeeping Force from 1987 to 1990. More disturbing, is the self-aggrandizement of factions within military and paramilitary forces as a result of the militarization of the economy. As Steve Hollingworth of CARE International put it in early 2002: ‘This war [in Sri Lanka] is not an event, it is an institution. – by which he meant a pattern of systematic and regular behaviour. As an institution, there are foundational structures which sustain it – economic, social, political, and so on. As noted above, the high degree to which the society, economy, and politics have become dependent

62 Introduction

directly and indirectly on war-fighting, is striking. A well-placed worker in multilateral organization in regular contact with the President’s office noted the tacit acceptance by politicians of shakedowns at the ground level by soldiers and officers. A humanitarian worker, discussing the ‘inefficiencies’ and obstacles to the transportation of aid into war zones, described the trip by the organization’s convoys as ‘the voyage of a thousand envelops.’ However, the scope and level of corruption appears to apply to all levels, from the foot soldier to the military brass where allegations are rampant concerning misappropriations and kick backs in the procurement of large scale military equipment such as air craft.67 The legal action in early 2002 taken against Anuruddha Rattwatte (former Deputy Defence Minister and Minister for Power and Energy under the PA regime) and his sons may shed light on some of the structures and processes of self-aggrandizement within a political economy of militarized violence – at least at the highest levels. A particular concern of the government during the JVP resurgence in the late 1980s was the infiltration of that extremist group into the military. This was facilitated by the rapid increase in the size and role of the military. Even prior to the start of the second JVP insurrection (with the signing of the India Sri Lanka Agreement), newspapers sporadically reported the dismissal of soldiers for alleged JVP connections and large-scale thefts of military hardware undertaken with inside collaboration.68 Despite the crushing of the JVP capabilities in 1989, some JVP–military linkages remain in place.69 As early as March 1992, there was a growing concern about the morale of the army following a military report which revealed that over 3,000 soldiers had deserted in the preceding four years. According to the Colombo Press at the time, many of the deserters were from the front line infantry in the north east. There has reported to have been a ‘dramatic’ (but unspecified) rise in murders, rapes and robberies which is conveniently attributed to the deserters. By 2001, depending on who one asked, the estimated number of deserters varied between the official 20,000 to an unofficial 40,000. And now, there is little doubt that the patterns of violence – in the South East of the island in particular – illustrate an increasing militarization of criminalization as T56 Assault Rifles and deserters enter gray and black economies. This is especially evident in the ‘Deep South,’ the southern coastal areas which are historically the support bases for the JVP. The last ten years have seen the military suffer a series of set backs beginning with the assassination of the Commander of the Navy in


Colombo (November 1992), the assassination of its Defense Minister (March 1991), and the August 1992 land mine explosion on the armycontrolled Kayts island (West of the Jaffna Peninsula) which killed the Northern Commander of the Sri Lankan Army, the Commander for Jaffna, the Commander of the Northern Naval area, and three senior Lieutenant Colonels, among others. These assassinations have hindered the military’s fighting ability in the North and East and exacerbated the on-going friction among career officers seeking advancement (Karunatilleke 1993: 30). The death of Lt. Gen. Denzil Kobbekaduwa in the Kayts land mine explosion and the assassination of Athulathmudali (the former Minister of National Security) have removed two individuals who commanded the loyalty of a significant number of troops. LTTE at the gates of Jaffna In late 1999, the LTTE routed Government Forces in the Wanni, and won back, in a few days, what had taken the SLA two years to capture – 1000 square kilometers of territory. 70 Janes Defence Weekly called it ‘the most humiliating battlefield debacle for the ruling People’s Alliance Government of President Chandrika Bandaranaike Kumaratunga,’ at a cost of an estimated 1000 combatants on each side. By the following May, the LTTE laid siege to Jaffna, before being pushed back. Then in August 2001, there was the devastating LTTE attack on Bandaranaike Katanayake Airport just outside of Colombo. In two waves, LTTE strike teams penetrated the high-security complex at Katunayake at 3.30 a.m. and destroyed a total of 11 aircraft and damaged three. Three passenger aircraft – two A 330s and an A340 – of Sri Lankan Airlines and eight Air Force military aircraft – two Israeli built Kfirs, a Ukrainian MiG-27, two Mi-17 helicopter gunships and three Chinese K8 advanced training aircraft – were destroyed. Two other passenger aircraft – an A320 and an A340 – were badly damaged. In the history of aviation terrorism, this attack is considered the worst.71 Thus, despite the size of the armed forces (measured in manpower and budget size), it is an institution which is rife with both leadership and performance problems. The JVP The JVP is a Sinhalese extremist group which first surfaced in 1971 when it attempted to overthrow the government of Sirimavo Bandaranaike. The insurrection was brutally put down, leaving an

64 Introduction

estimated 8,000–10,000 dead and many more in prison and ‘rehabilitation camps.’ When the UNP came to power in 1977 under Jayewardene, he lifted the proscription on the JVP. There appeared to be a dramatic change in the JVP ideology at that time as reflected in its policy declarations at the time. It rejected armed revolution and ‘extra-electoral struggles’ and adopted a ‘militant parliamentarianism’ similar to that of the traditional Communist Parties in Sri Lanka. The domestication of the JVP appeared to be under way when it was again proscribed along with a number of other leftist groups in 1983 after being accused (implausibly) of initiating the 1983 ethnic riots. Within only four years, the JVP’s Che Guevarist-type revolutionary program of two decades ago was resurrected in the second insurrection of 1987–90. The JVP was described as ‘anarchic rather than classically Marxist, nihilistic and as violently anti-Indian as they are anti-government [and] they appeal to a broad spectrum of young Sinhalese who provide an endless stream of initiates’ (Matthews 1989: 425–6). The signing of the India–Sri Lanka Agreement in 1987 galvanized general opposition to the government of J.R. Jayewardene; it also accelerated the JVP recruitment campaign with its revolutionary call for the overthrow of the state. The membership of the JVP was varied. It included a mixed crosssection of society from professionals to unemployed rural youth, from Buddhist monks (particularly those attending university) 72 to military personnel (see below for elaboration). The strong association of the JVP with Sinhalese Buddhist culture was reflected in the fact that it was virtually non-existent in the Sinhalese Christian areas along the coast north of Colombo; but as will become clear below, this association with Sinhalese Buddhism was conveniently variable. The second insurrection left tens of thousands of people dead in the Sinhalese South of the island and for a time, the most serious threat to civil society in Sri Lanka emanated not from the Tamil paramilitaries in the North and East, but from the JVP. For example, in August 1987, there was an unsuccessful grenade attack on Jayewardene and Premadasa in the very heart of the parliamentary complex allegedly by the JVP. The would-be assassins escaped although an individual was later charged. It is estimated that at the height of the second insurrection the Sinhalese government deployed over 50,000 security forces in the southern half of the island in an effort to contain the JVP threat (Weaver 1988: 81). During the parliamentary elections in February 1989, it is reported that some 70,000 troops and policemen were deployed in the south in an attempt to contain the violence (FEER,


2 March 1989: 13). The 1987 to 1990 period was characterized by the systematic disappearances and murders by the JVP, government security forces, and government-linked vigilante squads such as the ‘Black Cats’ (International Alert 1989; Matthews 1989; Chandraprema 1991; Gunaratna 1990). Despite the Sinhalese nationalist chauvinism that provides the ideological underpinning of the JVP, the organization’s relationship to Tamil sub-groups is decidedly ambivalent. While rumors of possible linkages between the JVP and Tamil paramilitaries have been neither confirmed nor disconfirmed, it may be observed that the JVP has never attacked the LTTE, the principal Tamil paramilitary group. Further, the JVP appeared to respect the fighting skills of the Tigers (Matthews 1989: 430) and there was a shared understanding that each was fighting the same enemy. During the second JVP insurrection, an undetermined number of Up-Country Tamils were active in the local JVP leadership and rank-and-file in the Hill Country. 73 A Plantation Tamil labor organizer explained that, as self-proclaimed Marxists, there was an ideological resonance within some elements of the JVP to support the Up-Country Tamils’ right to struggle for selfdetermination.74 However, the JVP also depended on support mobilized from rural Sinhalese villages. While JVP mobilizers in these areas built support by pragmatically rallying and harnessing anti-Tamil sentiment, this was not uniform and revealed an ideological tension within the JVP ranks. Finally, it should be noted that the targets of JVP violence had been almost exclusively Sinhalese rather than Tamil – from the roadside vendor who makes the mistake of selling a soft drink to a police officer, to the relatives of security personnel, as well as any representative of the state apparatus. The brutal government measures, eventually resulting in the deaths of the highest levels of JVP leadership, have been credited with ‘decapitating’ the movement. Sri Lankan optimists and military officers claim that the JVP threat has been extinguished. However, community workers and those individuals on the front lines of the conflict in Sri Lanka disagree. Instead, it is argued that the violent wing of the JVP has merely entered the same ‘latent phase’ as the interim between the 1971 insurrection and the 1987 resurgence as it entered the formal arena of Party politics. 75 JVP potency remains to be reharnessed in the future because none of the underlying social problems that provided it with sustenance has been addressed – poverty, unemployment, lack of social and economic mobility, land-

66 Introduction

lessness, isolation, and disillusionment with government (Matthews 1989: 427; Ivan 1993). JVP-like tactics are evident within the university setting and there appears to be significant support among the working poor. The transformation the JVP from street revolutionaries to parliamentarians led by Wimal Weerawansa has resulted in the ‘domestication’ of the formal face of the movement, leading for example to a parliamentary coalition being struck between the People’s Alliance of Chandrika Kumaratunge and the JVP. The inability of the JVP to galvanize opposition to the February 2002 ceasefire, suggests a weakening of a mobilizational capacity which was so evident in 1971 and 1987. However, given that the underpinning grievances which provided the fuel for earlier episodes of JVP-led violence have not been addressed (especially youth unemployment and disillusionment), the potential for sequels of related violence should be seen to rest in the youth population, rather than the current leadership within the formal political system.

Tamil paramilitaries and internecine feuding The feuding among Tamil paramilitary groups is an especially bloody manifestation of one axis of division within the Tamil community (see for example, Hoole et al. 1992 and University Teachers for Human Rights (UTHR) Reports). Feuding between the major Tamil paramilitaries (particularly the LTTE, EPRLF, TELO, and PLOTE)76 has been a common feature of paramilitary politics since the creation of these groups in the late 1970s and early 1980s – at times spilling into the streets of South Indian cities. Increasingly, the feuding is spilling over into the adopted cities of the Tamil diaspora – the largest populations being in Canada, Australia, the UK, the US, and the Scandinavian countries.77 Violence became especially vicious following the gradual withdrawal of the Indian Peacekeeping Force (IPKF) from the Northern and Eastern Provinces in 1989. This created a power vacuum and provided incentive for the paramilitaries to intensify their battle for territorial control of the traditional Tamil regions. As importantly, the groups were also fighting to win militarily the right to be accepted as the representative of ‘the Tamil People’ in the eyes of both Tamil and Sinhalese constituencies. In the context of the current discussion, it should be noted that the withdrawal of the IPKF and subsequent reassertion of LTTE territorial control in the north and east sent many EPRLF supporters (voluntary and nonvoluntary) fleeing LTTE retribu-


tion. This was reflected in a surge of applications for refugee status in Canada and elsewhere.78 Since the mid-80s, the LTTE has been the overwhelmingly dominant Tamil paramilitary organization in Sri Lanka, and has become ‘one of the most proficient and dangerous guerrilla/terrorist groups in the world’ (Chalk 2000). It possesses a very sophisticated global network which is able to raise millions of dollars a month through ‘donations,’ extortion, money laundering, drug running, and a full range of black market activities. According to reliable intelligence sources,79 the LTTE fighting force in 1997 was 13,200 (which includes between 5,280 and 7,920 child soldiers). The LTTE has acquired and used a range of lethal, technologically advanced military hardware with devastating effect. This includes: surface-to-air missiles,80 Rocket-Propelled Grenades (RPGs), light anti-tank weapons, and long-range artillery. It is reported to possess at least 50 tonnes of TNT and 10 tonnes of RDX explosive (Janes 2000; Chalk 2000), and has captured tens of thousands of automatic weapons and light weapons from SLA military camps, in addition to the assorted military hardware of government forces. The bomb suit technology developed by the LTTE has earned it the dubious distinction of being the most effective suicide bombers in the world – having killed two heads of state and hundreds of others including politicians, military personnel and civilians. The LTTE is said to have exported this technology to other organizations using similar tactics, in particular to groups active in Israeli occupied territories (Gunaratne 2000).81 LTTE superiority is not limited to the battlefield. It is also maintains a well organized political and diplomatic lobby in major Western (donor) countries. In the late 1990s, the LTTE hired a top US law firm run by Ramsey Clark, a former Attorney General in the Lyndon Johnson administration, to fight Washington’s 1997 decision to label the group as a terrorist organization. Although the effort subsequently failed, it nevertheless amply demonstrated the LTTE’s ability to access the very highest echelons of the US legal establishment–no mean feat for a jungle-based insurgent force located on the other side of the world (Chalk 2000). In Canada, an LTTE fund raising organization succeeded in having the Canadian Minister of Finance and Minister for International Cooperation attend a fund-raising dinner in Toronto, despite the vigorous discussion in local newspapers leading up to the event.82 Indeed, the Canadian government is reported to have granted Canadian $11 million in federal aid to an organization with membership in a group identified by a special Senate committee on security

68 Introduction

and intelligence as a ‘political and benevolent’ front group for the Liberation Tigers of Tamil Eelam.83 To compete with the battle-hardened and militarily superior LTTE, the other Tamil paramilitaries have attempted to draw resources from the Sri Lankan and Indian governments. Indeed, the major non-LTTE Tamil paramilitaries (TELO, PLOTE and Eelam People’s Democratic Party (EPDP))84 have been allied along side the Sri Lankan military against the LTTE since 1987.85 The paramilitary cleavages have been adroitly harnessed by both the IPKF and the Sri Lankan Army to pit Tamils against Tamils. At times, the government’s manipulation of paramilitary animosities has subsidized the violence in unexpected ways. For example, the unsuccessful impeachment proceedings against President Premadasa in September 1991 produced wellfounded allegations that he was responsible for providing the LTTE with arms and supplies for use against the IPKF and the India-backed Tamil National Army (TNA) in November 1989.86 (Incidentally, these same supplies were then used by the LTTE against the Sri Lankan Army, East Coast Muslims, Tamil civilians, pro-government Tamil paramilitaries, and Sinhalese settlers in the North). It appears that the President was attempting to capitalize on the divisions between the Tamil paramilitaries in order to buttress his political position vis-à-vis challengers within the Sinhalese political arena. Covert support to the LTTE allowed him to circumvent the constraints of the 1987 India-Sri Lanka Agreement in an attempt to push the IPKF off the island. If successful, Premadasa would have been able to defuse a serious challenge from an array of opponents which had coalesced around the issue of expelling the IPKF. At the time, despite Sinhalese heterogeneity, a diverse range of groups opposing the Agreement had coalesced against the UNP government: political parties led by the SLFP, community groups, the media, academics, segments of the Buddhist clergy, Sinhalese cultural groups, and Buddhist patriotic organizations. Perhaps most dangerously, it provided an impetus for mobilizing the discontent of segments of the unemployed Sinhalese youth which was ultimately expressed in the JVP insurrection of 1987–89. Similarly, despite the fact that the EPRLF and the Sri Lankan government are supposed to be ‘on the same side’ in the fight to defeat the LTTE, there are credible allegations that Colombo also provided support to the LTTE in its feud against the EPRLF.87 At one level of analysis, the feuding is rooted in intra-group power politics. But at another level, the feuding is exacerbated by the rivalry between


Colombo and Delhi. In the latter (regional) context, intra-group feuding assumes the features of a war by proxy – with Delhi backing the EPRLF and the TNA, and Colombo backing (however counter-intuitively) the LTTE (v. Gunaratna 1993). This paramilitary feuding has implications for both current and future patterns of conflict in the North and East. The Tamil paramilitary coalition against the LTTE is very clearly a coalition of convenience rather than commitment. It has been maintained by the organizational glue provided by Delhi and Colombo and by a shared understanding among the militants that they are threatened by the LTTE. If the LTTE threat is weakened or removed, the unravelling of the coalition is likely follow and the conveniently overlooked intraparamilitary tensions and conflict is likely to spill into more violent feuding in the north and the east – and perhaps even into the streets of Colombo since that is where pro-government Tamil paramilitaries have now located their offices.88 The fueling of intra-group antagonism by Colombo and Delhi for short-term, short-sighted political goals contributes to the brutalization of civilians in the North and East and will inhibit (or at least complicate) movement toward accommodation in any effort to construct a post-war settlement. Paramilitary versus regional allegiances The divisions between the East Coast Tamils and the Jaffna Tamils are evident in the fact that the LTTE leadership remains Jaffna-dominated. At times it has appeared that the different Tamil paramilitaries on the East Coast have had a closer affinity with each other than with their respective organizations in Jaffna. In the past, various Tamil paramilitaries had often shared meals and camps together on the East Coast. It is interesting to note that in 1986 when the bloody LTTE–TELO feud broke out in Jaffna, the Batticaloa leader of the LTTE called together all the paramilitary leaders in the area for a meeting at which they issued a joint statement that the problems of the East were different from those in the north and therefore should be treated differently. The Batticaloa LTTE leader then gave his personal assurance that all other paramilitaries in the area would be protected. However, the Jaffna leaders of the LTTE radioed to two Batticaloa commanders of Jaffna origin and had them carry out the assault on TELO members in the East. Thus, the feud was carried on in the East (Hoole et al. 1991: 97; and personal interviews). The tension between the Jaffna leadership and the East Coast cadre was equally evident in the lead up to, and the

70 Introduction

fall out after, the signing of February 2002 Ceasefire Agreement (see Chapter 9).

On powerful neighbors The central reality of South Asian geopolitics is its Indo-centric nature. If any international or domestic conflict threatens to upset regional stability, the attention of the government of India will inevitably be aroused and often intervention will result. India has always been perceived as a threat by the government of Sri Lanka regardless of which political party was in power. The dominant political dynamic within India is the tension between the central government and the state governments, exacerbated by the persistent Indian phenomena of regionalism, factionalism and communalism. This is an important consideration in a discussion of ethnic conflict in Sri Lanka because it sheds light on actions of the state government of Tamil Nadu which do not always correspond to central government directives. Tamil Nadu has long been the southern pole of a north-south regional split in India. Nonetheless, the state government of Tamil Nadu has been a critical intermediary between the Central Government of India and various Tamil militant groups of Sri Lanka. This was especially evident during the pre-negotiation phase of the India-Sri Lanka Agreement. During the colonial period, fear of India was mitigated by the presence of the colonial power. However, misgivings about Indian aggression returned at independence in 1948 and, accordingly, a Ceylon–UK Defence Agreement was immediately signed. Suggestions made by some Indian leaders in 1948, for the creation of a regional confederation to ensure India’s security, only served to exaggerate Sri Lankan fears. Furthermore, India’s intervention in the 1971 Bangladesh crisis illustrated to Sri Lankan politicians that India was not averse to flexing its muscles as the hegemonic state in the region. The 1987 Accord was viewed in the same light by opponents within Sri Lanka. The Jayewardene government wanted to counteract Indian influence in Sri Lanka mindful of the fact that the Indian government had covertly accepted the presence of militant Tamil training bases in Tamil Nadu since the early to mid-1980s (Gunaratna 1993). Thus, the Sri Lankan government requested and received the assistance of a number of other countries, which may have served to make India more nervous about the intrusion of foreign powers into South Asian affairs.


In addition to obtaining defence equipment from South Africa and intelligence and counter-insurgency training from Israel, Sri Lanka received training and equipment from two of India’s most threatening neighbors – Pakistan and China (Rao 1989). It is, therefore, not surprising that the government of India constructed the 1987 Accord to explicitly prevent external powers from meddling in ‘India’s back yard.’ The Government of India is acutely aware that Sri Lanka could be used by foreign powers to gain access to India’s exposed southern flank (including its submarine base in Cape Comorin and air bases in Kerala). Because of these national security concerns, the Government of India wants domestic conditions in Sri Lanka stabilized at least to the point where external powers could not possibly exploit the ethnic conflict to gain a foothold in the region. The signing of the India-Sri Lanka Peace Accord 1987, and the introduction of 60,000–80,000 Indian troops into Sri Lanka, enabled India to block foreign power involvement in the Sri Lankan conflict. However, it also introduced Indian troops into a quagmire between 1987–90 which had divisive repercussions within India. ‘India’s Vietnam’ gave rise to an Indian ‘Vietnam Syndrome’ which raised questions about its ability to exercise influence in neighboring states (Austin and Gupta 1988). This chapter has introduced the major actors in the Sri Lankan conflict. The following chapters apply the analytical framework presented in Chapter 2 to critical junctures in trajectory of ethnicized conflict in Sri Lanka. As explained above, a critical juncture is a turning point in a conflict marked by an event or episode which has the potential to fundamentally affect inter-group or intra-group relations. The critical junctures examined in this book are judged to be the most important ones since independence. While there may be others that could be included (for example, the assassinations of Rajiv Gandhi in 1991 and President Premadasa in 1993), 89 the six here are sufficient for an investigation of the impact of a set of mediating factors on the intra-group and inter-group dimensions of ethnic groups in conflict. It is expected that the incorporation of other episodes would reinforce and embellish the study, but add only marginally to the overall theoretical argument. By examining in greater detail the critical junctures in the conflict, it is possible to formulate a clearer understanding of the inter-connections between the intergroup and intra-group dimensions of the conflict.

72 Introduction

The critical junctures which constitute the following chapters are as follows: I. 1948 Independence and the Disenfranchisement of the Plantation Tamils II. 1956 Election and the Premiership of S.W.R.D. Bandaranaike III. 1971 JVP Insurrection and 1987 JVP Resurgence III. 1983 Riots IV. 1987 Indo-Sri Lanka Agreement V. 2002 Ceasefire Agreement

Part II Critical Junctures

4 Critical Juncture I: 1948 Independence and the Disenfranchisement of the Plantation Tamils

The context As noted in the Chapter 3, Sri Lanka’s transition from the status of colony to independent state was conspicuously untraumatic when compared with the decolonization experience of India. The transfer of political power to the ‘Model Colony of Ceylon’ was technically challenging, but free of the political turmoil or social catharsis evident in other cases of decolonization (de Silva 1982). Academic accounts of the transition illustrate the elite character of the process, that is, political power was passed from the British to an Anglicized indigenous elite (de Silva 1982, 1986; Manor 1989). As the next chapter illustrates more fully, the ascendant post-colonial political elite strikingly resembled the departing colonial elite – in political form and content (Oberst 1985). In retrospect, the multi-ethnic appearance of the political elite is especially noteworthy. Personal and professional relationships transcended ethnic differences. Tamil and Sinhalese elites shared a common socialization process which included attendance at elite English-medium schools in Sri Lanka and prestigious universities in Britain. As Wriggins explains: ‘From this experience, the cultural diversities were in many ways bridged. Intimate friendships and a substantially common view of life, public purpose and private enjoyment came to be shared by many English-educated in both Sinhalese and Tamil communities’ (Wriggins 1960: 230). The multi-ethnic elite character of the government coalesced in what was said to be a government of national unity in August 1948 when the Ceylon Tamil Congress led by its leader, G.G. Ponnambalam, entered the ruling coalition with Prime Minister D.S. Senanayake. 75

76 Critical Junctures

Yet, within a year of independence, parliament passed legislation which rendered the majority of Plantation Tamils both ‘stateless’ and ‘voteless.’ At first glance this episode poses a puzzle. If the government was genuinely multi-ethnic, how could it have passed this legislation? An adequate explanation requires a shift from an inter-group to an intra-group analytic lens.

The episode At independence the constitution of Sri Lanka was based on the 1944 Soulbury Commission report which had deliberately avoided the contentious issue of defining the criteria for citizenship (de Silva 1986: 145–149). By skirting the issue, Britain side-stepped a messy political issue in the concluding years of its rule. However, following independence the issue quickly and prominently inserted itself onto the political agenda. The parliamentary response was the Citizenship Act (No. 18) in August 1948 which stipulated that citizenship was to be determined either by descent or by registration. The conditions of registration included proof of three generations of paternal ancestry in Sri Lanka and were so stringent that they applied to no more than a few resident Indians. Citizenship was tied not to one’s birth in the country but to the birth of one’s ancestors.1 Wilson argues that the provisions of the Act were intended not to cover all resident Indians (Wilson 1974: 30). To ‘accommodate’ the remainder (that is, the vast majority of Plantation Tamils), a second Act was passed in 1949 for those Indian residents who wished to become Sri Lankan citizens (the Indian and Pakistani Residents (Citizenship) Act (No. 3 of 1949)).2 These two pieces of legislation ‘conferred automatically and without formality the status of citizenship by descent on the indigenous elements of Ceylon’s population, i.e., on Sinhalese, Ceylon Tamils, Ceylon Moors and Burghers, but the number of Indian Tamils which could claim citizenship under the Act was infinitesimal in proportion to the Indian population in the island’ (Kodikara 1965: 112–3). The third and final piece of legislation in this series was the Ceylon (Parliamentary Elections Amendment) Act (No. 48 of 1949) which amended electoral laws so that only citizens could be placed and retained on the electoral register. The sum effect of the three Acts was the disenfranchisement of those Plantation Tamils who up to that date had possessed the right to vote (Wilson 1974: 31) and the imposition of significant restrictions on Indians and Pakistanis seeking Sri Lankan citizenship.3

1948 Independence and Disenfranchisement 77

These Acts set in motion a protracted process of bargaining and negotiation between the governments of India and Sri Lanka over the status of more than 800,000 Plantation Tamils. This has been a discontinuous process of haggling over how many Plantation Tamils would be granted citizenship and the conditions under which it would be granted.4 In compliance with the 1948/49 citizenship legislation, 237,034 applications were filed for the citizenship of 825,000 ‘Indian Tamils.’ The processing of these applications took 13 years, and in 1962 it was announced that 134,188 were granted citizenship – approximately 16 per cent of the total (Wilson 1977: 32). Forty-four years after the initial legislation, the UNP government could still announce as front page news that it was ‘planning to repatriate’ 42,000 ‘Indian Tamils’ (Daily News (Colombo), 18 June 1992: 1). One overall effect of this jockeying has been to decrease the Plantation Tamil population through repatriation to India. Hollup (1992) estimates that out-migration from 1968 to 1983 totalled approximately 400,000 people. Thus, whereas in 1953 Plantation Tamils constituted 12 per cent of the total population, by 1981 this share had dwindled to 5.5 per cent (de Silva 1986: 417). This process has isolated the Plantation Tamils from parliamentary politics and forced them to rely on other channels to represent their interests. Thus, Plantation Tamil demands have been filtered through and shaped by the trade union channels of interest articulation and resource mobilization. Until recently, their interests have been framed as labor and employment issues. However, changes within the Plantation Tamil community – specifically increases in literacy, education, and political awareness – are altering the form and content of demands. As a result, Plantation Tamils are shifting from issues of citizenship and labor rights to issues of political representation and governance. Although the pro-UNP trade union, the Ceylon Workers’ Congress (CWC), still dominates plantation politics, the emergence of a Plantation Tamil political party called the UCPF is a development of the 1990s which suggests the political mobilization taking place within the Plantation Tamil community. In an event reminiscent of the election of the imprisoned hunger strikers in Northern Ireland, the leader of the UCPF was elected to political office from his jail cell in the May 1994 Provincial Elections.5

Model I A Model I analysis of this episode frames disenfranchisement as a straightforward act of discrimination by Sinhalese politicians in order


to weaken the parliamentary power of Tamils. This analytic lens highlights two particular facets of the episode: Sinhalese ethnic chauvinism and the potential political challenge posed by the Plantation Tamils. Ethnic chauvinism The communalist elements within much of the debate in the legislature and the newspapers are conspicuous. For example, the following statements by Sinhalese MPs: It is unthinkable that we should give … full rights of citizenship to people who have not made Ceylon their permanent home. The vast majority of the Indians in Ceylon consider India to be their home and Ceylon their place of occupation … they are here only to earn and make money and to take it all away to India… . Unless we stem the tide of the growing domination of Indians in Ceylon in our economic and social life, our extinction as a Ceylonese nation is inevitable (D.S. Senanayake, 1940, quoted in Ponnambalam 1983: 77). If some of the members of the opposition had the same depth of feeling for their own people as they have for their cochchi sahodarayas (Malayali comrades) they will agree with me that 40 males living in a house with 4 or 5 Tamil or Sinhalese women are hardly the type of people are to be encouraged to become citizens of Sri Lanka (T.F. Jayawardena in Hansard, 10 December 1948, quoted in Jayawardena 1990: 88). If for the protection of the interests of the nationals of this country, the interests of hundreds of thousands of workers in this country, I have to be racial, then I am indeed racial, and I shall continue to be racial (A.E. Goonasinha, Minister Without Portfolio, quoted in Jayawardena 1990: 89). Political challenge of Plantation Tamils This was not the first time that a predominantly Sinhalese government took action against the potential voting power of Plantation Tamils. The first concerted attack had taken place in the late 1920s in response to a recommendation of the Donoughmore Commission that the franchise be granted to all men and women over the age of 21, rather than be restricted to the 4 per cent of the population which had been entitled to vote in the early 1920s. Although this opened the franchise


to women and to the working class, it was the potential electoral power of the Plantation Tamils which attracted the full attention of Sinhalese politicians representatives who vented fears that ‘Kandyan electors would be politically ‘swamped’ by the Indian vote.’6 Although modification to the Donoughmore recommendations reduced the number of eligible Plantation Tamil voters, the 1931 election was the first in which a section of Plantation Tamils were able to vote. They succeeded in electing two candidates of Indian origin in that election as well as the next one in 1936. Sinhalese political leaders expressed apprehension about the growing size of the Plantation Tamil electorate (145,000 in 1936 and 225,000 in 1946) and successfully took action in 1937 to restrict the Plantation Tamil vote in village-level elections (Jayawardena 1990: 75). In the 1941 State Council election, the strict interpretation of the electoral laws effectively disenfranchised approximately 120,000 Plantation Tamil voters from a voting population 141,899 (Tinker 1977: 38). Yet, by the 1947 election, despite the machinations of 1941 and considerable intimidation, seven Indian MPs were returned from the Central Province constituencies (Manor 1989: 189). This afforded Plantation Tamils parliamentary representation but still left them with the worst per capita representation in the House, with one MP to represent 104,608 constituents.7 The argument that fully enfranchised Plantation Tamils would pose a challenge to Sinhalese political control was applied at both a national and local level. At the island-wide level, the argument exhibits classic intergroup logic. By lumping the Plantation Tamil vote in with that of the ‘Sri Lankan Tamils,’ it was mathematically concluded that the total would add up to a politically threatening 23 per cent of the total voting population (de Silva 1986: 417; Ponnambalam 1983: 76, 78). At a local-level scale, the 1948–49 actions are usually understood as a result of the fear by up-country Sinhalese land owners that the electoral power of over a million plantation Tamils could pose a threat to their economic and political interests (Wilson 1974: 28; Kurian 1989: 349). Accordingly, the Kandyan Sinhalese areas benefitted greatly from the disenfranchisement of the Plantation Tamils. It enabled (pro-UNP) Kandyan Sinhalese politicians to annex the seven parliamentary seats previously held by Plantation Tamils (Wilson 1974: 43; Manor 1989: 190). In 1959, the Kandyan politicians benefitted again from the redrawing of constituency boundaries which apportioned seats on the basis of population which included the Plantation Tamils although the vast majority of them had no voting rights (Wilson 1974: 43). There is a clear connection between the


disenfranchisement of Plantation Tamils and the strengthening of Kandyan political elite. Kandyan control was increased at the expense of Plantation Tamil political representation. In addition to the fear of the potential electoral leverage by Plantation Tamils, Sinhalese political leaders became increasingly nervous about the rise of trade unionism within the plantation sector. In light of the dismal conditions of the Plantation Tamil workers – indebtedness, inhumane working conditions, abject poverty, lack of medical and social facilities, and collusion of the state in the maintenance these conditions (Kurian 1989) – the Sinhalese political leaders had good reason to fear the rise of militancy of Plantation Tamils in the Plantation sector. With the rise of the labor movement elsewhere in urban and rural areas, there were militant strikes in 1939 and 1940 on the plantations. Generally, labor organizations were met with violence by employers. When the Ceylon Indian Congress was organized (with Nehru’s patronage) to represent Plantation Tamil interests, the hitherto ‘docile coolies’ came to be seen as a threat being mobilized against the state. In the view of many Sinhalese politicians, the General Strike of 1947 – which involved both plantation and urban labor – fused the ‘Indian Menace’ with the ‘Red Peril.’ Thus, the Plantation Tamil threat was perceived by Sinhalese political leader as twofold: electoral attack in parliamentary politics and/or revolutionary attack in extra-parliamentary politics. The answer to some was simple: disenfranchisement to solve the first threat, and intimidation and force to solve the second. Ponnambalam attributes both class and ethnic motivations to the Sinhalese actors in this episode: ‘The 1949 amendment deprived [Plantation Tamils] of their vote and they became a million stateless and voteless people. Both of these steps were taken because they were Tamils who bolstered the Tamils’ strength in parliament and because their working class solidarity with their Sinhalese counterparts was a constant danger to the upper-class control of the state’ (1983: 78).8 Ponnambalam down-plays the inconvenient fact that disenfranchisement was supported by the Tamil MPs who had been (in his words) ‘successfully wooed’ into the first government of independence. Instead, emphasis is placed on the protest resignation of two Tamil Congress leaders to form the Federal Party which became the dominant political party of Sri Lankan Tamils in the North and East of the island. However, the Federal Party had limited success articulating Plantation Tamil interests. The next section, discusses the intra-group dimensions of events and argues that a full understanding of the episode requires that the Model I explanations be supplemented with Model II explanations.


Model II A Model II analysis of this episode does not deny the instrumental role of Sinhalese chauvinism in the public and personal politics of Sri Lanka. While acknowledging the presence of inter-group tensions, a Model II analysis incorporates the intra-group nuances of the episode. In particular it asks: if this episode is a manifestation of a continuous expression of Sinhalese chauvinism, what explains the support given to the legislation by Tamil MPs? Although some members of the Ceylon Tamil Congress (CTC) criticized the legislation, others members of the Congress endorsed it. Furthermore, many other minority members (including Tamils, Muslims, Burghers, and Europeans) also voted in favor of the legislation.9 In other words, the three pieces of legislation which deprived Plantation Tamils of both citizenship and the vote were supported by MPs from the other minority communities including Sri Lankan Tamils. Without this support, the government would have had to force the legislation through parliament which would have made it difficult to justify its actions. A full understanding of the event requires an appreciation of the intra-group relations within both the Sinhalese and the Tamil communities. Within the Sinhalese community the episode illustrates efforts by Kandyan Sinhalese to maintain economic and political position within the economic heartland of the plantation economy. Within the Tamil community, it points to the general lack of solidarity between the Plantation Tamils and the other sub-groups. Manor argues that the Sri Lankan Tamil politicians found themselves in a dilemma. They could protect Plantation Tamil interests and watch the legislation get forced through. Or they could collaborate with the UNP and ‘hope that Sinhalese leaders and voters would remember their cooperation and not turn on them. In the event, most of them chose the latter approach’ (Manor 1989: 190). The decision of the Ceylon Tamil Congress appears to have been based on pragmatic political calculations: short-term support for long-term self-protection and/or benefit. This summarily undercuts any claim of political homogeneity and solidarity among Tamil sub-groups and highlights the political self-interest which facilitates tactical inter-group alliances and frequently underpins provocative ‘ethnic acts.’ One result of this episode was an attenuation of already weak Tamil intra-group relations and a consequent diminution of the willingness of sub-groups from this ethnic category to respond in concert to intergroup challenges. By ‘wooing’ Ponnambalam’s Tamil Congress into the UNP government, a wedge was driven between it and the Ceylon


Indian Congress supported by Plantation Tamils. An important consequence of this process was a splintering of the Tamil Congress and the creation of the Federal Party. By weakening the Tamil Congress, the UNP weakened a voice for moderation and compromise and created the conditions for the rise of a more hardline Tamil political party. It is important to consider the process by which a multi-ethnic government passes communalist legislation. Although one can identify Sinhalese chauvinist elements from the start of parliamentary politics in 1931, it is nonetheless accurate to portray inter-group relations (that is, ‘Sinhalese–Tamil relations’) as relatively harmonious in 1948 at the level of the political elite. While the Ceylon Tamil Congress existed to represent and agitate for group-specific interests, the UNP was composed of both Sinhalese and Tamil MPs and ministers. To characterize the parliament as multi-ethnic, as Oberst (1985) and Wriggins (1960) have done, is accurate only if the ‘multi-ethnic’ designation refers exclusively to the political elite and only if it is conceived in strict Model I terms, that is, according to homogeneous Sinhalese and Tamil categories. An application of Model II analysis points out intra-group divisions and stresses which are often ‘played out’ in the inter-group arena. This episode shows the considerable distance between the Jaffna Tamil political leadership and the socially, economically and politically marginalized Plantation Tamils. More importantly, it shows that subgroup self-interest may facilitate inter-group arrangements which compromise potential intra-group allies. The episode illustrates the tactical and political maneuvering by Jaffna Tamil politicians to mobilize political resources for future use in the parliamentary arena. The inter-group collaboration in the disenfranchisement of the Plantation Tamils was also facilitated by the shared socialization process of the Jaffna Tamil and Sinhalese political elites. This leads Jayawardena to argue that the salient axis of identity affecting voting behavior in the years following independence was class rather than ethnicity (1990: 92–95). The episode can be represented as follows (see Figure 4.1): Elites + Sub-Group B1 (Sinhalese)

+ +

Figure 4.1 Tamils

Non-elites Sub-Group C1 (Muslims) Sub-Group A1 (SL Tamil) Sub-Group D1 (Burgher)

versus

Sub-Group A2 (Plantation Tamil)

Elite-non-elite conflict and the disenfranchisement of plantation


Yet by playing a tactical inter-group communalist game, Sri Lanka Tamil politicians contributed to the ethnicization of the political system which eventually boomeranged on them. Thus, the eventual consequence of this episode may be represented as follows (see Figure 3.2): Sub-Group B1 (Sinhalese)

versus

Sub-Group A1 (SL Tamil) Sub-Group A2 (Pl. Tamil)

Figure 4.2 Ethnicization of the elite conflict and the disenfranchisement of plantation Tamils

The disenfranchisement of the Plantation Tamils was undertaken within the Westminster model of parliamentary democracy. This was the first instance in independent Sri Lanka where the parliamentary system was used to manipulate the electoral balance of ethnic subgroups to enhance the relative strength of the Sinhalese electorate. A centralized system of government in which there was no formal protection of minority rights, and in which the state dispensed goods and services on the basis of political expediency to build up electoral constituencies, has led to the communalization of parliamentary politics (Warnapala and Woodsworth 1987; Shanmugaratnam 1993: 5). By the early 1980s, the manipulation of the political system had led to reduced legitimation of the system and a steady decline in the rule of law (see Chapters 7 and 8, in particular, discussion of amendments to the constitutions and the 1982 referendum in lieu of elections).

5 Critical Juncture II: 1956 Election and the Premiership of S.W.R.D. Bandaranaike

The context The year 1956 is pivotal in the ethnic politics of Sri Lanka. It is the year in which ethnic chauvinism was first employed by a political party to contest and win a general election and to form a national government. In the past, ethnic sentiments had been stirred and harnessed by politicians, but by and large, ethnic chauvinism was either an unstated element in elections or a relatively insignificant sideshow.1 This changed in 1956 with the campaign and election of S.W.R.D. Bandaranaike and his SLFP in an electoral coalition called the Mahajana Eksath Peramuna (MEP). The election politicized issues which continue to resonate, antagonize and mobilize today: the ‘rehabilitation and restoration’ of Buddhism to a pre-eminent position in society; the use of the indigenous Sinhala language; and the fostering of Sinhalese national identity by the state. Bandaranaike did not create the Sinhalese chauvinist support base which swept him to power. He temporarily consolidated a wide range of Sinhalese Buddhist sub-groups by manipulating the powerful cultural symbols of language, religion, and ‘race.’ His most effective rallying call was for a ‘Sinhala Only’ language policy. He offered existing disgruntled groups a means of articulating (often legitimate) grievances and promised a swift and satisfactory response once he was in power.2 Yet, once chauvinism was galvanized in the Sinhalese intra-group arena, similar sentiments could expediently be rallied by Tamil politicians also seeking to build intra-group support. Bandaranaike’s shrill rhetoric on the campaign trail was more than matched by Sinhalese Buddhist activists whose concerns had been neglected by elected 84

1956 Election and S.W.R.D Bandaranaike 85

officials.3 In the end, Bandaranaike’s assassination by a monk (and onetime supporter) illustrates that the wielding of extremist rhetoric for short-term advantage inflates the expectations of a support group and invites political disaster4 when these expectations are not seen to be met. As discussed further below, the episode also illustrates how successful mobilization of untapped political resources by one actor can stimulate competitors to seek to tap those same resources. Those who are unable to do so are either politically marginalized or forced to innovate and mobilize still other resources. After 1956, ethnic sentiment became a necessary ingredient for mobilization in electoral politics. Similarly, after 1988, the late President Premadasa’s populist politics which mobilized the support of the poor and marginalized has made it necessary for political competitors to similarly cultivate this support base. The 1956 election – more than any election before or since – presented the electorate with a dichotomous choice between an Anglicized, Westernized elite represented by the incumbent UNP government of Prime Minister Sir John Kotelawala,5 and ‘SinhaleseBuddhist activists’ represented by the leader of the opposition, S.W.R.D. Bandaranaike. Up until 1956, the party loyalties of voters and candidates had been fluid. Candidates from the same party would frequently run against each other in the same constituency; there were instances in which candidates ‘dis-affiliated’ from a party after being elected; and voter choice was based on the politics of personality, patronage and intimidation as much as on party allegiance (Manor 1989: 84–6, 120, 180). The 1956 election was the first time that issues and party programs outweighed the personalities involved (Manor 1989: 253). The post-colonial rule under the UNP was not only highly Anglicized, it was anglocentric. Requests by bhikkhus (Buddhist monks) for greater government support for Sinhala language and for Buddhism – ‘lest the Sinhalese race be destroyed’ – were frequently treated with ridicule by politicians (Manor 1989: 196). Most politicians had neither the political will nor the personal inclination to shift away from Anglicized forms of government and social relations. Thus, the manners, symbols and styles of the colonial era persisted inside and outside Parliament.6 The parliament and legislative elite between 1924 and 1948 was disproportionately Christian, mostly high caste, highly urbanized, Western educated, largely engaged in Western-type occupations, and of the highest economic and social class (Singer 1964: 56–60).


Although the Western lifestyles of the Sri Lankan elite (both Sinhalese and Tamil) dominated political and economic life, a Buddhist revivalist movement had been under way since the late 19th and early 20th centuries (Tambiah 1992; Kemper 1991; and Matthews 1988/89, 1985; Smith 1978). Yet, indicators of the potential political significance of Buddhism in Sri Lanka were largely overlooked in the pre-1956 period. Studies of Sri Lankan politics published before 1956 rarely mention Buddhism (Bechert 1978: 202). Bandaranaike had formed a communal political organization in the mid-1930s called the Sinhala Maha Sabha but it was poorly institutionalized and lacked political presence in the legislature. It was primarily a vehicle for Bandaranaike’s political career rather than an independent, self-sustaining organization. While Buddhism may have been neglected in the political realm, politics was by no means being neglected in the Buddhist realm, as bhikkhus vigorously debated whether it was appropriate for them to engage in political activity. In the past, Buddhist interests had been represented in politics by the Buddhist laity. A mutually beneficial relationship had developed in ancient times whereby the state would materially support the Sangha (the community of Buddhist monks) which would in turn endorse and legitimate the state (Smith 1978). During the colonial period this relationship fell into disrepair,7 and bhikkhus – especially young, left-leaning ones – began gravitating from the periphery to the center of political activism. A driving force in the politicization of the Sangha was Walpola Rahula8 who argued that there is both a moral obligation and long historical precedent for the involvement of bhikkhus in politics.9 Soon after independence, the UNP government of D.S. Senanayake had heard and rejected demands from a variety of poorly organized Sinhalese–Buddhist interest groups which sought to revive Buddhism and ‘rescue’ the Sinhala language and the Sinhalese ‘race.’ Buddhist teachers and parents complained of the threat from well-funded Christian schools to their children’s beliefs and sought government funds for Buddhist schools; practitioners of ayurvedic (traditional) medicine sought higher status and government patronage; Buddhist activists campaigned against alcohol, gambling and animal slaughter; teachers in vernacular schools sought parity in status and salary with English medium teachers. Throughout the 1950s, various Sinhalese Buddhist interest groups voiced their complaints, but their lack of organization and coordination with like-minded actors constrained their impact beyond their immediate localities (Manor 1989: 219). The number of discontented groups continued to increase to and included


vernacular writers; small shop keepers and entrepreneurs who felt threatened by large (UNP-favored) Sinhalese, British or Indian firms; consumer groups; poorer peasants and members of depressed classes such as the Wahumpuras and Bathgamas. The upsurge in cultural and linguistic concerns was fuelled in part by the UNP’s inertia and bungling on several major issues which served to underscore the Anglicized character of a political system operating in a majority Sinhala Buddhist society. In 1953 basic laws had still not been translated from English into the vernaculars; the oaths sworn in courts of law were unchanged from colonial times; it was still illegal to transact business at banks in Sinhala or Tamil; the Education Department was slow to introduce vernacular teaching as promised by the UNP (Manor 1989: 229). [Sinhalese Buddhist] grievances were not new. Most of them had been quietly simmering for years, in some cases for decades. Nor was it new that people should sporadically articulate them and take action to dramatize them. What was new was the incidence and intensity of their appeals and the organization of issue-oriented voluntary associations for sustained pressure. People were beginning to understand and operated within the logic of open, representative politics and, in so doing, were changing what had been an alien imposition into something more fully their own (Manor 1989: 228). During the 1950s these groups began to assume greater cohesion and forcefulness in the presentation of their grievances and demands. While some members of the UNP – most notably, J.R. Jayewardene – counselled the government to accommodate and defuse such demands, Kotelawala and his colleagues failed to take the advice or the challenge seriously. By late 1955, UNP MPs finally recognized the electoral danger posed by Sinhalese Buddhist interest groups and resolved to adopt ‘Sinhala Only’ as official policy ‘at the earliest possible occasion.’ However, Kotelawala left the country on a month-long tour just as language demonstrations were heating up. As Acting Prime Minister, Jayewardene fully endorsed the adoption of a ‘Sinhala Only’ language policy. In contrast, Kotelawala’s initial reaction upon his return was: ‘If there is a problem, it will fix itself’ (cited in Manor 1989: 236). The 1956 election was the occasion for the problem ‘to fix itself’ though not in the way that Sir John would have hoped. The UNP was routed.


The episode As the election results below show, the UNP was soundly defeated. For the first time since independence, a non-UNP government assumed political power. Despite the attention focused the language issue, Bandaranaike and his associates had not been especially optimistic about the outcome of the election. As in past elections, the UNP seemed to have all the advantages: it possessed more financial resources; it received more press coverage; it had again (illegally) mobilized government employees; it had a huge advantage in the transportation facilities necessary to get voters to the polls;10 and it had a large supply of thugs to provide intimidation including the laborers from Kotelawala’s graphite mines and his ‘green-shirts’ (a UNP youth wing) which had been used with effect on previous occasions to break strikes and provide political intimidation. Since the advent of popular elections in the 1931, electoral politics had been a combination of crude material incentives and equally crude intimidation (Manor 1989). Consequently, there was no perceived need for building a continuous network of support, just a little tactical planning prior to elections. Also contributing to SLFP uncertainty about the outcome of the election was its awareness that it did not have a well-developed support network reaching out to those Sinhalese sub-groups which had been neglected by the UNP – the sub-groups which eventually constituted the core support for Bandaranaike’s electoral bid. However, in this Table 5.1

Results of the 1956 general elections

Party MEP UNP NLSSP CP FP TC Independents Others

Seats won

% of seats won

votes

% of votes

49 8 14 3 10 1 10 0

51.6 8.4 14.7 3.2 10.5 1.1 10.5

982,228 718,164 274,204 119,715 142,036 8,914 352,928 18,510

37.5 27.4 10.5 4.6 5.4 0.3 13.5 0.7

Source: Manor (1989: 242). Notes: MEP (Mahajana Eksath Peramuna – led by the SLFP); UNP (United National Party); NLSSP (Nava Lanka Sama Samaja Party – Trotskyist); CP (Communist Party); FP (Federal Party); TC (Tamil Congress).


regard, the SLFP was on equal footing with the UNP since it too lacked a network of popular support. As the UNP Minister of Local Administration (1936–46), Bandaranaike had attempted to cultivate a grassroots network by creating local councils, but had failed to establish anything enduring due to the lack of financial resources and the in-fighting of other UNP ministers (Manor 1989: 125–67). Although Bandaranaike failed to cultivate or institutionalize these grassroots linkages, he did take advantage of available channels of communication and influence. His political message was relayed to a grassroots audience through five principal groups: local-level government officials; ayurvedic practitioners; vernacular teachers; the youth with limited prospects; and bhikkhus (Buddhist monks) – groups which, by virtue of the UNP’s neglect, constituted willing and able conduits for Bandaranaike’s political message. Each of these sub-groups were in positions to influence the voting behaviour of a larger network of people. The bhikkhus and Buddhist laity were crucial in the election because the SLFP was able to tap into a network ready-made for political campaigning. The Eksath Bhikkhu Peramuna (EBP or United Front of Monks) was a recently formed organization of ‘progressive monks’ which actively promoted Bandaranaike, the SLFP, and the MEP.11 The EBP was adamantly anti-UNP, anti-Westernization, and anti-Catholic. It was equally adamant in its support for the Sinhala-Only language policy and for the implementation of democratic socialism. Tambiah explains the role to the EBP as follows: The EBP monks, working through a network of local sangha sabhas [the assembly or council of a Sangha], proved to be formidable and untiring election campaigners – making personal house-to-house visits and distributing pamphlets. It is said that between three thousand and four thousand monks – about a fourth of the national total – participated as campaign workers. Although the UNP did enlist on its side some establishment monks, even the leaders of first-rank pirivenas [monastic educational institutes], who tried to prevent the monk activists from electioneering, by and large their support was not effective. Thus, it is no exaggeration to claim that the 1956 election, which swept Bandaranaike and the MEP to power, was the climactic and singular moment in twentieth-century political life, when a significant number of monks temporarily organized to win an election.12 Never again in the ensuing decades would the sangha show this much purpose and action. As Phadnis


put it, the EBP’s decisive contribution lay in its role in the support mobilization of Buddhists and in providing a country-wide bhikkhu cadre to a party with very little organization and projecting its image as the ‘party of the common man’ (Tambiah 1992: 44)13

Model I A Model I analysis of this episode highlights the detrimental impact of escalating Sinhalese chauvinism on inter-group relations. The political manipulation of ethnic symbols alienated many Tamils from the national political process and encouraged reactive parochialism. Like the SLFP, the UNP adopted ‘Sinhala Only’ as official party policy (albeit belatedly) and, increasingly engaged in its own provocative communalist acts. Both Sinhalese and Tamil commentators have analyzed the episode through a Model I lens (often with each blaming the other for stoking conflict).14 Yet, even a cursory reading of the episode illustrates that the run-up to the election was dominated by intra-group competition between the UNP and the SLFP. Prior to the election it was the way Tamils were characterized by Sinhalese politicians rather than Tamil actions per se (and vice versa, the characterization of the Sinhalese by Tamil politicians) which was so politically volatile. While it is possible to identify an inter-active dynamic at the inter-group level, in which the chauvinism of one group evoked a response in kind from the other, it is equally important to recognize the source of that conflict: the intra-active dynamic in which sub-groups competed against each other for intra-group support by increasing the level and pitch of their communalism.

Model II Despite Bandaranaike’s rhetoric, his political battle was not with ‘the Tamils’ or even with the Federal Party, the main Tamil political party. His primary political battle was with the UNP for the control of government. The intra-group origins and location of this battle should not be obscured by the divisive inter-group consequences of Bandaranaike’s instrumental communalism. Further, G.G. Ponnambalam of the Federal Party was mobilizing Tamil intra-group support using similar tactics – at times making the two parties (SLFP and FP) seem more like allies than arch-enemies. The mutual benefits of well-calculated displays of inter-group communalism for mobilizing intra-group support were recognized by both men. At the personal


level, Bandaranaike’s relationship with Ponnambalam, his ‘archcommunalist’ counterpart, was reasonably cordial. But their public encounters assumed the level of spectacle as they squared off against each other in communalist jousting matches (Manor 1989: 131). Manor describes the mutual advantages as follows: Since [the mid-1930s] neither had achieved pre-eminence in their own community and both saw communal polarization as a means to that end, each found it advantageous to exaggerate the threat that the other posed. So although the two men taunted each other mercilessly – in one exchange Bandaranaike associated Ponnambalam’s ancestry ‘with a simian prelude,’ as Governor Caldecott put it – they were in some ways covert allies. At times they joined forces overtly and this, like their amiable personal relations, raised doubts about their devotion to communalist cause. … In 1942 they united in a call for a national government containing all communities and the following year in a bizarre episode Ponnambalam urged Tamil estate workers to vote for the [communalist Sinhalese] Sinhala Maha Sabha candidate in a State Council by-election. These things indicated to many that for Bandaranaike, Sinhalese chauvinism was more a tactical ploy than a matter of conviction. They were essentially correct, but his chauvinism was no less dangerous for that (Manor 1989: 131–32). Biographical incongruities and the instrumentalities of communalism Jiggins (1979: 4) describes Bandaranaike’s political image as follows: ‘S.W.R.D. appeared as a slight figure who walked to election meetings and who wore the village cloth and shirt; a Buddhist who promised radical social change and the redistribution of economic opportunity.’ Yet, there are biographical incongruities which complicate the Sinhala Buddhist personae of Solomon West Ridgeway Dias Bandaranaike – the English-speaking, Westernized, Oxford-trained Anglican named in part after his God Father, the British Governor of Ceylon. Bandaranaike’s late conversion to Buddhism in his mid-30s led many to note the political expedience of this decision at a time when universal suffrage was creating a massive untapped Buddhist constituency.15 His mother tongue was English and, although some of his oratory skills could be employed in his Sinhala speeches, he admitted in 1955 that ‘I can read in Sinhalese (sic), but not very fluently’ – a remark for which he was bitterly attacked by linguistic revivalists (Manor 1989: 230).16


These biographical incongruities are often overlooked in the broad overview of Sri Lankan politics. Yet, ‘the personal’ is ‘political’ and the instrumentalism which underpinned the biographical incrongruities of Bandaranaike’s personal life are mirrored in his political career. In order to draw theoretical insights from this episode, it is important bear this in mind while analysing the intra-group political dynamics which propelled Bandaranaike’s rise to power. The communalism of 1956 and afterwards17 originated not in an inter-group conflict between the Tamils and Sinhalese, but in an intragroup conflict within the Sinhalese elite. It was not, t however, a conflict between the Anglicized Sinhalese elite and the traditional, vernacular elite, as Horowitz (1973) has argued. Rather, it was a conflict between two Westernized elite factions in which one more effectively and instrumentally mobilized the support of the traditional, vernacular elite.18 Bandaranaike’s battle with the UNP dates back to the party’s inception in 1946 and continued throughout his tenure as a UNP minister. It was not a surprise when in 1951 Bandaranaike split from the UNP to form the Sri Lanka Freedom Party; more notable is that he stayed in the Party for so long. As a minister in the first independent Government of Sri Lanka, he had been in constant conflict with the upper echelon of the UNP leadership, most of whom closely guarded their parochial interests and stifled any initiative which encroached on their self-claimed jurisdictions.19 The UNP ‘family’ protected its own, and Bandaranaike was an outsider.20 Partisan patronage and nepotism was the mainstay of political power and the UNP earned its alternative moniker, the ‘Uncle–Nephew Party,’ when the prime ministership passed from D.S. Senanayake (1947–52) to his son, Dudley (1952–53), to his nephew, Sir John Kotelawala (1953–56). In terms of the structure of the political system, the creation of the SLFP might be viewed as a positive development because, by effectively creating a two-party system, it offered the possibility of increased vitality and responsiveness in a system which had been dominated by the increasingly inertial UNP. Instead, however, the resulting responsiveness pandered to ethnic chauvinism and was consequently particularistic and fragmenting rather than inclusive and integrating. Resource mobilization Bandaranaike recognized early in his political career that the judicious manipulation of communalism could yield political rewards. Yet, it was only with practice that he mobilized and won the ‘Sinhalese


Buddhist vote.’ In the 1930s and 1940s, he had difficulty attracting the support of Sinhalese revivalists. Their emphasis was on cultural issues, while his was on the economic dimensions of communal politics, in particular the Indian Tamil presence on the island (Manor 1989: 248n). However, by the mid-1950s, he had expediently modified his political tactics to highlight cultural issues. This tactical shift suggests the instrumentalism driving Bandaranaike’s political program. While there were economic dimensions to Sinhalese-Tamil sub-group relations, the cultural issues were more politically potent. Thus, his strategy shifted accordingly. As Bandaranaike is reported to have said in an interview in 1956: ‘You know my dear fellow [smiling broadly], I have never found anything to excite the people in quite the way this language issue does’ (cited in Manor 1989: vi). In the context of the theoretical interests of this thesis, the shift in emphasis from economic to cultural issues highlights the efficient modification of resource mobilization techniques. The concluding chapter will further discuss the process whereby the opening up of certain political resources by one group leads others to compete for those same resources. In this episode, Bandaranaike reached beyond the constituency which constituted the UNP support base and thereby ‘discovered’ new political resources which, once mobilized, swept him to power. After his election, Bandaranaike was an awkward political position. Having embraced and articulated communalist concerns in order to mobilize electoral support, he had to deflate the inflated Sinhalese Buddhist expectations while pacifying the politically alienated Tamils in order to staunch escalating communal tensions. In March 1957, with the approval of leading bhikkhus, Bandaranaike unveiled a threepoint plan to ease Tamil concerns. It included: a proposal for devolution and increased autonomy in Tamil majority areas by creating regional councils; constitutional amendments to safeguard constitutional rights of minorities; and provisions for the ‘reasonable use’ of Tamil within the provisions of the Sinhala-Only Act. In a role reversal from the pre-election period, the UNP launched an antiBandaranaike campaign using precisely the same communal belligerence which had served Bandaranaike so well. An official resolution of the UNP executive accused Prime Minister Bandaranaike of perpetrating a ‘colossal fraud’ in ‘now showing extreme concern for the exaggerated claim of a small Tamil clique of the most virulent communal minded and violent anti-Sinhalese politicians … under whose complete influence Mr Bandaranaike has now come’ (quoted in Manor 1989: 269).


Despite intra-group attacks from the UNP and extremist Sinhalese elements and, despite the political posturing by the main Tamil political party, the Federal Party, an agreement was struck which accommodated much of the Tamil demand for increased regional autonomy (the B–C Pact). Part of the impetus for Bandaranaike to come to an agreement was the impending deadline set by the Federal Party for launching a mass protest over the neglect of ‘Tamil’ concerns. Like the three-point proposal noted above, the B–C Pact sought to devolve powers to regional councils and thereby increase Tamil autonomy. While the Sinhala-Only policy was left untouched, Tamil was recognized as the ‘language of the national minority in Sri Lanka’ and would be the language of government adminstration in the Northern and Eastern Provinces (with provisions for Sinhala-speakers in the areas). In short, the Pact was a statesmen-like compromise that seemed to possess the necessary ingredients for healing the distrust and antagonisms which had built up at the inter-group level as a result of Bandaranaike’s intra-group power struggle. There appeared to be general acceptance of the agreement (Manor 1989: 271). However, ultimately the UNP was able to rally sufficient opposition to pull the Pact down by mobilizing resources from the same sub-groups which had enabled Bandaranaike to rise to power. Central to the UNP assault on the Pact was the symbolically explosive protest march by J.R. Jayewardene from Colombo to Kandy in October 1957.21 In the face of attacks from within his party, from the UNP, and from mobilized Sinhalese Buddhist sub-group, Bandaranaike finally acquiesced and abrogated the Pact. At this stage, conflict moved from the intra-group to inter-group arena. A Tamil civil disobedience campaign evoked a vicious Sinhalese backlash and the so-called ‘race riots’ of 1958 (Vittachi 1958; Manor 1989: 286–99; and Tambiah 1992: 46–57). SLFP MPs began to defect from the party until a clear majority of elected legislators eventually stood against him, leaving him ‘in an exceedingly precarious, undignified position’ (Manor 1989: 305). He managed to barely defeat a vote of no-confidence, but it seemed only a matter of time before he was politically defeated. An indication of the escalating tensions and Bandaranaike’s sinking political fortunes was the decision by his aides, as a precaution against trouble, to station armed men off-stage during the Party’s annual conference which turned out to be a raucous affair (ibid.). Two months later, he was assassinated by a bhikkhu acting in concert with the high priest of the Kelaniya Temple, a prominent figure in rallying support for Bandaranaike during his 1956 campaign.


Opportunity structure A crucial factor which accelerated and synchronized the ethnic mobilization of discontented Sinhalese Buddhist interest groups was timing. The decision to hold a general election in 1956 served as the final straw which activated and politicized many Buddhists. The year 1956 marked the Buddha Jayanthi – the celebration marking 2500 years from the death of the Buddha, and the arrival of the first Sinhalese to the island. A number of popular traditions and expectations were associated with the event, such as the prophecy that a universal king would appear at that time and reestablish a Buddhist polity (Malalgoda 1970: 38–39). The idea of the ‘sacred rights’ of the Sinhalese was reintroduced into public discussion in an influential, openly chauvinistic book published in 1953 by Vijayavardhana, entitled The Revolt in the Temple: The history of Lanka is the history of the Sinhalese race. … The Sinhalese people were entrusted 2500 years ago, with a great and noble charge, the preservation of … Buddhism … in 1956 will occur the unique three-fold event – the completion of 2500 years of Ceylon history, of the life of the Sinhalese and of Buddhism. The birth of the Sinhalese race would thus seem to have been not a mere chance, not an accidental occurence, but a predestined event of high important purpose. The nation seemed destined as it were, from its rise, primarily to carry aloft for fifty centuries the torch that was lit by the great World Mentor twenty-five centuries ago (Vijayavadhana (1953), quoted in Jayawardena 1990: 104–5). While many members of the sangha were still suspicious of Bandaranaike’s Buddhist credentials and commitment, it was clear to most of them that the excessively Anglicized UNP government of Kotelawala was even further from the Buddhist ideal. The first phase in the campaign against the UNP in the 1956 Election rallied around the slogan, ‘No elections before the Buddha Jayanthi.’ Contributing to the disapproval by Buddhist activist, was the fact that the elections could have been put off until May 1957. In February it was decided that satyagraha (protest of passive resistance) would be performed if the general elections were not postponed. Appropriating the vocabulary with which they were most familiar, bhikkhus began to describe the general election as Mara (satan, or the temptress), from whom Buddhism needed to be protected. When the UNP proceeded


with the dissolution of the Parliament in preparation for the election, bhikkus carried out satyagrahas in towns and centers throughout the island. This included a fast in Colombo by 250 monks.22 The second phase of Buddhist involvement was the active and open campaigning on behalf of the MEP against the UNP. Buddhist clergy were prominent at the MEP rallies. They actively supported the anti-UNP candidates (usually the MEP). Bhikkhus published pamphlets of arguments why the UNP should be defeated. At public assemblies bhikkhus argued that Buddhists had suffered under colonial rule and continued to suffer under the UNP government. The conclusion of many bhikkhus was the need for direct political action. According to one bhikkhu: The bhikkhus should be present in every polling booth. They should explain to the people how to use the vote correctly. … A government that will work for the country, religion and its culture should be elected. The end of the Sasana will not be very long if we remain in silence. … We appeal to bhikkhus to visit every Buddhist home and to direct them on the right path. You may have to confront many difficulties. But be ready to sacrifice your life to restore a Buddhist Ceylon (Newspaper passage quoted by Wriggins 1960: 345). The second event which helped Bandaranaike’s election bid was the release of an extremely critical report on the state of Buddhism in Sri Lanka entitled ‘The Betrayal of Buddhism’ by the ‘Buddhist Committee of Inquiry’ (the ‘Buddhist Commission’). The Committee was composed of prominent ‘scholar-monks’ and prominent lay persons and represented the views of some of the island’s foremost scholars and educators. The report dealt less with the political role of bhikkhus, than with the decline of Buddhism because of direct colonial attacks, particularly in the area of education where Christian Schools were supported to the virtual neglect of Buddhist Sinhalese schools (Tambiah 1992: 34). The report called for the withdrawal of all grants in aid to Christian missionary schools and affirmative action on behalf of Buddhism and the sangha in future state action. The most striking feature of the report was its angry, condemnatory tenor. It concluded with an exhortation which conveniently parallelled Bandaranaike’s political platform: ‘Education in Ceylon today should be oriented towards bringing forth a generation with an intimate awareness of its national [i.e. Sinhala] language, history and culture and capable of


enriching that national heritage’ (quoted in Tambiah 1992: 34). While initially the Committee may have been stimulated by anti-Christian sentiment, it easily (and soon) incorporated anti-Hindu, anti-Tamil sentiment. In one of the best studies of the period, and specifically of Bandaranaike’s place in it, Manor (1989) argues that Bandaranaike was predisposed to instituting the use of both vernaculars even in late September 1955 when ‘Sinhala-Only’ was adopted as official party policy. Manor illustrates Bandaranaike’s long history of oscillation between contradictory positions. This, combined with the expedience of many of Bandaranaike’s past personal and policy decisions, leads Manor to argue that Bandaranaike was hoping to finesse the language issue, that is, to ride the ‘Sinhala-Only’ wave into office only to maneuver a de facto dual language policy into place in order to assuage minority concerns. Indeed, the SLFP resolutions adopted in 1955 along with the ‘Sinhala-Only’ resolution were carefully worded to allow for the substantial use of Tamil.23 The counter-factual argument could be made that prudent cultural policies which balanced the needs and fears of both Sinhalese and Tamil communities might have avoided the subsequent spiral of violence. Manor makes an important observation of the 1956 election: ‘It marked the first cycle in a pattern which has recurred as a central and poisonous feature of the political process at critical junctures. The party in power strives to foster communal accommodation. The major party in opposition manipulates Sinhalese parochialism to wreck that attempt’ (Manor 1989: 269). The final word in this episode goes to Manor, who argues that the UNP must bear some responsibility for the emergence of anti-Tamil bigotry (1989: 253). He is sweeping in his allocation of blame for the flare up of chauvinism: The failure of the UNP to deal sensibly with those discontents, or with the Sinhala revival when it arose, places far more of the responsibility on the emergence of anti-Tamil bigotry upon the old ruling party than has been attached to it. Bandaranaike, however, also bears a considerable share of that responsibility. He never abandoned his desire to aid both vernaculars, as his sotto voce comments during the campaign on the use of Tamil suggest, but the main thrust of his speeches contributed to the fervour and to the dangerously unrealistic expectations of Sinhalese chauvinists. Bandaranaike, the opportunistic candidate, did much to encourage


the emergence of anti-Tamil bigotry with which Bandaranaike the curiously naive, utopian Prime Minister was unable to cope (Manor 1989: 253). At an inter-group level of analysis, there is no doubt that ethnic chauvinism was harnessed and politicized by S.W.R.D. Bandaranaike. Nor is there any doubt that this consequently polarized the political system. However, it would be a mistake to explain Bandaranaike’s actions based solely on their outcome for this would overlook the intra-group dimensions of the episode. Inter-group animosities were harnessed by Sinhalese and Tamil leaders for use in the intra-group arena rather than the inter-group arena. Bandaranaike’s young political party was challenging the UNP for control of the Sinhalese community. By fostering ethnic chauvinism and capitalizing on existing networks and social institutions, Bandaranaike succeeded in wresting power from the UNP. A similar dynamic was evident in the Tamil community. The fear of Sinhalese domination was adroitly exploited by Ponnambalam in order to rally Tamil intra-group support. While it is important to recognize the inter-group consequences of this episode, it is equally important to recognize the intra-group compulsions as well. The following episode illustrates a similar dynamic, whereby fears of the opposing ethnic community were manipulated by competing ‘ethnic entrepreneurs’ to rally intra-group support for their intra-group battles.

6 Critical Juncture III: 1971 JVP Insurrection and 1987 JVP Resurgence

In the history of Sri Lanka, the only threat to state authority1 which exceeded the JVP insurrection of 1971 is the JVP resurgence of 1987–90. Although the conflict in the North and the East involves victims and combatants from all ethnic communities, it has never fundamentally threatened the state to the extent of the two JVP episodes. Both the insurrection and resurgence were essentially intra-group conflicts among state actors, the JVP, and a wide range of disgruntled Sinhalese sub-groups. Furthermore, the violence was geographically limited to the Sinhalese-majority areas of the island. The actors and the geography of violence thus ground these episodes more firmly in a Model II than in a Model I analytical framework. While this book argues the need to consider intra-group dynamics affecting critical junctures in ethnic conflicts, these episodes illustrate the need to consider also the inter-group dimensions of intra-group episodes of conflict.2 The strategy of the leadership of the JVP – like that of S.W.R.D. Bandaranaike (see preceding chapter) – intensified and harnessed the inter-group fears of the Sinhalese as a means of mobilizing resources for use in the intra-group arena. A central feature of JVP rhetoric in the late 1980s was the representation of, and response to, Tamil3 political and economic concerns such as language use and access to employment and educational opportunities. Representations of inter-group relations were central points of reference in intra-group mobilization by the JVP. This was frequently couched in JVP appeals to ‘patriotism’ and the ‘protection of the motherland.’ In addition to the rhetorical connection between the inter-group and intra-group arenas, there was a nuanced and generally neglected relationship between the JVP and 99


Tamil paramilitary organizations. Both of these facets will be addressed in this section in order to place the necessary emphasis on the Model I dimensions of the episodes.

1971 JVP insurrection The context It is possible to identify six distinct phases in the political life of the JVP: Phase I: Mid-1960s–19714 – a formative phase characterized by internal struggle for control, ideological coherence, and structure and building capacity to try to overthrow the state. Phase II: 1971–77 – the 1971 insurrection followed by government crack-down, imprisonment, political hibernation and ideological reformulation. Phase III: 1977–83 – the legalization of the JVP and a period of mainstream political activity and relative ideological moderation. Phase IV: 1983–87 – proscription, government crack-down, and political atrophy. Phase V: 1987–89 – the JVP resurgence, precipitated largely by the signing of the Indo-Sri Lanka Agreement, characterized by mass terror and unprecedented intra-group violence by the JVP and government forces. Phase VI: from 1990 – the killing or imprisonment of the JVP leadership by Sri Lankan security forces in November 1989 staunches the worst violence and pushes popular dissent back underground. It needs to be emphasized that a particular constellation of political factors shaped the character and actions of the JVP in each phase. The two most significant factors are the JVP’s ideological position at the time (the product of power struggles among internal factions)5 and its legal status in the political arena (whether or not it was proscribed by the party in power). The JVP was most notable for its mobilization of a large population of disaffected youth. From its inception in the mid-1960s, its leader, Rohana Wijeweera, attempted to mobilize support from some of the same sources cultivated by Bandaranaike, such as bhikkhus,6 ayurvedic practitioners, rural Sinhala-medium teachers. This was a rational strategy given the newly-activated political power of these groups which

JVP 1971 and 1987 101

was demonstrated in the 1956 election. Having been mobilized effectively by Bandaranaike, their support was sought by subsequent political actors attempting to capture political control. The demographic composition of the country in 1971 was conducive to a youth-driven insurrection.7 In 1969, 60 per cent of the total population of 12.5 million were under the age of 25. In the 14–25-year-old category, 45 per cent were out of school and unemployed. Approximately 74 per cent of the population were literate – the majority literate in only one language as a result of S.W.R.D. Bandaranaike’s policy of vernacular education. Despite the high Sinhala literacy rate, English remained the language of status and the key to getting employment in higher paying jobs. Most of the JVP activists in 1971 were products of the Sinhala medium education system. They possessed secondary education qualifications but had minimal opportunity for gainful employment. Most were drawn from rural areas (where more than 75 per cent of the population lives) and particularly from the ‘hinterland’ which had been neglected by government. Forty-one individuals were identified and charged before the Criminal Justice Commission (CJC) as the most important leaders of the 1971 Insurgency (Chandraprema 1991: 2). All 14 Politburo members of the JVP were Sinhala Buddhists, mainly from the strongly nationalist regions of the South. ‘With one exception, they were in their twenties, and had grown up during the years of the “Sinhala Only” agitation of the mid-1950s, the ethnic riots of 1958, and the mounting chauvinism against the minorities even of the Left parties in the early 1960s’ (Jayawardena 1990: 142). They also tended to be from the lower castes.8 In general, the JVP was a fluid, secretive, underground organization which lacked centralized control and incorporated a wide range of participants and groups that often emerged out of the particular local conditions. Chandraprema identifies the support base more specifically as a ‘middle-layer of Sinhala non-poor in the countryside’ – ‘a social stratum notorious for its small proprietor mentality, ideological narrowness and conservatism. Their comparatively favoured economic position enables them to imbibe the ideological heritage of the land and turn out to be its most ardent protagonists’ (1989: 20). In the context of the theoretical concerns of this book, it needs to be noted that just as Bandaranaike’s tapping of unmobilized Sinhalese sub-groups spurred subsequent political actors to draw support from the same sources, so did the 1971 insurrection consolidate the youth, and especially students, as a politicized sub-group in the marketplace


of potential political support. For example, some factions within the SLFP saw the rise of the JVP not as a potential threat, but as an opportunity to tap into new resources which might be harnessed for different political ventures. Anura Bandaranaike, son of S.W.R.D. Bandaranaike and leader of the ‘Anura faction’ of the SLFP (until he crossed the floor to the UNP in late 1993), is reported to have said that he organized the ‘Front to Assist Released Activists’ in the mid-1970s ‘to gather the finest Buddhist activists – the intelligentsia of this country the JVP managed to attract’ (Gunaratna 1990: 130). The danger of neglecting student concerns is illustrated in the student unrest which followed from the government’s banning of the student assemblies in the mid-1980s (Matthews 1989: 432–33; and Gunaratna 1990: 47). In addition to opportune demographic factors, the economic crisis of the late 1960s created conditions conducive to discontent: increasing unemployment; decline of foreign exchange earnings (which led to import controls); deteriorating terms of trade; and a growing trade deficit which was not bridged by foreign aid. As the government borrowed heavily from abroad, subsidies and welfare services were cut back. In November 1968, under IMF pressure, the rice subsidy was reduced and the rupee devalued (Jayawardena 1990: 141; Ivan 1989: 15–20). Such economic flux led to rising prices and diminishing standards of living. Overall, the period is characterized by ‘continuous, militant student agitation in the universities and the youth of the country’ (Jayawardena 1990: 141–42) and increasing labor unrest in urban and plantation sectors in private and state-owned firms. Perhaps the most important factor contributing to the 1971 insurrection was the politics of exclusion which predominated on the island. Jiggins describes it as follows: The ‘Government’ was paternalistic. Its activity was seen as something done by ‘them’ for ‘us’ rather than something in which the ordinary individual could participate effectively. If one was of the wrong family, the wrong caste, and had no finances, there was no chance of standing for political office. Since they could not change the system from within, because they could not enter it effectively, they would attack it from without. … In April 1971 the insurgents attacked the formal political structure as an inadequate parliamentary game played in the interests of elite families and bourgeois parties that had skilfully manipulated traditional social relationships to their own exclusive advantage (Jiggins 1979: 126).

JVP 1971 and 1987 103

The episode The events leading up to the 1971 insurrection and details of the insurrection itself have been adequately addressed elsewhere.9 In brief, the insurrection was premised on the belief that it is possible to launch and win a one-night revolution. By all accounts, the JVP in 1971 was well-armed with grand ideas and youthful enthusiasm, but poorly armed with explosives, weapons and training. It was also riven with factional in-fighting.10 Nonetheless, the nine senior members of the JVP decided to launch a mass, concerted attack against what was viewed as the most threatening state actor – the police force.11 However, when JVP insurgents prematurely attacked a police station in Monaragala in the early morning hours of 5 April, the crucial element of surprise was lost. This allowed the government to rally its forces and respond to the large but uncoordinated insurrection as it broke out across the island in attacks on 92 police stations (Gunawardena 1990: 104). Prime Minister Sirimavo Bandaranaike declared a state of emergency and received military assistance from a variety of countries (including Britain, the United States, India, the Soviet Union, Egypt, Yugoslavia and Pakistan). By the end of the month, the insurrection was put down at considerable physical, political, and economic cost.12 According to the official account, 53 security forces personnel were killed and 323 were injured. The most likely estimates of the number of people killed during the April insurrection are between 8,000–10,000. But it will never be possible to obtain or determine the precise number of fatalities and casualties.

1987 JVP resurgence13 The context The failure of the 1971 Insurrection resulted in the deaths and imprisonment of thousands of youths. Some of them were released after their ‘rehabilitation’ but many remained imprisoned under inhuman and often brutal conditions (Chandraprema 1991: 39–43) until the 1977 election when the new Prime Minister, J.R. Jayawardena, released all JVP activists from jail. The reasons given for the release of the JVP vary. Some have argued that the release was precipitated by a bureaucratic oversight: that the emergency regulations of the Public Security Act (under which most were being held) could not be extended once parliament was dissolved in order to hold a general election (Gunaratna 1990: 136). Wijeweera is reported to have explained: ‘we were not


released from the prison because of the magnanimity of this government or that of the capitalist class but because of the steadily mounting pressure from the working class for the release of the members of the JVP and other political prisoners’ (Gunaratna 1990: 141). Echoing Wijeweera in more diffuse terms, former President Jayewardene is reported to have told Gunaratna that ‘if I did not release him, he would have become like Nelson Mandela. There was feeling growing among our people, and I could not have resisted it’ (ibid.). Once legalized (1977–83), the JVP entered a period of mainstream political activity which culminated in Wijeweera’s 1982 presidential bid. This was a period of relative ideological moderation in which the JVP officially rejected revolution and ‘extra-electoral struggle.’ The salient point preceding the 1987 resurgence is that the moderating trend which might have led to the ‘domestication’ of the JVP was cut short by a decision by the UNP government of Jayewadene to proscribe the JVP again on the unlikely charge that it had been involved in organizing the 1983 riots (see Chapter 7).14 It was further alleged that the JVP had initiated the riots as part of a ‘three-stage plot to overthrow the government’ (Gunaratna 1990: 189). However, as the 1983 Episode in that chapter describes, there is considerable evidence that the 1983 riots were heavily organized, if not instigated, by specific government actors. A more likely explanation for the proscription is that Jayewardene was concerned with the growing political power of the JVP as reflected in its electoral victories in the 1981 District Council elections (Gunawardena 1990: 152) and the 1982 Presidential Election in which Wijeweera achieved third place, ahead of all other Leftist candidates.15 A militaristic May Day Parade by the JVP was also identified by Jayewardene as a significant factor in his decision to proscribe the organization (Gunawardena 1990: 187). In retrospect, the proscription was not only a mistake (because it pushed the JVP underground and left ‘extra-electoral struggle’ as the only option available to it), but it reflects a pattern of political decay in which the members of legal opposition political parties are harassed and intimidated while legal recourse is simultaneously closed to them due to political manipulation. If such measures do not sufficiently stifle the activities of the political parties, then proscription of one form or another may be imposed. This was the general pattern for the suspension of the civic rights of Mrs Bandaranaike (1980), the expulsion of the TULF from Parliament (1983), and the proscription of the JVP in 1983. The economic, demographic and political conditions which proved conducive to the rise of the JVP in 1971 was similar in 1987.16

JVP 1971 and 1987 105

Economic indicators for 1989 show an unemployment rate of 20 per cent which was not evenly distributed in the population. Seventy per cent of those unemployed were youths between the ages of 15–20 years. In the regions which proved to be strongholds of JVP support in 1971 and 1987 (Southern Province, North Central Province and Uva Province), unemployment was particularly severe. Seventy-five per cent of the entire district of Hambantota subsisted on food stamps, and industry and agriculture had atrophied from the minimal levels they had attained. Matthews describes the two periods as follows: Although much has changed in Sri Lanka since the JVP-led insurrection of 1971, most of the social and economic causes of that trauma remain – notably, widespread poverty, unemployment, lack of social and economic mobility, landlessness, village isolation and alienation from a political system found by many to be dishonest and opportunistic. To these must be added several new factors, such as the secessionist war in the north and the east and the negative effects of the UNP’s ‘open economy’. The JVP has been able to monopolize much of the economic and political discontentment undermining the traditional left-of-centre [parties] in these matters (Matthews 1989: 427). Thus, the heartland of 1971 JVP ‘patriots’ remained the basis for the 1987 JVP ‘patriots.’ The episode Although proscribed in 1983, the JVP began preparing itself militarily only from late 1984 when the government crack-down began to wane and as attention was redirected to the increasingly violent ‘Tamil Problem’ in the North and the East (see Chapter 7). Although economic, social and political grievances were widespread, it was the signing of the July 1987 Indo-Sri Lanka Agreement which catalysed the JVP resurgence.17 A description of events from 1987–90 reads as a uninterrupted list of escalating atrocities.18 From 1987 to early 1990, the Sinhalese South of the island was paralysed politically, socially and economically by the seemingly limitless levels of brutality and violence committed by security forces and the JVP. No one was immune; children, the elderly, women and monks were all victims. The JVP attacked all levels of government and public employees, ‘collaborators,’ and anyone who violated JVP orders (such as curfews, boycotts, communicating with or selling food to security forces, and so on).19 The


definition of legitimate targets later expanded to include ‘social deviants’ such as illicit alcohol sellers, alcoholics, and husbands who beat their wives. The specification of targets by security forces in antiJVP operations were similarly elastic: anyone with possible JVP linkages. It was not uncommon to have 10–20 individuals abducted simultaneously by men armed with Sri Lankan Army weapons dressed in civilian clothes. Disappearances were rampant. Women and children were held hostage until their fathers, sons or brothers presented themselves to the captors – often never to be seen again (Gunaratna 1990: 288).20 ‘Show killings’ were instrumentally undertaken by JVP, state and quasi-state forces (such as vigilante squads or death squads composed of out-of-uniform government forces) to instill fear and to command acquiescence or compliance.21 The effectiveness of ‘show killings’ was tied to the level of terror and revulsion they could induce. Thus, mutilation and desecration were central communicative elements in the murders. Meeting places and crossroads were frequent sites for the display of horribly mutilated corpses: women with breasts cut off; corpses without limbs or genitals; gouged or burnt out eyes, ears, noses, and lips. Headless bodies and severed heads were placed in central locations with messages forbidding their removal – in one incident in Kandy, the severed heads of Sinhalese youth were placed around the lake in the center of town and in another incident the heads of 18 students were displayed neatly around the Peradeniya University pond (Chandraprema 1991: 296–7). Often desecration applied not only to the death of victims but to burial rites as well – sometimes leaflets left on the corpse(s) specified how the last rites should be carried out. Failure to obey such rules summarily initiated further disciplinary murders. Such symbolic prohibitions were intended to demonstrate that JVP power extended beyond even life and death and thereby instill the fear necessary for control.22 Not all violence during this period was directly JVP-related. The conditions of lawlessness and chaos provided the opportunity for the settling political and criminal scores. Particularly important to note are the allegations by opposition parties, particularly the SLFP, that the UNP government of the late President Premadasa organized death squads which murdered SLFP MPs and supporters under the cover of JVP activities. In the spring of 1992, P. Udugampola, the former Deputy Inspector of Police for the Southern Province, claimed that he was aware of a UNP-sponsored death squad called the ‘Black Cats’ which had killed over a thousand political opponents, including SLFP supporters in 1988–89.23 As Deputy

JVP 1971 and 1987 107

Inspector, Udugampola was in a position to know of vigilante operations and he provided affidavits which included lists and details of people who had been murdered. Following this disclosure, the government charged him with bringing the government and certain individuals into disrepute and Udugampola, fearing for his life, immediately went into hiding.24 The beginning of the end of the resurgence was precipitated by a tactical error on the part of the JVP. Before August 1989, the JVP had actively recruited and sought the support of ‘patriotic soldiers’ within the Sri Lankan Armed Forces with some success (e.g. Gunaratna 1990: 221–5). Off-duty soldiers were sometimes used in JVP operations – just as they were used in pro-government vigilante operations. The disbanded Rajarata Rifles Regiment have been identified as the founding members of the military wing of the JVP (Marino 1989: 9).25 As noted in Chapter 3, there are newspaper reports of the dismissal of soldiers for having JVP connections and of ‘inside help’ in the thefts of weapons from military installations. The JVP rhetoric had been very careful to emphasize that the organization viewed soldiers as victims of political pressure from an unpatriotic state. JVP leaflet campaigns stressed, ‘that they [the JVP] mean the armed forces no harm, and that they have been raiding army camps to obtain firearms. … The armed forces are seen as fellow “patriots” because they have been at the forefront of the anti-Tamil struggle’ (Chandraprema 1989: 22).26 The JVP sought to win them over (preferably fully armed and capable of delivering more weapons). In August 1989 the JVP made a critical mistake when it changed its tactics and began issuing death threats against all security forces and especially their families (Gunaratna 1990: 269; Chandraprema 1991: 293–303). The JVP Politburo apparently thought that this would induce the ‘patriotic sections of the armed forces’ to defect en masse to the JVP (Chandraprema 1989: 294). But the tactic had the opposite effect; it repelled potential supporters and increased rank and file antagonism towards the JVP. In response, a thinly veiled Army death squad issued posters announcing ‘Ape ekakata thope dolahak!’ (‘Twelve of yours for one of ours!’); in other words, twelve to one retaliation for JVP killing of security forces or their families.27 The twelve to one retribution was met and often exceeded, as illustrated in the reprisal slaughter of 200 people in a village in the Menikhinna-Kundasale area after the JVP hacked to death the families of three servicemen (Chandraprema 1991: 296). In August-September 1989, JVP murders and government force reprisal killings would leave hundreds of corpses scattered throughout the island every morning.


Chandraprema suggests that the sheer volume of slaughter by the government forces weakened JVP support and caused thousands of JVP activists to surrender (1989: 297). It is also true that, as the top members of the JVP Politburo were arrested, the JVP leadership began to crumble. One leader after another was arrested, led the police to more leaders, and was then summarily killed while allegedly trying to escape or a similar pretence (Chandraprema 1991: 304–311). On 12 November 1989, Wijeweera was captured and fully cooperated with the police – as he had following the 1971 Insurrection.28 Wijeweera ‘volunteered’ to make a videotaped statement appealing ‘JVP membership and friends to eschew violence and help maintain law and order at a time of severe stress.’ But by the early morning of the day following his capture, Wijeweera was dead; killed while allegedly trying to escape from his captors. The JVP violence festered into early 1990 but petered out as the secondary and tertiary levels of JVP leaders were killed by government forces. Model I In contrast to the discussion of other critical junctures identified in this thesis, the two periods analysed in this section illustrate the need to consider the inter-group dimensions of intra-group conflicts. The current discussion focuses on two particular facets of inter-group relations: first, the shifting position of the JVP on ‘the Tamil Question,’ and second, JVP links with Tamil sub-groups, particularly Plantation Tamils and Tamil paramilitary organizations. Through an examination of these two matters some important nuances of the conflict become evident. The ideological oscillation of the JVP on ‘the Tamil Question’ and its shifts in its definition of root political problems opened or closed channels of potential resource mobilization in the intra-group arena. Thus, for example, the recasting of the JVP program from a ‘class struggle’ to an anti-Indian, anti-Tamil, anti-state ‘patriotic struggle’ expanded its potential support constituency to any sub-group whose members felt that the state and its actors were not defending the interests of the ‘Sinhalese nation;’ no longer was it confined to a classdefined support constituency.29 The shift in the representation of inter-group relations had direct implications for the intra-group conflict. JVP links with Tamil sub-groups introduce the awkward question of explaining the participation (albeit limited) of Plantation Tamils and Tamil paramilitary organizations before and during the 1987 JVP resurgence. An explanation must draw on both Model I and Model II analy-

JVP 1971 and 1987 109

sis. The fact that a number of Plantation Tamils were directly involved in the JVP struggle and that there were some connections between the JVP and Tamil paramilitaries does not strip the JVP of its Sinhalese chauvinist elements, but it does illustrate that the inter-group battle lines of ethnic conflict are not so rigidly hard and fast as often suggested. The instrumentalities of ideology The examination of JVP ideology is an important starting point for understanding the JVP phenomenon but it is necessary to compare this with JVP actions. The mechanism by which it initially attracted followers was a series of five lectures presented in small discussion groups and on audiotapes.30 The ideological basis of the JVP was articulated in these lectures. In the pre-1971 period, the lecture on ‘Indian Expansionism’ underscored the staunchly anti-Indian posture of the JVP’s ideology and, by extension, its mobilizational strategy. Included in this lecture was the JVP view of the Plantation Tamil Workers. In brief, the presence of Plantation Tamils on the island was portrayed as both a legacy of colonial rule and a product of the continued pursuit of ‘imperialist interests by foreign powers.’ The Plantation Tamils were portrayed as living in the ‘best parts of Sri Lanka’ and enjoying ‘benefits like housing, education and health facilities.’ Their conditions and living standards were construed as being superior to those of Sinhalese peasants and their cultural and social links were said to reside in their ‘homelands’ in India (Jayawardena 1990: 145). The incongruity of this image was as striking then as it is now, since the Plantation Tamil workers continue to live in some of the worst conditions in the country. This particular representation of Plantation Tamils prior to 1971 not only excluded them from participating in the JVP political program, it placed them as a group in opposition to the Sinhalese. First, the JVP conspiratorially (and anachronistically) implicated Plantation Tamils in a process of continuous Indian expansionism beginning in the 4th Century B.C.31 And secondly, it threatened to destroy the livelihood of most Plantation Tamils by promising to dismantle the plantation sector in order to build a self-sufficient Sri Lankan economy. Whatever this might have meant, it resonated with disgruntled and unemployed Sinhalese youth. In the pre-1971 lectures, Plantation Tamil workers were presented only as a component of Indian expansionism. A commentator in the Lanka Guardian accurately expressed the inconsistency of this formulation:


This posture of the JVP was so absurd that the most exploited segment of the Sri Lankan proletariat was portrayed as an agent of the Indian monopoly bourgeoisie, while simultaneously counterposing it to the other exploited sections of our society such as chena cultivators. So much for the worker peasant alliance (15 January 1979, quoted in Jayawardena 1990: 146). This chauvinist outlook was similarly applied to other non-Sinhalese groups. They were treated as ‘one homogeneous group with an extraterritorial loyalty. No attempt was made to distinguish between Tamils on a class basis or even on the basis of ‘Sri Lanka’ and ‘Indian’ Tamils’ (Jayawardena 1990: 146–7). Following the failure of the 1971 insurrection, the JVP shifted its position on the ethnic question and refocussed more broadly to incorporate class dimensions. The lecture on ‘Indian Expansionism’ was dropped from the Five Lectures and program of instruction. This opened the opportunity for the JVP to cultivate linkages with Tamil sub-groups. By 1977, dominant factions within the newly-legalized JVP adopted the position that ‘the Tamil people’ were an oppressed ‘nation’ and therefore were entitled to the right of self-determination even to the point of secession. During the 1977–83 period, Lionel Bopage, the JVP Secretary General, advocated a slightly more centrist position on the ethnic question arguing that territorial autonomy was the only viable solution to the ethnic conflict (Chandraprema 1989: 10). This led to the development of linkages between some Tamil subgroups and the JVP; for example, Tamils appeared in the lists of candidates put forward by the JVP for district council and municipal elections in 1980 and 1981; as well, the JVP candidate for deputy mayor of Colombo was a Tamil (Jayawardena 1990: 149). During this period, JVP statements on language and religion became quite liberal. All ethnic and caste discrimination was criticized and condemned (Samaranayake 1987: 282). Gunaratna writes: ‘Wijeweera also said that the State should not make any pronouncement on state religion or state language. The JVP would give freedom of worship, but would not allow temples, churches or mosques to possess acres and acres of land. Students would be taught in all three languages, up to university level. They would be free to choose the medium they wished to study in as far as the JVP government was concerned. If a person wrote in French, he would be replied to in French’ (Gunaratna 1990: 141). The new liberalism and moderation of the JVP stretched to considerable limits when it expressed dissatisfaction with the 1964 and 1974

JVP 1971 and 1987 111

Plantation Tamil repatriation agreements because they were formulated without any form of consultation with the Plantation workers. ‘Equal rights for Tamils on the basis of language, employment and land settlement were clearly advocated and all forms of discrimination based on race or caste was (sic) firmly condemned’ (Samaranayake 1987: 284). This ‘new openness’ was reflected organizationally as well, and the JVP began to work in the hitherto neglected Tamil dominated areas, especially Jaffna (ibid.). Although the public image of the JVP during the 1977–83 period was relatively accommodating and liberal, there was a widening of ideological opinion within the leadership (Matthews 1989: 429). This became evident in rallies, party meetings and the party’s official paper. Particularly contentious was the debate over the JVP’s position on the ‘ethnic’ or ‘national’ question. Despite the JVP General Secretary’s accommodating stand on the issue, others in the organization marginalized the secessionist struggle and focused instead on broader socioeconomic problems (Matthews 1989: 429). In general, the JVP absorbed such diversity and attempted to avoid taking a public stand which would alienate particular segments of its support base. This noncommittal position disappeared following its proscription in 1983. In light of the apparent moderation by the JVP while in the mainstream political arena, its transformation following proscription in 1983 is striking. However, given the 1971 insurrection and the JVP’s origins with a hybrid revolutionary ideology, it is not as surprising as suggested by Chandraprema when he writes: ‘The transformation of the JVP from an accepted left wing formation to a blatantly chauvinist vanguard group is probably the most bizarre political volte face in recent Sri Lankan history’ (1989: 10). Nonetheless, the accommodating position toward Tamils in the North, East, and Up-Country disappeared following proscription. Centrists represented by Lionel Bopage were purged from the organization, and a new, more antagonistic, position on the Tamil Question was overtly adopted, though as the next section illustrates, linkages with Tamil sub-groups were covertly cultivated.32 By equating political devolution with the first stage of Tamil secession, Wijeweera foreclosed the possibility of increased Tamil autonomy as a means of managing inter-group conflict. In 1984, he melodramatically stated: ‘The JVP and I are totally opposed to any imperialist attempt to divide the country. As long as the JVP exists, as long as I live, we shall not allow any imperialist force to divide the country.’33 The United States was identified as one of the principal ‘imperialist


forces’ manoeuvring to separate the country so that it could more easily control both halves (Gunaratna 1989: 13). Accordingly, Tamil paramilitary organizations were all viewed as ‘instruments of the CIA.’ Despite this public position, the JVP maintained an expedient relationship with Tamil groups, particularly when it facilitated the acquisition of arms and training. The 1983 riots had a significant impact on the JVP. As an organization, the JVP was probably not involved in instigating or organizing the riots, although individual JVP members may have participated. Nonetheless, the riots provided the UNP government with the pretext for proscribing the JVP. Importantly, for the future direction of the organization, the riots suggested to Wijeweera a way to mobilize resources. Premachandra observes that when the organization went underground in 1983, ‘an important change took place within the JVP within the first four or five months of its proscription. The habitually opportunistic and impatient leadership of the JVP, and Wijeweera in particular, had seen the enormous power of Sinhala chauvinism for the mobilization of the otherwise apathetic masses’ (1989: 13). At this stage, the political platform of the JVP shifted to rest solidly on an antagonistic approach towards both Plantation Tamil and Sri Lankan Tamil claims as a means of mobilizing Sinhalese intra-group support. The characterization of the post-1983 JVP as a Sinhalese chauvinist organization follows from a reading of its stated ideological platform. Although it mobilized intra-group support on the basis of anti-Tamil sentiment, this did not preclude the mobilization of intra-group resources arena through surreptitious linkages with Tamil sub-groups. JVP–Tamil sub-group linkages Chapter Two briefly discussed the ‘delicate’ relationship between the JVP and the various Tamil paramilitary organizations. Matthews suggests circumstantial links between the JVP and the LTTE: ‘The JVP never attack the Tigers and appear to have respect for the way in which LTTE leader V. Prabhakaran has faced great military odds over the years with skill and courage. There may even be communication and transfer of information and explosives technology between the two groups’ (Matthews 1989: 430). The ‘explosives technology transfer hypothesis’ of Matthews is further suggested (but not confirmed) by the JVP use of the same powerful landmines that the LTTE has used with such impact in the north since 1983 (Matthews 1989: 431). Gunaratna (1993) provides more specific details of the JVP-Tamil militant nexus. As discussed below, Sinhalese and Tamil ‘revolutionaries’ have trained

JVP 1971 and 1987 113

together in South India – offering further substance to allegations of information transfer, if not direct technology transfer. The relationship between Tamil paramilitary organizations and Sinhalese revolutionary organizations was kept low-key. The Sri Lankan government, Amnesty International, and various news agencies have suggested that cooperative arrangements have been forged between such sub-groups across inter-group boundaries. Gunaratna offers details of some of the connections between JVP splinter groups and Tamil paramilitaries organizations (1990: 127–35). Details of these linkages were made available to the author during interviews with a range of individuals in Sri Lanka (May–July 1992) including a series of meetings with the ‘co-organizer’ of an organization called Vikalpakandayama which he described as ‘a politico-military organization dedicated to the revolutionary path to social change.’ According to Gunaratna, Vikalpa is a ‘significant clique which attracted many revolutionaries from the JVP’ (1990: 131). The ‘co-organizer’ explained that Vikalpa explicitly sought to link the paramilitary actors in the North and the East with those in the South.34 It was predominantly Sinhalese in membership but included initially a Plantation Tamil organizer and K. Padmanabha, leader of the EPRLF.35 The EPRLF not only provided Vikalpa with weapons and explosives, it also provided military training for some of its Sinhalese cadres alongside the EPRLF in Southern India (personal interviews, Sri Lanka, May–July 1992). An interview with one of the highest leaders of the PLOTE also revealed direct JVP–Tamil paramilitary linkages. It was explained that beginning in 1983 PLOTE provided the JVP with military ‘help’ (i.e. training and supplies).36 The PLOTE-JVP relationship is also discussed by Gunaratna: It is alleged that the [PLOTE] and the JVP enjoyed close links till 1988 where transfer of weapons, explosives and the transfer of landmine technology took place. PLOTE was given 12X10 lorry engines by a Sinhala group which they could use in their boats. Government intelligence reveals that members of the JVP were trained in the North and PLOTE ran camps for members of the JVP in Southern Sri Lanka. It is on record in the late 1980s a prominent politician in Colombo helped [to foster] the JVP–PLOTE friendship … through an appeal made to him by an articulate PLOTE representative D.P. Sivaram, now underground following Uma’s assassination [Uma Maheshwaran, PLOTE leader]. It is on record that the first landmine that killed a police officer in the South was the work of a


PLOTE operator. … The JVP mine technology was no different to those (sic) of the northern group (Gunaratna 1990: 133).37 There is sufficient evidence to make the case that there were sustained, though covert, contacts between the JVP (and JVP-related organizations) and Tamil paramilitary organizations. However, these contacts did not coalesce into an inter-group, anti-state, revolutionary coalition. These linkages are significant because they underscore the permeability of both inter- and intra-group boundaries even under conditions of extreme violence. Specifically the collaboration highlights the following points: intra-group divisions permit cooperation and coalition-formation among sub-groups across the inter-group divide when intergroup resources can be employed in intra-group conflicts; ‘primordial’ ethnic chauvinism becomes salient through instrumental mobilization; and the Kautilyan dictum applies to ‘ethnic’ as well as conventional conflicts – ‘the enemy of my enemy is my friend.’ Model II The presentation of the two episodes illustrates the intra-group character of the 1971 insurrection and 1987 resurgence. The discussion in the Model I section above places needed emphasis on two important facets of the inter-group dimensions of the episodes. This section will discuss some intra-group issues not yet introduced. Resource mobilization: redefining problems, finding new resource bases There was a significant ideological shift within the JVP which enabled it to seek support from a broader network: instead of advocating revolutionary class struggle, the JVP advocated revolution through patriotic struggle (Gunaratna 1990: 232). This is a crucial ideological reorientation which expanded the potential set of intra-group resources available to the JVP. It enabled Wijeweera to build a deeper, more varied resource base – a prerequisite for the protracted conflict of 1987–90. By reframing the insurrection as a ‘patriotic struggle,’ the conflict was redefined – although, importantly, ‘state actors’ remained central targets (but by no means the exclusive targets) of the JVP in both episodes. In 1971, the attack was narrowly class-based, and the state was conceived as the representative as self-aggrandizing class interests. In 1987, while some of the class rhetoric persisted, the state was more broadly attacked as illegitimate – not because it represented the interests of a particular class, but because it did not represent or defend the

JVP 1971 and 1987 115

interests of ‘the nation.’ By shifting the terms of conflict, the JVP shifted the basis of its mobilization and sought support and resources across class boundaries. To the extent that the resources it cultivated were voluntary, the JVP were thus able to elicit the support of any individual or actor who was disgruntled with the state, rather than having to rely narrowly on disaffected Marxists from the fragmented left. State action and student dissent The 1987 resurgence, like the 1971 insurrection, mobilized the frustration and anger of the youth. Despite the passing of years and changes in government, the youth faced the same limited prospects as their cohorts in 1971. The 1971 insurrection was an illustration of what could happen when the concerns of a large segment of discontented youth are neglected. Yet the message was not heard or the lesson was not learned, and the frustration and disillusionment of the youth was left to smoulder until it could be re-mobilized and expressed through reinvigorated JVP channels. The Jayewardene government aggravated this process in 1982 when, in response to growing student organization and militancy, it banned all student councils and thereby closed formal channels for expressing student grievances and dissent. The assemblies and action committees which subsequently were formed by the students were also not recognized by the State. They therefore had no formal status or impact. The closing of official student organizations provided an opening for JVP infiltration. Unofficial student organizations became an important institutional conduit for JVP mobilization of support from the student sector (Gunaratna 1990: 47; Matthews 1989: 432). In 1987–90, as in 1971, the universities became centers of JVP support and anti-government activity.38 JVP infiltration allowed a relatively small group to effectively control the universities. The JVP successfully forced the universities to shut down until the government met a series of demands (Matthews 1989: 436). Had the government not silenced the traditional means by which students expressed their concerns and grievance, the JVP would not have had such convenient access for the mobilization of the student bodies.39 Opportunity structure Prior to 1983, the Tamil Eelam struggle was relatively low in intensity and therefore not central on the political and psychological landscape of most Sinhalese. This lessened its salience as a mobilizational tool within the Sinhalese community. After the 1983 riots, however, the


escalating conflict with Tamil paramilitaries in the North and East increased the visibility and salience of ‘the Tamil Question’ in the Sinhalese intra-group arena. This served to make segments of the Sinhalese community receptive to an organization mobilizing on an anti-Tamil platform. The JVP was not the only actor trying to harness anti-Tamil sentiment; government actors, particularly the UNP and the SLFP, were also engaged in anti-Tamil activities at the same time (see Episode IV). The government’s signing of the extremely unpopular Indo-Sri Lanka Agreement provided the JVP with the opportunity to challenge the legitimacy40 of UNP and its claim to be the genuine representative and protector of Sinhalese interests and the dhammadipa (‘the island of the teaching of the Buddha’). This allowed the JVP to mobilize support from elements of the public which expressed antiIndia and anti-government sentiment that resulted from the signing of the Indo-Sri Lanka Agreement. Without the Agreement, it is difficult to imagine that the JVP would have been able to resurface which such destructive force. The Agreement transformed the opportunity structure and thereby made available to the JVP resources that increased the effectiveness of its anti-government activities. The Accord ‘encourag[ed] unrepresented groups to protest in the belief that the costs of insurgency [had] been lowered’ and provided a focal point which ‘enabl[ed] them to overcome their habitual disunity’ (Tarrow 1989: 36). Conclusion: the utility of inter-group antagonism in intra-group competition Sinhalese chauvinism in Sri Lankan politics is not simply a function of the interaction between Sinhalese and Tamil sub-groups. As a Model II analysis of the JVP espisodes illustrate, even in the JVP’s post-1983 anti-Tamil phase, its contact with Tamil sub-groups was more cooperative or coalitional than conflictual. The conspicuous chauvinist element of the JVP – often wrapped in the vocabulary of ‘patriotism’ – was more a function of intra-group rather than of inter-group relations. Sinhalese chauvinism was a mobilizational tactic in the intra-group arena. While such mobilization has the potential to sour inter-group relations, its origins are in intra-group conflict. By implication, a focus on the elimination of chauvinism as a means of managing ethnic conflict must consider both its intra-group origins, its inter-group consequences, and the interaction between the two. To the extent that the JVP was seen to be successful in mobilizing support on the basis of anti-Tamil sentiment, its competitors, in partic-

JVP 1971 and 1987 117

ular the state, sought to mobilize on the same basis – setting in motion a spiralling dynamic of intra-ethnic group competition which pulled violent extremism from the margins into the mainstream. For example, despite the devastating violence of the 1987 resurgence, both the SLFP and the UNP refrained from directly criticizing the JVP in the run-up to the 1988 presidential elections. In April 1988, Premadasa offered to vacate some of his Party’s seats in Parliament in order to give the JVP parliamentary representation as an inducement for the JVP to lay down their arms (Gunawardena 1990: 304). Similarly, the SLFP went so far as to offer the JVP three ministries in the government it hoped to form at a time when the intra-group violence had escalated to unprecedented levels (FEER, 27 October 1988: 30). Both Premadasa and Bandaranaike publicly stated that there was ‘no proof’ that the JVP was behind the violence – blaming instead the DJV (Deshapremi Janatha Viyaparaya, or ‘Patriotic People’s Organization’), the military wing of the JVP. Further, Premadasa released a number of JVP detainees as a conciliatory gesture to the JVP (ibid.; FEER, 22 December 1988: 23) and offered to dissolve parliament in preparation for a General Election if the JVP would participate in a national caretaker government throughout the presidential and parliamentary elections (FEER, 24 November 1988: 38). The extremist strain of the Sinhalese nationalism is illustrated in the JVP’s proposed solution to ‘the ethnic problem’: the ‘reallocation of the country’s population so that the Sinhalese and Tamils would be dispersed all over the island’ (Jayawardena 1990: 147). This position echoes a proposal by the late Lalith Athulathmudali, the former Security Minister, who said that if he won the presidency he would apply a ‘solution’ derived from the communist insurgency in Malaysia: the ‘translocation’ of the Tamil residents in the Northern Province to the Sinhalese South – particularly the youth.41 Athulathmudali’s proposal is but one example of the movement of extremist claims and programs from the margins into the political mainstream. The phenomenon is also seen in S.W.R.D. Bandaranaike’s Sinhala-Only policy and the ‘counter-insurgency terror’ of the security forces in the South in 1987–90. In parliamentary debates, Cyril Mathew, the former Minister of Industries, is reported to have said: ‘Terrorism cannot be stopped and has never been stopped by the means of the law. Terrorism has been stopped by terrorism. In no other way is it possible. … Terrorists have to be killed because they are terrorists. They are like mad dogs and no better than that’ (quoted in MIRJE 1980: 19).

7 Critical Juncture IV: 1983 Riots

All studies of contemporary Sri Lankan politics identify the ‘anti-Tamil riots’ of 1983 as a critical marker in inter-group conflict. The episode was followed by an increase in the abuses and excesses that have come to characterize the conflict in both the inter-group and intra-group arenas, most evident in human rights violations by military and paramilitary forces. The episode was also followed by a rapid militarization of the conflict1 and a consequent increase in the size, number and strength of Tamil paramilitaries; further decline in the rule of law and democratic institutions; and an increased censorship and propaganda (Rotberg 1999; Gamage and Watson 1999; Hyndman 1987, 1988; Leary 1981; Hoole et al. 1990; Rubin 1987; Marino 1987).

The context The 1977 landslide victory of the UNP of J.R. Jayewardene over the Sri Lanka Freedom Party was followed by an outbreak of post-election violence. Although it began as a confrontation between UNP and SLFP supporters, it quickly was directed toward both Sri Lankan Tamils resident in the Sinhalese south and Plantation Tamils in the Central Province. Perhaps because the defeated SLFP had earlier extended its rule under unpopular emergency legislation, Jayewardene was hesitant to declare a state of emergency. The island-wide attacks continued over a period of two weeks resulting in an estimated 300 deaths and the displacement of 35,000 people – most of whom were Tamils (USCR 1991: 6). Thousands of Tamil refugees were pushed northward in search of security. The violence contributed to Tamil distrust of UNP promises to redress to Tamil grievances as promised in the UNP election manifesto. 118

1983 Riots 119

The frustration of Tamil youth2 in and around Jaffna was gradually mobilized by the angry young men who would become the leaders of the Tamil paramilitary organizations – such as the high school dropout named Velupillai Prabhakaran who in the early 1970s organized 30 fellow teenagers into a group calling itself the Tamil New Tigers (later the Liberation Tigers of Tamil Eelam). With the murder of the pro-government mayor of Jaffna in 1975, the Tigers began to establish their ‘credentials’ and make their presence felt. The Tigers and other nascent Tamil paramilitaries (in particular PLOTE and TELO) maintained a relatively low-level disturbance through a succession of attacks on Tamil policemen and Tamil public officials in the Northern Province. The 1976 ‘Vaddukodai Resolution’ by TULF calling for a ‘free sovereign, secular socialist state of Tamil Eelam,’ served to further bolster the radicalism of the Tamil youth3 and heighten fears of many in the Sinhalese community. In 1978, Tamil militant groups killed a number of police officers and carried out a series of bank robberies. Hoole et al. (1992: 23) describe the state of affairs in the North at that time: ‘As a purely security problem, the Tamil militancy had gone beyond routine policing, but as a political problem, it was well within control. The TULF was willing to settle for a fairly modest grant of autonomy for the Tamil areas which included some compromise on land settlement. The militants at this point in time respected TULF and were not challenging it. But the government decided to play tough, and given some of the racist attitudes of some of its leading members, every action of the government’s began to be seen as punitive.’ Thus, in July 1979, the government declared a state of emergency in Jaffna and passed the Prevention of Terrorism Act (PTA). Although the PTA was enacted in 1979, it could be applied retroactively to any act committed before that date – in contravention to the International Covenant on Civil and Political Rights. Emergency regulations progressively destroyed normal patterns of life in many parts of the North and East. The consequent hardship and deprivation led Tamils from these areas to seek opportunities and asylum in India and elsewhere.4 The wide-ranging emergency regulations facilitated systematic human rights abuses by state actors. State authorities were permitted the following: arrest without warrant; incommunicado detention; the taking of persons ‘to any place’ for interrogation; the restriction of an individual’s movements and activities; detention without trial for up to 18 months in any place, under any conditions, directed by the ‘Secretary of the Minister;’ security force disposal of corpses without inquest or post-mortem; and seizure


of property and vehicles.5 Thousands of Tamils were rounded up and brutally interrogated. Human rights abuses by security forces proliferated (Movement for Inter-Racial Justice and Harmonies, MIRJE 1980, 1981, 1983; Leary 1981; Amnesty International 1983). Sri Lankan NGOs and international human rights monitoring organizations have thoroughly examined and criticized the formulation and implementation of such legislation.6 In what appears to have been a two-track policy (or a ‘carrot and stick approach’), the UNP government also pursued the idea of District Development Councils (DDC) as a means of responding to Tamil demands for increased regional autonomy through the decentralization of government power.7 A Presidential Commission appointed to develop the DDC system adopted the scheme proposed by the dissenting member, the TULF appointee. This was a positive development as it was ‘the first time that a predominant Sinhalese Government had implemented any measure of decentralization which had been approved by the Tamil political leadership, though not as a solution to their long-term demands’ (Coomaraswamy, quoted in Leary 1981: 41). Initial District Development Council elections were scheduled for June 1981. However, the security situation in the North rapidly deteriorated. In early 1981, 27 young Tamil males were arrested under the PTA and held incommunicado following a bank robbery in Neerveli in which two policemen were killed. On 24 May, the Tamil who headed the UNP list of candidates for the DDC elections was shot dead by Tamil militants. In addition, a UNP organizer was similarly killed. On 31 May, two policemen were wounded and two were killed under disputed circumstances at a TULF Distinct Development Council (DDC) election rally. In retaliation, from 31 May to 2 June, a group of 100–200 police men went on a rampage in Jaffna town and burned down the central market, the offices of a Tamil newspaper, the TULF headquarters, the home of the TULF MP for Jaffna, and the Jaffna Public Library (located only 50 yards from the police station).8 The government admitted that policemen had mutinied and run amok, attributing this to the killing of their two colleagues.9 Present in Jaffna throughout the police rampage were the Minister of Lands and Land Development, Gamini Dissanayake, and the Minister of Industries, Cyril Mathew, perhaps the most virulent anti-Tamil Sinhalese chauvinist in the UNP government. The presence of these two Ministers, in addition to a Brigadier General and the Inspector General of Police, and the flooding of Jaffna with an externally

1983 Riots 121

recruited police force only days before the rampage, led many observers to conclude that the attack was organized and planned in advance (e.g. MIRJE 1983: 6; MIRJE 1981: 12). The fact that not a single suspect was charged with any of the violence in Jaffna (MIRJE 1983: 16) further suggests either state complicity or the abrogation of state responsibility for security and the rule of law in the North.10 When TULF MPs made their submissions to Parliament concerning the June violence – including the Jaffna MP whose house had been burned to the ground – they were ‘repeatedly interrupted by spurious denials, mocking laughter, and unseemly jeering and heckling’ (MIRJE 1983: 17; Sri Lanka Resource Centre 1981).11 Further adding to the existing inter-group tension was the misrepresentation of events by the Sinhalese media (MIRJE 1983: 17; Spencer 1984; Leary 1981: 6). Under these conditions of disorder, the government announced its intention to hold DDC elections on 4 June as scheduled. On the eve of the election, without consulting with the Speaker of Parliament or giving notice to the Head of State, security forces rounded up and arrested the Leader of the Opposition and three TULF parliamentarians on the charge of ‘disrupting the democratic process.’ They were only released the following morning on direct orders from the President. Despite these conditions and allegations of vote rigging and malpractice, the TULF won every seat in the Jaffna District. However, lack of government implementation and provision of financial resources to the DDCs later led to the failure of the scheme, further adding to Tamil distrust of Government commitment to devolution and accommodation (Matthews 1987, 1993; Islam 1987; Hoole et al. 1992: 27–29). Conditions continued to deteriorate in the North in tit-for-tat violence; Tamil militants attacked security forces which in turn cracked down on the civilian population. On 28 July 1981, a group of armed Tamil militants attack a police station just outside of Jaffna – the first direct attack on a police station since the 1971 JVP insurrection. Two policemen were killed and the militants escaped with weapons and ammunition. In response, the government replaced police in the area with the Army. Armored vehicles and Army trucks became conspicuous sights in the peninsula, and the security clamp-down produced the appearance of an army of occupation. Coincident with the militarization of the North, anti-Tamil sentiment was being stoked in the South. In July 1981, in an unprecedented parliamentary maneuver, the UNP majority held a no-confidence vote against the Leader of the Opposition, TULF leader A. Amirthalingam. The Opposition (including the SLFP and a CP Member as well as the


TULF) walked out in protest and, according to one account, ‘the government members feasted themselves on a disgusting verbal orgy of communal provocation and personal abuse’ (MIRJE 1983: 20).12 The speeches were reported widely in the Sinhala press and allegedly were made available through the printing and distribution of thousands of additional copies of the speeches in Hansard. Anti-Tamil posters and slogans appeared on walls over the island.13 This added to the incitement to further violence against the Tamil community which broke out in August in Ampara14 and spread to Plantation areas, Ratnapura, and Negombo (MIRJE 1983: 22–34; Leary 1981). The August 1981 anti-Tamil violence was organized and systematic (MIRJE 1983: 28; Leary 1981: 21; Obeyesekere 1986: 20; USCR 1991: 8). Obeyesekere (1984) identifies numerous instances throughout the island in the late 1970s and early 1980s in which government actors had a clear and direct involvement in intimidation and acts of violence. He notes the pattern in these activities: ‘… the gangs were organized, they came in government vehicles, they were sometimes accompanied by MPs and for the most part, they belonged to the JSS, the trade union arm of the party in power. This almost certainly accounts for police inaction’ (1984: 164). The JSS (Jathika Sevaka Sangamaya) is the largest trade union in the country and was headed by Cyril Mathew, (then Minister of Industries). These activities received little public comment because almost all newspapers were directly or indirectly controlled by the government. Reports of such actions by the Communist Party newspaper Aththa were stifled by government closure of the presses at different points in time. Tamil sources claim that 1981 attacks had been organized by ‘people close to the government’ and that the police and armed forces did not intervene until a state of emergency was declared days after the violence had begun.15 President Jayewardene stated: ‘I regret that some members of my party have spoken in Parliament and outside words that encourage violence, and the murders, rapes and arsons that have been committed’ (quoted in The New York Times, 11 September 1981, A 2). Despite Jayewardene’s expression of regret, the government did nothing to be able to avoid a repetition of the anti-Tamil violence that erupted in July 1983. The events of 1981 set in motion a provocation–retaliation dynamic which fuelled the paramilitaries strategy: Tamil paramilitary attacks on soldiers, policemen, and others16 were met with unchecked reprisals by the ‘forces of law and order’ (for examples, see Hoole et al. 1992: 35–62). These included harassment, intimidation, assaults and a range

1983 Riots 123

of human rights abuses including torture (MIRJE 1983: 50–76; Amnesty International 1983). Tamil militants (in particular the Tigers and PLOTE) continued to murder UNP politicians in the North and, significantly, the LTTE began threatening TULF officials (Hoole et al. 1992: 56–7). According to Meyer and most other independent analysts: ‘Police reaction [to paramilitary assassinations] was extremely violent’ (Meyer 1984: 143). This excessive and indiscriminate response was promoted by the blurring of the distinction between ‘terrorists’ and ‘civilians’ and was sanctioned by both government actors and the Sinhalese media.17 The 1983 riots must be understood in this context of on-going hostility in the North between Tamil civilians and security forces and escalating anti-Tamil animosities in the rest of the island.

The episode The proximate trigger for the 1983 riots was the killing of 13 Sinhalese soldiers in an ambush by the LTTE close to Jaffna.18 The patrol had apparently been lured out with false information when it hit a Claymore land mine.19 The morning after the ambush, the army went on a rampage in the North and East. Sixty or more civilians were killed in and around Jaffna in the immediate aftermath of the ambush, and yet more people were killed by security forces in the following days (Hyndman 1988: 9). This includes an incident in Manipaya in which soldiers stopped a civilian bus and selectively murdered all Tamil males, and another incident where the same soldiers stopped a minibus and indiscriminately fired into it (Hyndman 1988: 9). The killing of the 13 soldiers was reported immediately by the media and their names were published, but the killing of the civilians in the Army rampage the following day was not reported (Hyndman 1988: 10; Bastian 1993: 22).20 It was decided within the military that rather than have the soldiers buried in their respective birth places, they would be buried with full military honours in Colombo’s central cemetery of Kanatte in Borella. Crowds began to gather at the cemetery in anticipation of the funeral, but there were delays in transporting the bodies to Colombo ‘on account of conditions prevailing in Jaffna, where the government was finding it difficult to control the army that had gone on a rampage against Tamil civilians’ (Bastian 1993: 22). The crowd of an estimated 10,000 (Hyndman 1988: 11) became increasingly restive as ‘certain segments of the crowd … deliberately fanned’ resentment and hostility against Tamils (ibid.). Ultimately, the bodies were not buried, but brought to army headquarters and returned to the


relatives. On the evening of 24 July, mobs spilled out from the cemetery and attacked Tamil shops and houses in the Borella area, adjacent to the cemetery. Mobs then spread outwards, even attacking Tamil shops only a few hundred metres from the presidential residence. But no curfew was announced (Meyer 1984: 139). According to Hyndman’s (1988) account, the violence of the first night was relatively localized, and the following morning many Colombo Tamils went off to work as usual. As the day progressed, however, anti-Tamil violence spread throughout the city.21 When a curfew was finally imposed after over 24 hours of rampaging mobs, it was not enforced with much authority. Eyewitness accounts reported the armed forces being present during acts of violence, and either standing by and ignoring, or actually participating in, the attacks on Tamils and their property (Hyndman 1988; Tambiah 1986: 21–33; Obeyesekere 1984: 153–74). The Times (of London) reported: Army personnel actively encouraged arson and looting of business establishments and homes in Colombo and absolutely no action was taken to apprehend or prevent the criminal elements involved in these activities. In many instances army personnel participated in the looting of shops. (5 August 1983) President Jayewardene said in a television interview yesterday that troops and police had sometimes encouraged the anti-Tamil violence. The President told a BBC interviewer: ‘I think there was a big anti-Tamil feeling among the forces, and they felt that shooting the Sinhalese who were rioting would have been anti-Sinhalese; and actually in some cases we saw them encouraging them.’ (9 August 1983)22 According to Tambiah (1992: 74), ‘the victims in Colombo were Tamil shop keepers; Tamil home owners, especially of the middle class and administrative, clerical and professional categories; large Tamil business capitalists and entrepreneurs; and Indian merchants, both Tamil and non-Tamil.’ On 25 and 27 July, 53 Tamil political prisoners, detained under the Prevention of Terrorism Act, were murdered in a riot at Welikade Prison in Colombo while soldiers and prison officials stood by (Hyndman 1988: 17–26).23 Anti-Tamil violence spread out from Colombo to those areas with significant concentrations of Tamil sub-groups.24 In Trincomalee, 130 sailors went on a rampage through town and are reported to have damaged 175 buildings and to have

1983 Riots 125

injured several people with one reported death (Hyndman 1988: 13; Bastian 1993: 22). Official estimates of the number of persons killed in the violence of the last week of that July is 387. Conservative, unofficial estimates are around 2,000 (Matthews 1984: 189; Bastian 1993: 23; Hyndman 1988: 5–52, esp. 26–7). According to official estimates, over 100,000 people were displaced the violence. However, others have argued that this was the number of displaced Tamils in Colombo alone, and that there were 175,000 more displaced people outside of Colombo (Bastian 1993: 23). More than 18,000 Tamil homes were destroyed along with 5000 shops and businesses. Economic losses are estimated to have been US$300 million or more (USCR 1991: 8).

Model I There can be little doubt that the 1983 riots followed a general pattern according to which Sinhalese mobs attacked Tamils from all subgroups, their homes, property and businesses. Neither is there any dispute that the riots were preceded by a deterioration of inter-group relations. A dynamic of provocation and retaliation was set in motion between the Tamil militants and the government security forces in the North and East. Tamil militant attacks on Tamil security personnel contributed to the ‘Sinhalization’ of the armed forces and sharpened the prevailing perception of the conflict as a dichotomous inter-group confrontation. Thus, Tamil civilians suffered the brunt of the extra-legal retaliation and ‘quasi-legal’ counter-terrorist measures25 by a security force which appeared increasingly Sinhalese in complexion. The Tamil civilian population was caught in the middle between the government military and the incipient Tamil paramilitaries,26 yet local Sinhala media coverage served to worsen the situation by their inaccurate or distorted reporting of events to the Sinhala language audience (MIRJE 1983: 16–22).27 The deterioration of this facet of inter-group relations on the island is crucial to understanding the volatile pre-conditions which enabled the ambush on 23 July to ignite such unbridled anti-Tamil violence. A notable example of a Model I representation of the 1983 riots may be found in de Silva (1986: 337–41). He argues that the riots were a manifestation of unharnessed mass psychological hysteria and asserts that while the security forces initially were negligent in playing a law enforcement role, there is no evidence to suggest that the government was directly involved in the organization of the riots: ‘Anyone who


saw [the rioters] at work would have sensed the operation of something like a mass of visceral antagonisms, a frightened force fed on a diet of rumours, tensions, fears and paranoia, and a fearsome rage directed against the Tamils – any Tamil for that matter – on the assumption that they were all communally responsible for the terrorist outbreaks in the north, and in fact for the incident which sparked off this vengeful fury’ (de Silva 1986: 339). While de Silva’s assertion that the riots were not organized is dispelled below, he is correct that one of the distinctive features of the 1983 riots is that for the first time, the riots were directed against the Tamils as an undifferentiated group regardless of social class or economic standing.28 Previously, middle- and upper-class Tamils had been relatively insulated from anti-Tamil disturbances (though occasionally their shops would be burnt down if they located in a lower class Tamil neighbourhood). Thus, the 1983 riots episode is a critical juncture in the conflict because, by breaching the protective effect of class, a differentiating intra-group boundary which had dampened the overall impact of previous outbreaks of ethnic violence broke down. The Tamil community was defined by others wholly in terms of its ‘Tamil-ness.’ Both the violence and the response by Sinhalese political leaders initiated a functional homogenization of ‘the other.’ When the middle class Tamils became victims in 1983, there was a tumbling down of protective psychological barriers. There was a breakdown in middleclass Tamil complacency and a sharpening of the sense of one’s Tamilness, despite the many Sinhalese friends which many middle-class Tamils had. As Coomaraswamy puts it: ‘the fabric of society ripped [along two] seams:’ the Sinhalese-Tamil seam of ethnic identity and the seam of class.29 De Silva’s Model I assessment of the July 1983 riots is identical to the UNP Government representation of events as articulated in the speeches by UNP ministers (including the President and Prime Minister) following the outbreak of hostilities. The President was conspicuously silent during the riots. There was no official address to the nation until his speech (in Sinhala and English) on 28 July which was broadcast on radio and television and published in the newspapers. The tone of Jayewardene’s speech and those by other government leaders was the same: they were marked by a lack of compassion or sympathy for the Tamil victims of the riots – a fact that was referred to even 15 years after the event.30 ‘These speeches were designed to placate the Sinhalese community’ (Obeyesekere 1984: 167). Jayewardene’s speech neither condemned the rioters nor consoled the

1983 Riots 127

victims. In effect, it appeased the Sinhalese chauvinists by representing the riots as a legitimate expression of anger: ‘because of this violence by the [Tamil] terrorists, the Sinhalese people themselves have reacted’ (quoted in Hyndman 1988: 35). The ‘reaction’ by Sinhalese mobs was not denounced but presented as an understandable consequence of Tamil ‘terrorist’ actions (Nissan 1984: 178). In a shift in focus, Jayewardene noted early in his speech that: ‘the Sinhalese will never agree to the division of the country, which has been a united nation for 2,500 years’ (ibid.). Aside from being historically wrong or at best a representation of Sinhalese Buddhist mythic history, it shifts attention from the violence by Sinhalese mobs against innocent Tamils throughout the island reinforced the very Sinhalese chauvinism which had contributed to the disintegration of inter-group relations and expedited the outbreak of the riots. Jayewardene’s speech presented an undifferentiated view of the Tamil sub-groups, thus exacerbating intergroup polarization and conflict. His statement justified the Sinhalese violence against Tamils and endorsed a Model I (inter-group conflict) representation of the situation by blurring the boundaries between Tamil ‘terrorists’ and Tamil civilians. The public statement by the President was the occasion to announce the government’s intention to propose a constitutional amendment banning the promotion of separatism (see Chapter Seven). Tamil separatists – not just Tamil militants – were cast as the initial aggressors and therefore the root cause of the riots. This further blurred intra-group Tamil distinctions by placing TULF in the same camp as the Tamil paramilitaries.31 In addition, the distinctions were blurred between those few Tamil sub-groups which advocated and participated in violence and the vast majority which did not. The end result was that attention was diverted from atrocities committed by Sinhalese mobs during the riots. The constitutional amendment obstructed both intergroup and intra-group communication links and marginalized moderate Tamil political actors by effectively banning the democratically elected TULF. If the intention of the amendment was to reduce the impact of separatism, the effect was the exact opposite, as Tamil paramilitary organizations were provided with the political space and the opportunity to rapidly expand.

Model II De Silva asserts: ‘despite the comments of some contemporary observers of these incidents suggesting a link between the mobs and


influential government politicians, no firm evidence has yet emerged to support this contention’ (1986: 339). The absence of a government inquiry into the episode ensures that ‘firm evidence’ will not emerge from this quarter (Hyndman, in Russell 1984: 613). Furthermore, the stringent censorship regulations imposed by the government successfully silenced many domestic sources which might have publicly presented ‘firm evidence.’ However, the international media, international fact-finding missions, human rights monitoring organizations, and foreign observers have been able to assemble facts and assess evidence independent of government interference.32 De Silva’s assessment aside, the overwhelming judgment by independent sources is that the 1983 riots not only illustrate patterns of violence, they illustrate systematic patterns of organized violence in which government actors played a central role. Hyndman (1988: 40) and others (Tambiah 1992; Hoole et al. 1992; Piyadasa 1984) report eyewitness accounts of rioters being transported in state vehicles and army trucks to different parts of Colombo and the island where they were unloaded to initiate and inflame anti-Tamil violence.33 ‘It was alleged by very many people, some eye witnesses and some repeating information which they believed to be firmly authenticated, that persons known to be active supporters of the UNP … were on many occasions leaders in the gangs perpetrating violence’ (Hyndman 1988: 40). Piyadasa identifies by name specific UNP councillors and UNP leaders who led mobs. These ‘mob captains’ were ‘experts in arousing a mob’ (Tambiah 1992: 75). ‘Certain Sinhala politicians and their local managers and bosses, entrepreneurs of organized crime and smuggling, small businessmen (the mudalalis) and their henchmen figure prominently as directors and manipulators of mass violence’ (Tambiah 1992: 75).34 Particularly noteworthy in the accounts of the 1983 riots is the role attributed to the Jathika Sevaka Sangamaya (JSS) and its leader, Cyril Mathew. Obeyesekere describes the JSS as an institutionalized and organized ‘band of thugs’ which served its political master, Cyril Matthew: ‘They … effectively controlled government offices and corporations and wielded enough power to transfer and intimidate even high officials. It also adopted and promoted the Sinhalese-Buddhist ideology. The precarious identity of marginal people was thus given a new reality and meaning: a political and nationalist ideology’ (1984: 161). ‘Mathew’s ideology is doubtless shared by a vast number of Sinhalese, but the JSS has given it an unprecedented militancy. Moreover, the union has spread its tentacles into other areas of the

1983 Riots 129

country. Thus today, MPs have created through the JSS and through other local groups in small market towns a complex, powerful network of organisations that can be put to political use’ (Obeyesekere 1984: 162). Piyadasa (1984) specifically identifies Mathew and his JSS members as the leaders of gangs responsible for the violence in parts of Colombo, especially in the Tamil neighborhood of Wellawatte.35 According to Eric Meyer’s account of the riots: ‘The operations that I witnessed were methodically organized. Their leaders often dressed in European clothes and had written instructions and lists of places to attack. Groups of five or six youths in sarongs armed with Molotov cocktails and clubs would empty houses and shops of part of their contents and set fire to them, continue on their way forthwith, often by car. The looters from the nearest shanty town would then arrive whilst the generous distribution of arrack would help to maintain the excitement’ (Meyer 1984: 139). Hyndman (1988: 13) explains the ‘written instructions and lists’ seen by Meyer: ‘Senior members of the government, members of opposition parties, lawyers, members of citizen groups, people affected by the violence and international aid workers interviewed were all consistent in stating that, from the beginning of the disturbances, many people in the mobs in the streets possessed the election lists containing the names and addresses of all those who lived in particular streets. The lists indicated the houses in occupation of Tamils and also whether the owner of the house was Tamil, Sinhalese or Muslim. The possessions and houses of Tamil people were then systematically attacked. If a Tamil family was living in a house rented from Sinhalese owners, the house itself was not damaged but the furniture and the property of the Tamils within would be destroyed. In many streets the Tamil owned shops were destroyed but those owned by Muslims or Sinhalese were spared. The same thing happened with homes.’ As the violence spread out from Colombo into such areas as Kandy, Matale, Nuwara Eliya, Badulla and Bandarawella, it exhibited a similar systematic structure. According to Hyndman (1988: 13), the incidents were initiated by people coming in from outside the districts. Again, lists were used to identify Tamil property which was then systematically attacked. The local inhabitants were then encouraged by the outsiders ‘to follow with further depredations.’ The 1983 riots were, therefore, not a spontaneous inter-group phenomenon. While the killing of 13 soldiers in Jaffna certainly provided the occasion, or proximate trigger, for the riots, the ensuing violence was purposefully exploiting the opportunity this incident


provided. While the initial bout of violence around the central cemetery may have been relatively unplanned, the subsequent riots were largely organized, driven and exacerbated by Sinhalese sub-group actors. An examination of the intra-group elements organizing the July 1983 violence sheds light on the breadth, depth and dynamics of the violence. The importance intra-group factors in outbreaks of mob violence is seen when one compares the July 1983 episode with the August 1992 incident in which an LTTE land-mine explosion on the Kayts island killed the Northern Commander of the Sri Lankan Army, the Commander of Jaffna, the Commander of the Northern Naval area, and three senior Lieutenant Colonels, among others.36 Despite the gathering of tens of thousands mourners at the Central Cemetery in Colombo and the fact that the victims were the most senior officers in the security forces (rather than foot soldiers), there were no equivalent anti-Tamil riots throughout Colombo or elsewhere. A Model I analysis would lead one to expect a similar outcome in both episodes given that inter-group context in both events is roughly equivalent, that is, both periods are characterized by high tensions and violent confrontation between the LTTE and government security forces. An explanation of the different outcomes draws attention to the intragroup differences between the cases. It remains to be explained why the 1983 riots were organized. There are two complementary and mutually reinforcing explanations, one political and one economic. The economic explanation identifies the UNP’s economic liberalization program as an irritant in inter-group relations and a precipitant to the riots (Herring 1998). The expansion of economic liberalization caused both an increase in the volume of trade and a narrowing in the profit margins as Sinhalese and Tamil merchants became locked in more heated competition: In many sectors the proliferation of small businesses led to bitter competition. … With the reduction in the flow of foreign capital, upon which this, in many respects artificial, expansion directly depended, many financially insecure businesses were threatened by bankruptcy. In such a serious situation, it became tempting to eliminate dangerous, often more efficient, rivals. In addition, a call to loot one’s competitor, whom one could characterize as an exploiter or race enemy was a cheap way of making a reputation amongst the lower classes, and there was always the possibility of making an easy financial killing in buying in at low prices the booty which had

1983 Riots 131

become burdensome to the looters. The activities of the mudalalis do not in themselves explain the organized manner in which the riots began, but they were in large measure responsible for their extension (Meyer 1984: 151). This was certainly not the first time that business competitors might have employed violence to improve the position of one enterprise over another. Furthermore, as Obeyesekere notes, in the past ‘racial violence was often sparked by business competition. Merchants employ the dispossessed proletarians of these towns to eliminate business rivals, especially during periods of post-election violence’ (1984: 159). The conditions created by economic liberalization from 1977 onward created new commercial and business opportunities. Shops throughout the island displayed an array of imported and local goods. New business premises and houses had been constructed in Colombo and higher rents became possible. ‘At the same time, this context had aggravated competition, so that the riots could be the occasion for Sinhalese businessmen to wipe out their competitors, for landlords to get rid of unwanted tenants and so on’ (Tambiah 1992: 73). Thus, during the riots there are numerous accounts of landlords attempting to intimidate unwanted tenants out of their homes and businesses (Tambiah 1992: 73; ‘Case 1’ in Kanapathipillai 1990: 324–9; Hyndman 1988: 51). The locations of violent attacks in Colombo appear to support this argument: the central market and business zones; areas of new industrial development stimulated by the new ‘liberalization policy;’ and middle class residential areas (Tambiah 1992: 75). It does not explain, however, why Muslim businessmen, often richer and more notorious competitors, were not similarly victimized. This directs attention towards political rathern than exclusively economic factors. The direct involvement of government actors in orchestrating the riots begs the question why elements within the ruling party would embarrass and destabilize the very government of which they were a part. The answer to this question directs attention towards the factions and power dynamics within the ruling UNP. A political explanation of the riots is developed by Meyer: ‘if … one admits the existence of a serious internal crisis in the ruling party, the mechanisms [which caused the riots] become more intelligible’ (Meyer 1984: 142). Meyer argues that Jayewardene lost control of both his Party and his armed forces. The loss of control of the Party was a result of internal feuding: ‘it is certain that at the heart of the UNP there was a latent crisis between those favoring opposing strategies to deal with Tamil terror-


ism, the Prime Minister being, amongst the president’s entourage, more favourable to negotiations, and the Minster for Industries [Cyril Mathew] taking a hard line’ (Meyer 1984: 143). Manor and Segal are more explicit about the logic driving the factions: [I]n 1983, his Sinhalese extremist faction [which orchestrated the riots] found itself to be a minority within the government and at a considerable disadvantage in relation to more moderate groups (led by Prime Minister R. Premadasa who, in Sri Lanka’s Gaullist system, seemed the likeliest successor to 76-year-old President Jayewardene). If they could provoke serious anti-Tamil Violence that would polarize communal relations further and discredit moderation in the eyes of many Sinhalese, they would undermine their factional rivals within the government. They therefore organized the systematic destruction of many Tamil homes and businesses, including firms belonging to Tamil supporters of the Prime Minister. In the process, they succeeded in strengthening their own position with in the ruling party and forcing the prime minister, other cabinet ministers, and President Jayewardene himself to adopt more uncompromising pro-Sinhalese positions. Since the riots, the Sinhalese chauvinists have also used their now-found leverage to thwart attempts to negotiate accommodation with the Tamils, so that the situation has remained perilously polarized (Manor and Segal 1985: 1175, emphasis added). The second point of dissension within the ruling party was a result of jostling for position to replace the ageing president; two major antagonists identified by Meyer are Prime Minister Premadasa and the ‘young bloods of the UNP, notably, Gamini Dissanayake’ (Meyer 1984: 143).37 According to this argument, the riots were orchestrated by actors within the government – rather than by ‘the’ government per se – as a means of pushing a hard line military response to the ‘Tamil Problem in the North.’ In the post-1983 riot period, the UNP government also took the opportunity to exercise extremely restrictive control of all news reports (Hyndman 1988: 31).38 It exercised its power to seal the presses and shut down a number of local newspapers. This is an effort to stifle dissent in both the inter-group and intra-group arena. In the Sinhalese intra-group arena, it sealed the presses of the three parties proscribed on 30 July as well as the newspaper of the SLFP – which had neither been proscribed not implicated in the riots. In Jaffna, the government also closed two of the most prominent newspapers, a Tamil-language paper

1983 Riots 133

and an English-language publication (Saturday Review). Such closures illustrate the government’s efforts to use the occasion to strengthen its position in political battles in both the Sinhalese intra-group arena and the Sinhalese-Tamil arena. Nissan’s (1984) study of the speeches of the top government leaders during the riots offers insights into the efforts by the government to manipulate the blame for the 1983 violence onto opponents in both the inter-group and inter-group arena. On 30 July, Minister of State de Alwis introduced the ‘Master Plan conspiracy theory’ of the riots. As with the other public speeches by UNP leaders, de Alwis did not express concern or sympathy for the Tamil victims and instead emphasized how the Sinhalese themselves had been adversely affected by the violence. He asserted that: ‘Some of the factories that have been burned down employed thousands of people. Ninety per cent of the employees in these establishments were Sinhalese.’ Unidentified ‘Sinhala Parties in the South’ and ‘Tamil Parties in the North’ were said to be planning a three stage plan to over-throw the government by instigating conflict between: 1. Sinhalese and Tamil; 2. Sinhalese and Muslim; and 3. Sinhalese Buddhists and Sinhalese Christians. The argument that the intention of the violence was to overthrow the government implied that the riots ultimately were not communally based.39 Nissan writes: ‘Images of great danger were compounded in this speech: northern and southern plotters, aided by “the minds of certain foreign elements;” groups of trouble makers “unconcerned with Sinhala or Tamil or any other race issue;” looters waiting on the side lines to benefit from the destruction. Only by refusing to be misled, and by supporting the president, could such multiple threats be averted, and majority interests be ensured’ (Nissan 1984: 179). The riots thus provided a convenient pretext for some intra-group political house cleaning. The government took the opportunity to blame the riots on an internationally-aided Communist plot to take over the country (see Chapter 6). It thus promptly proscribed the JVP on 31 July 1983, along with two other Marxist organizations, the Sri Lanka Communist Party (Moscow) and the Trotskyite Nava Sama Samaja Party. It was alleged that the JVP had initiated the riots as part of a ‘three-stage plot to overthrow the government’ (Gunaratna 1990: 189). Opponents of the government and independent observers stressed the improbability that these three parties would cooperate in this undertaking. On the one hand, Tamils were themselves blamed for provoking the violence. On the other, the situation in the north was seen merely to


provide an excuse for others to instigate violence designed to undermine the integrity of the state. Through this contradiction, a series of identifications were made. Any Tamil might be a terrorist or a sympathizer with separatism. Anti-state activity in this form was interpreted as inherently anti-Sinhalese, threatening the survival of the Sinhalese people, their religion and their civilization. Similarly ‘the hidden hand of southern extremists’ anti-state activity (whether it existed or not) was also identified as being opposed to the principles of Sinhalese national identity. The two sides became one. The UNP government insisted that it would not forget the special needs of the Sinhalese people in Sri Lanka. It appealed to the very sentiments that it claimed the plotters against the state were manipulating. National chauvinism and violence were implicitly endorsed through being condemned only because they were being used for the wrong ends. Anti-UNP parties could be proscribed, with no evidence produced against them, and even talk of separatism could be banned, in the name of protecting majority Sinhalese interests, equitable with the continuing power of the UNP (Nissan 1984: 184). The 1983 riots are a conspicuous critical juncture in the trajectory of inter-group conflict in Sri Lanka. They were preceded by a deterioration of inter-group relations, and resulted in the polarization of the intergroup arena as Sinhalese mobs attacked Tamils as an undifferentiated group, unlike in the past when higher social class and economic standing acted as a protective barrier to such incidents. Statements by UNP members of Cabinet and the Sinhalese media justified the attacks and endorsed the view that they were an understandable carthasis perpetrated by amorphous mobs. However, this Model I interpretation of events conveniently disregards the Sinhalese intra-group tensions which sheds light on the origin of the episode, the motivations of Sinhalese mobilizers, and the systematic pattern of the riots. The intensity and dynamics of the riots are better understood as the result of 1. a dispute among factions within the UNP over how to deal with the ‘Tamil problem;’ and 2. the negative impact of the UNP program of economic liberalization which induced business rivals to harness communal tensions in order to offset the resulting economic dislocation of the new program. An exclusive Model I analysis fails to recognize or explain the underpinning dynamics of the episode. If the riots originated principally in a Sinhalese intra-group conflict, then explanation or prescriptions limited to the inter-group level of interaction are deficient.

8 Critical Juncture V: 1987 Indo-Sri Lanka Agreement

On 29 July 1987, India and Sri Lanka signed an agreement which had as its stated purpose the establishment of ‘peace and normalcy in Sri Lanka’ – that is, to end the Sinhalese–Tamil conflict. The Indo-Sri Lanka Agreement was wholly unanticipated until just before it was signed in Colombo.1 It formalized the indo-centric reality of South Asian politics, most conspicuously by providing for the introduction of an IPKF on to the island ‘to guarantee and enforce the cessation of hostilities’ including the disarming of all Tamil militant organizations. However, the Agreement contained one overwhelming and ultimately fatal flaw: it was premised on the compliance of non-signatories.2 The LTTE, the most powerful Tamil paramilitary organization, had been excluded from the drafting of the Agreement and was presented with a fait accompli. Its initial acquiescence and long-term commitment to the Agreement were questionable from the start. Put another way, the Agreement was a bilateral solution to a problem which was multilateral in structure (incorporating Sri Lanka–India conflict, Sinhalese-Tamil inter-group conflict, Sinhalese intra-group conflict, and Tamil intragroup conflict). The initial IPKF of 6,000 to 7,000 rapidly grew to 60,000–80,000 troops, and by October 1987 it was militarily engaged with the LTTE in both the North and East.3 The Indo-Sri Lanka Agreement was an international accord signed by the Heads of State of two sovereign countries. While it is useful to examine the geopolitical dimensions of the Agreement4 – because they illuminate a process by which ethnic conflicts are ‘internationalized’ – in the current discussion they do not constitute the central point of reference. The purpose of this section is to examine the Agreement’s impact on intra-group and inter-group relations. Geopolitical factors are considered only where they have a direct bearing on these two sets 135


of relationships. In this context, the Agreement exacerbated both intergroup and intra-group violence in Sri Lanka. It further aggravated matters by forcing the Indian Government into a role that it was illsuited to perform.

The context Escalating Tamil paramilitarism, 1983–87 The political disillusionment and alienation of Tamil sub-groups resurfaced following each bout of communal violence (1956, 1958, 1977, 1978, 1981). Up to 1983, the Tamils of the Northern Province expressed this principally through electoral channels in their support for Tamil political parties such as the Ceylon Tamil Congress, the Federal Party and, since 1976, almost exclusively, the TULF. Following the 1983 riots, political alienation increasingly was expressed through extra-electoral channels. This is most evident in the massive increase in the number, size and strength of Tamil paramilitary organizations during this period (Hoole et al. 1992: 71–98; Gunaratna 1990a: 18–27). It should be emphasized that the 1983 riots were a necessary, but not a sufficient, t cause for the escalation of Tamil paramilitarism. There are a number of other factors which contributed to this escalation and set the broader context for the signing of the Indo-Sri Lanka Agreement. The December 1982 referendum Claiming that a general election would be too disruptive while the island was under a state of emergency, Jayewardene held a referendum to extend the life of Parliament for another six years. Earlier, a constitutional amendment had been passed sanctioning such a manoeuvre. Thus, the referendum which allowed him to cancel general elections until late 1988 was technically legal. The result was close but affirmative (54.66 per cent for, 45.34 per cent against). While the referendum suffered from voting ‘irregularities,’5 it was dubious in more fundamental way: it allowed Jayewardene and the UNP to stay in power with a massive parliamentary majority (84 per cent of the seats in Parliament) on the basis of an affirmative vote by only 37.8 per cent of the total electorate (‘Samarakone’ 1984: 84).6 Jayewardene thereby averted the danger of losing his party’s large parliamentary majority and its ability to amend the constitution unilaterally – which led to the next factor contributing to the alienation of Tamil sub-groups, particularly in the North and East.

1987 Indo-Sri Lanka Agreement 137

The sixth amendment Since its formation in 1976, the TULF had won virtually every Tamil majority constituency in the North and East. In the aftermath of the 1983 riots, the UNP majority in Parliament passed the Sixth Amendment to the Constitution which effectively banned the democratically elected members of TULF from the legislature. The Amendment states: ‘No person shall directly, or indirectly, in or outside Sri Lanka, support, espouse, promote, finance, encourage or advocate the establishment of a separate State within the territory of Sri Lanka.’ It also specifically requires that all members of parliament, office-holders of various kinds and even every attorney at law, take an oath in support of the Amendment. Refusal to do so would entail the loss of parliamentary seat, office or profession. Because of the TULF’s formal commitment to a separate state, it was effectively banned. By curtailing freedom of thought, conscience, speech and expression, (principles also guaranteed in the Constitution), the amendment contributed to the erosion of democratic framework of governance (Tambiah 1986: 45; Hyndman 1988: 95). Most TULF members left the island to operate temporarily from Madras – a decision which was also motivated by security concerns emanating from an increase in attacks on TULF members by Tamil paramilitaries, particularly the LTTE. However, Tamil intra-group violence relocated to Madras as well.7 By eliminating TULF from Parliament (which, at the time, was the Official Opposition Party), the UNP not only barred Tamils from participating in parliamentary democracy, it delegitimized that system of governance. This left open only extra-parliamentary channels for the articulation of Tamil political concerns. The path was clear for the establishment of Tamil paramilitary rule in the North and East. In effect, both the UNP government and the Tamil paramilitaries participated in a common campaign: the elimination of TULF and the dismantling of democratic institutions of governance in the North and East. New emergency regulations In addition to the emergency regulations discussed in the preceding chapter, new emergency regulations were implemented which increased political alienation and Tamil paramilitarism. Particularly noteworthy is the November 1984 establishment of ‘prohibited’ and ‘security’ zones in the North and East. One zone covered the coast line from Mannar to Mullaitivu and prohibited all unauthorized entry in a


corridor which started 100 yards inland and extended 100 yards out to sea. As a consequence, more than 20,000 families of fisherman were deprived of their livelihood (Hyndman 1988: 81–3).8 As thousands of coastal Tamils were deprived of their traditional means of livelihood, conditions became conducive to the generation of potential paramilitary recruits and the mounting of a paramilitary response. Assistance to Tamil paramilitaries From the early 1980s, Tamil paramilitary organizations enjoyed safe haven in South India along side the growing refugee community from all Tamil sub-groups. There were accusations by Sri Lankan government officials – and denials by Indian government officials – that Tamil militants were being trained on Indian territory. Despite the diplomatic posturing by Indian politicians, there was a growing volume of documented evidence that Tamil paramilitaries were receiving extensive training and assistance from both the Central Government of India and the State Government of Tamil Nadu9 as well as from the international terrorist network.10 These factors contributed to the alienation of Tamils in the North and East while simultaneously closing legal and parliamentary channels for political expression. They added to the sense of the declining rule of law, and the diminishing legitimacy of the state and its representatives. The covert Indian support for Tamil paramilitary organizations served to develop the extra-legal and extra-parliamentary channels for the expression of political dissent. After 1983, the size, number, and strength of Tamil paramilitary organizations increased rapidly. While the overall pattern indicates an intensification of inter-group conflict, it is difficult to obtain exact data on violence. In some cases, information is not available; in other cases, government and paramilitary information is divergent or contradictory – even dates and locations may not correspond. For example, in 1985 according to government sources, 885 civilians and 194 members of the security forces were killed, whereas according to Tamil sources at least 2,000 civilian were killed and security forces went on a rampage on an average of once a week somewhere on the island (Amnesty International 1986). In light of the difficulty of relying on ‘official sources’ for information, it is necessary to counter-balance assessments of the level of violence with information provided by independent sources (based in Sri Lanka and internationally). In the period from 1 March 1985 to 31 January 1986, Amnesty International representatives confirmed 2,578 killings, 12,105 arrests,


and 547 disappearances, in addition to the displacement of some 55,000 people (30,000 from the Trincomalee area alone) (ibid.). The majority of these casualties would be the product of inter-group conflict, though Tamil paramilitary feuding also claimed significant number of lives. Perhaps more indicative of the escalation in intergroup violence is the increased odiousness of the violent acts:11 the LTTE massacre of over 80 Sinhalese children, women and men at the Kent and Dollar Farms in the Eastern Province (November 1984);12 the LTTE massacre of 70–75 Sinhalese (again including children and women as well as monks) in Anuradhapura (May 1985) apparently in retaliation for the murder of 43 young Tamil males in Jaffna by the Army five days earlier – which was itself retaliation for the death of an officer in a land mine explosion (9 May 1985); the bludgeoning deaths of 26 Tamils on a ferry by out-of-uniform naval personnel (15 May 1985); the shooting deaths of 13 Tamil passengers on a bus in Trincomalee by Home Guards (2 June 1985); the massacre of as many as 100 Tamils after an army patrol narrowly escaped injury in a land mine explosion in Vavuniya (16 August 1985); the killing of four police constables in Batticaloa in a land mine explosion and the subsequent shooting deaths of nine local civilians by their colleagues. In addition to ambushing army patrols, Tamil paramilitaries publicly executed 61 Tamils in 1985 (usually for alleged collaboration with ‘the enemy’) by tying them to a lamp post or telephone pole, hanging a list of alleged crimes around their necks, and then brutally murdering them. During the 1983–87 period, children, women and men were hacked to death with machetes, burned to death with gasoline-filled tires, and tortured to death by military and paramilitary forces. The list here provides only a small sample of atrocities. Media accounts, human rights reports and fact-finding missions attest to these particular events and catalogue many more.13 A pattern of violence and retaliatory violence appears clearly in the inter-group dynamic of conflict in this period: the murder of the members of one group elicited a response in kind from the violent sub-groups within in the victim community. 14 This is an important feature in the intra-group process of resource mobilization and consolidation by Tamil paramilitary organizations. Government security force retaliation alienated the civilian Tamil population and created the next generation of combatants. This provocationretaliation dynamic is a fundamental principle in any guerrilla war. The three stages of the dynamic may be represented as follows (see Figure 8.1).

140 Critical Junctures Provocation Tamil Sub-Groups

Sinhalese Sub-Groups

Sub-Group A2

Sub-Group B1 Sub-Group A1 (LTTE) versus Sub-Group B2 (Sinhalese Villagers)

Sub-Group Ax

Sub-Group Bx

Retaliation Sinhalese Sub-Groups

Tamil Sub-Groups

Sub-Group B2

Sub-Group A1 Sub-Group B1 (Sinhalese Villages) versus Sub-Group A2 (Tamil Villagers)

Sub-Group Bx

Sub-Group Ax

Polarization and Consolidation Tamil Sub-Groups

Sinhalese Sub-Groups

Sub-Group A2 + Sub-Group A1

Sub-Group B1 + Sub-Group B2

+ Sub-Group Ax

+ Sub-Group Bx

Figure 8.1

Provocation–retaliation–consolidation dynamic

The provocation–retaliation dynamic may diminish the support of Tamil sub-groups for the government, but that does not ensure the intra-group success of any particular Tamil paramilitary or political organization. The passing of the Sixth Amendment helped to ensure that the dominant Tamil sub-group would not be the democratically elected TULF; the emergency regulations and the decline of both the rule of law and democratic institutions made it unlikely that the process would be democratic or participatory; and covert Indian government support to the major Tamil paramilitary organizations increased the likelihood that political control would be won through intra-group paramilitary violence. Thus, at the same time that Tamil paramilitaries were attacking government security forces and defenceless Sinhalese civilians in the inter-group arena, they were also attacking each other in a battle to determine who would control the North and the East. There was a surge in killings of dissenters and opponents within and outside paramilitary organizations. According to Hoole


et al. (1992: 76), by 1986 this Tamil intra-group violence had reached ‘epidemic proportions.’ Although the LTTE ‘won’ this intra-group competition and established itself as the most powerful Tamil paramilitary organization, such feuding continued after the signing of the IndoSri Lanka Agreement. The stated goal of the LTTE campaign has a more sinister relevance in 1993 than it did in early 1987: ‘a One-Party State on the Yugoslavian model’ (Pfaffenberger 1988: 138).

The immediate pre-agreement political–military landscape The LTTE established itself as the dominant Tamil paramilitary organization in the North and East through a strategy which combined the elimination of all challengers and the provision of state-like services in the increasingly worn-torn North, such as mail service, licensing and vehicle registration, ‘law and order,’ ‘taxation,’ the rationing of scarce goods, the management of health and social services, and so on. In the period 1983–87, conditions deteriorated to the point where, in January 1987, the Jayewardene government imposed a fuel and economic blockade on all of the Tamil areas of northern Sri Lanka, essentially placing the entire area under siege. The military simultaneously increased its military assaults in an effort to alienate the LTTE from its support constituency.15 This included regular artillery shelling and air raids of the Jaffna Peninsula.16 The LTTE widely publicized the many instances in which residential areas containing hospitals, schools, temples, and Tamil homes came under government fire, contributing greatly to the mobilization of international and regional sympathy and support from South Indian Tamils, the State Government of Tamil Nadu and the Central Government of India. Not surprisingly, the LTTE was less forthcoming in explaining the role it sometimes played in provoking government forces to fire into residential neighbourhoods (especially in and around the Jaffna hospital) by using these areas as shelters from which to snipe at and bomb security forces (Hoole et al. 1992). Nor did the LTTE refer to its alleged use of civilian hostages as shields from enemy fire (Pfaffenberger 1988: 138).17 Two particular incidents by Tamil paramilitaries significantly added to the deterioration of Sinhalese-Tamil inter-group relations in the immediate pre-Agreement period. On 17 April 1987, the LTTE stopped a number of vehicles on the Trincomalee–Colombo Road and massacred over a hundred Sinhalese passengers.18 A few days later, a bomb blast in the Colombo bus station killed 100–200 people and injured hundreds more. The bomb was later attributed to EROS, a close ally of


the LTTE. A government-imposed curfew defused potential anti-Tamil riots this time. It also cancelled an anti-Government protest planned by a group of bhikkhus and generally dampened the intra-group opportunities to criticize Jayewardene through public demonstrations. Nonetheless intra-group criticism of Jayewardene and calls for his resignation increased (Pfaffenberger 1988: 138–9). These two LTTE provocations were classic components of the provocation–retaliation dynamic noted above. They were successful in provoking the outrage of Sinhalese sub-groups, precipitating responses in kind, and ultimately polarizing communities in the inter-group arena. The Government retaliated with local military operations and, at the end of May, it launched ‘Operation Liberation’ (O’Ballance 1989: 71–87; Hoole et al. 1992: 119–32). The attacks and counter-attacks continued – including the 2 June 1987 hijacking of a bus near Ampare in which 29 novice monks, most aged 14–16, were taken into the jungle and hacked to death with machetes (Rubin 1987: 40). This ‘outrageescalation’ hindered chances for constructive inter-group exchange. The immediate beneficiaries of the consequent breakdown of intergroup relations were the LTTE and the UNP. It allowed each to rally anti-Other chauvinism in their respective intra-group struggles to establish their legitimacy and credibility as the sub-group that could best represent and protect the whole community. The government justification of its military offensive was intended for a Sinhalese audience and did little to alleviate the worries of Tamils in the North and East who were caught between the Tigers and the Army. In response to criticism of the ‘collateral damage’ of its military assaults in the North (i.e. the killing of innocent Tamil civilians), an official government spokesperson said: ‘The government reserves the right to attack such targets ’till such time as the killing of civilians is stopped, and the peace process is allowed to continue’ (The Hindu, 22 April 1987: 1). This statement followed the bus massacre and the Colombo bus station bombing and clearly referred to Sinhalese civilians in the South rather than Tamil civilians in the North. Similarly, the increasingly antagonistic rhetoric of Sinhalese political leaders appeared to be directed both to the Government of India and Sinhalese sub-groups – as illustrated in a speech in parliament by Prime Minister Premadasa on 24 April 1987: ‘When the lives of our people are in danger, we are not prepared to go in for a political solution … any friend [i.e. the Government of India] who tells us to find a political solution will be considered as the biggest enemy’ (The Hindu, 25 April 1987: 1).


Until this stage in the conflict, the central government of India had been erratically trying to play the role of mediator through the offices of the state government of Tamil Nadu in Madras. However, the political jostling between the State Government and the Central Government ruled by Rajiv Gandhi’s Congress (I) Party hindered communication and mediation efforts (Bush 1989). As the Jayewardene government prepared for the media-labelled ‘final military solution,’ the Indian government found itself unable to influence events within Sri Lanka. Consequently, India’s credibility as a mediator rapidly deteriorated. As the Sri Lankan Government blockade and military offensive dragged on and rumours of disease and starvation circulated in the press, the Indian parliament and media responded by stepping up their calls for direct intervention by Indian troops.19 On 2 May, the Chief Minister of Tamil Nadu announced a ‘gift’ of 40 million Rupees (then about $US3.2) in humanitarian aid for Tamils in Sri Lanka – to the embarrassment of the Central Government in India and the criticisms by the Sri Lankan Government that this would be a violation of the country’s sovereignty and integrity. Similarly, the first step towards direct Indian intervention came not from the Central Government in New Delhi, but from the State Government of Tamil Nadu in the form of a ‘humanitarian flotilla’ of aid from the Tamil Nadu port of Rameshwaram to Jaffna on 3 June 1987. Not surprisingly, the flotilla was quickly blocked by the Sri Lankan Navy and forced to return. However, the following day Indian Air Force cargo planes, flanked by Mirage 2000 fighter jets, succeeded in dropping twenty-two tons of emergency relief supplies into Jaffna (Joshi 1987: 2). There were questions whether the supplies went where they were supposed to go, but there was no doubt that the Indian action was a flagrant violation of Sri Lankan sovereignty.20 Following the air drop there was a period of intense negotiation between the governments of India and Sri Lanka (de Silva 1993; Bush 1989). No doubt a central consideration in Jayewardene’s mind was the fear that India might be planning a more long-term intervention strategy if the conflict continued to threaten regional stability.21 Both Gandhi and Jayewardene were compelled to act by both regional and domestic pressures. Gandhi saw rising ‘anti-Delhi,’ anti-Congress-I sentiment in Tamil Nadu as the AIDMK government in Madras strengthened itself politically through its adroit harnessing South Indian Tamil sympathies for the besieged Sri Lankan Tamils; he also feared that the Sri Lankan Tamil separatists might rekindle South Indian Dravidian separatist sentiment at a time when anti-Hindi language extremists


were reappearing in Tamil Nadu;22 he also feared that Sri Lankan violence would increasingly spill into South India;23 and perhaps the most important geopolitical consideration was the growing foreign involvement in Sri Lanka including Pakistan and China (Rao 1989; Gunewardene 1986; Gunasekera 1988). Following the Indian Air Force incursion, Jayewardene was keenly aware that India might chose to launch an extended unauthorized intervention. Domestically, Jayewardene felt pressured by Sinhalese sub-groups to gain military control over the Northern and Eastern Provinces. Yet, intensified military actions increased the possibility that India might again intervene.24 Early signs of a resurgent JVP added a further domestic compulsion: the threat to intra-group stability and UNP security. An Indo-Sri Lanka Agreement would allow Jayewardene to divert military resources to ‘disturbances’ in the South (i.e. to the Sinhalese intra-group arena) while committing the Indian Government to a specified (if minimal) set of rules limiting its military engagement in Sri Lanka. It was in this context that the Indo-Sri Lanka Agreement of 29 July 1987 was formulated and signed.

The episode At a regional level of analysis, the signing of the Agreement underscored the hegemonic status of India. Similarly, the failure of the Agreement and the subsequent IPKF withdrawal tarnished that same status. The Agreement enabled the Government of India to exclude outside power involvement in the Sri Lankan conflict. Under the terms of the Accord, the Government of India acquired substantial control over Sri Lankan foreign policy. The Government of Sri Lanka had to ensure that foreign military and intelligence personnel on its territory do not ‘prejudice Indo-Lankan relations’ and that all foreign broadcasting facilities on its soil ‘be used solely as public broadcasting facilities and not for any military or intelligence purposes.’ It had to ensure that Trincomalee and other ports ‘not be made available for military use by any country in a manner prejudicial to India’s interests.’ For its part, the Government of India undertook to restore and operate the Trincomalee oil tank facilities jointly with Sri Lanka, to provide training facilities and supplies to the Sri Lankan military, to ‘deport Sri Lankan citizens found to be engaging in terrorist activities or advocating separatism or secessionism,’ to disarm the Tamil militants (within 72 hours), and to ‘establish peace and normalcy in Sri Lanka.’ The success of the Accord rested on the ability of the Indian government and the IPKF to convince, or coerce, the Tamil militants to


comply with an Agreement they neither drafted nor signed. The 72hour deadline for the Tamil paramilitaries to surrender their weapons expired with only a small proportion being turned in. The deadline was then extended, but the response was unenthusiastic. LTTE leader Prabhakaran’s reluctant initial acceptance of the Agreement was evaporating having been granted in early August while he was held incommunicado by the Indian government in a five-star hotel in Delhi ‘for meetings’ with Indian officials. In contrast, the smaller Tamil paramilitaries saw compliance as a means of gaining the political and military resources needed to defend themselves from attack by the LTTE. Thus, they cooperated fully with the IPKF and later with the Sri Lankan military – more as an effort to defeat the LTTE than out of a particular commitment to the Agreement. The most significant Tamil paramilitary cooperants are PLOTE, TELO, EPRLF and the EPDP.25 Despite the IPKF presence, the LTTE continued its campaign to become the sole representative of ‘the Tamil Nation’ by attacking rival Tamil paramilitaries. Factional fighting broke out with particular ferocity in September when the LTTE killed over 150 rivals while the IPKF allegedly did little to prevent the massacre (de Silva 1993: 137).26 The occasion for the LTTE to break completely from the Agreement came on 4 October when the Indian Navy apprehended 17 Tamil Tigers off the Jaffna coast, including one suspected of organizing the slaughter of over a hundred Sinhalese bus passengers in April. IPKF officers decided that since the group was fully armed – and therefore in violation of the Indo-Sri Lanka Agreement – it would be handed over to the Sri Lankan authorities for interrogation. Twelve of the 17 Tigers committed suicide en route to interrogation in Colombo.27 On 5 October, Prabhakaran declared a LTTE policy of non-cooperation with the IPKF. There followed a week of violence in which the LTTE killed over 200 people; this included the ‘execution’ of eight Sinhalese soldiers who had been previously captured by LTTE,28 two senior officials of the state-owned Sri Lanka Cement Corporation and four Sinhalese employees of the national television station (Pfaffenberger 1988: 144; Rubin 1987: 50; O’Ballance 1989: 98). By mid-October, with Sri Lankan Government prodding, the Indian Government decided to forcibly disarm the LTTE and to pursue a military option in order to assert control over the Northern and Eastern Provinces. The strength and tenacity of the LTTE surprised the IPKF. As in all guerrilla wars, there was no quick overall victory. By 15 December 1987, the IPKF had lost 350 men with 1,100 more wounded.29 By the time the Indian troops left on 24 March 1990, it was estimated that 1,155 soldiers of the IPKF had been killed and 2,984


wounded (India Abroad, 30 March 1990: 1). The quick foreign policy victory that Prime Minister Rajiv Gandhi may have been hoping for (in part, to buoy up his sagging domestic popularity) had backfired.30

The intra-group dimensions of an inter-group alliance The Accord marks an critical juncture in the Sri Lankan conflict because it restructured the axes of confrontation within and between groups – while also adding a new axis of military confrontation between the LTTE and the IPKF. The diagram below illustrates how the boundaries of intra-group conflict also constituted the boundaries for the inter-group alliance in support of the Indo-Sri Lanka Agreement. The signing of the Agreement intensified conflict in both the Tamil and Sinhalese intra-group arenas and polarized sub-groups into proAgreement and anti-Agreement camps. Pro-Agreement Tamil and Sinhalese sub-groups overlapped and joined forces in the inter-group arena. However, this inter-group alliance further exacerbated intragroup conflict as anti-Agreement subgroups in both Tamil and Sinhalese communities accused pro-Agreement sub-groups of being ‘traitors’ and ‘anti-patriotic.’ Thus, while existing boundaries of intra-group conflict contributed to the formation of an inter-group alliance, the Accord intensified violent conflict within both the Sinhalese and Tamil intragroup arenas. As a result, the intra-group boundaries of the inter-group alliance were maintained by the intensified intra-group conflict precipitated by opposition to the Accord. The dynamics and boundaries of the inter-group and intra-group conflicts became inextricably linked as a direct consequence of the Indo-Sri Lanka Agreement. Sinhalese

Tamil

Oppose

JVP SLFP MEP MSV

Muslim

Support

Oppose PLOTE EPRLF TELO EPDP TULF CWC

UNP

LTTE

SLMC

Inter-group alliance Intra-group conflict

Intra-group conflict

Figure 8.2 Sub-group position on the Indo-Sri Lanka agreement inter-group alliances and intra-group conflicts


Model I The IPKF became a buffer between the Tigers in the Tamil-dominated North and the Government security forces in the Sinhalese South. By cutting off contact between the two communities there was a reduction in inter-group violence in the Northern Province.31 The IPKF thus dampened the inter-group violence between two sets of ethnic antagonists only to replace it with IPKF–LTTE military confrontation and, as discussed below, increased intra-group confrontation. While the IPKF separated inter-group antagonists in the North, it was a very different situation in the ethnically mixed areas of the East where it was unable or unwilling to place itself between ethnic antagonists.32 There, violence accelerated in both the inter-group and intragroup arenas as a result of a central component of the Agreement: the conditional joining of the Northern and Eastern Provinces into one ‘administrative unit’ (Clauses 2.0–2.7). The Agreement stipulated that an Interim Council would be elected to govern the region33 until a referendum was held in the Eastern Province to determine whether unification would be permanent. Thus, the future of unification lay in the hands of those who would vote in the referendum on the East Coast. A simple majority would determine the result. Yet, even during periods of relatively lower levels of violence, non-Tamil communities in East – and even East Coast Tamils – were apprehensive about the economic, political and social domination of Northern Tamils. This sentiment and the ethnically mixed composition of the Eastern Province cast significant doubt on whether the referendum would endorse continued unification. The LTTE unconditionally demanded merger since potentially this would allow it to exercise control over a much larger area than the Northern Province. Under these conditions, with the referendum on the immediate political horizon, the LTTE thus responded to dissent with what has become its standard operating procedure; it violently attacked opponents and launched its campaign to alter the demographic balance of the East by terrorizing Sinhalese, and later Muslims, out of the area.34 Many of these attacks were reported in the international media. For example, in early October: 42 people were killed when a train was ambushed and set on fire with petrol bombs near Batticaloa; another 40 were shot dead in their homes in that city; 29 were killed in a bus ambush at Lahugala (South of Batticaloa); while 30 Sinhalese settlers were killed at Pabavikulam (north-west of Trincomalee); 21 others were killed in


Kuchchaveli as were 35 fishermen at Eravu (sic) south of Batticaloa. By the end of the week over 5,000 Sinhalese settlers and refugees had sought the security of Temples, army camps and even police stations. (O’Ballance 1989: 99) By October 1 [1987] what had appeared to be … a spontaneous communal clash [in Trincomalee] took on the appearance of an organized pogrom, with the LTTE members taking the lead in a seeming effort to drive the Sinhalese out of the Eastern Province. … The reported toll of homeless rose to … 5000 on October 4, with about 500 houses and shops burnt and 18 people killed. … LTTE gunmen stopped the Batticaloa-Colombo night mail, separated the 40 or so Sinhalese (by some accounts Sinhalese and Muslims) passengers, and killed them all. They likewise executed about 20 bus passengers in Amparai district. They killed 27 settlers in a village outside of Trincomalee and 38 in a village outside of Batticaloa. By the end of the week of violence, about 200 Sinhalese had been killed. The ultimate death toll may prove to have been higher, as many people were missing. About 20,000 had fled their homes and were flooding into the Sinhalese areas of the country, including Colombo. Much of the city of Trincomalee had been gutted by flames (Rubin 1987: 51–2) As noted above, in the initial period, the IPKF exacerbated LTTE violence on the East Coast by not intervening when attacks or disturbances took place (de Silva 1983: 140). Members of the Tamil Madrasi Regiment were reported to have expressed their support for local Tamils. The regiment was subsequently withdrawn on 11 October for refusing to participate in the Jaffna offensive against the LTTE (Rubin 1987: 52). Sinhalese who had been displaced by violence in Trincomalee accused the IPKF of partiality towards the Tamils and complicity in the attacks (ibid.). The response to the LTTE-instigated inter-group violence was threefold. There was an exodus of Sinhalese villagers from the Eastern Province. The IPKF redoubled military efforts to defeat the LTTE. And, most interesting in the context of the current research study, both the Indian Government and the UNP-government attempted to strengthen the intra-group position of the Pro-Agreement, anti-LTTE Tamil paramilitaries. The response by the Indian Government and the UNP response warrants further discussion. The deteriorating situation in the North and East postponed the holding of Interim Council elections until November 1988. Although


the Tigers ordered a boycott and threatened to murder anyone participating in the elections, there was a significant turn-out for elections in the Eastern Province (The Economist, 26 November 1988: 32; O’Ballance 1989: 121).35 The EPRLF won an overall majority in the Eastern Province. However, rather than hold a election in the Northern Province, the UNP government agreed to allow the two main anti-LTTE rivals (EPRLF and ENDLF) to split the 36 seats between them (ibid.). EPRLF leader A. Perumal was appointed Chief Minister of the ‘Provisional Council’ and he immediately declared Trincomalee the Provincial capital rather than Jaffna – as anticipated given that much of the Northern Province was under direct Tiger paramilitary control. The undemocratic deal dividing the Northern Province between anti-LTTE paramilitary organizations was an effort by Jayewardene to politically buttress pro-Agreement allies in their intra-group conflict with the LTTE, a common enemy (under most circumstances – see below) and the principal anti-Agreement subgroup in the Tamil arena. The Indian Government pursued a similar strategy, only it militarily buttressed pro-Agreement allies in their intra-group conflict with the LTTE. As noted above, Tamil paramilitaries had arisen with much strength following the 1983 riots, in part, due to the support provided to the militants by the Government of Indian and the State Government of Tamil Nadu.36 Prior to 1987, the Indian Government tacitly acknowledged LTTE’s dominant paramilitary and political position and treated it as the leading representative in the Tamil intragroup arena.37 The patron–client relationship between the Indian Government and the LTTE probably contributed to Gandhi’s belief that the LTTE could be induced to support the Agreement. When it became evident that the LTTE would not be induced to support the Accord, the Indian Government redirected its resources to a replacement client in Sri Lanka – the EPRLF. As well, despite the disapproval of the UNP Government, in late 1989 the IPKF began to train and arm a ‘new’ local military force called the TNA composed of anti-LTTE paramilitaries and ‘fresh recruits.’ The intention was to replace the IPKF with a new client force of 30,000 ‘local soldiers.’38 Both the political strategy pursued by the UNP government and the military strategy pursued by the Indian Government were premised on the harnessing of intra-group divisions and animosities. Such strategies are inherently liable to backfire – as the original Indian support for the LTTE illustrates. The government-endorsed Tamil paramilitary coalition against the LTTE was a coalition of conve-


nience rather than commitment. It has been maintained by the organizational glue provided from Delhi and Colombo and by the common understanding among the non-LTTE militants that they are threatened by the LTTE. If the LTTE threat is weakened or removed, the unravelling of the coalition is likely to follow, and the conveniently overlooked intra-paramilitary tensions and conflict may spill into more violent feuding in the north and the east – and perhaps even into the streets of Colombo since that is where pro-government Tamil paramilitaries have located their offices.39 The fuelling of intragroup antagonism by governments in Colombo and Delhi for shortterm, short-sighted political goals contributes to the brutalization of civilians in the North and East and will inhibit (or at least complicate) movement toward accommodation in any effort to construct a post-war settlement. Model I: LTTE–UNP coalition The instrumentalist underpinning of conflict in Sri Lanka is nowhere more evident than in the periods of LTTE–UNP cooperation. Two examples will suffice, one overt and one covert. As discussed below, a major opponent to the Indo-Sri Lanka Agreement was Prime Minister Premadasa. He publicly expressed his objections and dramatically refused to attend the signing ceremonies and related functions. When Premadasa assumed the office of President in late 1988, the JVP resurgence was at its peak. He immediately undertook to abrogate the Agreement, aware that it would remove the central impetus to conflict in Sinhalese intra-group arena. He also expressed concern about losing control over the Indian-backed EPRLF in the Eastern Province. The LTTE was similarly interested in having the IPKF removed from the conflict as this force had become its chief adversary in the North and East. The LTTE was also aware of the growing power of its intra-group rival, the EPRLF, which was shielded from attack by the IPKF. The Indian Government, however, expressed reluctance to leave on Sri Lankan terms, seeking to prolong the foreign policy gains which had been acquired through the Agreement. To expedite the Indian departure, Premadasa launched ‘talks’ with the Tamil Tigers. In May 1989, LTTE leaders were flown to Colombo by the Sri Lankan air force for meetings with a UNP negotiation team of four cabinet ministers. In the first four days of talks, the IPKF lost at least 80 men in three separate incidents with the LTTE (Far Eastern Economic Review, 25 May 1989: 25). Over a year of negotiations, both


Premadasa and Prabhakaran made public concessionary gestures in an effort to rally international public opinion against the IPKF presence in Sri Lanka. Both the LTTE and the UNP argued that the conflict was an internal matter which should therefore be resolved ‘by Sri Lankans.’ Incredibly, the LTTE declared an end to its 17-year war and formed a political party ostensibly to contest elections following the IPKF withdrawal.40 The immediate slaughter by the LTTE on the East Coast which followed the IPKF withdrawal puts these statements in context.41 This incident illustrates an overt inter-group alliance struck for the common objective of expelling the IPKF from the island. The covert alliance between the Premadasa and the LTTE was noted in Chapters 2 and 3: the charge in 1991 that Premadasa had been responsible for supplying weapons and material to the LTTE (in November 1989) in an effort to encourage attacks against the India-backed Tamil National Army and to apply indirect pressure on the IPKF to ‘quit the island.’42 This scandal provided one of the bases for the impeachment procedings launched against Premadasa in October 1991. 43 This is a prime example of how mutual interests may lead to overt or covert inter-group ethnic alliances between otherwise bitter enemies, challenging the applied logic of Model I. The LTTE acquired weapons and material and Premadasa applied pressure on India to pull out while publicly maintaining the appearance of diplomatic opposition to the Agreement signed by his predecessor. While both the UNP and the LTTE entered into an inter-group coalition in order to expel the IPKF, they each had different reasons for doing so. However, for each, the motivations derived from the intragroup arena. For Premadasa, it was to quell the JVP resurgence and for the LTTE it was to reassert itself as the supreme Tamil paramilitary organization. In the context of the current discussion, it should be noted that the withdrawal of the IPKF and subsequent reassertion of LTTE territorial control in the north and east sent many EPRLF supporters and government collaborators (voluntary and nonvoluntary) fleeing LTTE retribution. The fear of retribution was well-founded in light of the summary massacre of 200–300 Tamils by the LTTE as it entered Batticaloa in December 1989; as well as the massacre of around 600 police officers who had surrendered to the LTTE in June 1990 after an extensive series of attacks on police stations all along the East Coast, breaking a ceasefire that the LTTE had maintained with the government.44


Model II Sinhalese intra-group arena In the Sinhalese intra-group arena, wide-spread opposition to the Agreement was the catalyst for the resurgence of the JVP (see Episode III) and it precipitated the most intense anti-government attack ever experienced on the island. The Sinhalese supporters of the Agreement were a numerical minority in the intra-group arena (Samarasinghe and Liyanage 1993: 158).45 Yet with the UNP government as the principal backer, supporters were buttressed by the complement of resources commanded by the state – especially its coercive resources. To the extent that dissent was expressed within the UNP, it was largely curtailed by Jayewardene’s authoritarian powers of influence and intimidation – though, there were high-level opponents within the UNP who could not be muzzled,46 most notably, Prime Minister Premadasa, National Security Minister Athulathmudali, and Agriculture Minister Jayasuriya. Recognizing the opportunity created by dissention within the cabinet and the UNP ranks more broadly, Mrs. Bandaranaike, leader of the SLFP, announced that her party would back Premadasa if he were to maintain his opposition to the Agreement (The Hindu, 12 August 1987: 1). This was a tactic intended to capitalize on internal UNP dissent, which did not have much impact – most likely because of the very tight control maintained by Premadasa within UNP rank and file, generally. Sinhalese intra-group post-Accord violence As early as 22 July 1987, it became apparent that an Indo-Sri Lanka Agreement would be signed imminently. Sri Lanka’s foreign policy concessions to India provoked attack from a wide range and number of Sinhalese sub-groups. Reaction was especially strong among the Buddhist clergy and the Sri Lanka Freedom Party. The SLFP organized a ‘mass campaign of agitation’ using the range of disruptive tactics (rallies, symbolic protests, Buddhist ceremonies of worship, and so on). The bhikkhus of the ‘Movement to Protect the Motherland’ organized a large assembly on 27 July which provided a platform for the expression of opposition to the Agreement. These meetings and protests took a violent turn on 28 July. Three buildings, including a Government Ministry and a state-owned newspaper office, were burnt down and scores of buses, trucks and cars were gutted. Police opened fire on a rampaging mob killing at least 19 people and injuring 120 others. Perhaps most symbolic of the depth of Sinhalese opposition to the


Agreement was the incident which occurred while Prime Minister Gandhi was reviewing a Sri Lankan honor guard before leaving the island on 30 July; a Sri Lanka sailor struck him with his rifle butt in full view of the international media. From 28 July to 2 August, over 2,500 acts of violence were reported, including at least 20 deaths and over 150 injuries. In addition, around 2,000 vehicles and over 700 buildings were destroyed or damaged (Samarasinghe and Liyanage 1993: 166–7). According to some analysts, the situation deteriorated to the point where Rajiv Gandhi deployed an Indian Navy destroyer in Colombo harbor in case the President and his Ministers need to be evacuated (de Silva 1993: 134). On the 18 August, during a government parliamentary group meeting on the Agreement in the heart of the Parliament Complex, a grenade was thrown at the President allegedly by a member of the JVP. The explosion killed a junior minister and injured several other members including Athulathmudali, Minister for National Security. Sinhalese anti-Agreement campaigns escalated and intensified, led by the JVP and the SLFP. The introduction of Indian troops into the north of the island allowed Jayewardene to redeploy Sinhalese troops to the south in an attempt to stem the tide of rapidly escalating violence by Sinhalese extremists, the JVP. Jayewardene made instrumental use of the Agreement to staunch both inter-group and intra-group challenges by military means. De Silva (1993: 135) observes that within three days of the signing of the Agreement, ‘more than half the Sri Lankan troops in the Jaffna peninsula were airlifted to Colombo in Indian airforce transport planes. Their presence and that of the Special Task Force of the Police helped quell the riots much more expeditiously than seemed likely at first.’ In was estimated that in early 1988 the government had over 50,000 security forces tied up in the southern section of the island, trying to contain the JVP threat and other anti-government activities (Weaver 1988: 81). During the parliamentary elections in February 1989, an estimated 70,000 troops and policemen were deployed in the south in an attempt to contain the violence (FEER, 2 March 1989: 13). Further adding to an already militarized situation, the government announced a plan to allow UNP Members of Parliament to recruit private militias or Home Guards of 300 supporters to provide protection from a resurgent JVP (Rubin 1987: 3–4). No other Political Party was officially permitted the same protection. This 30,000 man force was under the personal control of politicians. As private armies with state sanction, they were free to act without effective legal or judicial restraint. In essence, the government response to the escala-


tion of intra-group violence was provision of the means to contribute to further escalation. Tamil intra-group arena In the Tamil intra-group arena, LTTE confrontation with the third largest army in the world perpetuated the existing guerrilla war zone in the North and the East. As the IPKF assumed the role previously played by the Sri Lankan military forces, it too was accused of similar human rights abuses (FEER, 25 May 1989: 28; Rubin 1987; Hoole et al. 1992). As suggested above, more significant in the context of the theoretical concerns of this book is the polarizing effect of the Agreement on the Tamil paramilitaries. In the Tamil intra-group arena, the Indo-Sri Lanka Agreement reconfigured paramilitary power relations. The feuding between the LTTE and the other Tamil paramilitaries had been ongoing since the early 1980s.47 As a result, coalitions had been formed among non-LTTE paramilitaries for intra-group defensive purposes48 and inter-group offensive purposes. Although LTTE deserters and antiLTTE paramilitaries had on occasion cooperated in security force campaigns against the LTTE (e.g. O’Ballance 1989: 75), the Sri Lankan government and its military had remained an enemy common to all groups. In early 1987, however, all major non-LTTE Tamil paramilitary organizations (with the possible exception of EROS) came to view the government security forces as less of an enemy than the LTTE – which had been especially successful in its attacks against the competing Tamil paramilitary organizations (O’Ballance 1989: 75). With the signing of the Agreement and the introduction of Indian troops, the anti-LTTE Tamil paramilitaries allied themselves with the IPKF and began to participate in joint operations against their common enemy, the LTTE, in the North and East (Samarasignhe and Liyanage 1993: 159). This conditionally pro-government, anti-LTTE coalition of convenience (TELO, PLOTE, EPRLF, and the EPDP) continued joint operations with government security forces following the eventual withdrawal of the IPKF in May 1990. This is a clear example of the formation of an inter-group alliance in order to appropriate resources for use in the intra-group arena. The anti-LTTE paramilitaries acquired additional resources (in particular, funds, weapons, and logistical support) from their alliance with the Government that could then be applied directly to their intra-group feud with the LTTE. The signing of the Indo-Sri Lanka Agreement exacerbated intra-group violence in both the Sinhalese and Tamil arenas, and it formally49


internationalized the conflict by introducing Indian soldiers onto the island. The deployment of the IPKF in the North insulated the LTTE from direct attack by the government forces and vice versa but added a new axis of military confrontation in the conflict. And finally, far from inhibiting inter-group violence in the east, it exacerbated it. This episode offers useful insights into the way in which inter-group coalitions can be formed in order to pursue intra-group objectives. Progovernment or pro-Agreement Tamil paramilitary organizations recognized the opportunity to mobilize resources from the inter-group arena to strengthen them in their battle against the LTTE. Pro-Jayawardene factions within the UNP also saw the opportunity to draw on the military resources offered by the Government of India – which, not surprisingly, had its own geo-strategic and domestic political reasons for getting further involved in the conflict. Jayawardene saw the ‘India option’ as a means of advancing his government’s interests in both the inter-group arena (the replacement of Sri Lankan troops for Indian troops in the anti-LTTE campaign) and the intra-group arena (the reallocation of Sri Lanka troops to the Sinhalese majority areas of the south to quell growing intra-group agitation by the JVP). Yet when the India option became more of a liability than an asset, Jayawardene’s successor was not averse to offering support across the ethnic divide to his arch enemy, the LTTE, if it would help him to push the Indian Peace Keeping Force off the island. The tactical character of such coalitions, the prominence of subgroup self-interest and the conspicuous mobilization of resources points to an underlying instrumentalism which ethnic entrepreneurs may employ to animate sub-groups. Despite such interest-based behaviour, it would be a mistake to equate ethnic groups with interest groups. Ethnic sub-groups clearly have interests as well as mobilizers working to rally support in pursuit of these interests. Yet, ethnic subgroups are not reducible to interests. As Epstein notes, ‘to describe an ethnic group as having interests is one thing, but to define it in these terms is quite different’ (cited in McKay 1982: 400). McKay continues: ‘Nor does the existence of political and economic interests negate the continued existence, even in a modified form of primordial interests’ (ibid.).

9 Critical Juncture VI: February 2002 Ceasefire Agreement (CFA)1

The episode analysed in this chapter differs from the others in this book, because it was still being played out as this book was being completed. While the selection of this episode allows for the testing of a two-level analysis on current events, the chapter may lack a more definitive sense of closure that is possible in historical cases. However, the contemporary character of the episode allows more latitude for probing inter- and intra-group political dynamics, for example, through direct interviews with observers and actors in Sri Lanka, and through the use of electronic sources of information absent in all previous cases considered in this book.2 As the year 2002 dawned, Sri Lanka found itself in a position it had not been in for some time. The election of Ranil Wickramasinghe as prime minister in December 2001 stimulated a wave of cautious optimism among Sri Lankans and international observers alike. However, it was an optimism tinged with an ominous sense of déjà-vu, having had similar hopes and aspirations dashed following the election of Chandrika Kumaratunge as president in 1994 when her People’s Alliance Coalition swept to power on a much touted platform of peace and constitutional reform. President Kumaratunge, too, had declared a unilateral ceasefire and initiated peace talks with the LTTE within two months of her inauguration. However, the high expectations and optimism of 1994/95 were quickly dashed when a re-armed and reinvigorated LTTE attacked and destroyed two naval craft and two SLAF Planes in April marking the start of what came to be known as Eelam War III. The president employed the Orwellian argument that the state needed to launch a ‘a war for peace’ thereby setting in motion an escalating spiral of ferocity and destruction. Optimism was dashed by the fire brand of war as the island was 156

2002 Ceasefire Agreement 157

plunged again into economic, social, political, and most especially, human turmoil. If the 1994/95 Kumaratunge peace initiative is one important point of reference for setting the context and analysing the 2002 CFA, the second referent is signing of the Indo-Lanka Peace Agreement in July 1987. As discussed in Chapter Seven, the signing of that agreement served as the lightening rod for the second JVP insurrection – an explosion of violence unprecedented in colonial and post-colonial history. A critical question at the time of the CFA was whether it would detonate a chain reaction of similar violence. Could political entrepreneurs or opportunists manipulate and harness the fears of vulnerable and volatile Sinhalese sub-groups, such as Sinhalese Buddhists, ‘armed stakeholders’ (including 30,000–40,000 armed deserters), disgruntled political parties (or factions within those parties), and perpetually disillusioned youth? The ingredients for such a possibility certainly existed. But such an outcome did not occur (at least, not within the timeframe of the writing of this book). The Wickramasinghe regime was able to manage the tensions with notable finesse.3 However, sporadic attacks, such as the bomb attack on the residence of the Sarvodaya leader and high profile peace activist Dr A.T. Ariyaratne4 and others discussed below, illustrated the potentially deadly expression of opposition and dissent within the Sinhalese constituency.

The context From the mid-1990s the military confrontations between the LTTE and Government forces intensified. Territory passed back and forth between combatants. In 1995, the SLA recaptured Jaffna from the LTTE, but by May 2000 the Tigers marched back to (but not in through) the same gates. The LTTE had demonstrated its ability to launch large-scale lightening operations against dug-in, and undermotivated, government troops – with devastating impact. In late 1998, one battle saw an estimated 2000 troops slaughtered in a 36-hour period. During this same period, the LTTE brought terror to the economic and political heart of the island. Colombo was wracked by assassinations (including a near miss on President Kumaratunge in which she lost an eye), and bomb attacks carried by trucks, cars, bicycles, and ‘Black Tigers’ wearing LTTE-engineered suicide jackets. In 1996, a massive bomb destroyed the World Trade Centre and adjacent Galadari and Hilton hotels (killing 86 people and injuring 1,400). In October


1997, the LTTE bombed the Bank of Ceylon – successfully gutting the entire Central Business District, and draining it of economic vitality. In 1998, LTTE targets expanded to include the highest profile politicalreligious site on the island when a truck bomb ploughed into the most sacred Buddhist shrine in Sri Lanka, the Dalada Maligawa in Kandy. Then, in August 2001, the LTTE launched a devastating attack on what should have been one of the most fortified military targets on the island: the military airbase immediately next to the International Bandaranaike/Katanayake Airport. It was one of the worst attacks in aviation history.5 It was clear that the LTTE could not only launch successful attacks, it could launch spectacularly successful attacks. The lethality of LTTE attacks against military and civilian targets had not been blunted in the least after almost two decades of military confrontation. If anything, it had been sharpened. By the start of the new millennium, the economic outlook for Sri Lanka had never looked worse. Years of bloated and unaffordable defence spending, combined with negative fluctuations in oil prices bit deeply into Sri Lanka’s economic maneuverability by virtually depleting its foreign currency reserves. Analysts were predicting that the end of the textile quota system in 2004/05 would lead to the likely collapse of 50–60% of the garment industry – one of the principal sources of foreign earnings.6 This was exacerbated by the unsustainable military and security budget that maintained over 250,000 men, women, and children under arms.7 In effect, successive governments had been using military employment as a grand youth employment scheme cum poverty alleviation programme – to the point where the economy of the rural poor in the South was three times more dependent on ‘military remittances’ than official poverty alleviation programmes (Janasaviya and Samuradhi); and more dependent on Army recruitment and compensation than on overseas remittances (32 per cent versus 30 per cent in 1997) (Dunham and Jayasuriya 1998). The implication of the shifting dependencies within the rural economy is articulated by Dunham and Jayasuriya as follows: ‘the emerging picture is one of a fragile and highly brittle peasant economy, where poverty may have been staved off (for the moment) by transfers and remittances, but where there is no identifiable and sustainable way out of poverty’ (ibid.). While poverty may have been staved off (for some segments of the Sri Lankan population), it was being bought at the cost of continued military violence. The unsustainability of this situation at the household level was paralleled at the national level. The election of Ranil Wickramasinghe was followed by initiatives which had immediate, visible, and positive impacts on lives and liveli-


hoods throughout the island, for example: the de-escalation of military confrontations through the maintenance of parallel unilateral ceasefires; improved humanitarian cooperation which allowed easier movement of goods and people between cleared and uncleared areas; the opening of large sections of formerly blocked roads (most notably the A9 highway to Jaffna and the main thoroughfares of Colombo); and the rapid clearing of landmines to allow ‘normalization’ of civilian life. The dismantling of the formal pass system for Vavuniya, combined with the drastic reduction and removal of check points throughout the island reduced indiscriminate interrogation and had positive ripple effects (economically in the movement of goods and people, as well as psychologically). Most importantly, they reduced contact with many of the quotidian structures of authority that subsidized and permitted human rights abuses. Particularly notable was Wickramasinghe’s skillful ability to undertake important symbolic unifying acts – such as his trip to the Jaffna Peninsula (14 March 2002), which was the first visit by a Sri Lanka Prime Minister since 1983. The suggestion that a bridge be built to link Northern Sri Lanka to Southern India is of the same symbolic magnitude (despite the obvious technical, engineering, and financial challenges inhibiting the movement from idea to action). Other initiatives arose from within civil society and the private sector, as illustrated in the visits to Jaffna by delegations from the Ceylon Chamber of commerce, and their expressions of interest in investing in specific business opportunities in Jaffna (in the banking sector, supermarkets and factories). There appeared to be systematic efforts to initiate a gradual and calibrated process of moving from ‘reversible good will gestures’ to non-reversible actions designed to nurture the peace process; in effect, efforts to move from the ‘putty’ of good will to the ‘hard stone’ of institutionalized self-, joint-, and common interests. The rapid action by Wickramasinghe stood in contrast to the earlier gestures by Kumaratunga which encountered difficulties and obstacles in implementation. Thus, for example, in August 1994 she announced that her government would eliminate the embargo on most goods being transported to Jaffna – a move which would have been a significant step in redressing a deep and well-founded grievance of the Tamil population. However, this publicly announced gesture was never fully implemented. The US Ambassador to Sri Lanka during this period suggests that this was the result of poor organization and follow-up rather than doubtful bona fides of the President (Schaffer 1999: 136). In light of difficulties that humanitarian and development actors have


had in the past in getting supplies in to the North, the disconnection between the policies in Colombo and the practices in military controlled areas must be acknowledged. The end-result was a loss of public confidence in the North in the government’s commitment to peace. Also in contrast to the Kumaratunga initiative, the Wickramasinghe regime did not enter discussions with the LTTE with a ‘pre-prepared’ proposal to change the constitution prior to the signing of a potential agreement. According to G.L. Pieris, the government was entering with no pre-conceptions or preconditions – with one exception: the partition of Sri Lanka into two separate states was not on the table. The implication was that constitutional changes were something that should follow from rather than precede the talks with the LTTE.8 However, in 1994/95, the government ‘grew increasingly exasperated with the LTTE’s unwillingness to review its constitutional proposals’ (Schaffer 1999: 133). It is particularly striking how disconnected the political front of the Kumaratunga peace initiative was from the military action that dominated the headlines of national media at the time. In stark contrast, the Wickramasinghe initiative appeared to have reigned in the military forces quite tightly. Indeed, many civilians in the North and East prior to the signing of the formal Ceasefire, felt that the Government Forces refused to intervene in incidents of human rights violations and criminal activity attributed to the LTTE for fear of disrupting the delicate political discussions taking place. The question of whether President Kumaratunga could not, or would not, reign in the military was a very important one. Maintaining the military within its constitutionally defined roles under governmental oversight is critically important in any post-conflict scenario in which demobilization and demilitarization will be essential requirements for the war-to-peace transition. This discussion underscores the importance of leadership capacity. It appears clear that Ranil Wickramasinghe was much more politically adept than Chandrika Kumaratunge who entered office with minimal formal political experience in 1994, despite the political pedigree of her family. While her lack of political baggage may have been appealing to the electorate, it may also have increased her vulnerability to manipulation by the LTTE who exploited her need to be seen to be responding to the Sinhalese peace constituency. By contrast, Wickramasinghe was a veteran politician who was able to approach and negotiate with the LTTE with a hardened, interest-based, perspective. This pragmaticrealist perspective on the current peace process was articulated by G.L. Pieris, one of the principal architects of the Wickramasinghe’s


strategy.9 To the question, ‘why should the Government trust the LTTE in the current discussions,’ he responded that the government was ‘not entering this in a spirit of distrust, but one of reality’ based on the recognition that no party wins – or can win – through war-fighting. He felt that the government had to enter talks with optimism so as not to undercut the process. More pointed, was his response to the assertion that the LTTE was only using the calm to re-group and re-arm as it has done in the past. Pieris is said to have argued that the government is certainly regrouping and rearming, and that it was unrealistic to expect the LTTE not to do the same until there is a ‘real peace.’ Such are the motives and realities in a non-peace context. This logic informed behaviour on both sides of the battle line. From an inter-group perspective, when the political capacity of a new leader is coupled with dire economic forecasts, military stalemate, a war-weary population, and the presence of neutral Third Party facilitator, then there governments appear more receptive to the idea of war termination – if not full blown peace. Whether the transition from war to peace is possible or successful, as we see below, is dependent on intra-group conditions.

The episode A formal ceasefire agreement was signed between the Government of Sri Lanka and the LTTE on 22 February 2002 after sustained and diplomatic facilitation by the Government of Norway privately and publicly. The behind-the-scenes role of Norway as an international facilitator was a new and important factor contributing to the establishment of conditions to enable the initiation of a formalized cease fire and peace process. The discreet Norwegian role generated and harnessed sufficient trust from the principal stakeholders to ease the process forward – including India, which was consulted each step of the way. Particularly noteworthy, was the ability of Norway to employ the appropriate political resources necessary to constructively affect the cost-benefit calculations of both the GSL and LTTE. The signing of a joint ceasefire document followed a three-month period in which the SLA and the LTTE maintained parallel but separate ceasefires. The Agreement included a clear timetable and a range of confidence building measures, in addition to a monitoring mechanism – the ‘Sri Lanka Monitoring Mission’ (SLMM) composed of Sri Lankan and Scandinavian members. The other particularly noteworthy part of the CFA the was Article 2, which details the commitments of the signatories to undertake specific


‘measures to restore normalcy’ with explicit reference to the ‘[abstention] from hostile acts against the civilian population, including such acts as torture, intimidation, abduction, extortion and harassment’ (Article 2.1). Very importantly from a legal context was the prohibiting of ‘search and arrests under the Prevention of Terrorism Act,’ requiring instead that all arrests ‘shall be conducted under due process of law in accordance with the Criminal Procedure Code.’ This was critical for beginning the process of removing one of the most important institutional structures subsidizing the abuse of human and civil rights. According to a colleague who was to meet up with the LTTE two days after the signing of the Agreement, the meeting was cancelled because the participants from the other side did not yet have the appropriate passes or permission required under the terms of the Agreement to travel from uncleared to cleared areas. In that instance, it appears that the Agreement was influencing behaviour in the immediate period following the signing. Subsequent violations by both sides (the majority being by the LTTE – see below) indicate the variability in the honouring of the agreement. It was unclear whether such violations were systematic or incidental. Ultimately, it will be the Sri Lanka Monitoring Mission which determines the degree of adherence by the signatories. In mid-July, there was a particularly worrying incident in Palk Bay, South of Jaffna peninsula in which two LTTE trawlers were intercepted by the Sri Lankan Navy (SLN). The subsequent stand-off involved the LTTE holding two Sri Lanka Monitoring Mission Naval Monitors hostage as they were engaged in their official duty both in inquiring into the incident, and in also settling the dispute.10 But the agreement continued to hold.

Model I With the application of an inter-group level of analysis, we can begin to identify some of the dynamics compelling the Government of Ranil Wickramasinghe and LTTE Leader Vellupillai Prabakaran to move towards a formal ceasefire and peace talks. Military stalemate Until the election of Wickramasinghe, the dominant form of communication between the Government and the LTTE was military violence – spilling over into, and dominating, the civilian sphere in the North and the North East.11 However, as detailed above, neither side was able to deliver a definitive military victory. Particularly sobering for the GSL


was the SLA loss – in three days in early 2000 – of territory that had taken them years to win inch by bloody inch. On the other side of the inter-group divide, some residents of LTTE-controlled areas suggested that the Tiger leadership was aware of the gaps between (1) its possession of the military resources that allow it to make territorial gains (2) the lack of the military resources to hold new ground, and (3) the lack of political resources to succeed on the political front domestically and internationally.12 For a number of years, the LTTE has intimated that it would be willing to accepted something short of an independent state of Tamil Eelam. This rhetoric was certainly prevalent in the period surrounding the signing of the CFA. According to a religious leader in Batticaloa, there was an acute sense within the civilian population that ‘the sacrifice has been too great, and the goal still so far away’ – a sense, perhaps, echoed within the LTTE. In the over-used language of William Zartman, this is sometimes expressed as a ‘hurting stalemate’ in which (1) the balance of military [and non-military] forces are such that each side sees clearly that neither can win, and neither can lose; and (2) conditions will get ‘precipitously’ worse if the war continues. Most commentators in Colombo pointed to the ‘war weariness’ of all segments of the population. This was a common assertion in 1994/5 as well, suggesting (if true) that this may be a necessary but insufficient condition for peace/peacebuilding. As one observer from the East put it diplomatically: ‘The Tamil community is very, very tired. We have lost many people to both the LTTE, and to the Sri Lankan military. They are grateful to the LTTE, but are now calling for something else.’

Model II Intra-Tamil LTTE financial and human resources When the CFA was signed many representatives of the donor community and the Colombo-based Sri Lankan research community, felt that the terrorist attacks in Washington and NYC – and more specifically the concerted military–security response by the USA which began with its bombing campaign on 7 October – was a factor influencing the LTTE decision to enter into peace discussions with the government. Having already been proscribed in the countries containing large diaspora populations (representing large voluntary, and involuntary, support bases), they argued that the LTTE leadership was maneuvering


to avoid a further clamp-down by the US and its allies in the so-called ‘war on terror.’ Others argued that the efforts by some of the major powers to choke off funding to terrorist organizations were starting to have an impact on the external fund-raising activities of the LTTE. These efforts were given additional force in the post-7 October world. A member of the TULF however, felt that the attacks of 11 September/07 October did not have an impact on ‘the LTTE mind set,’ arguing instead that the change in thinking occurred in May 2000 when the LTTE were within inches of capturing Jaffna and the US, India intervened/sent a strong signal to LTTE that it would have to get involved if the LTTE began to slaughter the 40,000 surrounded soldiers.13 A crusty International NGO worker came to the same conclusion, albeit using a different argument: ‘How much pressure is the LTTE under, really? Is their money really drying up? They can always adopt other modes [of fund raising]. The September 11 issue is bullshit. After the attacks [on September 11], Sri Lanka went from 500th to 5000th on the International Conflict “hit parade.” ’ Nonetheless, the LTTE itself has made an effort to justify the increase in the scope and scale of ‘taxation’ extortion in the East in 2002 as a necessary response to the reduced flow of funding from expatriate Tamils because of the bans imposed the organization in the US, Canada, the UK and Australia. The East Coast Tiger leader Karuna, ‘admitted that the movement was charging money for its upkeep.’14 This included very heavy taxes on goods being transported to LTTE controlled areas. In late June the Tigers released a list of goods liable for taxes. It included some 112 items – from a box of matches to gasoline – and excluded several items including children’s milk foods and biscuits. Further, all government employees were asked to pay 8 per cent of their salaries while shopowners and businessmen paid amounts specified by the LTTE. This meant that they were subject to a form of double taxation (CPA 2002). Under the terms of the CFA, these activities are wholly prohibited. It was widely agreed that during the period prior to the CFA, the levels of forced recruitment, kidnapping and extortion rose markedly in the East. Those with access to the North and North East at the time, reported that this was evident from Mannar to Batticaloa to Jaffna.15 As during the 1994–95 ceasefire period, the LTTE moved fluidly across the borders from uncleared to cleared areas, leading (according to some observers) to increased militarization of the East from the North. It appeared that the permeability of the borders, combined with the reluctance of Government forces to engage in confrontation, enabled the broadening of the geographical scope of regular fundraising activi-


ties.16 It was widely reported for example, that the LTTE systematically entered Banks in ‘cleared’ areas to collect the names and bank account balances of all customers. Kidnapping and ransoming have always been a LTTE fund-raising tactic. However, the period preceding the signing of the Ceasefire Agreement saw an increase in the scale and scope of this activity: scale in terms of the excessively exorbitant demands for ransom money and scope in terms of the number and status of those being kidnapped – which included some of the most prominent businessmen and community leaders who had previously always been able to negotiate a mutually agreeable arrangement. This was especially evident during Fall 2001. According to a human rights worker who approached the military authorities concerning the kidnappings in the Eastern Province at the time, he was told that there was nothing they could do, because they feared destabilizing the security environment during the precarious period of parallel unilateral ceasefires.17 After the signing of the CFA, the LTTE continued these activities. Complaints of violations were lodged regularly with the SLMM. The number of complaints concerning extortion, abduction and forced child recruitment by the LTTE was high. At the end of June, the SLMM reported that it had received 380 complaints, 270 of them have been made against the LTTE and 110 against the government. The most common violation by the LTTE was abduction or kidnapping followed by harassment and extortion. Concern was also repeated regarding child recruitment. On the government side, violations listed include: harassment, restriction of movement, restriction on fishing, and occupation of civilian land and property.18 In the immediate post-CFA Sri Lanka, some concern was expressed by donors about the potential destabilizing impact within LTTE-controlled Tamil areas due to the speed with which the Wickramasinghe regime moved the peace process forward, e.g. the lifting of sanctions on the movement of goods and people in and out of the Wanni.19 The risk of destabilization was seen to be directly related to the challenges to the control of, and support for, the Wickramasinghe and Prabhakaran regimes within their respective communities. As one international aid worker pointed out, by decreasing barriers to the movement of goods, services and people to the North and North East, the GSL was increasing economic opportunities for civilians. As a result, the space was created for them to address some of the economic grievances themselves – reducing the role for the LTTE in managing the economic affairs of the populations within territory under its control.


In LTTE-controlled areas, the removal of restrictions on movement of goods and people went a considerable way towards addressing one of the central grievances of Tamil civilians – and to the extent that such grievances translated into support for the LTTE, then their removal entailed a commensurate reduction in the LTTE’s support constituency. Further, one would expect that if push and pull factors are sufficiently strong, Tamil civilians in LTTE-controlled areas may ‘vote with their feet’ by leaving for areas outside of the realm of Tiger control – again, undercutting the authority and resource base of the LTTE. In the assessment of one head of an international organization who had met with LTTE leadership, the LTTE were concerned that their absolute control was starting to crumble. They feared being trapped by their own rhetoric. Any increased movement of people was threatening to them. ‘The LTTE may be willing to accept this for a few months as long as there is a political gain/compensation. But if the gain diminishes and the control/rule of LTTE is threatened, then it may be induced to break the ceasefire. Paradoxically, the speed of government concessions may in fact undermine the peace process by its destablilization of the LTTE. If this is a government strategy, then it is doomed.’ One of the ways of reducing the chances of LTTE ‘defection’ from the peace process, was by increasing the public profile of International Community support for the Ceasefire Agreement and peace. This was a tactic pursued actively by international agencies based in Colombo which, for example, arranged a visit by 30 Colombo-based foreign diplomats to Jaffna in late March,20 as well as a steady stream of high level delegations by the UN, the Asian Development Bank, the World Bank, as well as leading donor countries like the US, Britain, and Norway. A land mark press conference by Prabhakaran in Kilinochchi in April attracted a frenzy of 300 local and foreign journalist further – further committing the LTTE to the peace process in a very public way. According to field worker from a humanitarian organizations working in uncleared areas, the LTTE seems overwhelmed by the attention and the many diplomatic missions coming through. This served to increase the difficulties (political costs) for the LTTE to be seen to be stepping away from the peace process. LTTE motivations In the wide-raging discussions concerning incentives and disincentives for war or for peace within the various constituencies on the island, perhaps the most important, and most uncertain factor, was the ques-


tion of the motivations and reliability of LTTE participation in the ceasefire and subsequent peace talks. There were many hypotheses and suggestions as to why the LTTE, and specifically its leader V. Prabhakaran, might reject war-making. A psychological explanation offered in both Colombo and the Eastern Province was that Prabhakaran may be motivated by a desire to see, and to enjoy the fruits of, victory in his life time. Comparisons were made to Chairman Arafat’s life-long (and at the time still unrewarded) struggle to attain the goal of statehood for the Palestinian people. It was also suggested that he may have matured and mellowed over the years in the change from teenage revolutionary to husband and father. Although the military successes of the LTTE have been impressive, it was also pointed out that recruitment appeared to rely increasingly on coercion and force, leading many to note the diminishing quality of cadre. Not unrelatedly, a number of observers pointed to increased incidents of resistance and dissent by Tamil civilians towards LTTE heavy handedness, from the school principal who refused to allow it to recruit on school property, to street demonstrations by whole villages protesting acts of extortion, kidnapping, and ‘taxation.’ One field-based NGOs worker, mentioned an incident in Valaichchenai a week prior to the signing of the CFA, when the community banded together to stop the LTTE from kidnapping a business man. There is a second, equally critical, intra-group factor that affects the prospects of war or peace. It is one that is very rarely discussed: the question of whether there are rogue factions within the LTTE which would continue armed struggle in direct opposition to Prabhakaran if he was to formally commit to non-violent political competition. Some observers suggested that the escalation of extortion in the East (in which the LTTE denies involvement) were the acts either of factions building up ‘war chests’ to finance continued armed struggle, or of individuals bent on self-aggrandizement. Throughout the history of the LTTE, Prabhakaran has always brutally eliminated dissent within his ranks and within the Tamil Community.21 However, in the simplified logic of bargaining and negotiation theory, the more parties to a game, the more difficult it is to reach and maintain mutually acceptable solution. Dissenting armed factions in an volatile, ethnicized, violence prone environment, would not bode well for peace processes. The official LTTE response to the accusation of increased extortion was delivered in publicly circulated notices in which it denied responsibility, and accused others ‘like them’ of undertaking the activities


(implying, pro-government Tamil paramilitaries such as the TNA, Radzek, EPRLF, TELO).22 The notice stated that the LTTE was not raising taxation levels. It reminded citizens that LTTE Cadres are always identifiable by their dog tags and cyanide amulets. And, it reminded the public that the LTTE always gives receipts (a fact confirmed by those who have suffered extortion/kidnapping by the LTTE). The notice provided the address of the LTTE Financial Office if any one has concerns or complaints. One individual in the East at the time, suggested that many of these activities might have been the actions of LTTE individuals rather than the organization as a whole. The motive behind them might be to collect the resources necessary to allow them to leave the country before the next phase of the military–political game. A kind of pension scheme that fits the modis operandi of the organizational culture. It was also widely reported that there had been a very intensive increase in recruitment by the LTTE. In Batticaloa District, the following areas were noted in various discussions to have lost youth and children to LTTE recruiters: Kumarthurai; Kiran; Eruvur; Sithandee; Vundharumulli; Sandhivelli; Sowkuddi; Kuddiyurripu; Mavadivambu; Vallaichennai; Chankaladi; Thirukoil; Kalluvangikuddi; Pouruthivu.23 The children were said to have been taken to Tharavai, the main LTTE training base for Batticaloa District.24 The official LTTE response to accusations of increased forced recruitment is also outlined in a public notice, which states bluntly that ‘Priests’ (of all denominations) are giving false reports to the press. One resident of Batticaloa – with no love lost for any of the armed stakeholders – suggested that there was in fact a voluntary increase in youth joining the LTTE in ‘the hope of being part of the movers and shakers and job [recipients] of the future if the LTTE joins the political process.’ Joining the LTTE ‘voluntarily’ (as with the SLA) is not unrelated to sustained and systematic efforts glorify military culture. The LTTE creates and nurtures a popular culture of militarism and martyrdom through, for example, war videos, speeches, and verbal haranguing by its political wing. Even children’s play grounds are militarized with toy automatic weapons mounted as handles on see-saws. In Batticaloa, one father told me that his son had brought home from school the LTTE video of the attack on Bandaranaike/Katanayake Airport in August 2001. He figured that all A-level students in town would probably have already viewed the video. According to Father Harry Miller of the Batticaloa Peace Committee, although all of these activities were continuing at the time, the pace


slowed down as a result of ‘publicity’ and consequent international pressure on the LTTE. One long-time and well-known human rights observer in the east expressed a widely held doubt that the LTTE would be able to make the ‘democratic transition.’ It was felt that its ability to cultivate or harness those internal political capacities which might enable the transformation from a military to a political project are limited by the rigidly topdown-greatman structure of the organization, that is ‘all things come from the mind of Prabahkaran.’ Even Anton Balasingham, the lead political ideologue, is said to vet all statements through Prabahkaran. ‘If they have any brains,’ said the observer, ‘the LTTE would use the existing political system and network of administrators in any effort to manage the N/NE.’ However, given the structure, history, and culture of the LTTE, it will likely try to insert its own (bureaucratically incompetent) men, and thereby ensure the failure to develop effective structures of governance. In any case, the execution of dissenters or competitors to Prabahkaran (such as the murder of deputy leader Mahathaya) has culled many individuals who might have possessed the necessary skill set to administer and manage an LTTE political entity. The micro-management style of the LTTE allows very limited space for choice, innovation or learning. One example offered to illustrate this point was what was called by one observer, the nascent ‘Talibanization’ of the East – public dictates by the LTTE that appear to be designed to enforce a conservative Tamil nationalism (such as: prohibiting women from wearing pants and visiting the beach, and forcing women to wear only saris). There is no evidence from past experience that the LTTE might be flexible enough to move forward in a way other than the closed Prabahkaran-driven fashion. As the human rights observer put it: ‘There is absolutely no reason to believe that Prabahkaran could be an enlightened statesman.’ This sentiment was echoed by a development worker who said, ‘ “community empowerment” not in the LTTE dictionary.’ Indeed, for weeks after the signing of the CFA, the LTTE prohibited local development workers for International NGOs from leaving LTTE controlled areas in the Wanni for reasons which were unclear, but which nonetheless underscored its authoritarian style. Intra-Muslim tensions and Tamil–Muslim tensions25 The unhealed wounds of the Kattankudi Mosque Massacres, just south of Batticaloa, on 8 March 1990 – in which 103 people were murdered –


sustain deep levels of distrust between the Muslim and Tamil communities in the East. Massacres in Eravur (121 killed) as well as in border villages are central points of reference in the Muslim community’s relations with the LTTE. Each case of kidnapping and extortion alienated the Muslim community not only from the LTTE, but from the Tamil community as well. It needs to be underscored that extortion and kidnapping is not restricted to the Tamil population. In Kattankudi, Muslim businessmen and civilians also face extortion, theft, and kidnapping by the LTTE. Indeed, many in Kattankudi reported that they were affected more intensely than the Tamil civilian populations because the LTTE forces them to pay additional support because they do not provide their children as warriors. Further, on the other side of Muslim inter-group relationships, members of the Muslim community in Sri Lanka also felt strongly that they were neglected in the Government’s peacebuilding, development, and economic-strengthening efforts. Muslim leaders felt they had been overlooked in discussions of civil society organization, in the media, and in human rights activities. Discontent and insecurity was growing within the community at the time of the CFA and its immediate aftermath. There was a danger that the government and LTTE would further aggravate this sense of alienation if their concerns were not acknowledged. If non-violent political interaction was felt to be ineffective then the danger looms of armed factions26 articulating their political demands through the use of militarized violence – the principal form of political dialogue on the island over the last two decades. Such scenarios were expressed in discussions within the Muslim community at the time. However, most community leaders in Kattankudi made it clear that they were not seeking special rights, but equal rights. As importantly, the majority of those were not calling for separate states for Muslims and for Tamils. Nonetheless, they were adamant in their refusal opposed opposition to the unification of the Northern and Eastern Provinces under a single political unit that would be controlled by the LTTE – one of the most contentious issues to be settled if the peacebuilding process is to take root. As was pointed out by Muslim leaders in Kattankudi, the LTTE demand for a unified North and Eastern Province administrative unit would decrease the proportion of the Muslim population from 32 per cent (in the East) to 17 per cent overall. In narrow, mathematical terms, this is seen to be a diminishment of their political voice – unless appropriate guarantees and corresponding mechanisms were put in place.


The precedent to which Muslim leaders referred was the 1987–91 period of the N/NE Provincial Council which was argued to have been directed by Jaffna Tamils in their own interests. This leads to the conclusion that any peace agreement must satisfy not only the LTTE and the Government, but the Muslims (and potential spoilers) as well. One message that was delivered at the time from the East was that it would be a mistake for either the GSL or the international community to lump the North and the East together. Eastern Tamils and Muslims alike fear domination by Jaffna Tamils. It was repeated many times in the East that it ‘is dangerous’ to accept the LTTE’s claim to be the sole representative of ‘the’ Tamil people. Any and all constructive peacemaking and peacebuilding interventions must build from an understanding of the complete and complex ethnic-political puzzle. Conscious of Muslim-Tamil tensions (and their potential negative impact on LTTE relations with the donor and international communities), the LTTE took efforts to be seen to be addressing them. In April, a Memo of Understanding was signed between the SLMC and LTTE – which among other things contained encouraging statements from the LTTE that it would welcome back displaced Muslims returning to their homes in LTTE controlled areas. At the same time, the chief political ideologue, Anton Balasingham admitted that the 1990 expulsion of Muslims from Jaffna by the LTTE was a political mistake. This was followed by an apology to the Muslim community by Prabakaran. Then in early June, the head of the SLMC, Rauf Hakeem met with the leader of LTTE’s political wing Thamilchelvan in Kilinochchi, and agreed to set up a ‘special LTTE–SLMC joint committee’ to monitor the full implementation of the Memo of Understanding. Perhaps nowhere in the current episode is the salience of intra-group dynamics more evident than in the eruption of violent clashes between Muslims and Tamils in Batticaloa and in Valaichchenai, in late June (2002) involving grenade throwing, shootings, murders, abductions, disappearances, stoning buses, theft, and the burning of shops, homes and a local municipal building.27 Police and security forces appeared to stand aside as mob violence escalated. Even once a curfew had been imposed, a row of shops located next to the police station were burned to the ground. The region was brought to a virtual standstill. Over a 100 people were injured and 11 people were killed. A total of 96 shops are on record as having being burned. Of them, sixty one belonged to Muslims and thirty five to Tamils. This included established shops containing millions of rupees worth of goods as well as small scale cadjan tea shops.


The targeting of shops underscores some of the economic tensions that exist between the Tamil and Muslim communities in Batticaloa District. The general sentiment is that the Muslims own most of the shops and businesses – 90 per cent of the shops in Valaichchenai; and 80 per cent of the shops in Batticaloa town (CPA 2002). They also own extensive paddy land and cattle herds. And because Muslims also dominate the local level money-lending sector, many farmers and fishermen are indebted to them. This leads to the Muslims purchasing produce – rice and prawns, for example – from the producers (usually Tamils) at a price which is much lower than the market price. Thus, there is a strong feeling of economic exploitation of Tamils by Muslims, expressed by Tamils. This was never denied by the Muslims interviewed for the CPA Fact Finding Mission, they said that since the CFA, the LTTE was encouraging Tamil farmers and fishermen not to pay their debts to the Muslims and not to trade with the Muslims in the area. And finally, the Muslims contend that although they are 28 per cent of the population of Batticaloa, they occupy only 2 per cent of the land. Even if this is not a factual statement, it indicates a sense of insecurity within the Muslim community of the District that needs to be addressed. It appears that the Batticaloa and Valaichchenai incidents were related to the violence in Muttur a few days earlier when tensions between the local Muslim community and the LTTE sparked an attack on the local LTTE office allegedly by Muslims. The LTTE in the Batticaloa area called a hartel (mass strike) in protest of the attacks in Muttur – thus instilling an anti-Muslim sentiment in the hartel. No advance notice was given to the Mosque Committee which is the traditional forum for disseminating such information, and as a result, on the appointed day of the hartel, many Muslims (and Tamils) who were unaware of the action wandered out and found themselves in areas dominated by members of the other community. Although it is unclear what incident triggered the unrest, it is clear that mobilizers on both sides were armed with grenades, and that both sides used them against each other. When asked as to how civilians came to be in possession of grenades and guns, the Muslims said that the LTTE was supporting the Tamils while the Tamils said the police was supporting the Muslims (CPA 2002). It was clear that on both sides, some outside forces had been involved in fomenting the clashes. Efforts intended ostensibly to dampen tensions served to exacerbate relations within and between Muslim and Tamil communities. On 29 June, Minister and SLMC leader Rauf Hakeem visited Valaichchenai and spoke to representatives of the Muslim community, visited several of


the homes where Muslim people had been killed. However, he did not meet with or visit any Tamils in the community which consequently stoked the disgruntlement of the Tamil community (CPA 2002). On the same day, the LTTE leader Vishu came and met exclusively with the Tamil community. Later that day the LTTE distributed a leaflet blaming those who were opposed to the LTTE–SLMC accord as being responsible for the troubles. On 30 June, an organisation calling itself the Organisation of Ethnic Unity in Kattankudy issued a leaflet accusing Mr Hisbullah, a Muslim political leader, and his supporters of being involved in the clashes. The homes attacked during the violence included that of Deputy Minister Abdul Cader. According to the fact finding mission: Everyone was clear that this last incident was a consequence of tensions within the SLMC and attributed it to groups in the Muslim community who opposed Mr. Cader’s support for the SLMC leader Mr. Rauf Hakeem. There is a strong feeling that there is group of young militant Muslims behind the clashes although nobody was able to identify them. The existing tensions between Mr. Hakeem, the SLMC leader and Mr. Hisbullah of the SLMC were also contributing to these tensions. From an intra-group perspective, this incident reveals political dynamics which may prove to be the Achilles heel (or perhaps one of two intra-group Achilles heels) to the peace process. The escalating tensions between (1) the Muslim community and the LTTE, and (2) the Muslim and Tamil communities more generally, are exacerbated by (3) schisms within the Muslim community and specifically the competition between Muslim factions to be seen as the representative of the Muslim community. One of the principal intraMuslim schisms is that between supporters and opponents of Minister Rauff Hakeem, leader of the SLMC. His main opponent is Minister M.H. Mohamed who has attempted to stir up support for his faction, and undercut support for Hakeem, by accusing the latter of being unable to reduce the extortion and violence by the LTTE against the Muslim community despite having a signed agreement. Following the clashes, the media became more explicit in their reporting about the fissures appearing within the SLMC over the government’s handling of the crisis. A group of SLMC parliamentarians even threatened to sit with the opposition if ‘the safety of Muslims in the east were (sic) not assured.’28 In response, Hakeem began to take a much more public oppositional and hard line stance towards the LTTE,


rejecting for example, not only a unified Northern and Eastern province, but ‘LTTE leadership in the Eastern Province’29 in an apparent effort to regain community support being cultivated my M.H. Mohamed. However as long as LTTE harassment continued, the more politically exposed Hakeem would become. LTTE’s eastern leader V. Karikalan, muddied the waters by announcing that the LTTE ‘would talk to all parties representing the Muslim interest, not just the SLMC,’ thereby (1) undercutting Hakeem’s position as representative of the Muslim community – thereby weakening his position within the Muslim community and between the Muslim community and the Government since Hakeem is also a Cabinet minister within the ruling UNP government, (2) weakening the SLMC–LTTE Memo of Understanding, and (3) increasing the incentives for other Muslim sub-groups to mobilize support for their factions and claims to speak on behalf of Muslim interests. According to an American academic in the East at the time of the Mutthur, Batticaloa, and Valaichchenai incidents and the aftershocks, there appeared to be a strong feeling that the Muslim response was in part a reaction to escalating LTTE extortion:30 that is, Muslim civilians vs LTTE (over issue of excessive taxation). A second suggestion (which was said to hold ‘considerable credence’) was that the Muslim backlash was driven by factions within the Muslim community who oppose Minister Hakkim’s meeting and subsequent Agreement with LTTE leader Prahbakaran. In this case, this would be an example of intra-group competition (pro-Hakeem’s vs anti-Hakeem’s factions) being fought out in the inter-group arena (Muslim-Tamil) which may have implications for the larger inter-group arena (LTTE–GSL): that is, Pro-Hakeem’s vs anti-Hakeem’s factions Æ to scuttle Hakeem-Prabakaran Agreement. A third suggestion, introduced within the Sinhalese constituency in The Island newspaper, was that the clashes were organized by the LTTE in an effort to scuttle the Peace Talks with the Government scheduled for August – an example of narrow inter-group analysis: that is, LTTE vs Muslims Æ to scuttle LTTE-GSL peace process. What becomes clear is that the dynamics of conflict within groups appears to be tied to those between groups. Intra-Sinhalese political dynamics The elections to Sri Lanka’s 12th Parliament in December 2001 saw the UNP displace the People’s Alliance (PA). The election of Ranil Wickramasinghe as prime minister reflected some significant shifts


within the Sinhalese electorate. Election results showed an upswing in popularity not only for the United National Party but also most remarkably in favour of the radical Marxist JVP, whose popularity increased by 296,579 votes over the course of a year, mostly at the expense of Kumaratunge’s PA.31 Both the JVP and the PA appealed largely to Sinhala Buddhists during the elections, by persistently attempting to stoke nationalist sentiments among them. Electorally then, we see significant intra-group support for the proposed UNP peace initiative; the rejection of the PA’s floundering War for Peace; and, most interestingly, increased support for the harder and more chauvinistic line of the JVP. Many observers in early 2002 felt that discontent would fester in the South if the North was seen to be gaining disproportionately from ‘the peace.’32 The question of the equitable allocation of state resources – as well as donor resources – may serve as a rallying call for the mobilization of dissent. Thus, one of the challenges for both the Government and the donor community is to maintain equity in the allocation of the so-called peace dividend. An interesting balancing act is required between responding to hard needs on the one hand (which are greater in some sectors in the North and the East), and avoiding the politicization of aid by pandering to the ethnicized demands of political entrepreneurs – in the North no less than the South. If the intra-Muslim conflicts are one possible Achilles heel of the peace process, then the other lies in the intra-Sinhalese conflict configured according to the antipathies between the President and the Prime Minister whose personal and political relations are characterized by considerably acrimony. Many have accused the leaders of monkeywrenching each other’s programmes and initiatives out of personal vindictiveness – despite a British brokered agreement in 2000 that specifically committed each party not to undercut the peace efforts of the other out of the usual inter-party rivalry. The threat of Kumaratunge to use her Presidential prerogative to dissolve parliament and call a snap election when it best suits her political self interest was rattled like a sabre in the direction of Wickramasinghe from his election in late 2001. The impact of such a move on the peace process would be devastating.

Conclusion As noted in the introduction to this chapter, this episode differs from the others in the book because it was on-going at the time of writing.


For this reason, the episode offers an opportunity to assess the utility of a two-level analysis as events unfold. The analytical utility of the chapter is not dependent on whether or not the formal cease fire and subsequent peace talks lead in to the long, slow, transition from war to peace. Whether Sri Lanka moves down the road of peace or the road of war, this chapter offers important clues as to where to look for explanation. Even more importantly, it points to intra-group issues to be addressed in an effort to increase the likelihood that the path of peace is chosen. The main point to be underscored in this chapter is that the intergroup dynamics that appear to have enabled the signing of the Ceasefire Agreement (such as financial constraints, and military stalemate) are quite different from the intra-group issues that need to addressed if the CFA is to move towards a sustainable peace process. There are very important policy implications that follow from this observation. Most significant, is the danger that a narrow inter-group focus by international actors seeking to nurture and support the peace process, may not only be ineffective, it may be counter-effective. That is, the flood (or promised flood) of development assistance into reconstruction and development of the North and North East, may serve to sharpen a sense of neglect among volatile sub-groups within the Sinhalese south. Self-interested political mobilizers are always ready to stoke the flames of discontent. Further, a peace agreement between the LTTE and the Government which does not adequately protect the human rights of individuals from all communities in the North and East (i.e. Tamil, Sinhalese and Muslim) sows the seeds of further violent conflict.

Part III Conclusion

10 Fitting the Pieces Together

Summary of critical junctures Each of the preceding chapters above has analysed a critical juncture in the evolution of the conflict in Sri Lanka. The following summarizes the key elements in the cases studied in this chapter and categorizes each episode according to the relative impact of inter-group (Model I) and intra-group (Model II) dynamics: Type 1: A conflict which is essentially inter-group with minimal intra-group influence. Type 2: A conflict in which inter-group factors are dominant, but intra-group factors are notable. Type 3: A conflict in which inter-group and intra-group factors have roughly comparable influence. Type 4: A conflict in which intra-group factors are dominant, but inter-group factors are notable. Type 5: A conflict which is essentially intra-group with minimum inter-group influence. None of the episodes can be identified as wholly inter-group or intragroup in nature. While Type 1 and Type 5 situations may exist, they were not observed within the critical junctures identified in this study. Once one begins considering Model II factors, it becomes apparent that that most situations of ethnic conflict arise from some combination of inter-group and intra-group factors. In all of Sri Lanka episodes considered in this chapter, intra-group dimensions proved to be significant. During the 1983 riots and the insurrection and resurgence of the JVP, intra-group factors proved appeared to be dominant. Even in Type 2 179

180 Conclusion

and Type 3 episodes, the intra-group dimensions need to be considered in order to make sense of events. These chapters demonstrate that the episodes – and the conflict more broadly – can not be understood fully using only a Model I analytical lens. The application of both Model I and Model II analytical lenses enables us to assess the relative importance of inter-group and intra-group factors in an ethnic conflict (see Figure 10.1). Episode I: 1948 Independence and disenfranchisement of plantation Tamils Model I: Disenfranchisement of Plantation Tamils was a means of weakening the parliamentary power of the Tamil community as a whole (23 per cent of the total voting population) especially in the estate areas where Plantation Tamils could challenge Kandyan electoral power. Model II: Sri Lankan Tamil MPs voted in favor of disenfranchisement as a means of protecting and securing intra-group power. Type of episode: Type 2. Remarks: Without the support of other minorities, it would have been difficult, if not impossible, to justify disenfranchisement of the Plantation Tamils. Episode II: 1956 Elections and the premiership of S.W.R.D. Bandaranaike Model I: The electoral mobilization by Bandaranaike politicized ethnic chauvinism and polarized the political system. Model II: Inter-group animosities were mobilized by both Tamil and Sinhalese leaders for use in intra-group, not inter-group, competition. Type of episode: Type 3. Remarks: Actions cannot be explained solely by outcome. It is unclear whether Bandaranaike's chauvinism was a short-term tactic or long-term strategy. The former is more likely. It is necessary to recognize the mutual benefits of communalism for both Tamil and political leaders in intra-group political struggles for control and legitimacy. Episode III: 1971 JVP Insurrection and 1987 Resurgence Model I: The JVP intensified and harnessed the inter-group fears and antiTamil sentiments of Sinhalese sub-groups – even as it maintained covert linkages with Tamil paramilitary organizations. Model II: The success of JVP mobilization on the basis of anti-Tamil, antiIndian chauvinism encouraged other Sinhalese sub-groups to similarly mobilize support, thereby pulling extremism from the political margins to the mainstream. Type of episode: Type 4. Remarks: Model I illustrates that the representation of inter-group relations may be instrumentally modified over time in order to mobilize resources from different sub-groups in the intra-group arena. A Model II analysis locates the principal actors and geography of violence of these episodes within the Sinhalese intra-group arena. Intra-group competition precipitated a spiral of Sinhalese extremism in the intra-group rather than inter-group arena. Figure 10.1

Summary of findings at the inter- and intra-group levels in Sri Lanka

Fitting the Pieces Together 181 Episode IV: 1983 riots Model I: The riots were preceded by a deterioration in inter-group relations. The riots by Sinhalese mobs were directed against Tamils as an undifferentiated group – regardless of social class or economic standing. The UNP government and the Sinhalese media justified Sinhalese violence against Tamils and endorsed an inter-group representation of the episode by blurring the boundaries between Tamil civilians, Tamil separatists, and Tamil ‘terrorists.’ Model II: The riots exhibited patterns of directed violence. Government actors were central in orchestrating the riots, which were not a spontaneous, mass phenomenon. The intensity and dynamics of violence are better understood as a result of: 1) a dispute among factions within the ruling UNP over how to deal with the ‘Tamil Problem;’ and 2) the negative impact of the UNP program of economic liberalization which induced business rivals to harness communal sentiment in order to off-set the economic dislocation of new program. Type of episode: Type 4. Remarks: The UNP government encourages a Model I representation of the riots. The episode enabled it to attack opponents in both the inter-group and intra-group arena. It subsequently passed a constitutional amendment which eliminated the democratically elected Tamil party (TULF) from Parliament. The government also proscribed three political parties, including the JVP, and clamped down on the media. Episode V: The Indo-Sri Lanka Agreement Model I: Pro-Agreement inter-group alliances formed. The Agreement aggravated inter-group violence in the Eastern Province. Model II: Opposition to the Agreement and Pro-Agreement inter-group alliances aggravated intra-group violence. Type of episode: Type 3. Remarks: The Agreement restructured the axes of confrontation within and between groups and added a new axis of military confrontation (LTTE vs IPKF). Episode VI: 2002 Ceasefire Agreement Model I: At an inter-group level, military stalemate, economic constraints on both sides, leadership changes, and the changing international context following 11 September/7 October compelled the principal antagonists to the discussion table Model II: Permissive factors existed within the Sinhalese intra-group arena (such as war fatigue and a disorganized opposition) which allowed the Wickramasinghe regime to take steps towards negotiated peace (pushed further by declining economic conditions). However, within the Tamil community, we see divisions between the LTTE and populations under its control. And most conspicuously, we see serious strains between the LTTE and Muslim leadership, and between Muslim and Tamil communities (exacerbated by intra-Muslim tensions). Type of episode: Type 3. Remarks: Inter-group politics may lead parties to sign agreements, but intragroup politics are critical for sustaining and expanding them. Figure 10.1

Continued

182 Conclusion

The preceding chapters analyse patterns of conflict in Sri Lanka using the two-level critical juncture methodology introduced in Chapter 2. This book demonstrates that a number of mediating factors condition intra-group relations and their impact on inter-group conflict and vice versa (identity mobilization, resource mobilization, opportunity structure, and state actors). Consideration of these factors helps us to identify and better understand the conditions under which relations between groups in violence-prone settings are likely to improve or worsen. The ‘stories’ and histories of conflict in Sri Lanka are thick, nuanced, multi-layered, and often, contested. This study has analysed a number of critical stories within a framework which highlights certain events while self-consciously muting others in an effort to locate intra-group actors more centrally in our understanding of these types of conflict. In this way, analytical continuity and historical structure are generated from a mass of stories that have been told and re-told at different times, in different ways, by different people. A brief review of the ebb and flow of intra- and inter-group relations over time underscores the need to appreciate the variations of, and the connections/disconnections between, different manifestations and types of violence. In the Sri Lankan case, this would include: LTTE–SLA violence; legacies of JVP violence; simmering Plantation Sector violence; election violence and inter-Political Party violence; militarized criminal activity and corruption. It is clear that the patterns of conflict in Sri Lanka – indeed, in all cases of violent conflict – are indelibly shaped by circumstances, particularities, individuals, and sometimes plain dumb luck (be it good or bad). None of the contemporary manifestations of violence in Sri Lanka are sui generis. They follow the trajectory of societal and political developments in the country over time. An understanding of the present requires a consideration of its linkages to the recent and distant past, as well as immediate events – whether these linkages are empirically verifiable or manufactured out of political convenience. Because of the case-focussed methodology employed in this study such generalizations are necessarily ‘middle range’ and ‘contingent’ (George 1982). As in all case-focused studies, there is a tension between those findings which are particular to a case, and those which are nomothetic (that is, those which are generalizing and rule-seeking). This chapter discusses both types of findings. The first section explores some of the factors affecting levels and patterns of violence in Sri Lanka while the second section broadens its scope to consider more

Fitting the Pieces Together 183

generalized propositions concerning the ways in which, and the conditions under which, intra-group and inter-group factors weave together in ethnic conflicts. Given the complexity and evolving nature of ethnic politics in Sri Lanka, it is impossible to write the definitive end or conclusion to the story. Ultimately, this will be written by Sri Lankans themselves, and then re-written by their children and grandchildren. However, the following generalization seems reasonable: while every ethnic conflict possesses its own unique combination of structural and triggering factors, there are some general relationships between intra-group and inter-group factors that appear to hold ceteris paribus – although the relationship is not necessarily monotonic or linear.

Accounting for differences in levels and patterns of violent conflict Search for settlement It is important to bear in mind the distinction between the ‘resolution’ and the ‘management’ of conflict. The former refers to the efforts to resolve or eliminate the underpinning grievances and irritants which sustain a conflict, while the latter applies to efforts to control or ‘deescalate violence’ without necessarily eliminating the root causes of the conflict.1 In some cases, conflict management efforts inhibit the broader conflict resolution process; for example, when ‘anti-terrorist’ legislation succeeds in reducing attacks (at least in the short-term), but does so through the use of tactics which weaken the foundations of democratic governance.2 This is particularly true for cases of ‘protracted social conflicts’3 in which antagonists interpret the world in zero-sum terms. Any perceived gain by one side is viewed as a loss by the other. While a measure may dampen expressions of overt conflict in the short-term, it may also sustain, exacerbate, or entrench conflict in the long-term. Nonetheless, conflict should not seen to be a necessarily negative or destructive phenomenon. In essence, development processes in all countries are inevitably conflictual, destabilizing, and subversive because they challenge established economic, social, or political power structures which inhibit individuals and groups from pursuing their full potential. However, there is a need to maintain a clear distinction between violent and non-violent conflict. While there is a tendency to focus more on violent, rather than non-violent, conflict, it is important

184 Conclusion

to examine those liminal moments at which non-violent conflict ‘turns’ (or ‘re-turns’) violent and is militarized. This study also illustrates how, inter-group violence may be employed instrumentally in non- or limited-violent conflict in the intra-group arena. Typically, the presence or absence of conflict mediating mechanisms and institutions4 are central factors influencing whether a conflict passes the threshold into violence – this might include representative political systems, a transparent and fair judicial system, an equitable social system, and so on. At a national level, an important indirect indicator of the presence of such conflict mediating mechanisms is simply the prevalence of efforts by state and non-state actors to nurture a peace process.5 The presence and intensity of peace initiatives (at all levels) serve as a crude litmus test of the ‘P–C Balance’ (Peace–Conflict balance) of a conflict. Some have argued that violent conflict is the ultimate expression of the breakdown of a society’s systems of governance, and that reconstruction therefore rests primarily upon the renegotiation and refashioning of new systems of governance at the community, sub-national and national level. Thus, ‘reconstruction’ requires strategies and interventions to promote institutional arrangements that can facilitate and sustain the transition from violent conflict to sustainable development. An appealing feature of this kind of approach is the way its analysis of the problem is tied directly to an understanding of the nature of solutions. In the case of Sri Lanka, the search for a political solution has frequently been eclipsed by the explicit promotion of an exclusively military response by government actors – both within the ruling political party and the Sinhalese opposition parties – and most certainly by the LTTE. This unilateral militarized approach may be contrasted with the case of Northern Ireland, where the British government has employed a mix of political and military strategies since the early 1970s (Darby 1990; Rose 1990; Flackes and Elliott 1989, 361–76). In Sri Lanka however, the Government’s military effort evoked an escalating response in kind from Tamil paramilitary organizations. The interaction of government actors and Tamil residents of the North and the East was typically restricted to military or paramilitary contact in one form or another – from surveillance and searches to intimidation, interrogation and ‘collateral damage.’ To the extent that the Sri Lankan Government exclusively pursued an exclusive military option, it abrogated its responsibility as a national representative government to continue the search for political settlement. Yet Premadasa’s tactical


arming and financing of the LTTE highlights the point that unexpected inter-group ‘arrangements’ may be undertaken in an effort to gain advantage in both the inter-group and intra-group arenas. Importantly, such arrangements were still limited to the military realm; that is, they failed to fully examine the scope for equivalent arrangements in the political or diplomatic realms. Shortly after the assassination of President Premadasa, his successor, D.B. Wijetunge, continued to cast the conflict in Sri Lanka to domestic and international audiences as a ‘terrorist problem, not an ethnic conflict’ (Pravada, August/September 1993: 1–2; Sri Lanka Information Monitor, September 1993). For all the differences between the UNP and the SLFP/PA, one point of similarity was President Chandrika Kumaratunge’s continuation of this strategic perspective – as illustrated in her ‘War for Peace’ rhetoric following her election and failed peace initiative in 1994. This particular representation of the conflict is not new. The same sentiments were historically voiced by UNP leaders during their rule from 1977 to 1994 – for example, former Security Minister Athulathmudali,6 former Defence Minister Wijeratne,7 former President Premadasa,8 and former Minister Cyril Mathew. As discussed in the preceding chapter, the interesting exception appears to be Prime Minister Ranil Wickramasinghe following his election in late 2001. Whether or not this approach will serve to finally restore peace to the island remains to be seen. Nonetheless, the pattern since 1981 has been to cast the conflict as a terrorist or military problem within the Sinhalese intra-group arena. The government’s characterization of the conflict may be partially valid insofar as both state and non-state ‘armed stakeholders’ (the LTTE, in particular) employ terror to control local populations and to pursue secessionist or counter-secessionist campaigns – perhaps most evident in the extensive use of suicide bombers. However, to dismiss the identity dimension of the conflict is to ignore a crucial element which contributes to the intransigence and intensity of the conflict. As importantly, this one-dimensional characterization of the conflict is used to justify military or ‘anti-terrorist’ ‘solutions’ at the expense of political solutions. The definition of the Sri Lankan conflict as a ‘terrorist problem’ – not unlike the Bush regime’s ‘War on Terror’ – isolates the effects of conflict from its causes. The basic flaw in the policies which follow from this logic is the assumption that it is possible to apply a military ‘solution’ to a socio-political problem. In the list of ‘lessons spurned’ (as opposed to ‘lessons learned’) this applies to Sri Lanka, no less than

186 Conclusion

the United States in a post-September 11/post-October 7 world.9 The Sri Lanka case illustrates the ways in which the application of inappropriate security policies may be ineffective at best, and may aggravate the conflict, at worst, thereby perpetuating violence and suffering within both LTTE-controlled and Government-controlled areas. The observation that the conflict has tended to be defined in narrow military terms however, only begs the next question of why this is so. As discussed further below, the answer is intimately related to intra-group politics. Even in those instances where intra-group leaders may wish to expand the range of strategic and tactical options at their disposal in the inter-group arena, they must first navigate the course within their own constituencies. Thus, as we have seen time and again in each of the critical junctures in this book, extremist elements have limited the latitude of inter-group actions. The moblization of popular dissent and protest in an effort to undercut the incumbent ruling regime, has been a staple of Sri Lankan politics, and indeed all cases of ethnicized politics. However, it is not the only political dynamic at work, as was illustrated in the final critical juncture, the 2002 Ceasefire. Intra-group politics may also create the space for expansion of strategic and tactical options towards peace, as well as toward escalating war. Identity mobilization and the politicization and manipulation of cultural symbols by state and non-state actors In Sri Lanka, beginning with S.W.R.D. Bandaranaike, the protection and maintenance of Sinhalese cultural identity shifted from the realm of non-state cultural institutions (such as the Sangha, Sinhala medium schools and so on) to the realm of the state. State actors have assumed a direct role in the maintenance of Sinhala Buddhist identity. Two policies are especially prominent: the designation of Sinhala as the official language in 1956,10 and enshrinement of state protection and encouragement of Buddhism in the constitutions of 1972 and 1978.11 The privileging of the language and religion of one ethnic community over another challenges a state’s neutrality and its claim to represent all citizens equally within its territorial borders. In Sri Lanka, the allocation of state resources to ‘protect and foster’ Sinhalese Buddhist culture contributed to the ethnicization of ‘the state’ and fostered the development of an exclusive, rather than inclusive, political process.12 This alienated non-Sinhalese Buddhist sub-groups and exacerbated inter-group polarization which led, in the most extreme manifestation, to the growth of secessionist movements in the North.


The privileging of Sinhalese cultural symbols and institutions was undertaken in tandem with the privileging of Sinhalese material interests – an explosive combination. Moore argues that the ‘use of state power for the benefit of the ordinary Sinhalese has been, and remains, the primary legitimation, implicit or explicit, of all governments elected since 1956 at least, and arguably, since 1931’ (1985: 29).13 The dual focus on Sinhalese cultural symbols and institutions and particularistic allocation of state material resources underscores the fusion of identity and interests which contribute to the intransigence of conflict – though, as discussed in the following section, there are additional factors that make this process especially volatile. The manipulation of cultural symbols and institutions is not restricted to the Sinhalese Buddhist communities by any means. The collateral impact of the politicization of ethnicity by dominant identity group, is that other collectivities are manoeuvred (internally and externally) into the use of the same categories of self-definition and interest articulation. In early 2002, a human rights worker on the east coast of Sri Lanka referred to what he called the nascent ‘Talibanization’ of the East expressed through public dictates by the LTTE which appear to be designed to enforce a conservative form of Tamil nationalism – illustrated, for example, in the dictate that Tamil women wear only saris (not pants) (Bush 2002).14 Similarly, Muslim cultural symbols and institutions have become the corner stone of community and politics in the North East – particularly following the campaigns of ethnic clensing of the early 1990s – as the mosque and Islamic welfare institutions have focused and galavanized a sense of collective identity. Tamil resentment and alienation were initially expressed though parliamentary structures by the Federal Party and TULF and only later through extra-parliamentary forms of violence by paramilitary organizations. However, it is not sufficient to identify the same set of factors as the cause (i.e. alienation) for two different outcomes (i.e. political struggle and paramilitary violence).15 There needs to be something more to explain why the same set of grievances led to the different forms of expression – one violent and the other non-violent. The triggers for violent, rather than parliamentary, opposition did not originate within the Tamil community and sub-groups’ experience of discrimination. Rather, it is the subsequent actions by state actors which delegitimized parliamentary options, closed conventional nonviolent options, and openned the way to large-scale violent opposition. In Sri Lanka, for example, this would include: dubious amendments to

188 Conclusion

the constitution; the role of parts of the government in the violence against Tamil sub-groups in 1981, 1983, and 1987; the belated speeches to the nation by the President and other UNP leaders following the 1983 riots which sympathized more with Sinhalese perpetrators of violence rather than the Tamil victims; the indiscriminate nature of military offensives; the decline of the rule of law; the rampant human rights abuses; the on-again, off-again proscription of political parties; the disappearances, and so on. Segregation Sri Lanka is an increasingly segregated society. Bouts of anti-Tamil violence since the election of the UNP in 1977 have displaced large numbers of Tamils (in addition to earlier riots in 1958). In the early stages, those who were displaced sought refuge in the Tamil-majority districts in the North, but after the riots of 1983, refuge was increasingly sought outside of the country (Refuge 1993; USCR 1994; USCR 1991; IRB 1992). The LTTE version of ethnic cleansing in the Northern and Eastern Province has similarly driven Sinhalese and Muslims out their homes.16 While such displacement has not always been permanent, the overall tendency has been towards the spatial segregation of communities. Communities were forcibly segregated by mob violence in the intergroup arena. The result has been the politicization of space and delineation of ‘ethnic’ territory. This has sharpened the divide between communities (i.e. increased polarization) and has introduced ‘interfaces’ into the political and geographical landscape in both cases. As Feldman explains in the Northern Ireland context, an interface is ‘the topographic–ideological boundary sector that physically and symbolically demarcates ethnic communities … from each other … a spatial construct pre-eminently related to the performance of violence’ (Feldman 1991: 28). Boal and Murray elaborate: ‘Because of the physical segregation of the two groups, most of the inter-group clashes have taken place along the boundaries. … The inner parts of these areas have remained relatively free, a haven for a victim of a second type of conflict manifestation, intimidation’ (1977: 370). Interface violence in Sri Lanka has taken place on the East Coast and contiguous or contested territory of the ‘Border Villages’ in the North and North West. In the ‘majority areas’ (i.e. the Sinhalese majority districts in the southern half of the island) the level of overtly violent inter-group conflict is lower than in the areas on the interface. In Sri Lanka, the government-launched military offensive into the Tamil-majority areas


of the North, has served to maintain rather than challenge geopolitical segregation. In short, segregation has been reinforced by the broader and immediate issues of displacement and on-going inter-group violence. One unpublished study, however, outlines the factors contributing to increased segregation. This includes legal pressures (such as the Citizen Acts of 1948 and 1949 and the Land Reform Acts of 1972 and 1975), political pressure (displacement from the political patronage system), administrative pressures, and social pressures (including destabilization due to violence) (Silva 1986). Thus, for example, throughout the 1990s, and continuing into 2001, all Sri Lankan Tamils in Colombo had to register with the local police station. All landlords were responsible for ensuring that their tenants were registered if they were Tamils. Similarly, any Tamils coming to Colombo – to conduct business, visit relatives and so on – similarly had to register with the police. Such legislation was reinforced by random, as well as mass, security checks. The contribution to the ethnicization and politicization of space, as well as to functional segregation is profound. Sub-group linkages across the inter-group divide In Sri Lanka, despite the presence of a number of exceptional local NGOs,17 the development of grassroots initiatives has tended to be stifled by governmental and paramilitary intimidation. Thus, for example, a government commission to investigate local and international NGO activities in the early 1990s became one way for the government to clamp down on voluntary sector activities. NGO workers were constrained by the looming fear that they might be hauled up before the commission and accused of anti-government activities.18 Although the NGO commission was terminated when Wijetunge assumed the presidency in 1993, the government continued to stifle NGO initiatives or any organization (particularly the media) critical of its policies or practices. In late December 1993, emergency regulations forced all NGOs19 to register all financial receipts and disbursements. The hindrance of linkages across the inter-group divide was perhaps most effectively accomplished by the rigid control of the movement of goods and people into and out of war-affected areas. In explicit violation of the Geneva Conventions, this included the blocking of medical personnel and supplies headed for civilian populations in conflict areas.20 The supply of food items has also been also subject to embargo and manipulation, especially during periods of active operations

190 Conclusion

between the main combatants. One closure in the Vanni, (which began on 26 June 1999), lasted more than six weeks, until the LTTE and the army reached a mutually acceptable arrangement to open the supply line. The functioning of humanitarian organizations was badly hampered and conditions for civilians, especially IDPs, deteriorated sharply throughout July (Human Rights Watch 2000; CHA 1999 & 1998). It is important to note that in addition to the logistical and security impediments which inhibit the entry of people and humanitarian supplies into the war zones, there are also ‘inefficiencies’ and obstacles due to corruption and kick backs all along the supply route taken by the convoys in both cleared and uncleared areas – prompting one humanitarian worker to describe the trip as the ‘voyage of a thousand envelops.’ The restriction of inter-group contact is mirrored by the LTTE. Even following the signing of the ceasefire in February 2002, the LTTE was not allowing local staff of the major International NGOs to leave the Wanni (including OXFAM, CARE, and FORUT) for reasons that were not entirely clear. It was also reported recently that the LTTE were restricting the movement of people under 35 years old in the Wanni21 – a practice which has been maintained consistently through out the conflicts. Through the use of emergency regulations, each ruling regime from 1983 to the 2001, circumvented and thus further weakened the legitimacy of democratic institutions in Sri Lanka. It is becoming clear that any transition from the present to a future political dispensation – if it is to be just, democratic, and peaceful – will have to address an increasingly conspicuous clash of governance structures that reinforce intergroup divisions: governance structures on the Government side characterized by patronage, and on the LTTE side characterized by violence and coercion.22 Nonetheless, over the last few years there has been a significant and positive improvement in the organizational skills and resources among local-level and national organizations and associations in Sri Lanka, resulting in a ‘strong and healthy mutual respect’ between NGOs, community organizations, researchers, donors, and increasingly, some parts of government. This was especially evident and important in the collaboration in the preparation and drafting of the National Framework for Relief, Rehabilitation and Reconciliation in Sri Lanka (‘3R Framework’) from 1999 to 2001. Such developments provide positive counter-pressure on the divisive forces of the competing governance structures and processes noted above.


India A sudden shift in the South Asian geopolitical balance is unlikely. In one way or another, the Indian Government will likely play a role in Sri Lankan politics for reasons of prestige as well as regional and domestic security. It will strive to maintain its status as the regional hegemon. It will continue its efforts to block foreign powers from gaining a foothold in Sri Lanka, thereby protecting its southern naval and air bases from possible attack. And it will continue to limit conflict in neighboring countries from spilling into India. However, the means by which these objectives will be obtained may vary. The India–Sri Lanka Free Trade Agreement signed in 2000 may be a particularly important mechanism for pursuing these goals. While the Indian Government is unlikely to ignore Sri Lanka, it certainly may – and does – shift its support from one sub-group and organization to another within both the Tamil and Sinhalese communities. The Indo-Sri Lanka Agreement was an attempt by the Indian Government to employ both carrots and sticks to get the principal actors to the bargaining table. But the signing of the Agreement and entry of Indian troops did not affect the cost–benefit calculations of the chief actors significantly enough to entice the relevant actors to consider alternatives to war-fighting. When the Tigers found themselves fighting Indian troops, they simply made alternative arrangements for the securing and transporting the resources previously acquired from India. In the end, the Government of India proved unable to deliver on the promises listed in the Agreement and one of the largest armies in the world was unable to defeat the paramilitary force it had supplied and tutored since the early 1980s. The presence or absence of the Indian government, its pursuit of one policy or another, may affect the relative position of one sub-group over another, but thus far it has been unable to limit or control the conflict as a whole. Such control has always been impossible, simply because there are too many variables beyond its immediate control. Ultimately, the set of mechanisms which could regulate and dampen conflict is the product of national and local level rather than regional structures and processes. Thus, they need to be cultivated from below. They cannot be imposed from above. A recognition of the Indo-centric reality of the region has been the systematic way in which the Norwegian Government has kept the Indian Government informed of developments in it covert and overt peace support efforts.

192 Conclusion

State actors and structures: internment in Northern Ireland and arrest operations in Sri Lanka It is useful to compare the Sri Lanka experience with that of Northern Ireland in order to high light salient dimensions of the ‘story.’ In Northern Ireland, the policy of internment without trial contributed substantially to the consolidation of Catholic intra-group support for the IRA from 1971–75. From 1981 to early 2002, a similar practice has been an increasingly conspicuous component of internal security operations in Sri Lanka. Cordon, search and arrest operations are sanctioned by emergency regulations such as the Prevention of Terrorism Act and affect Tamils from all sub-groups throughout the island, including Colombo and plantation areas. The operations in Sri Lanka are the equivalent to internment without trial in Northern Ireland in terms of logic, objectives and consequences. The practice rests on the suspicion of guilt rather than the assumption of innocence. Similarly, in both cases the operations were undertaken to meet surveillance and information-gathering functions. The consequences in Sri Lanka, as in the early stages of the conflict in Northern Ireland, are: the militarization of relations between the government and ‘minority’ sub-groups;23 inter-group polarization; and increased minority Tamil support for violent paramilitary actors. In Sri Lanka, these negative consequences are exacerbated by widespread police extortion of money from detainees in return for their release,24 rampant disappearances, beatings and torture. A number of generalizations may be drawn from these two episodes: (1) Specific security policies may have a profound impact far beyond the security arena and may exacerbate the very violence they are meant to alleviate; (2) they may encourage an inter-group ‘homogenization process’ according to which individuals define themselves in mutually exclusive identity categories; (3) by contributing to the militarization of inter-group relations, it is more difficult to pursue political solutions or develop genuine confidence-building measures; (4) the security policies may increase the sense of insecurity in the minority communities and contribute to the perception by some majority subgroups that the minority community is composed only of terrorists and terrorist sympathizers;25 and (5) to the extent that a security policy is seen to be disproportionately applied to members of one ethnic group rather than another, it diminishes the legitimacy of the government and the neutrality of the security forces, i.e. it reinforces the ethnicization of state actors. In December 1975, the Secretary of State for Northern Ireland released the last of the ‘detainees’ and the policy of internment fell


into disuse – though it has not explicitly been repudiated or removed from the menu of legal policy options. By contrast, in Sri Lanka security force cordon and arrest operations continue unabated up to early 2002. At certain points in time, such operations appear to have escalated. For example, between 1 June and 31 December 1993, according to government statistics, 15,000 arrests were made under the cover of emergency regulations in Colombo alone – almost all of those arrested were Tamils (Amnesty International 1994a).26 In Northern Ireland, internment had a lasting impact on the conflict by virtue of the stimulus it provided to IRA recruitment from 1971 to 1975, as well as its role in the consolidation of a collective sense of victimhood within the Catholic community. While the impact of the policy has been felt long after it ceased to be enforced, it has diminished over time. As long as an equivalent policy is in place and vigorously enforced in Sri Lanka, there is no diminution of impact. While the negative consequences have lessened over time in Northern Ireland, they have not done so in Sri Lanka. Thus, the 1990s phase of the Sri Lanka conflict parallels the early phase of the Northern Ireland conflict in terms of high levels of inter-group conflict and the other characteristics noted above. Moderate voices in the intra-group arena were drowned out by the extremist reaction to the security policies, closing off possible bridges for inter-group communication and moderation. The following difference between Sri Lanka and Northern Ireland is also noteworthy. The B-Specials in Northern Ireland and the Home Guard and other militarised groups in Sri Lanka have both exacerbated conflict and participated in sectarian violence.27 However, in Northern Ireland, the B-Specials were disbanded in 1969 and the security apparatus has been purged of the worst overt links to sectarianism. In Sri Lanka, on the other hand, the sectarianism of state-sponsored or state-condoned actors was still evident up to 2002 (evident in continuing human rights abuses and disappearances). For example, the Home Guards – i.e. the village militias armed by the Government of Sri Lanka ostensibly to protect communities in border areas from LTTE attack have been implicated in instances where they fled from LTTE attacks on their villages, but returned from the jungles to massacre neighbouring Tamil villages. According to the Human Rights Watch World Report 2000: ‘The security forces’ use of home guards and armed ex-militant Tamil groups as auxiliary units to aid in military operations continued to draw criticism from human rights defenders. These groups have engaged in illegal detention, murder, abduction, extortion, assault, torture, forced conscription, and forced eviction.’28 While there are still sectarian incidents in Northern

194 Conclusion

Ireland, they no longer enjoys official state sanction and lack the state channels of expression. Sri Lanka, however, may be entering a new phase, with the signing of the Ceasefire Agreement in February 2002, which involved the demobilization of the pro-government Tamil paramilitary organizations. While the impact of this move on tensions within the Tamil intra-group is not yet evident, it represents a step on the part of the government to distance itself from one source of systematic human rights abuses. As importantly, it forces these groups to shift their energies and resources into their political parties active in parliamentary politics. It is hoped that a similar process will be undertaken within the LTTE, as it is pulled (from outside by the international community and elements within the intergroup arena) and pushed (from within the Tamil intragroup arena) from the bullet to the ballot box.

The theoretical contribution of this study In this study, the critical junctures of conflict in Sri Lanka have been examined through Model I and Model II analytical lenses. The discussion of each episode highlights the impact of identity mobilization, resource mobilization, opportunity structure and the role of state actors on inter-group and intra-group relations.29 While the critical junctures approach needs to be further tested in other cases of conflict, its utility may be briefly assessed as follows: 1. Comprehensiveness: It assesses the relative impact and importance of inter-group (Model I) and intra-group (Model II) factors at specific critical junctures in a conflict, and in the evolution of the conflict, more broadly. 2. Transparency: Generalizations about the conflict are clearly and empirically grounded. 3. Consistency: Although the assessment of particular episodes within a conflict may be questioned or challenged, the overall assessment is a composite of all the episodes. This can ‘balance out’ possible variation in the interpretation of episodes. 4. Transferability: A critical junctures methodology may be applied and tested in other cases of conflict. 5. Specificity: This approach facilitates the comparison of different types of identity conflicts, where battle lines may be drawn along different axes of group identity such as language, caste, religion, tribe, and so on. The use of a common analytical framework allows for the identification of differences as well as similarities in patterns of conflict between the cases.


The principal disadvantage of a critical junctures approach is its time-consuming and labor-intensive character. It requires an intensive examination of cases drawing on a disparate range of issue areas. It also requires an extensive examination of events over considerable period of time. One needs to know the whole history of a conflict to be able to determine what are the critical junctures; one cannot rely on a random sampling procedure. The detail required to explain the impact of the intervening variables may become quite daunting. For example, the examination of the mobilization of identities demands consideration of anthropological complexities which are often neglected by traditional approaches to conflict studies in political science. Overall, however, the critical juncture approach appears robust enough to be applied to other cases. The summary table above categorize episodes according to an assessment of the relative impact of inter-group and intra-group factors on the dynamics of conflict. In keeping with the overall objectives of this study, this format locates intra-group factors more centrally in the understanding of ethnic conflict. This is a necessary corrective in a field in which billiard ball (i.e. inter-group) analyses still predominate. While this approach underscores the importance of intra-group factors in inter-group ethnic conflict, there is a need to also examine and understand when, why and how inter-group and intra-group factors are inter-related. The final section of this chapter responds to the question of whether there are generalizable relationships between intergroup and intra-group factors that might apply beyond the cases in the study. Identity and conflict In some cases, it is somewhat easier to sketch the patterns of violence than to understand the role played by ‘identity’ in these cases. While violent conflicts inevitably produce ‘countables’ (such as fatalities and displaced populations), the connection between identity and conflict is far less tangible and less amenable to quantification. Unfortunately, the adjective, ‘ethnic,’ is often brandished crudely as an explanation, rather than as a description of a conflict. ‘Identity’ does not mobilize individuals and groups to act one way or another as primordialist proponents would argue. On the contrary, the Sri Lanka case illustrates clearly that individuals and groups mobilize and animate identities. That is, identity does not mobilize individuals, individuals mobilize identity – from among a menu of possible identities within limits set by the broader opportunity structure and the specific political, economic and social

196 Conclusion

context within which identity is articulated. Typically, such mobilizers are part of a larger organization, or they may be seeking to create or reinforce a larger organization. At an important level of analysis, the ‘content’ of that mobilized identity (specific beliefs, language, and so on) in ethnic conflicts becomes incidental to the function of that identity. Identifiable markers stake out the boundaries of a group so as to clearly distinguish the in-group from out-groups. This demarcation is not an end in itself, but part of a broader mobilizational process which binds individuals to a particular idea of ‘the group.’ Central to this idea of the group is a sense of security and belonging. Those within that group who are seen to be able to represent or deliver that sense of belonging are thereby able to leverage further material and non-material resources. In effect, identity becomes a special kind of resource to be mobilized like other resources. But unlike material resources, identity needs to be constantly re-presented and validated in order to maintain its political efficacy. Not surprisingly, a particular definition of the group implicitly, and automatically, confers legitimacy, credibility, and status upon those who have done the defining. It is universal that all ethnic mobilizers seek to legitimate their particular representation of the group as continuous, unbroken, and primordial. But the acceptance of these claims of primordialism at face value disregards the inevitability of change in culture and identity over time and in different settings. More importantly, the acceptance of a fundamentally political representation of the past endorses the political objectives its sponsors seek to achieve in the present. Sometimes the political agendas of sub-groups in ethnic conflicts are more obvious than others, for example, that of the Orange Order and the various associations of Ian Paisley in Northern Ireland, or the Buddhist Patriotic organizations and the MEP of S.W.R.D. Bandaranaike in Sri Lanka. However, because identity has become politicized and volatile in ‘ethnics’ conflicts, all sub-groups are unavoidably enveloped in the process of group definition – if not actively, then reactively and contextually. For example, if elements within a community successfully mobilize supporters around an idea of identity which is exclusionary and chauvinist, even non-chauvinist sub-groups from the same community are forced to respond because their participation in the same arena is implicitly compared with that of competing sub-groups. Group identity is constantly being negotiated and nurtured. Its form at any particular point in time is the product of the conjunction and constellation of a variety of factors, including state structures and


societal processes as well as the activities of ethnic mobilizers. Different sub-groups within the same ethnic community may have differing conceptions of the overall group – differences which may be vast or slight, but potentially salient politically, either way. Ethnic conflict (whether intra-group or inter-group) is a product of the same factors which condition the articulation of identities: state and societal structures and processes as well as the strategies and tactics of ethnic mobilizers. However, the connection between identity and conflict is not constant. Each case needs to be analyzed on its own terms. The occurrence of conflict, and the specific location of the battle lines, cannot be definitively determined in advance. Yet, it does not take a crystal ball to predict a conflict that is already underway. Based on the experiences in Sri Lanka, once violent conflict breaks out, there appear to be a number of patterns which are likely to appear. These are addressed in the next section. Intra-group competition and conflict intransigence Intransigence in inter-group ethnic conflict is deepened when subgroups within the intra-group arena are able to effectively portray concessions or accommodation – no matter how small – as a compromise or ‘sell out.’ Any constructive initiatives across the inter-group divide are exploited by competitors to undercut the legitimacy of their opponents in the intra-group arena. There are perceptual and power dynamics which sustain this situation. Sub-groups must continue to perceive the opposing ethnic group as a threat (politically, economically, socially, physically). To this end, ethnic mobilizers actively cultivate the perception of inter-group difference and threat. In Sri Lanka (and in other cases), inter-group animosities have been instrumentally and purposefully manipulated in order to consolidate internal control. Furthermore, the representation of inter-group animosities have been instrumentally modified over time in order to mobilize resources from different sub-groups in the intra-group arena. Relatedly, inter-group ethnic violence can serve as a proxy for other types of disputes (e.g. economic and political disputes). This is illustrated in the orchestration of the 1983 riots by hawks within the Sri Lankan government to push their hard line position to the forefront of government policy. As discussed in Chapter 7, the patterns of looting and destruction and the identities of organizers also point to economic and business rivalries masquerading as ethnic conflict. This episode points to the political salience of contingent identity in ethnic conflict. Ethnic, socioeconomic, and ideological axes of identity were all

198 Conclusion

evident in the battle lines of violence over the course of mob violence. The government’s efforts to highlight only the ethnic dimensions (and thus, obscure the other identity dimensions) underscores the contingency and contextuality of identity. Some axes of identity are more or less politically salient than others in particular circumstances. As explained in Chapter One, these identities are not manufactured out of thin air. Existing ‘packages of attributes’ are harnessed for political purposes. As discussed further below, intransigence and conflict are promoted by a fluid intra-group arena in which no single group or coalition is able to definitively exercise control due to a lack of sufficient resources, in particular the lack of sufficient intra-group support. The linkage between the intra-group and inter-group arenas are evident: the intransigence of inter-group conflict is sustained by competition among subgroups for control of the intra-group arena. Intra-group competition and conflict escalation The case studies have pointed to instances where inter-group antagonisms were harnessed by ethnic entrepreneurs in an effort to advance their position in the intra-group arena.30 There is an inherent danger that other intra-group competitors may also attempt to rally support on the same parochial basis, thereby setting in motion a dynamic of competitive bidding in ethnic chauvinism. If intra-group opponents appear to be gaining the political upper hand by inflaming inter-group tensions, then state and non-state actors are pushed into adopting an increasingly intransigent position in the inter-group arena. Once a political entrepreneur successfully mobilizes support on the basis of a particular facet of group identity, other political entrepreneurs are faced with the choice between three options: (1) pandering to the same particularisms and parochialisms activated by their intra-group opponents; (2) innovatively mobilizing support on the basis of a competing axis of identity or interests; or (3) doing nothing and thereby risk political isolation and atrophy through the loss of resources. Under these circumstances there is a danger that intra-group competition might precipitate a spiralling dynamic of intra-ethnic group competition and draw violent intergroup extremism from the margins to the mainstream of the political agenda. Intra-group consolidation due to inter-group polarization Dramatic polarizing events such as the 1983 riots in Sri Lanka cleave relations in the inter-group arena. During these period, intra-group


politics become the engine of inter-group relations. The prospects for a reduction of tension under these conditions rest heavily in the intragroup arena. Inter-group polarization provides ethnic mobilizers with the opportunity to pander to fears of other ethnic groups as a means of mobilizing internal support. Such entrepreneurs have an interest in maintaining or even aggravating the polarization between groups as long as it serves their intra-group interests. Inter-group status quo and ethnic conflict intransigence Inter-group intransigence may also be subsidized by the development of a status quo between the principal ethnic antagonists. This may result from the construction and institutionalization of coping mechanisms. However, the conflict itself may become institutionalized, for example, through the modification of society and economy to suit the exigencies of changed circumstances which gradually becomes the ‘norm.’ While some combatant groups atrophy or are displaced, others become entrenched in the intra-group arena and acquire direct and indirect benefits from continued inter-group conflict.31 These conditions – which are a result of inter-group conflict – institutionalize political and economic intra-group structures to the benefit of the principal combatants. The incentives for these sub-groups to consider intergroup accommodation must be higher than what is gained through continued conflict. In the absence of special arrangements to alter the incentive structure to ‘make it worth the while’ of combatants to choose the ballot box over the T-56 Assault Rifle, then change is unlikely. It should be stressed that this contributes to the intransigence of the conflict, not necessarily its escalation. To the extent that the status quo is based on the implicit understanding by the principal antagonists that neither side is likely to make decisive gains, there is likely to be a scaling down of violent interaction to a ‘sustainable’ level, that is, a level at which resources are not wasted or used counter-productively, and future options are not foreclosed. As a result, militarized conflict in parts of Sri Lanka settled into a pattern of ‘lingering stalemate.’ It is under these conditions that we find the otherwise bizarre cooperative arrangements between combatants on opposite sides of the ethnic divide, such as those between the late President Premadasa and the LTTE. Ethnicization of the state The term ‘state’ has been used throughout this study to refer to a set of administrative and coercive institutions headed by an executive

200 Conclusion

authority. It highlights those elements which enable these institutions to play a significant role in civil society: coercion, institutional capacity, leadership, authority and some element of territorial control. ‘Ethnicization’ of the state is the process by which the particular interests of one set of ethnic sub-groups come to be unfairly promoted by state actors at the expense of other ethnic sub-groups. In Sri Lanka, state structures and processes have been ethnicized at different points in time. It would be both parsimonious and satisfying if it could be established that the ethnicization of the state triggers an increase in conflict, while de-ethnicization has the opposite effect. However, the case in this study does not sustain this generalization. In terms of the levels of inter-group violence, the impact of the ethnicization of the state is variable. As discussed above, in Sri Lanka S.W.R.D. Bandaranaike set out to ethnicize the state as a means of mobilizing political support. Coincident with this process in Sri Lanka was the intensification of inter-group conflict – which became increasingly violent in the late 1970s. In other cases, such as Northern Ireland, the state was similarly ethnicized during the Stormont regime from 1922 to the rise of Terence O’Neill in 1963. Yet, changes to the levels of violent inter-group conflict was minimal. This suggests that ethnicization of the state, on its own, is not a sufficient trigger for violent ethnic conflict. A closer examination of these two episodes uncovers an additional factor which helps to explain why ethnicization of state actors, sometimes does, and sometimes does not, lead to overt inter-group conflict: the constellation of actors involved and, specifically, the balance of power within both the intra-group and inter-group arenas. The rise (and fall) of Bandaranaike is characterized by a frantic scramble for intra-group hegemony. The leaders of ruling UNP, with the belated exception of J.R. Jayawardene, did not realize that Bandaranaike was a serious threat to the established order of the postcolonial ruling elite until his election victory in 1956 when the dominance of the UNP in the Sinhalese political arena was broken by the SLFP. Bandaranaike’s victory upset the existing Sinhalese intra-group balance of power and precipitated an internal competition for control among sub-groups – many of which self-servingly adopted similar chauvinist strategies. This was the context in which the Sinhala Only language policy sparked mob violence against Tamils throughout the island, and set the post-colonial precedent for future episodes. The subsequent inter-group violence was less the direct result of the ‘ethnicization’ of state policies, than the competition among Sinhalese


sub-groups using the opportunity to win intra-group control when the rules of the game seemed to be up for grabs. The study suggests that inter-group violence is not generated by state ethnicization per se, but the transition to or from an ethnicized set of state-society relationships. During these periods, relations between and within groups are characterized by fragmentation and fluidity. In the intra-group arena, the rules of the game are in flux and sub-group mobilizers see themselves in competition with each other for control of the intra-group arena, while at the same time, the established modus operandi guiding relations at the inter-group level is also in flux. During such periods of crisis, politics becomes a struggle over the basic rules of the game rather than allocation within an established set of rules. As a result, there is a sense of acute insecurity both within ethnic communities, as sub-groups jostle for position and control of the intra-group arena, and between ethnic communities, as established relations are ‘renegotiated’ (for example, dominant and subordinate relations). However, once the crisis passes, the arrangements tend to rigidify and moderate behavior between state and societal actors.

Conclusion Inter-group relations should be seen as resting precariously on a shifting constellation of intra-group relationships. Thirty years ago, Enloe suggested one important implication of this: ‘the width of the gap between intra-communal status groups affects the community’s political capacity to deal with outside pressures’ (Enloe 1973: 18). On closer observation, one quickly realizes that whole ethnic communities do not interact monolithically with each other. Rather, sub-groups bond or battle across inter-ethnic boundaries. In other words, not only do subgroups compete against each other in the intra-group arena, but they often form alliances with ‘opposing’ sub-groups across the inter-ethnic divide as well. The case of Sri Lanka illustrates a dynamic analogous to what Daniel Druckman has labelled the ‘boundary-role conflict’ (BRC) in the bargaining and negotiation process: in this ‘the negotiator is obliged to be responsive to the competing claims of his own and the other side. This dual responsiveness may be regarded as the defining feature of negotiation … ’ (Druckman 1978: 89). According to Druckman, the negotiator plays two roles: bargainer and representative. As a bargainer, a negotiator interacts with, and reacts to, opposing negotiators. As a representative, the negotiator must weigh and

202 Conclusion

balance the competing interests within the group he represents. As both bargainer and representative, the negotiator ‘attempts to build a package that will be acceptable to both the other side and his bureaucracy’ (ibid.). In cases of ethnic conflict there is an analogous dynamic which conditions political entrepreneurs and mobilizers in the intra-group arena. The potentially wide spectrum of heterogeneity within ethnic groups (e.g. variance according to political allegiance, caste, regionalisms, ideology, class, religious practice, dialect, kinship, and so on) presents a formidable challenge to the mobilizer who must be responsive to these differences in order to construct a support base. However, in responding to the different demands and expectations of these sub-groups, the mobilizer may begin to lose latitude of action in the inter-group arena. This dynamic may either inhibit or encourage the tactical use of ethnic extremism by the mobilizer. When extremist tactics do not resonate (or lose their resonance) with the ethnic constituency (for example, due to changes in balance of power, ‘war weariness,’ hurting stalemate, collective learning or ‘re-perceptualization’)32 mobilizers find that the continued advocacy or use of such measures will undercut their basis of support – most notably reflected in a reduced ability to mobilize resources (broadly defined). Conversely, mobilizers who counsel moderation or accommodation risk undercutting their support base if the potential constituency prefers a hard-line extremist stance. In this case, moderate mobilizers are forced either to: (1) stick to their initial stance and risk sliding into political obscurity (perhaps until a conducive constellation of conditions resurrects the message, if not the messenger) while other groups articulate an extremist program of action; or (2) shift their stance to meet – and harness – the extremist demands of the constituency and thereby maintain a mobilizational strategy that increases the likelihood of strengthening their support base. The rise of S.W.R.D. Bandaranaike and his assassination highlight the dangers and volatility of mobilizing extremist sentiments. Having mobilized communalist passions in order to get elected, he found himself in the awkward position of having to dampen escalating antiTamil sentiment once he was elected prime minister. His politicization of ethnic institutions and symbols had catapulted him to power, but they also came to be harnessed by his competitors in the intra-group arena and were turned against him. In the end, he was assassinated by a monk who had been an ardent supporter during the 1956 election campaign because Bandaranaike was perceived to be too conciliatory towards Tamil political demands.


This book has drawn more lessons for what not to do, than for what to do. But even the implementation of this set of general proscriptions would likely lead to a consequential improvement in the immediate living conditions of thousands of Sri Lankans. Further, the mere insertion of a self-consciousness about the variability of the linkages between inter- and intra-group relations, and their consequent impact on the dynamics of peace or conflict, should have a profound impact on the way we think about, and more importantly, in the way we work in areas affected by ethnicized conflict. Most important of all, a twolevel focus may allow us to anticipate and defuse tensions which facilitate the escalation and transformation of non-violent conflicts to violent ones. The fundamental implication of this book is that efforts to manage and ultimately resolve ethnic conflicts require that attention be explicitly and self-consciously directed toward the internal relations of ethnic groups as much as to inter-group relations. The intransigence and apparent insolubility of ‘ethnic conflict’ may well be subsidized by our analytical emphasis on the inter-group dimension and relative neglect of the intra-group dimension. Ethnic groups may indeed be polarized, but the reason may be related to internal factors more than to the action–reaction dynamic at the inter-group level. The incorporation of the intra-group dimension into our examination of inter-group ethnic conflicts pushes us in new directions for conflict management and resolution. ‘What is required is a form of intra-ethnic competition which is not wholly centrifugal, so that ethnic parties can find rewards in taking moderate positions on matters of inter-ethnic relations, hence making themselves available for coalitions of commitment’ (Horowitz 1985: 379). Even in the midst of protracted and intense ethnic violence ‘inter-action’ takes place for better or for worse (i.e. for socially constructive or destructive purposes). While overt and covert tactical inter-group alliances are not uncommon in ethnic conflicts, the challenge for those who seek peace is to defuse the socially destructive linkages while fostering the constructive ones. This task has yet to be systematically undertaken. The need to do so has never been greater.

Notes Preface 1. CHA Newsletter, Vol 6 Issue 6, December/ November 2001, p. 21. 2. It should be noted that in 2000 and 2001, that he Ministry of Defence backtracked and refused to allow the supply of electricity into uncleared (i.e., rebel held) areas in Batticaloa – for which Rs 10.0 million was already allocated two years previously under the village development programme. CHA Newsletter, Vol 5 Issue 5, September/October 2001, p. 24. 3. This episode is dealt with in greater detail in the text. Similar in bizarre-ness, though not significance, was the local level truce negotiated by the Red Cross in the jungles of Amparai District in June 2000, to allow a Government veterinarian into rebel-held territory to tend to an injured wild elephant. Dilip Ganguly, Associated Press, June 14, 2000.

1 Beyond Billiard Ball Analysis 1. This introduction is the reflection of an on-going discussion with Fuat Keyman of Bilkent University in Ankara, Turkey. It draws from Bush and Keyman (1997). 2. It is important to distinguish between conflict ‘resolution’ and ‘management.’ The former refers to the efforts to resolve or eliminate the underpinning grievances and irritants which sustain a conflict, while the latter applies to efforts to control or ‘de-escalate violence’ without necessarily eliminating the root causes of the conflict. Some analysts employ the term ‘conflict settlement’ to ‘indicate the formal ending of armed hostilities and the renunciation of the use of force. They believe the objective of conflict resolution to be unattainable, on the grounds that conflicts over fundamental values and needs will never be resolved to the complete satisfaction of the parties involved. In these circumstances, the best one can hope for is a settlement which ends the violence and in some small measure takes those conflicting values and priorities into account’ (Fen Osler Hampson and Brian Mandell, ‘Managing Regional Conflict: Security Cooperation and Third Party Mediators,’ International Journal, XLV, no. 2 (Spring 1990: 193). In some cases, conflict management efforts may inhibit the broader conflict resolution process. This is particularly true for cases of ‘protracted social conflict’ in which antagonists interpret the world in zero-sum terms. Any perceived gain by one side is viewed as a loss by the other. While a measure may dampen overt conflict in the short-term, it may also sustain or exacerbate conflict in the long-term. 3. For example, Buzan, People, States and Fear; r Klare and Thomas, World Security; and Seymon Brown, ‘World Interests and the Changing Dimensions of Security,’ in Klare and Thomas, World Security, pp. 10–26. 204

Notes 205 4. For example, Stephen Walt, ‘The Renaissance in Security Studies,’ International Studies Quarterly, 35 (1991): 211–39; and Kolodziej, ‘Renaissance in Security Studies?’. 5. For another attempt in this context, see Ole Waever, ‘Identity, Integration and Security,’ Journal of International Affairs, 48, no. 2 (1995): 389–431. 6. Jack Snyder, ‘The New Nationalism: Realist Interpretations and Beyond,’ in Richard Rosecrance and Arthur Stein, eds, The Domestic Bases of Grand Strategy (Ithaca and London: Cornell University Press, 1993), p. 180. 7. Ibid., p. 181. 8. Edward Azar, ‘Protracted Social Conflicts: Ten Propositions’, in Azar and John W. Burton, eds, International Conflict Resolution: Theory and Practice (Sussex: Wheatsheaf, 1986), pp. 28–41.

2 Learning to Read between the Lines 1. For example, the 1987 government offensive against the Tamil north of the island was preceded by prisoner swaps between government forces and the main Tamil paramilitary organization, the (LTTE, as well as by government offers to release hundreds of Tamils interned under the Prevention of Terrorism Act, in an effort to facilitate direct negotiations. A more recent example is discussed in Chapter 9 the 1994/95 peace initiative of Chandrika Kumaratunge which led to a formal ceasefire with the the LTTE. However, the high expectations and optimism of the initiative were ultimately dashed when a re-armed and reinvigorated LTTE broke the ceasefire agreement by attacking and destroying two naval craft and two SLAF Planes in April thereby initiating what came to be known as Eelam War III. The government’s response was equal in ferocity and destruction, described by Kumaratunge as ‘a war for peace.’ As discussed in greater detail in the text, both examples illustrate the ways in which conflict escalation or resistance to peacemaking is a function of intra-group politics within the Sinhalese and the Tamil communities. 2. Newcomers to Sri Lankan politics sometimes find it surprising that despite the escalating violence, which characterizes the Sinhala-Tamil ethnic divide, an array of Tamil paramilitaries have now allied themselves with the Sinhalese-dominated government – the same paramilitary groups which previously had been warring with the Sri Lankan security forces. The point made here is that the funnelling of resources across the inter-ethnic divide to sub-groups within the Tamil community has had a profound effect in the Tamil intra-group arena as these paramilitaries jostle to build a support base and increase their influence. Conspicuously, this process has included the terrorizing and brutalizing of other Tamils and Tamil-speaking Muslims by Tamil sub-groups. Personal interviews, Sri Lanka, 2002, 2001, 2000, 1999/ May-July 1992. See also Bush (1993), Hoole et al. (1992), and most publications by UTHR (Jaffna). 3. The perennially problematic issue of causality in the social sciences is increasingly being challenged and recast. For example, in a unique study of reforming an irrigation system in Sri Lanka, Norman Uphoff draws on chaos theory which ‘explores different kinds of order and causation which

206 Notes

4. 5. 6.

7.

8.

9.

10. 11. 12.

13.

14. 15.

16. 17.

are nonlinear and only loosely determinant, finding surprising patterns in the dynamics of open systems that match human realities better than the closed system reasoning of classical physics’ (1992: 14). See Chapter 6, Critical Juncture III: 1971 JVP Insurrection and 1987 JVP Resurgence. For example, the confrontation between the Liberation Tigers of Tamil Eelam and the Eelam Peoples’ Revolutionary Liberation Front in mid-1990. The distinctions between and within these groups will be examined in Chapter 3. At this stage it is sufficient to note that the Sinhalese are predominantly Buddhist and Sinhala-speaking; they constitute 74 per cent of the total population. The Tamils are predominantly Hindu and Tamilspeaking; they constitute 18 per cent of the total population. All demographic data is based on the most recent (1981) census. In Sri Lanka, this would include the strong inter-communal ties among English-educated elites, particularly in the early post-colonial period. In the arena of violence it would include the episode discussed later in this book when former President Premadasa supplied arms, money and materiel to the Liberation Tigers of Tamil Eelam (see Chapter 8 on the Indo-Sri Lanka Agreement of 1987). For such chronologies of events in Sri Lanka, see INFORM Sri Lanka Information Monitor (Colombo), and Anton Philip (1991), ‘Sri Lanka: Chronology of Events’ produced by the Sri Lanka Resource Centre in Oslo, Norway. Sources of information are similarly varied and include: newspapers; archival sources; government reports; chronologies of events; human rights reports; reports of fact-finding missions; and interviews. In this study, as in Allison’s essay, the use of the term ‘model’ without qualifiers should be read ‘conceptual scheme.’ Refer to summary tables at the end of Chapter 9. The identification of ‘mediating’ factors of ethnic conflict may prompt the question: what is being mediated? In general terms, these factors mediate and condition the social, economic and political relations between and within ethnic groups. For two useful discussions of the primordial-instrumentalist debate, see James Mckay, ‘An Exploratory Synthesis of Primordial and Mobilizationist Approaches to Ethnic Phenomena,’ Ethnic and Racial Studies, 5, no. 4, (1982): 395–420; and George M. Scott, ‘A Resynthesis of the Primordial and Circumstantial Approaches to Ethnic Group Solidarity: Towards an Explanatory Model,’ Ethnic and Racial Studies, 13, no. 3 (1990): 147–71. Donald Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985), p. 66. As suggested in extreme instrumentalist arguments, for example, Eugeen E. Roosens, Creating Ethnicity: The Process of Ethnogenesis (California: Sage Publications, 1989). Donald Rothschild, Ethnonationalism (New York: Columbia University Press, 1981), p. 8. This was the case in India when the state boundaries were redrawn in the 1950s (Horowitz 1985: 66). Lebow asserts that this was also the case in Northern Ireland (1974: 208–9) and, as discussed further below, Sri Lanka,

Notes 207

18. 19.

20.

21.

22.

23.

24.

too, reflects this shift in the politically salient axis of identity (Nissan and Stirrat 1990; Spencer 1990). Personal interviews, Colombo, Kandy, Batticaloa, May–July 1992; 1998; 2000). See also ‘Scott’ (1984). Uphoff (1990), building on Uphoff and Ilchman (1969) and (1972), provides a useful basis for conceptualizing and operationalizing political resources. He develops the following resource categories: economic resources, social status, information, force, legitimacy, and authority. The utility of this approach is that it incorporates both material and non-material resources within a single analytical (resource-exchange) framework. Regarding the motivation for ethnic mobilization, I agree with Esman that the most likely cause of ethnic mobilization is a serious and manifest threat to the vital interests or established expectations of an ethnic community, its political position, its cultural rights, its livelihood or neighborhood. However, as addressed below, while defence against threat may stimulate the mobilization of collective identity, the mobilization of collective action will be affected by, among other things, the opportunities and limitations arising from the political and economic environment (Esman 1990: 54). Like Tilly (1978), my use of ‘organization’ does not necessarily imply ‘formal organization.’ The more extensive the common identity and internal networks of the group, the more organized it is. ‘Mobilization processes do not have to start from scratch. They can build on pre-existing networks of informal relations as well as on preexisting networks of formal organizations, political and otherwise’ (Kriesi 1988: 362). By ‘institution’ I refer to those implicit or explicit principles, norms, rules and decision-making procedures around which actors’ expectations converge in a given area (Krasner 1983). It is important to note that while such institutions channel the expression of collective sentiments and the mobilization of resources, the institutions themselves are neither unchanging nor static. Institutions can be modified over time and used in ways that they were not originally intended. However, although institutions may be malleable, they are not completely fluid. They possess a ‘stickiness’ that can help or hinder intergroup accommodation (Krasner 1988). For details on the arms transfers, see A. K. Menon. ‘The Other Battle Field,’ India Today, 15 October 1991: 99; The Hindu (Madras), 6 September 1991: 1; Frontline (Madras), reprinted in Christian Worker (Colombo), 2nd & 3rd Quarter 1991: xvi. These same supplies are now being used by the LTTE against the Sri Lankan Army, East Coast Muslims, Tamil civilians, pro-government Tamil paramilitaries, and Sinhalese settlers in the North. For other instances of Sri Lankan Government assistance to the LTTE, see Gunaratna (1993: esp. pp. 357, 359, 363). Socially constructive examples of ‘cross-border excursions’ may also be found: the mutually supportive inter-ethnic relationships that developed as a byproduct of the Gal Oya Water Management Project (Uphoff 1992: 119–21); pilgrimage sites shared by Hindus and Buddhists, such as Kataragama and Siripada; even cultural symbols have been shared including (somewhat incongruously) the late M. G. Ramachandran, the South Indian film starturned-politician who was a ‘cultural hero’ not only among Tamils but also among the urban proletarian Sinhalese as well (Uyangoda 1989: 37–43).

208 Notes 25. When the notoriety of a particularly brutal Loyalist killer became a embarrassment and liability to Loyalist paramilitaries, the (Loyalist) UDA and UVF are reported to have facilitated an assassination operation by the Provisional IRA (Feldman 1991: 59–65; 289). 26. The use of ‘India’ as a short-hand designation should not obscure the heterogeneity of Indian actors. Particularly important is the tension between the Central government in Delhi and the State government of Tamil Nadu in Madras (a state with 55 million Tamil speakers) which has led to different policies toward Sri Lanka at various points in time. 27. Early work includes: Laitin 1985, 1986; Brown 1989; Brass 1984, 1990; Horowitz 1985, 1991. 28. Susan Strange’s The Retreat of the State is an effort to push the theoretical pendulum back in the other direction. 29. It should be noted, however, that the state may not even be the dominant authority within its own borders, particularly when considering that ‘the modern state’ is a relatively recent innovation in many Third World countries compared with other authority structures. As Migdal (1987, 1988) points out, the ‘real’ politics of Third World countries lies in the struggle between the state and societal authority structures. While the state may have more material resources at its disposal, it does not necessarily have the nonmaterial resources (such as ideological support or legitimacy) needed for maintaining its authority and power. This points to the need to incorporate both state and societal structures and processes into the examination of ethnic groups in conflict.

3 An Overview of Sri Lanka 1. Although Ceylon was renamed Sri Lanka in 1972, this study will use ‘Sri Lanka’ to refer to the country even in the pre-1972 period. Additionally, although the government changed the spelling of the country’s name to Shri Lanka in November 1991, this study uses the former version, which remains the convention in Western publications. 2. The Physical Quality of Life Index (PQLI) combines measures of life expectancy, infant mortality and literacy into a single index with a maximum value of 100. In the 1970s, Sri Lanka scored 82 despite a per capita income of $US179 (Lal Jayawardena in Tambiah 1992: ix–x). 3. Medicins Sans Frontier, reported in The Globe and Mail, 23 November 1992: A12. The human rights violations which accompany displacement are extensive. As a recent report puts it: ‘These people, whom the international legal jargon has reduced to the acronym IDP, do not enjoy the same rights as their fellow citizens. At best they are patronized, mostly ignored and left to their own mercy, at worse harassed, arrested, abducted, raped, tortured, executed, exploited, imprisoned’. Indeed, every single right spelt out in the UN Guiding Principles on Internal Displacement is violated in Sri Lanka. CPA/CHA/LST 2001. 4. The government has invoked the Prevention of Terrorism Act (patterned on the South African model) and emergency legislation, both of which violate Sri Lanka’s obligations as a signatory to the International Covenant on Civil

Notes 209

5.

6.

7. 8. 9.

10.

and Political Rights and the International Covenant on Economic, Social and Cultural Rights. Kelegama (1999) offers the following estimate for 1996: Army – 129,000, Air Force – 17,000, Navy – 21,000, and Police – 68,000 for a total armed forces strength of 235,000. See also Matthews 1989: 437–39. The estimate of deserters is based on newspaper reporters and interviews in Sri Lanka in 1999. Sri Lanka’s Directorate of Military Intelligence estimates that 60 per cent of the LTTE fighters are below the age of 18. Another assessment of LTTE fighters killed in combat put the figure at 40 per cent, both male and female. (Rohan Gunaratna, ‘Childhood – a Continuous Casualty of the Conflict in Sri Lanka,’ Jane’s Intelligence Review, July 1998.) If we accept the estimate from Jane’s Intelligence sources that the LTTE fighting force in 1997 was 13,200, then the number of active child soldiers on that side would be between 5,280 and 7,920. That the practice of (ab)using children as soldiers is continuing, was evident to those organizations and individuals working in the field, particularly in the LTTE controlled areas – though it is not uncommon to see Government armed Home Guards in border villages who are clearly under the age of eighteen. And, although not discussed publicly by those involved in the LTTE Sri Laukan Army–(SLA) body exchanges of soldiers killed in combat, the volume and size of the contents of the body bags add evidence the continued (ab)use of child soldiers. UNDP, Mine Action Pilot Project Jaffna, 1998. p.1 Cited in David Dunham and Sisira Jayasuriya, ‘Is All So Well with the Economy and with the Rural Poor?,’ Pravada, vol. 5, no 10&11 (1998) 24. Nisha Arunatilake, Sisira Jayasuriya, and Saman Kelegama, The Economic Cost of the War in Sri Lanka, Research Studies: Macroeconomic Policy and Planning Series No. 13 (Colombo: Institute of Policy Studies, January 2000) Unlike any other guerrilla or paramilitary organization, it has a large, and seemingly limitless, supply of suicide bombers and operational suicide squads. Suicide bombers are idolized as ‘martyrs’ in gaudy public memorials found in public spaces and training camps within LTTE-controlled territory. All LTTE fighters wear an amulet of cyanide around their necks, which they vow to swallow rather then be captured. It appears that in the LTTEcontrolled North, the culture of militarism and martyrdom feeds the labour needs of military machine which are replenished with the help of atrocious living conditions characterized by systemic human rights abuses by government security forces, a future characterized by hopelessness, and a noble escape through death for the Eelam cause (see Trawick 1999). A recruitment officer for the LTTE insisted that they were not pressuring youngsters to join the ranks of fighters: ‘It is not the Tigers who are recruiting young people, it is the government who are driving the children to join the Tigers’ (BBC 1998). In light of the the practice of the LTTE to use child combatants (including suicide squads), concerns have been raised about the use of these orphanages as a mechanism to fill the ranks of the so-called LTTE ‘Baby Brigades’ (Gunaratna 1998a). Even more numerous than the orphans created by the war, are the children who have lost one parent, rather than both, to the war. It was recently estimated there are 19,000 ‘war widows’ in Jaffna, of which 9,000 are below the age 35 years (CHA 2000). Both orphans

210 Notes

11.

12. 13.

14.

15. 16.

17.

18. 19.

and single-parent children are especially vulnerable in such precarious situations. The extent of the violence was variously presented. Amnesty International reported that 1,000 people a month were being slain in JVP violence in the latter part of 1989 (New York Times (NYT), T 14 December 1989: A 18). Sanjoy Hazarika of the reported that ‘at least thirty people die every day because of the civil war’ (‘In Sri Lanka, the Dainty and the Dead,’ NYT, T 5 September 1989: A 11). In August 1989, diplomatic sources estimated that 30–40 people were killed every day in Sri Lanka. More than 1,000 members and supporters of Premadasa’s ruling UNP alone are estimated to have been killed since Premadasa’s election in December 1988 (Far Eastern Economic Review (FEER), 17 August 1989: 17). By November 1989, James Clad reported that ‘in recent months a couple of hundred murders take place every week’ (FEER, 9 November 1990: 37). He later revised his estimate to ‘300-400 bodies a week’ (FEER, 16 November 1990: 59). The Wall Street Journal reports that 15,000 have died in the last six years (30 October 1989: A 6). Ninety-four political killings were recorded on the day of presidential elections alone (December 1988) and 417 killings within the 13 days after the election (Sunday Times (London), 15 January 1989: A 3). That is, vigilante squads often containing out-of-uniform security forces. As discussed further below, ‘legitimate targets’ for attack or ‘disciplinary action’ went beyond security personnel and public officials. At times, it included families of security forces and all levels of public employees (such as post men and bus drivers). It came to include any non-supporter of the JVP cause and anti-social elements (such as neighborhood drunks and petty criminals), somewhat similar to the purges employed by the Khmer Rouge in Cambodia. This disintegration has included the incorporation of massacres into the military repertoire of all combatant groups. For details on recent Army massacres, see INFORM, Sri Lanka Information Monitor Situation Report (September 1992: 14; February 1993: 14), US Committee for Refugees (1991: 16, 22), and British Refugee Council, Sri Lanka Monitor (March 1992: 3; January 1992: 1). For details on Tamil paramilitary massacres (i.e. the Liberation Tigers of Tamil Eelam), see University Teachers for Human Rights – Jaffna (1991), Amnesty International (1991); Amnesty International (1993), Pieris and Marecek (1992), and Abeyesekera (1992). This is evident in the work of Robert Kearney, Mick Moore, Gananath Obeyesekere, A. J. Wilson and K. M. de Silva. Spencer (1990) contains a selection of scholars who challenge the rigidly bipolar model I interpretations of Sri Lankan history. See also Kemper (1991) and Rogers (1987), and (1987a). Also called ‘Indian Tamils’, ‘Hill Country Tamils’ or ‘Estate’ Tamils. Some of the less rigorous discussions of Sri Lanka, categorize Up Country Tamils with Sri Lankan Tamil. Far Eastern Economic Review, 21 February 1985: 36. The LTTE is the largest and most powerful Tamil paramilitary organization. It has controlled significant portions of territory in the north and east of the island. The LTTE rose to power due to the organizational and military abilities of its leader Velupillai Prabakaran and its efficiency in murdering

Notes 211

20.

21.

22. 23.

opponents (individuals and other Tamil paramilitaries). The LTTE is extremely well armed and manufactures a number of its own weapons and munitions, such as mortars and rounds, rifle grenades, rocket-propelled grenade launchers, and improvised torpedoes. For discussions on the rise of the LTTE, see Hellman-Rajaanayagam 1994; Gunaratna (1990a); and Hoole et al. (1992). For details on weapons, see Gunaratna (1990a: 45–57). The town currently designated ‘Ampare’ or ‘Ampara’ was previously spelled ‘Amparai.’ The change from the Tamil to a Sinhala spelling reflected an influx on Sinhalese settlers in and around the town. According to the 1921 census for the district, the demographic composition was 8.2 per cent Sinhalese, 37.7 per cent Tamil, and 54 per cent Muslim (Kemper 1991: 145). This compares with the 1981 census which found 37.5 per cent Sinhalese, 20.1 per cent Tamil, and 41.6 per cent Muslim. Thus, the slight alteration in place spelling reflects a demographic shift which is politically unsettling to many East Coast Tamils. Just as the geographical designations ‘Derry’ and ‘Londonderry’ have been politicized, so have ‘Ampare’ and ‘Amparai,’ with the former connoting a pro-Sinhalese sentiment and the latter connoting a pro-Tamil sentiment. For example, according to the 1981 Central Budget, the capital expenditure in the Jaffna District (approximately 6 per cent of the total population) was only 2.6 per cent of the national capital expenditure. On a per capita basis, the capital expenditure in the Jaffna District was Rs. 313, while the national expenditure was Rs 656. In addition, foreign aid allocation to the Jaffna District for the period 1977–82 was nil (CRD 1984: 15). Furthermore, as late as 1975 almost 90 per cent of industry on the island continued to be located in the Sinhalese majority Western Province (Shastri 1990: 70). Of the 40 major government-sponsored industrial units, only five were located in predominantly Tamil areas – of these four were established in the 1950s and one in the 1960s (Manogaran 1987: 130–4, 139). All of these figures are based on a period which was relatively peaceful. While such economic discrimination added to the disgruntlement of Northern Tamils and was criticized by Tamil politicians, it is the full-scale warfare since 1987 which has decimated the regional economy. A development worker who had just returned from Jaffna in May 1992 described the Northern Province as having been bombed and brutalized (by state and LTTE forces) back into a pre-industrial economy. In April 1993, a Canadian academic reported: ‘Meanwhile, the Jaffna Peninsula is locked off by the army, and life there is worse than ever. You will know that Jaffna has lost half of its population. Those that remain under the de facto Prabhakaran [i.e. LTTE] government live hellish lives.’ Bruce Matthews, personal correspondence, 29 March 1993. The area has suffered under the double burden of economic discrimination and, more recently, war. For examples, see: in United States Committee for Refugees (USCR) (1991: 3), Hyndman (1985: 290) and Ponnambalam (1983: 268) In an 1992 interview with Lalith Athulathmudali, former UNP Minister of National Security and founder of the DUNF, he suggested that the solution to the ‘Tamil Problem’ had to start with the military annihilation of the LTTE and that if he won the 1993 presidential elections he would ‘translocate’ the residents of the North (particularly the youth). It was explained

212 Notes

24. 25.

26.

27.

28.

29. 30.

31. 32. 33. 34.

that they would live in ‘nice’ camps in the South and that ‘ideally’ they would have complete freedom. In this way, ‘the LTTE fish would be deprived of the life-sustaining water of public support.’ As a hurried postscript, Athulathmudali added that ‘of course’ this would be done only with the consent of the public. The assassination of Athulathmudali in April 1993 silenced one influential voice, but such sentiments continue in other extremist Sinhalese sub-groups. Personal Interview. Colombo, February 1992. Also called the interior or highland Sinhalese and the West and South Coast Sinhalese. Personal interviews, Sri Lanka, May–June 1992. Wet Zone Tamils live mostly in urban areas like Colombo, Kandy and Galle in the South-West quadrant of the country. This gets two monsoons a year while the rest of the country getting only one monsoon is referred to as the dry zone. Personal interviews, Batticaloa, June 1992. Determination of popular sentiment in a war zone is problematic, however. Attitudes and allegiances are context-dependent according to who is asking the question and which group is ‘in control’ at the moment. As a resident of the East Coast explained with only a hint of sarcasm: ‘there is always undying support for which ever group is holding a gun to your head.’ ‘Muslims sought and obtained membership and achieved positions of influence in all major national political (except Tamil) parties, particularly the [UNP] and [SLFP]. Indeed the link with the UNP has given that party the majority of the Muslim vote at every election since 1947. The UNP has always had more Muslim Members of Parliament than the SLFP. Within the Party, Colombo-based Muslims have been until very recently the dominant element’ (de Silva 1988: 208). LTTE attacks on Muslims took place before 1990, but they appear to be part of a two-track strategy which included efforts to build a common front against the Sri Lankan Government. Following 1990, LTTE strategy became exclusively antagonistic towards Muslims. Personal interviews. Batticaloa, February 2002 and December 2002. Somasundaram 1998: 42; Manogaran 1994: 114; UTHR (J)/University Teachers for Human Rights (Jaffna) 1994;Weaver 1988: 67; FEER 21 February 1985: 39; SLA Spearheads Sinhala Colonisation, TamilNet, 25 March 1998. Such as the Dollar and Kent Farm Massacres in 1984. Personal Interviews. Kattankudi, February 2002. http://www.hrw.org/wr2k/Asia-08.htm The Up-Country Tamils are the descendants of migrants from South India brought over by the British during the past century-and-a-half to work on the tea and rubber estates in the Central Province. Approximately 80 per cent are plantation workers on tea and rubber estates and a small but an important segment is engaged in trade and business in Colombo and upcountry towns (Hollup 1992: 320). Within the South Indian caste system, the majority of the Up Country Tamils are lower caste than the Sri Lankan Tamils in the north of the island. A rural development worker in the plantation area estimated that about 80 per cent of the Up Country Tamils would be low caste – of which, about 60 per cent would be harijans or

Notes 213

35. 36.

37.

38.

39.

40.

41.

42.

untouchables by the Indian ranking system, and the remaining 40 per cent would be from the next highest low caste. This, as well as spatial distance and political powerlessness, has contributed to limiting the interaction between the two groups of Tamils. This was recounted to me by a Jesuit Priest, Father Paul Casparsz, who worked with Bishop Nanayakara at the time. The term ‘Sinhalese’ means ‘people of the lion race.’ This mythical account plays down the existence and the rights of the original (aboriginal) population of the island, known as the Veddhas. There are a few thousand descendents of this population facing cultural extinction more imminantly than the aboriginal population of Australia. Although Max Muller initially defined ‘Aryan’ as a race, he later withdrew this definition arguing instead that ‘Aryan, in Scientific language, is utterly inapplicable to race. It means language and nothing but language, and if we speak of Aryan race at all, we should know that it means no more than X + Aryan speech.’ Max Muller, Bibliography of Words and the Home of the Aryans, (1888), cited in Committee for Radical Development (CRD) (1984: 42). There is a dearth of material specifically addressing the political dimensions of caste. This section benefits from an excellent unpublished paper by Matthews (n.d. (a)). Also useful are Roberts (1992) and Jiggins (1979). These are ritual occupations only. They originated in the pre-colonial period in connection with rajakariya (lit.: obligatory ‘king’s service’), casteregulated corvee work for those who did not own wet paddy land or were not involved in rice cultivation. E.R. Leach notes ‘the Washermen are only ritual washermen, the Drummers only religious drummers; in their ordinary life, Goyigama, Washermen, Drummers and the rest are all alike, cultivators of the soil’ (Leach (1953) cited in Matthews n.d. (a): 8). Although no political party represents the exclusive interests of a particular caste, caste is always a consideration in the choice of candidates and cabinet ministers. Further, as Matthews points out: ‘since Independence every government and every political party has had a Goyigama leader, except for a brief and freakish interim leadership of the SLFP by C.P. de Silva from March–May 1960, and the present United National Party administration’ (n.d. (a): 3). The low caste background of President Premadasa was a constant, if usually unspoken, issue in domestic politics. Some argue that it better enabled him to respond to Tamil concerns, while others argued that it inhibited him from forming a solid consensus within the Sinhalese community which is a prerequisite for inter-group dialogue. A.J Wilson argues that caste was a significant catalyst in JVP recruitment campaigns, especially among the Karavas and Duravas in the low-country southern districts (1974: 46). This is reinforced by Jiggins’ assessment of caste in the 1971 Insurrection. She points out that with within the ‘inner circle’ or ‘politburo’ of the movement, twelve of the fourteen were Karava (1979: 127). The Buddhist sangha in Sri Lanka has three principal Nikayas (‘sects’): the Siam, the Amarapura, and the Ramanna. Matthews explains that the divisions arose primarily as a result of caste differences. (Matthews n.d. (a): 9; Matthews n.d (b): 1; Matthews 1988–89: 621).

214 Notes 43. ‘Although it is somewhat of an over-generalization, it is still fair to say that Goyigama Christians tend to be Anglican, Karavas are frequently Catholic or Methodist and other still lower castes are more often than not members of the Methodist and other Protestant denominations. Ceylon Tamil Christian society is also informally regulated by caste tradition, particularly in marriage custom’ (Matthews n.d. (a): 11). 44. For a thickly detailed discussion of this process, see Jayawardenja (2000) and Jiggins (1979). 45. Brahmins do not figure prominently in Tamil politics. Although they rank higher than the Vellalars in the sacred framework, they rank decisively lower in the secular sense. On the whole, they are mere employees of the Vellalar who according to Banks (1971: 67) ‘do not hesitate to discipline Brahmins whom they consider to be behaving badly’ (p. 68) and even have ‘a recognized right to interfere in the details of the temple ceremonial, particularly in the matter of temple festivals and their organization.’ 46. This is the Tamil caste equivalent to the Sinhalese Goyigama. 47. Such reports increased as the LTTE gained exclusive territorial control of regions in the North and North East. (Sri Lanka Infromation Monitor, March 1993: 8; India Today, 15 October 1991: 91. Such practices were continuing in 2001 and early 2002. Bush 2002. 48. This action was undertaken under the cover of the extreme violence of the Second JVP Insurrection (1987–90). See Chapter 6: The 1971 JVP Insurrection and 1987 Resurgence. 49. Personal interviews. Sri Lanka, May–July 1992; 1998; 1999; 2000; 2001; 2002. 50. One NGO worker commented on pervasiveness of this phenomenon outside of the political arena as well, saying: ‘Factions, factions, everywhere. Even the International Youth Hostel Association is divided into warring factions.’ Personal interviews, Colombo June 1992. 51. For useful studies of the politics of patronage in Sri Lanka, see Jayasuriya (2000), Jiggins (1979) and Warnapala and Woodsworth (1987). 52. Bandaranaike’s assassin was a monk who was aided by Buddharakkhita, the chief incumbent of the influential Kelaniya Temple and the leader of the EBP (Eksat Bhikkhu Peramuna) or ‘United Monks Front,’ an ‘organization of political monks’ (Bechert 1979: 206) which had originally supported Bandaranaike in his 1956 election victory. 53. The B–C Pact was signed by Bandaranaike and S.J.V. Chelvanayakam, the leader of the Federal Party, the major Tamil political party at the time and the precursor to the Tamil United Liberation Front. 54. An incident noted by INFORM illustrates the intra-Party struggles: ‘Mrs. Bandaranaike left the island for Malaysia on March, leaving behind a letter nominating veteran SLFPer K.B. Ratnayake to act as the Leader of the Opposition, in her absence; the 45 MPs supportive of Anura Bandaranaike petitioned the Speaker to recognize Mr. Bandaranaike as the Acting Leader of the Opposition, flouting Mrs. Bandaranaike’s wishes. On [24] March, the Speaker said he would acknowledge Mr. Bandaranaike’s claim’ (Sri Lanka Information Monitor, Situation Report, March 1993). In November 1993, Anura Bandaranaike exited the SLFP and Joined the UNP as a Cabinet Minister.

Notes 215 55. Far Eastern Economic Review, 29 September 1988: 38. In the belief that it was going to win the 1988–89 elections, the SLFP is reported to have offered the JVP three ministries (FEER, 27 October 1988: 30). 56. Critics argue: ‘All power in the 1978 Constitution is concentrated nearly 100 per cent in the President. He is more an Executive President. He is the Head of State, the Commander in Chief of the Armed Forces, Head of the Government, Head of the Cabinet, source of appointment to the upper tiers of the judiciary. I don’t know what is left to concentrate on him… Both Parliament and Prime Minister have been so heavily devalued that the only thing that I can think of comparing them to is our rupee’ (Colin R. De Silva, quoted in Matthews 1992: 223). 57. Matthews observes the following shift: ‘Writing about Parliament in the early 1970s, Jupp [1978] remarked that the great majority of important politicians in Sri Lanka were drawn from the “oral professions” like law, and had Anglicized backgrounds. The social and educational background of most MPs in all political parties has now dramatically shifted. For example, in 1970 an estimated 75% of members of Parliament spoke English. Now I estimate that approximately only 40% have competence in this language’ (1992: 225). 58. For a review of events, see Matthews (1992) and Christian Worker, r 2nd & 3rd Quarter (November 1991: i–xviii). 59. US Department of State (2001) Department of State Human Rights Reports for 2000, February 2001, http://www.humanrights-usa.net/reports/srilanka.html 60. Lalith Athulathmudali was the first Asian President of the Oxford Union and spoke five languages (English, Sinhala, Tamil, French and German). He was a lecturer at the University of Singapore prior to launching his political career and was able to discuss knowledgeably topics ranging from colonial history and 18th century Church politics to military and education policy. 61. Such as Premadasa’s Janasaviya program (an acronym for several words, together meaning ‘a movement for spreading adaptable resources among the people’), a poverty-alleviation program under which he has pledged to give 1.4 million families Rs1,458 each month for two years in coupons (not cash) for necessities. At the same time Rs1,042 is to be deposited monthly in a compulsory National Savings Bank account so that a family may amass Rs25,000 at the end of that period. Far Eastern Economic Review, 27 October 1988: 30; Matthews (1989: 435); and Janasaviya Programme, Implementation Guidelines No. 1, Draft for Starting Work, National Housing Development Authority, Colombo, January 1989. 62. This is discussed in further detail in Chapter 9. 63. For successive failures in attempts by Tamil political parties to negotiate viable alternatives, see Wilson (1989). 64. By all accounts, the LTTE is responsible these and many other assassinations. See: Hoole et al. (1992); Law and Society Trust, State of Human Rights 2000; and regular UTHR (J) human rights reports. 65. This ethnicization has exacerbated tensions and Tamil mistrust of the security forces – helping to maintain inter-group stresses. 66. For example, the murder of Inspector Bastianpillai and a number of other police officers in 1978 (Hoole et al. 1992: 22).

216 Notes 67. Fredrika Jansz, ‘Where did the third C130 go?,’ The Sunday Leader, r 24 February 2002: 13. It was reported in many interviews in Batticaloa that in early 2001 the Sri Lankan Army extorted large amounts of money (Rs1.1 million by one estimate) from the local population to hold a ‘song and dance cultural event’ for the troops. Sources of funds included the Traders’ Association (61 leading businesses and shops) and 35 government-registered contractors, as well as schools and private individuals. Efforts were made to force locals to entertain the troops and TELO visited performers to remind them of their commitment. In the end, the LTTE called a hartel and cut electricity for the day of the performance. Personal Interviews. Batticaloa, January 2001. 68. See for example, ‘Colombo Sacks Soldiers Having Links with JVP,’ The Hindu, 20 April 1987: 1. 69. ‘An army officer has been charged under the PTA [Prevention of Terrorism Act] with helping JVP Committee member and present leader to escape to India last year. Capt. Dharmasiri Nissanka will appear before the Avissawella magistrate on April 27 for his involvement in the escape of Somawansa Amarasinghe, military sources said. Somawansa Amarasinghe and at least 200 other members of the JVP escaped to South India before some of them successfully sought political asylum in the West, reliable sources say’ (‘Army Officer Charged with Helping JVP Leader,’ The Island, 5 April 1992). See also, The Sri Lanka Monitor, December 1991: 4. 70. Sri Lanka Country Report on Human Rights Practices for 1999, Bureau of Democracy, Human Rights, and Labor, US Department of State, February, 2000, Jane’s Defence Weekly, January 12, 2000 71. See ‘A Wake Up Call,’ Frontline, vol 18: Issue 17, 18–31 August 2001. 72. Monks (bhikkhus) currently constitute 24 per cent of the arts enrolment, up from 8.8 per cent in 1982 (Matthews 1989: 482). 73. This was pointed out by a number of interviewees. However, Up-Country Tamil support for the JVP cause was the exception rather than the rule, motivated in a few cases by a common ‘opposition to the same oppressive system.’ Personal interviews, Kandy, May–July 1992. 74. Personal interviews, Kandy, June 1992. 75. Personal interviews, Colombo, Kandy, Batticaloa, May–July 1992. 76. EPRLF: Eelam People’s Revolutionary Liberation Front; TELO: Tamil Eelam Liberation Organization; PLOTE: People’s Liberation Organization of Tamil Eelam. 77. The paramilitary tensions are evident in most cities with a significant Sri Lanka Tamil population. Members of Tamil communities in Europe and North America are frequently harassed and forced to pay ‘donations’ to the war effort. The LTTE is particularly efficient in the collection of foreign funds both through such donations and a broader range of illicit activities (Chalk 2000; personal interviews). 78. This was brought to my attention by Dennis Cole in 1993, Coordinator of Intelligence, Immigration and Refugee Board of the Government of Canada. 79. Jane’s Intelligence (2000). Sentinel-South Asia, November 1999–April 2000, www.janes.com 80. This includes Soviet made SAM-7s ‘ purchased from corrupt government officials and insurgent forces in Cambodia, [and recently acquired] far more

Notes 217

81.

82. 83. 84.

85.

86.

87. 88. 89.

deadly and accurate, US-made Stinger missiles acquired from Afghanistan.’ Chalk 2000: 9. Even old data from August 1998 reveals the place of primacy of the LTTE use of suicide bombers. Chalk (2000) estimates that it had conducted 155 battlefield and civilian suicide attacks, compared to the 50 carried out by all other groups worldwide, including Hamas, Hizbollah, the Kurdish Worker’s Party (PKK) and Babbar Khalsa. ‘Martin Comes to Tamils’ Defence,’ National Post (Toronto), 31 May 2000. See also, National Post (Toronto), 3 June 2000. Ottawa Citizen, 4 October 2001. Eelam People’s Democratic Party is a Tamil paramilitary group which split off from the EPRLF in 1987. The EPRLF was a major actor while it received the full backing of the IPKF and Indian government. With the departure of the Indian troops, a large portion of EPRLF members and their families have re-located to South India fearing retribution from the LTTE. Other members sought protection by joining the government side to fight the LTTE with army support. However, according to the terms of the February Ceasefire, all of these groups are to be disarmed and disbanded. Art. 1.8: ‘Tamil paramilitary groups shall be disarmed by the GOSL by D-day +30 at the latest. The GOSL shall offer to integrate individuals in these units under the command and disciplinary structure of the GOSL armed forces for service away from the Northern and Eastern Province.’ Preliminary reports indicate that this is being done. Time will tell whether this is a permanent state of affairs. A. K. Menon, ‘The Other Battle Field,’ India Today, 15 October 1991: 99; The Hindu, 6 September 1991: 1; Frontline, reprinted in Christian Worker, 2nd & 3rd Quarter, 1991: xvi. Christian Worker, 4th Quarter, 1989: viii–x. This view is candidly shared by both military and paramilitary leaders. The assassination of Gandhi may be identified as a critical juncture because it appears to have radically altered the relationship of the LTTE with supporters in South India and in the Indian government. The impact of Premadasa’s assassination appears to have been ambivalent regarding major changes in inter-group or intra-group relations.

4 1948 Independence and Disenfranchisement 1. Relevant sections from the Ceylon Citizen Act, No. 18 (reprinted in Ponnambalam 1993: 75): 4 (1) a person born in Ceylon before the appointed date (15 November 1948) shall have the status of a citizen of Ceylon by descent, if (a) his father was born in Ceylon, or (b) his paternal grandfather and paternal great grandfather were born in Ceylon. 4 (2) a person born outside Ceylon before the appointed date shall have the status of a citizen of Ceylon, if (a) his father and paternal grandfather were born in Ceylon, or (b) his paternal grandfather and paternal great grandfather were born in Ceylon.

218 Notes

2.

3. 4.

5.

6.

7.

8.

5 (1) a person born in Ceylon on or after the appointed date shall have the status of a citizen of Ceylon by descent, if at the time of his birth his father is a citizen of Ceylon. Citizenship was possible if the following conditions were met: that the Indians or Pakistanis possessed ‘an assured income of a reasonable amount’, had no disabilities which made it difficult for them to conform to the laws of Sri Lanka, had their wives and dependent children residing with them during the required period of residence, and had renounced any other citizenship if they possessed it. The residency requirement was ten years if without a spouse and seven years if married (Wilson 1977: 30). For full details, see Kurian (1989), Jayawardena (1990: 64–98); Tinker (1977); and Wilson (1977: 28–38). Agreements along the way include: the Indo-Ceylon Agreement of January 1954, the Sirima–Shastri Agreement of 1964, the Indo-Ceylon Agreement Implementation Act of 1968, and the unilaterally enacted legislation in 1986 to repatriate 94,000 Plantation Tamils. I have argued elsewhere that: ‘many of the ‘new’ political voices among the Plantation Tamils are from those who are increasingly disgruntled with the ability of the pro-UNP CWC, the largest Plantation Tamil trade union, to deliver ‘the goods.’ That the Plantation Tamils have in the past articulated political demands through a pro-UNP political organization has also helped to maintain political barriers between them and other Tamil sub-groups. It was estimated that the CWC has captured approximately 25 per cent of organized worker support (Personal interviews with a range of union organisers and workers, Kandy, May to July 1992). The UNP linkages of the CWC have earned it the disparaging label, the ‘mini-UNP,’ by opposing unions. In the past, political dissent and demands could be pacified/defused by the adroit manoeuvring of the late leader, UNP Minister S. Thondaman, which included cracking down on opposing unions as well as delivering [some] political gains on paper such as citizenship for those Plantation Tamils who had been disenfranchised at independence. However, the death of Mr Thondaman has given rise to jostling for leadership within the CWC which has set in motion a battle for succession that challenges an element of stability within the Plantation Tamil areas. At this stage, a struggle for political control will likely take place both within the CWC and among the political groups in the Hill Country.’ (Bush 1993). The idea of being ‘swamped’ is a recurrent theme in these debates. Jayawardena (1990: ff. 69) offers a sample of Sinhalese chauvinist remarks and arguments drawn from Parliamentary debates. This compares with 1: 66,400 for the Sinhalese (68 MPs) ; 1: 63,831 for the Muslims (6 MPs); 1: 61,919 for the Sri Lankan Tamils (13 MPs); 1: 11,224 for the Burghers (3 MPs); and 1: 608 for the Europeans (4 MPs) (Manor 1989: 189). Kumari Jayawardena places similar emphasis on both the class and ethnic motivations (though in an intellectually more sophisticated way than Ponnambalam), concluding that the Citizenship Acts ‘[brought] to a conclusion the legal manoeuvres of the Sinhala majority to exclude the Plantation workers from citizenship, thereby disenfranchising the largest section of the working class’ (1990: 81).

Notes 219 9. See Jayawardena (1990: 92–95) for the voting record on the Indian and Pakistani Residents Citizenship Act (1948) and for an account of the ethnic backgrounds of those minority MPs voting in favor of the legislation.

5 1956 Election and S.W.R.D Bandaranaike 1. For example, in the economically hard times of the 1930s and the immediate post-WWII period, the political status of Plantation Tamils was represented as a threat to Sinhalese labor interests. 2. At that time, the Sinhalese as a whole were less advantaged than the Sri Lankan Tamils because of lower educational opportunities. Consequently, the Sinhalese generally had less status and income associated with good employment and commercial opportunities. 3. Bandaranaike denounced UNP leaders as ‘treacherous of their race and language’ and claimed that language parity ‘would mean disaster for the Sinhalese race’ (Manor 1989: 236). A bhikkhu (Buddhist monk) from Matara expressed a similar level of emotion: ‘It is not necessary to actually assault the Tamils. If ten to fifteen youths of Matara get together and with knives about on the public road, that alone will make the Tamils run back to the north’ (quoted in Manor 1989: 236). 4 Such as loss of political confidence and material resources, increased attacks from within, defections to competing groups, escalating extremism, and so on. 5. Sir John Kotelawala was known as ‘Asia’s foremost playboy politician, the card-playing two-fisted drinker in the white dinner jacket with an eye for the ladies’ (Manor 1989: 223). He relished and cultivated the image of ‘the strong man of Ceylon’ as a tough and virile man about the town as well as a thug politician able to intimidate and coerce with private cadre of goons (ibid.). As Manor explains: Sir John’s lifestyle and public indiscretions soon made him a symbol of westernized crudity in the eyes of campaigners for indigenization. His wellpublicized penchant for jodhpurs by day and dinner jackets by night, for cocktails and extravagant entertainments, seemed alien and offensive to many Ceylonese, particularly to the Buddhist clergy. He seemed game for anything. It took scathing editorials in the press to dissuade him from riding to a reception in a chariot pulled by sixty girls. In July 1954, he outraged pious Buddhist opinion by attending a barbecue where a calf was roasted on a spit, an incident which became part of the island’s folk memory and haunted him ever after. As the Buddhist upsurge gained momentum in 1955, Sir John blithely described it as ‘madness’ and blamed the monks for the decline of the faith. He seemed to represent all that the revivalists loathed, and this forced them to seek champions from outside the government (Manor 1989: 230). 6. Official photos of the first cabinet show only Bandaranaike and J.R. Jayewardene in ‘national dress.’ Death sentences for convicted Buddhist criminals concluded with ‘May God have mercy on your soul.’ Cabinet ministers gambled openly at race tracks and appeared in press photos, cocktails in hand. None of this endeared the UNP government to the Buddhist sangha or laity (Manor 1989: 199).

220 Notes 7. Many commentators have argued that the ‘enerving and castrating effect’ of foreign rule, which separated sangha from society was actively pursued by colonial rulers – particularly, the British. Rahula Walpola was one of the most influential exponents of this argument. See Tambiah (1992: ff. 28). 8. See Walpola (1974), especially Appendix I ‘What is Politics?’, Appendix II ‘Bhikkhus and Politics’, and Appendix III ‘The Kalaniya Declaration of Independence’. Note that this was originally published in Sinhala in 1946 and had an important effect in stimulating political debate within the sangha. 9. Walpola discusses the historical precedents in his Heritage of the Bhikkhu (1974) and his The History of Buddhism in Ceylon (1956). The moral obligation is represented as follows: ‘What the mind is to the body, religion is to politics. Politics bereft of religion becomes sin and evil’ (Rahula 1974: 122). This sentiment is given operational form in the 1946 declaration of the Vidyalankara Pirivena, the monastic collage where Rahula was a senior teacher at the time: ‘Even today bhikkhus by being engaged actively in education, rural reconstructions, anti-crime campaigns, relief work, temperance work, social work and such other activities, are taking part in politics, whether they are aware of it or not. We do not believe that it is wrong for bhikkhus to participate in these activities. We believe that it is incumbent on the bhikkhu not only to further the efforts directed towards the welfare of the country, but also to oppose such measures as are detrimental to the common good’ (emphasis added); reprinted in Rahula (1974: 131–3). 10. This was especially important since the election commissioner’s recommendation to increase the number of polls (in order to increase access) had been ignored by the ruling UNP government. This increased the transportation problem for SLFP supporters and because polling stations were located in UNP strongholds, it increased the opportunities to impede and intimidate them as well. 11. Ironically, the EBP would later be implicated in the assassination of Bandaranaike only three years later. 12. A similar phenomenon occurred in Burma in 1917 and 1938 (Bechert 1979: 204). 13. Tambiah cautions that ‘we should be careful not to credit the EBP with a strong organizational structure for systematic and long-term action’ (ibid.). The EBP was a extremely useful to the SLFP in the short-term but it lacked the formal linkages for longer-term staying power. 14. For a Tamil Model I analysis, see Ponnambalam (1983). Government publications and pro-government publications generally offer crude variants of Sinhalese Model I analysis, see Society for Ethnic Amity (n.d.). 15. Coincidentally, his ancestors also appear to have converted to Roman Catholicism under the Portuguese, Dutch Reformism under the Dutch, and to Anglicanism under the British (Manor 1989: 11, 110; Gooneratne 1986: 3–6). 16. Manor further notes that Bandaranaike was not always comfortable in his ‘national dress.’ In 1931, it was reported that he wore socks with his sandals – a breach of custom – and that they were held up by suspenders ‘of foreign origin’ (Manor 1989: 96). He was also frequently to be seen in European suits with stiff collars and occasionally he would appear in a newspaper

Notes 221

17. 18.

19.

20.

21. 22. 23.

photograph in a Western suit along side his prize greyhound, Billy Micawber (ibid. 130). In the 1950s, this ambiguity was captured in a cartoon which depicted Bandaranaike clad in national dress on the left side of his body and in a European suit on the right (ibid. 97) Most immediately, the violence of 1958. For details, see Vittachi (1958); Manor (1989: 286–299); and Tambiah (1992: 46–57). This distinction is important because each configuration of intra-group conflict leads to a different set of political outcomes. If the conflict had been between a traditional elite and Anglicized elites, as Horowitz (1973) erroneously suggests, and if the traditional elite had won the contest, then the resulting system of government probably would have deviated from the colonially bequeathed Western model. While it would unlikely be a theocracy, it would have positioned the Sinhalese Buddhist elite (the sangha, vernacular literati and educators, ayurvedic practitioners and so on) in a far more prominent and influential position in the affairs of state. But as long as the dissent of the indigenous elite was channeled through an instrumentally indigenized elite, the Western structures of government would be maintained, indeed reinforced, albeit in an ethnicized form. In other words, the set of outcomes in the first scenario would more likely entail a transformation of the system, whereas the second set of outcomes would more likely entail a reconfiguration of power within the existing system. Frequent and spectacular disputes between Bandaranaike and other ministers were reported in Senanayake-owned newspapers. The Byzantine infighting is well presented in Manor’s (1989) biography of S.W.R.D. Bandaranaike. This is literally as well as figuratively true. Bandaranaike was from a highcaste aristocratic Kandyan clan while Senanayake was from low-country aristocracy. The political significance of this is reflected in the competing networks of political and economic resources each was able to drawn upon. There was an on-going rivalry between the Senanayakes and the Bandaranaike–Obeyesekere clans; the latter had built up considerable wealth and influence collaborating with colonial authorities, while the former had led the nationalist and reform struggles against the British. Further tension was generated by the fact that the Bandaranaike– Obeyesekere clan members were long-time property owners while the Senanayakes were more recently propertied (Manor 1989: 108). The nepotistic structure of the UNP is outlined in Jiggins (1979: 96–110). For further details, see Manor (1989). For details, see Manor (1989: 271–272). For details, see Bechert (1978). ‘Both Sinhala and Tamil would be used in Parliament and all laws would be promulgated in both languages. Sinhala would be the language used by all courts, government offices and local boards except in the Northern and Eastern Provinces where the language should be Tamil and the medium of instructions in the schools would be the language of the local majority. Although the resolutions rejected parity for the two vernaculars, the SLFP had accepted something very close to it. The headlines read “Sinhala Only” but the small print contained a rather different message’ (Manor 1989: 233).

222 Notes

6 JVP 1971 and 1987 1. ‘Authority’ is used here in its Weberian sense: ‘the probability that a command with a specific content will be obeyed by a given group of persons’ despite resistance, regardless of the basis on which that probability rests (Weber 1947: 152). 2. Although it is possible to analytically and chronologically separate the 1971 and 1987 episodes, they are treated together here to enable the examination of shifts in ideology and mobilization over time. 3. At different points in time, with varying degrees of emphasis, JVP rhetoric distinguished between Plantation Tamils (particularly as workers) and Tamils from the North and East. See for example, Samaranayake (1987). 4. Chandraprema (1989: 6) dates the ‘inception of the JVP’ ‘around 1964 when a group of youthful dissidents broke away from the pro-Peking Ceylon Communist Party.’ Gunaratna (1990: 5) suggests that the JVP began in 1966–67. 5. For examples, see Jiggins (1979: 139–42), Gunaratna (1990: 37–61, 89, 127), and Chandraprema (1991: 33–43). 6. Interestingly, the JVP rarely made direct reference to religion. ‘It realizes that no anti-Buddhist party could survive in Sri Lanka and therefore does not discourage Buddhist ‘patriotic’ activism (in fact, it is assumed that some JVP opportunists have even donned robes, including perhaps even Wijeweera himself)’ (Matthews 1989: 430). Thus, although the JVP was able to mobilize support from the sangha, particularly the young bhikkhus, mobilization was based on revolutionary or ‘patriotic’ principles rather than Sinhalese Buddhist principles per se. During the 1987 JVP resurgence, the ‘Number 3 in the organization’ was D.M. Nandasena (alias D.M. Ananda), a monk who had given up the robes in his final year of study at Peradeniya (Chandraprema 1991: 10). 7. The following statistics are from Jiggins (1979: 123). 8. Jiggins (1979) makes a convincing case that the both leadership and followers were low caste. In the inner circle or ‘Politburo,’ 12 out of 14 (or 10 out of 11 depending on one’s source of information) were Karava. 9. See Alles (1977); Matthews (1989); Obeysekere (1974); Jiggins (1979); Samaranayake (1987); Jayawardena (1990); Gunaratna (1990); and Chandraprema (1989, 1991). 10. Two recent accounts of the JVP suggest that the premature attack on police stations which eliminated the element of surprise in the 1971 insurrection was the result factional in-fighting: Chandraprema (1991); and Gunaratna (1990). 11. At the time, the Sri Lankan military services were still largely ceremonial. Political leaders neither perceived the need nor commanded the resources to radically increase the military forces. In addition, following failed military coups in 1962 and 1966, governments were hesitant to build up the military apparatus, preferring instead to rely on the police forces to maintain law and order (personal interviews with Sri Lankan Army military officers, Colombo, May–July 1992). 12. For one estimate of costs, see Gunaratna (1990: 104). 13. The events of 1971 can be labelled an ‘insurrection;’ the JVP organized its members to launch a mass attack against police stations throughout the

Notes 223

14.

15.

16. 17.

country on a certain date. The preparation of the members may have been inadequate and the attacks may have been uncoordinated, but there was a sufficient level of organization and resources to enable identifiable actors to engage in a particular collective action at a particular point in time. In contrast, the events starting in 1987 lacked such structure. The JVP had only recently adopted a cell organizational structure and was in the process of reconstituting itself into an underground revolutionary movement (Gunaratna 1990: 200–201). It did not have sufficient resources to launch an insurrection. Instead, independently of JVP activity, events catalyzed political dissent within the Sinhalese intra-group arena and created ideal conditions for launching a sustained challenge to the status quo and the representatives of the state. The JVP responded to, rather than initiated, events. As discussed elsewhere in this chapter, the signing of the Indo-Sri Lanka Agreement proved to be the pivotal event which galvanized the opposition from many diverse sub-groups and enabled the resurgence of the JVP. Thus, while the events of April 1971 may be accurately labelled a failed insurrection, the period from 1987 to 1990 is more appropriately identified as a period of JVP ‘resurgence.’ Although the latter episode was a ‘resurgence’ rather than an ‘insurrection,’ it was much more destabilizing and bloody. On 31 July 1983, the JVP was proscribed with two other Marxist organizations, the Sri Lanka Communist Party (Moscow) and the Trotskyite Nava Sama Samaja Party. Opponents of the government and independent observers stressed the improbability of involvement by these groups. Meyer points to ‘the paradox of an attack directed against the Tamils by the JVP and the NSSP whose recent writings revealed a desire for rapprochment with the terrorists and an attempt to analyse the minorities problem’ (1984: 141). It is also unlikely that such an organized and concerted action as that in 1983 could have been planned without attracting the attention of the Central Intelligence Bureau which had claimed to have unearthed a Naxilite plot the day after the 1982 presidential elections (Obeyesekere 1984: 169). Wijeweera was, however, a distant third place finisher. By percentage of votes cast, the results were: UNP 52.91; SLFP 39,07; JVP 4.19; All Ceylon Tamil Congress 2.67; Lanka Sama Samaj Party 0.88; Nava Sama Samaja Party 0.26. The following statistics are from Matthews (1989: 427–8). The 1987 Insurrection had a very different leadership from that of the 1971 Insurrection. With the exception of Wijeweera, the 1971 leadership was not involved in organizing for the 1987 Insurrection. After the 1971 experience they entered other areas of activity. While some continued political activities, they did so through the legal political channels (Chandraprema 1991: 2; Gunaratna 1990: 143). Some of those who had been in leadership positions in 1971 became disillusioned with Wijeweera and joined anti-JVP political parties in 1977 when it was un-banned (Chandraprema 1991: 52). Premachandra’s (1991: 6–14) biographical sketches of the JVP leadership in 1987 indicate that 16 of the 35 leaders (for whom sketches were provided) had been involved in the 1971 Insurrection but none held positions in the Central Committee or ‘inner council.’

224 Notes 18. See, for example, the International Alert report, Political Killings in Southern Sri Lanka (Marino 1989) which includes 31 pages of the names of victims and the circumstances under which they were murdered in JVP-related violence. See also, Gunaratna (1990: 270–290). Chandraprema (1989: 30–31) writes: ‘Knives and swords are used in preference to guns to do the actual killing … guns being used only for intimidation of intended victims. Victims are very often cut to ribbons part by part … slowly. Whole families have been chopped to bits … the children often being burnt alive. Many members of the public – very ordinary people – like bus drivers, scavengers, small time shop-keepers have been shot for disobeying JVP orders. This has been done regardless of the fact that their orders were disobeyed only because of counter-orders by the armed forces.’ 19. As discussed below, the JVP initially avoided direct attacks on the military forces and, instead, attempted to recruit them to the JVP cause. 20. Each of these incidents, and those below, were confirmed in personal interviews conducted in Sri Lanka in May–July 1992. Abductions, disappearances, and hostage-taking were still occurring at that time on the East Coast and in the North. The culprits, however, were not the JVP, but government forces, pro-government Tamil paramilitaries, and the Liberation Tigers of Tamil Eelam (Amnesty International 1993; USCR 1991; INFORM Sri Lanka Information Monitor). r Discussions with an individual involved in a project documenting disappearances and human rights abuses on the East Coast, and an examination of a computer print out of over 3,000 missing persons from the Southern Province (made available to me by a Sri Lanka-based human rights monitoring group) reveal a striking fact: the majority of individuals (who had the courage to file affidavits) know who abducted their children, parents, siblings or friends. Details are frequently available on: the security force responsible; the identity of members of the abducting group; license plate numbers of vehicles involved in an abduction; and location where the abducted individuals were initially taken. Yet prosecution of those responsible for flagrant humans abuses has been almost non-existent. 21. With a specific focus on Sri Lankan and Mozambique, Nordstrom (1992) provides an insightful anthropological examination of the construction and the use of terror as a mechanism employed by state and non-state forces for gaining or maintaining socio-political control over a population. Such ‘dirty wars’ ‘seek victory, not through military and battlefield strategies, but through horror. Civilians, rather than soldiers, are the tactical targets, and fear, brutality, and murder are the foundation on which control is constructed’ (p. 261). 22. For example, JVP leaflets have placed the following restrictions on the burial of its victims: forbidding the burial to be a ceremonial occasion; limiting the service to only two bhikkhus for no more than a half hour; forbidding the body to be transported in a hearse or to be carried more than one foot above ground level; prohibiting notices of grief, photographs or announcements; prohibiting the cremation of the body and stipulating that the burial plot must be at ground level; and limiting funeral rites to no more than ten individuals (Gunaratna 1991: 299). 23. For a detailed discussion of Udugampola’s specific allegations, see the speech by the lawyer who convinced him to write down a full statement of

Notes 225 all he knew about government-related death squads: Batty Weerakoon in the Yukthiya (1 August 1993), reprinted as an insert in INFORM, Sri Lanka Information Monitor, August 1993. See also: Christian Worker (2nd & 3rd Qr.) 1992: i–ii; and the Ottawa Citizen, 11 April 1992: A 6. Udugampola, nicknamed the ‘butcher cop,’ was well-known for his brutality: ‘Udugampola’s mother, brother, sister-in-law and two children were killed and burnt [by the JVP]. Subsequently, the Udugampola brothers – a DIG, another an Inspector and an Army Major vied (sic) each other in heading the anti-JVP war with their own special teams. Udugampola, who was very active in anti-JVP operations transformed the strategy of combatting the JVP from cordon and search, to search interrogate and destroy. It was he together with Brigadier Laksman Algama who first started to fight the JVP unconventionally’ (Gunaratna 1990: 286). His brutality was frequently mentioned in interviews with development and human rights workers on the island (personal interviews, May–July 1992). Most felt that peacebuilding and reconciliation were not possible until human rights abusers were held accountable for their actions. 24. Following the assassination of President Premadasa in May 1993 and the subsequent change in UNP leadership, secret contact was made between Udugampola and the government which resulted in their respective withdrawal of allegations and criminal charges. In August 1993, Udugampola was appointed Acting Chairman of the Port Authority with a handsome salary and allowances for housing, entertaining, and fuel. Udugampola claims that the signatures on the incriminating documents are not authentic and the UNP government appears anxious to let the whole episode drop out of sight. What makes the Udugampola–UNP reconciliation and declaimers incredible is that when Udugampola passed his press statement (based on his affidavits) over to the lawyer Batty Weerakoon, the whole episode was videotaped by a BBC correspondent in Sri Lanka. The video shows Udugampola signing his name at the bottom of every page. Opposition party members maintain their accusations that the UNP government is implicated in vigilante murders. The government is unwilling to set up a commission of inquiry into the Udugampola allegations, e.g. verification of the lists of victims, political affiliations, and circumstances of death (Sri Lanka Information Monitor, August 1993) 25. The Rajarata Rifles regiment was deployed against Tamil militants in North in 1984/85 when it went on a rampage of indiscriminate killing of civilians following the murder of some of its members by Tamil militants. Following the incident there was a mass desertion of the regiment. The government responded with the complete disbanding of the regiment, the revocation of military honours and the regiment’s colors. Personal interview with Bruce Matthews, Colombo, June 1992. See also Marino (1989: 5). 26. The 21 April 1989 issue of the JVP newspaper, Ranabima contained the following passage: ‘The main reason why the respect of the people for the Army has deteriorated is not because the shortcomings and faults of the soldiers but because the Army has been utilized in committing various crimes against the people for political motives. Service chiefs like … are responsible for this’ (reprinted in Chandraprema 1991: 294).

226 Notes 27. Death threats were sent to the families of known JVP members. They concluded as follows: ‘Is it not among us, ourselves, the Sinhalese people that your son/brother/husband has launched the conflict in the name of patriotism? Is it then right that you who are the wife/mother/sister of this person who engages in inhuman murder or your children should be free to live? Is it not justified to put you to death? From this moment, you and all your family members must be ready to die!. … May you attain Nirvana!’ (reprinted in Chandraprema 1991: 296). 28. It should be noted that Wijeweera was arrested three weeks before the 1971 Insurrection and spent the duration in a Jaffna prison. During the legal prosecution of the insurgency leaders, Wijeweera conducted his own defence (resulting in a transcript of over 400 pages). His fiery defence of JVP actions served to imbue the movement with a romanticized tinge in the eyes of the youth of the mid-1980s; something which helped JVP mobilizational efforts. 29. The identification of the ‘class’ basis of the JVP is problematic. It was not based particularly on the urban working class or the peasantry. Premachandra (1989: 7) describes its support base as a ‘bitterly discontented layer of rural middle class youth.’ Wijeweera, however, specified the social location of this group more explicitly when he characterizes the unemployed youth as a group waiting to enter the rural proletariat (Jiggins 1979: 125). Jiggins explains ‘the movement was thus not among rural wage labourers, but characteristically, was the expression of those either waiting to sell their labour or engaged in improving its saleability’ (ibid.). 30. The original ‘Five Lectures’ were: ‘The Economic Crisis of the Capitalist System,’ ‘Indian Expansionism,’ ‘Independence,’ ‘The Leftist Movement’ and ‘The Path the Sri Lankan Revolution Should Take.’ For details, see Gunaratna (1990: 61) and Chandraprema (1989: 71–77). 31. In the form of Indian capital domination of the Sri Lankan import–export trade, cultural ‘infiltration,’ smuggling, and so on. 32. A Sinhalese development worker in the Plantation areas explained the tension faced by the JVP vis a vis Plantation Tamils. As self-proclaimed Marxists, the JVP cannot deny the Plantation Tamil right to struggle for self-determination – as this ‘resonates ideologically with the JVP.’ However, the JVP must also recognize that its support base is in the rural villages, that is, the Sinhalese villages. Thus, according to this observer, mobilization must pander to anti-Tamil sentiment; the difficulty is in determining the cost-effective balance between support for Plantation Tamil self-determination and the representation of Plantation Tamils as a facet of ‘Indian expansionism’ (personal interviews, Kandy, May–July 1992). 33. ‘A Message to the People of Sri Lanka’ (1984), quoted in Jayawardena (1990: 150). 34. Personal interviews, Sri Lanka, May–July 1992. See also Gunaratna (1990: 131–3.) 35. Subsequently murdered in Madras by LTTE in June 1990 with 15 others members of the EPRLF. 36. It was alleged in the interview that because the Sri Lankan government feared the PLOTE–JVP connection, it provided aid to the LTTE in an attack against a PLOTE camp in Wilpattu Wildlife Park in May 1989. The Sri

Notes 227

37.

38.

39.

40.

41.

Lankan military support is alleged to have included air, sea and land transport. The dynamics here are parallel to the allegations during the impeachment procedings that President Premadasa had supplied money and material to the LTTE for use against the IPKF and IPKF-supported defence forces (see Episode V). See also Gunaratna (1993: 333–4) for further instances of cooperation between Sinhalese and Tamil militant groups, including joint training in India and Sri Lanka and JVP efforts to use PLOTE connections to acquire arms from North Korea. It should be noted that the number of students directly involved in pro-JVP activity was actually quite small relative to their impact. Matthews (1989: 432) estimates that in 1989 only 10–20 per cent of the students were actively involved in violence and intimidation. A belief also held by A.C.S. Hameed, Minister of Education, who argued that the single primary factor contributing to student unrest was the abolition of student assemblies without providing an alternative arrangement for student representation (in an interview with Bruce Matthews (1989: 432). ‘Legitimacy’ is used in its Weberian sense: a conviction on the part of persons subject to authority that it is right and proper and that they have some obligation to obey, regardless of the basis on which this belief rests (Weber 1947: 324). Athulathmudali’s proposal is also mentioned in Chapter One. In the initial stages of ‘translocation,’ the Tamil youth could live in ‘nice’ camps in the South. Ideally, they would ‘have complete freedom’ and would enjoy ‘training programs.’ Athulathmudali argued that this way the ‘LTTE fish would be deprived of the life sustaining water of public support.’ As an after-thought, Athulathmudali added that ‘of course’ he would do all this only with the consent of the public (personal interviews, Colombo, May–July 1992).

7 1983 Riots 1. ‘Militarization’ is used here to refer to the tendency for intergroup relations and conflict to be defined in narrow military terms. This typically coincides with an increase in military-related expenditures and military crack-down on civilian and combatant groups. After 1983, the Sri Lankan government began to view the ‘Tamil problem’ exclusively as a ‘military’ one. The statement by President J.R. Jayewardene illustrates the phenomenon: ‘I shall have a military solution to what I believe is a military problem. After doing so, I shall tackle the political side’(quoted in The Times (London), 26 January 1986). It is argued below that the friction within the Jayewardene administration between ‘hawks’ (advocating militarization) and ‘doves’ (advocating political solutions) was a factor in the outbreak of the riots. In September 1993, President Wijetunga expressed similar sentiments when he stated his conviction that ‘there is not ethnic conflict … one only has to overcome a terrorist problem.’ The same position was articulated by Sirimavo Bandaranaike (Sri Lanka Information Monitor, September 1993), as well by her daughter in her role as president, most evidently in her ‘War for Peace.’

228 Notes 2. This frustration was aggravated by the perception that education and employment opportunities were being denied to them (CRD 1984: 7–13; Gunawardene 1975; Kapferer 1984; Kearney 1975; De Silva 1977; C.R. De Silva 1979). After the declaration of a state of emergency in Jaffna in 1979, the presence and activities of the ‘Sinhalese Army of Occupation’ added to the frustration of Tamil youth. For a discussion of the 1979 imposition of a state of emergency in Jaffna, see MIRJE (1980). 3. The text of the resolution is reprinted in de Silva (1986: 403–406). The Resolution concludes with the statement: ‘And this Convention calls upon the Tamil Nation in general and the Tamil youth in particular to come forward to throw themselves fully into the sacred fight for freedom and to flinch not till the goal of a sovereign socialist state of Eelam is reached.’ Although the TULF officially eschewed violent means to achieve this end, its public ambivalence and provocative speeches caused many in the Sinhalese audience to doubt its commitment to a non-violent pursuit of Eelam. A number of individuals interviewed in Sri Lanka (including a former TULF strategist and Tamil academics) felt that the TULF was tactically inciting the youth militancy as a means of extracting political concessions (such as devolution of political power) from the UNP government. It was a tactic, they pointed out, which back-fired. 4. Hyndman (1988: 83) estimates that by May 1985, there were over 100,000 Tamils from Sri Lanka in South India – an estimate confirmed in April 1985, by the Acting High Commissioner of India to Canada in a presentation at Carleton University (Canada). The Tamil diaspora is large and especially well-organized in Canada, Australia, the US, Britain, Norway, and Germany. The US Committee for Refugees (1991) estimates that by 1991 more than 210,000 had fled to the south Indian state of Tamil Nadu and at least 700,000 people from all communities had been displaced within Sri Lanka (News Release, 12 November 1991). For a general discussion of displaced persons in Sri Lanka, see CPA et al. (2001) and Thambayah (1992). 5. The full text of the Prevention of Terrorism Act and other emergency regulations are available through the Law Library of Congress. See The Democratic Socialist Republic of Sri Lanka, ‘Legislative Enactments of the Democratic Socialist Republic of Sri Lanka’ in force on the 31st Day of December 1980, Revised Edition (Unofficial), Volume II containing Chapters 25 to 30. Since the mid-1990s, Law and Society Trust has annually published Sri Lanka: State of Human Rights, which offers regular assessment of this, and related, security legislation. 6. Within Sri Lanka, NGOs include the Law and Society Trust, the Centre for Policy Alternatives, the Consortium of Humanitarian Agencies, the Centre for the Study of Human Rights, Lawyers for Human Rights and development, MIRJE, INFORM, the Civil Rights Movement of Sri Lanka, the University Teachers for Human Rights (Jaffna), and the Batticaloa Peace Committee. Outside Sri Lanka, the list of groups includes: Amnesty International, the International Commission of Jurists, International Alert, Lawasia, Asia Watch, and the British Refugee Council. 7. For details, see Tiruchelvam (1999); Matthews (1987, 1993) and Islam (1987). 8. The Jaffna Public Library contained over 95,000 volumes, many of which are irreplaceable and were of great cultural value (MIRJE 1983: 17; Obeyesekere 1984: 163; Leary 1981: 32).

Notes 229 9. In a Statement to Parliament (9 June 1981) on the violence in Jaffna, Gamini Dissanayake said: ‘ … there is no doubt whatever, that there was a very serious situation in Jaffna because the police force was on the verge of a virtual mutiny’ (reprinted in the Sri Lanka Resource Centre (1981: 57). 10. Gamini Dissanayake’s statement in Parliament suggests that the government lost control, but denies responsibility: ‘The duty of maintaining law and order is the responsibility of the Government, and the police force is an arm of that process, but when the police force is shot at and people are killed [by Tamil militants], the behavioral patterns of those officers who are shot at is something we cannot be responsible for’ (reprinted in Sri Lanka Resource Centre 1981: 68). 11. See for example: ‘Statement made in Parliament by the Leader of the Opposition, Mr. A. Amirthalingam on 9th June 1981’ (Sri Lanka Resources Centre 1981: 9–24) and ‘Statement in Parliament by Mr V. Yogeswaran, M.P. for Jaffna’ (ibid.: 25-30). 12. The communalist rancor in Parliamentary debates of this period is illustrated in excerpts reprinted in Sri Lanka Resource Centre (1981). See also, MIRJE (1983: 16–25). 13. One poster pasted throughout Colombo read: ‘Aliens, you have danced too much; your destruction is at hand. This is the country of we Sinhalas’ (USCR 1991: 8). 14. While the violence in Ampara is generally identified as the the starting point for this outbreak, the violence there was ‘spill over’ from a sporting event in Batticaloa. The violence in Ampara was ‘communalist’ in terms of the axis of confrontation, it may not have been communalist in terms of its ‘trigger.’ According to Norman Uphoff, the violence there and in parts of Gal Oya were spontaneous, in contrast to the more structured incidents which followed elsewhere (personal correspondence, January 1994). 15. President Jayewardene seems to have recognized this fact. According to Leary (1981: 21): ‘The International Herald Tribune reported that President Jayewardene, in an interview with a Reuters correspondent on August 14, stated that attacks on Tamils in Ratnapura appeared to have been organized. The Guardian (London) reported on August 15th that ‘it seems to have been established that an unnamed group is organizing the present violence for motives of its own.’ An editorial in The Hindu (India) of 18 August, 1981 stated that ‘a close look into the riots would show that behind them is a planned and systematic effort to aggravate racial animosity.’ It was widely reported that attacks in Negombo as well as attacks against passengers on a Jaffna to Colombo train were made by organized gangs. Tamil sources stated that it could not be ruled out that people close to the government were behind the organized violence. They also claimed that police and army forces did not intervene to prevent attacks until the declaration of the state of emergency many days after the attacks began.’ 16. There were three identifiable categories of targets attacked by Tamil militants at this time: 1.) the military and police personnel stationed in the North; 2.) moderate Tamil politicians who were seen to be cooperating with the UNP government; and 3.) those identified as informers by the militants. 17. In May 1983, a Defence Ministry spokesman announced: ‘The armed forces and the police in the North are to be given legal immunity judicial and wide ranging power of search and destroy’ (Hoole et al. 1992: 58). In early

230 Notes

18.

19.

20.

21.

22. 23.

June The Sun reported: ‘Under such circumstances soldiers were compelled to react as during a war particularly in their role of fighting armed terrorists who had no compunction about killing servicemen or members of the public. In view of this, it has been felt that police and service-men in the North should be given the freedom of the battlefield rather than have their morale sapped through conflicts with legal niceties. This is not a peacetime situation and the police and services must be provided with adequate safeguards when attempting to control the problem’ (ibid.). On 23 July the Sri Lankan Army began an operation to remove some of the Plantation Tamils from the predominantly Tamil areas of Mannar, Vavuniya, and Trincomalee where they had resettled following the antiTamil riots of August 1977 and August 1981. In the early hours of the morning, an estimated 600 displaced Plantation Tamils were herded into trucks and brought to the Nuwara Eliya District (Bastian 1993: 22). This event may, or may not, be directly related to the ambush in Jaffna. It was, however, an act that was sure to provoke a negative response from Tamil sub-groups in light of the political volatility of state-sponsored settlement of Sinhalese in Tamil majority districts. Hyndman (1988: 8) reports allegations that the ambush was in retaliation for the rape of several Tamil women by soldiers. The fact that the patrol was initially ambushed with a land mine is significant because it is a result of Indian government support for Tamil militants: ‘India had begun to supply the Tigers with Claymore land mines as a way of better withstanding the Sinhalese armies superiority in mobility and its increasing resort to a military solution’ (Tambiah 1992: 72). Hyndman (1988) suggests that reporting of the rampage by security personnel in the newspapers in the South might have moderated the Sinhalese backlash. There are many horrifying stories of the experiences of Tamil families (see for example, Kanapathipillai 1990; and Tambiah 1993). It should be stressed that there were many instances where Sinhalese and Muslim individuals and families took great risks to protect Tamil friends and strangers. Tambiah (1993) discusses some of these incidents. These statements by Jayewardene were confirmed by The Guardian, 9 August 1983. Both passages are reprinted in Tambiah (1986: 25). The magisterial inquiries into each incident are assessed by Hyndman (1988: 19–26). The prison authorities have been criticized for not removing the remaining Tamil detainees following the first attack. Many of the detainees, being held under the PTA, had not been charged with any criminal offence. Among the victims was a medical doctor who was the Secretary of the Gandhiyam movement. Prison officials claimed to be unable to identify any individuals involved in the massacre. The corpses could not be examined by an independent coroner because after both incidents they were claimed by the Detective Superintendent of Police (Colombo) under Regulation 15A of the emergency regulations which allows state authorities to dispose of corpses. Hyndman observes: ‘From the list of those participating in the inquiries it can be seen that lawyers for surviving Tamil prisoners (Mr. Yogarajah apart) were not present at the hearing. … On the second occasion [i.e. the 27 July massacre] the prisoners were all in the same build-

Notes 231

24. 25. 26.

27.

28. 29. 30.

31.

32.

33.

ing and many of them may have had the opportunity to see who their assailants were, or to be able to provide other evidence of assistance to the inquiry’ (1988: 23). While the deaths were found to be ‘all cases of homicide as a result of a riot in prison’ (ibid.: 22), no one has been charged. Such as Gampaha, Kalutara, Matale, Nuwara Eliya, Badulla and Bandarawella. ‘Quasi-legal’ because the measures were sanctioned by emergency regulations which were often in contravention to international legal conventions. The militant groups are more appropriately identified as ‘incipient paramilitary organizations’ because they had not yet acquired the size, structure, or material resources associated with full fledged paramilitary organizations. Concurring with the study of the 1981 communal violence by the Council for Communal Harmony through the Media, MIRJE (1983: 18) concludes: ‘As it turned out, not only the bulk of Sinhalese people were kept in the dark and even misled in regard to the truth about events in Jaffna, but they were also, at the same time, subjected to dangerously anti-Tamil propaganda.’ See also Siriwardene (1984: 226–228). This point was confirmed during a seminar of researchers at the International Centre for Ethnic Studies, Colombo, June 1992. Radhika Coomaraswamy, Director, International Centre for Ethnic Studies (ICES). Personal interview, Colombo, 28 May 1992. Lalith Athulathmudali, then Minister of Trade and Shipping, opened his speech to the nation as follows: ‘A few days ago, my friends, I saw a sight which neither you nor I thought we should see again. We saw many people looking for food, standing in line, greatly inconvenienced, seriously inconvenienced.’ Obeyesekere comments bitterly: ‘Here was the leading intellectual in the government speaking of the hardship faced by Sinhalese people queuing for food [because Sinhalese mobs had burned down Tamil shops] when 70,000 Tamils were in refugee camps. Equally astounding is the fact that neither the President nor any minister of the government made an official visit to single refugee camp to console the dispossessed’ (Obeyesekere 1984: 167). Many middle class Tamils were dismayed by the rationalization for the violence which they often heard from some Sinhalese: ‘if you call for separation, what do you expect?’ However, following the JVP terror of 1987–90, which coerced Sinhalese into not denouncing a repugnant political program, there was an increase in the sympathy by some Sinhalese for the victimization of Tamil civilians in the North caught between government forces and anti-government forces. Personal interviews, Sri Lanka, May–July 1992. See also Spencer (1984) and Nissan (1984). The work of external agencies should not be seen as minimizing or overshadowing the courageous role played by Sri Lankan development and human rights organizations which have consistently spoken out on behalf of victims of injustice from all communities. Despite the government clamp-down, Goonetileke (1984) lists over 400 articles written on the 1983 riots. See also Rupesinghe and Verstappen (1989) whose annotated bibliography includes over 2,300 entries. An up-dated version was subsequently published by Rupesinghe, Verstappen and Philip in 1993. Identification of such vehicles and their license plate numbers is presented in various publications, for example Piyadasa (1984) and in reports of earlier

232 Notes

34.

35. 36. 37.

38.

39.

instances of anti-Tamil violence (MIRJE 1981: 19). Frequently, affidavits reporting abductions by security forces also include badge numbers, license plate numbers and other pieces of ‘firm evidence’ identifying those responsible. In Sri Lanka editions of Hoole et al. (1992), the government censor blacks out the license plate numbers of the Sri Lanka Transport vehicle used to transport rioters in the 1983 riots (p. 64). Tambiah describes the rioters which were mobilized. They were Sinhalese males, ‘virtually all drawn from Colombo and its suburbs’ (Tambiah 1992: 74). More specifically, they were ‘the urban working class, particularly those in government factories, the laborers, small businessmen and others employed in the congested bazaars and markets, secondary school students and recent dropouts, the urban underclass of unemployed and underemployed, the residents of shanty towns. A more detailed enumeration would include the following ‘occupational’ categories: wage workers in factory mills; transport workers such as bus drivers and conductors, workers in railway yards and electrical installations; petty traders and workers in markets, including fish mongers and market porters; small shopkeepers and salesmen in government corporations; hospital workers and attendants; high school students and students of technical institutes and tutories including recent school dropouts’ (Tambiah 1992: 74). Discussions by Tambiah (1992: 73) and Obeyesekere (1984: 160) point to the same conclusion. For full details see, Christian Worker, r 2nd and 3rd Quarters 1992 (November). Dissanayake subsequently split with the UNP in 1991 (after Premadasa replaced Jayewardene as president) to form the Democratic United National Front with another former UNP Minister Lalith Athulathmudali. The censorship guidelines are presented and discussed in Hyndman (1988: 31–34). The guidelines established on 2 August 1983 include: #5. No statements will be permitted on any subject by political parties or political personalities other than statements arranged for broadcast through State media; and #6. No comment will be permitted by any person on the present security or political situation. The continued government repression of the media and freedom of expression generally gave rise to the ‘Free Media Movement’ in early 1993. A reminder of the government censor’s presence is offered in the blacked-out sections of ‘politically sensitive’ books in Sri Lanka, for example, the passages in Hoole et al. (1992: 47, 64, 66) dealing with the 1983 riots. Commenting on the conduct of the armed forces during the riots, Gamini Dissanayake noted, ‘It was only when they realized that this was not a mere communal outburst that they thought of doing their duty of maintaining law and order’ (Daily News, 6 August 1983, quoted in Nissan 1984: 180).

8 1987 Indo-Sri Lanka Agreement 1. The judgment that the Agreement was unanticipated is reinforced by a review of the literature and media reports written at the time. For example, Premdas and Samarasinghe (1988: 677) write that ‘the Accord came as a sudden thunderbolt.’ This view is expressed by Professor Virginia Leary

Notes 233

2.

3.

4. 5.

6.

7.

8.

(who headed the 1981 International Commission of Jurists fact-finding Mission to Sri Lanka) in an address to the Paterson School of International Affairs, Carleton University (Canada) in October 1987. And it is the view of the author based on field work in South India from 1986 to 1987 while on a Shastri Indo-Canadian Institute Fellowship (Bush 1989). Recently, K.M. De Silva (1993) has claimed that the Agreement was the product of a negotiating process which began in 1983. This assessment is untenable for many reasons, most obviously because the issues and actors have changed since 1983, making the identification of a single negotiating process impossible. For example, in Clause 2.9 of the Agreement the Governments of India and Sri Lanka committed the Tamil paramilitary organizations to the following: ‘ … A cessation of hostilities will come into effect all over the island within 48 hours of the signing of this agreement. All arms presently held by militant groups will be surrendered in accordance with an agreed procedure to authorities to be designated by the Government of Sri Lanka. Consequent to the cessation of hostilities and the surrender of arms by militant groups, the army and other security personnel will be confined to barracks in camps as on May 25, 1987. The process of surrendering of arms and the confining the security personnel moving back to barracks (sic) shall be completed within 72 hours of the cessation of hostilities coming into effect.’ The main IPKF offensive was concentrated in the Jaffna peninsula where military confrontation with the LTTE took the form of conventional military engagements. Military confrontations on the East coast consisted mostly of hit-and-run operations by the LTTE. See for example, de Silva (1991) and Kodikara (1989). Including, inter alia, ballot box stuffing, impersonation, and omissions from the voters list. (‘Samarakone’ 1984). The monitoring of election violence and irregularities has been undertaken effectively by a local NGO, The Centre for Monitoring Election Violence (CMEV). For past reports, see: http://www.cpalanka.org The unprecedented UNP majority won in 1977 was largely the consequence of voters’ overwhelming rejection of Sirimavo Bandaranaike’s SLFP government which had extended its mandate through emergency legislation. The rejection of the SLFP was so thorough that the TULF became the official opposition party. Regarding the possible discrepancy in a ‘first-past-thepost’ electoral system between percentage of popular vote won by a Party and the portion of seats allocated to it, ‘Samarakone’ observes: ‘it was to avoid such extreme incongruities that the government had introduced proportional representation in 1978. It was to reproduce such incongruity that it set aside proportional representation in favour of a referendum in 1982’ [emphasis in original] (1984: 84). The continued anti-TULF policy of the LTTE and its ability to infiltrate into Colombo are seen in by the LTTE’s assassination of the top TULF leaders in July 1989, Appapilai Amirthalingam and V. Yogeswaran (FEER, 27 July 1989: 10–11), as well as Dr Neelan Thiruchelvam in July 1999. INFORM reports that in September 1993 the government promulgation of a coastal security zone ‘barred to civilians [from] north of Mannar to Trincomalee’ resulted in over 90,000 fisher families losing their sole means of livelihood (Sri Lanka Information Monitor, September 1993).

234 Notes 9. There are many references to such training bases: Gunaratna (1990a: 48–54) Tambiah (1986: 109); O’Ballance (1989: 19, 31, 42) India Today, 31 March 1984, 15 July 1987; South, March 1985; and Far Eastern Economic Review, 21 February 1985. An especially detailed discussion of the covert involvement by Indian actors is provided by Kadian (1990: 98–109): ‘[The Sri Lankan Tamil Militant groups] activities in India were broadly under the aegis of the Research and Analysis Wing (RAW) of the Cabinet Secretariat [India’s foreign intelligence agency], i.e. directly under the Indian Prime Minister’s office’ (p. 98). 10. While I was in Jaffna in January 1981, a Catholic priest explained to me that funding, training, and arms were being supplied to ‘the Boys’ (a generic term for Tamil militants) by the PLO, the Soviet Union, Libya, and the IRA. In a June 1992 interview in Sri Lanka, a leader in PLOTE claimed that he had received training in Lebanon in 1984 from the Popular Front for the Liberation of Palestine (PFLP). He also identified the Palestinian Liberation Organization (PLO) as a source of support for EROS as early as 1976–77. 11. The following list draws from Marino (1987: 5). 12. The context of this episode does not detract from the heinousness of the massacre but it does point out the role of government policies in adding to ethnic tensions. Rubin (1987: 112) explains: ‘These farms had been the property of a well-to-do Tamil who donated them in 1978 for the resettlement of Indian Tamil plantation laborers who had been driven from their homes by anti-Tamil violence in 1977. The government subsequently took over the farms, expelled the Tamils, and settled Sinhalese convicts and their families there in a rehabilitation scheme. Tamils perceived the scheme as part of a government effort to settle Sinhalese in predominantly Tamil areas in order to counteract Tamil demands for control of those areas.’ 13. O’Ballance (1989: 35) claims that: ‘the government admitted that, to avoid attacks and ambushes [in early 1984], Sinhalese policemen had all been withdrawn from outlying police stations in the Jaffna Peninsula area to the two main police stations at Jaffna and Kankesanturai.’ The ultimate result of this decision could only be the strengthening of paramilitary control of these areas – which came to be known locally as ‘Tiger country,’ after the LTTE became the most powerful Tamil paramilitary organization. 14. Marino (1987: 2–8) provides a concise chronology of violence and counterviolence in the period from 1983 to 1986. 15. A comprehensive study of Sri Lankan politics from 1983 to the present has yet to be written. 16. See Frontline, Vol. 4, No. 12 (13–26 June), 1987: 4–16. For further details on the extent of the Sri Lankan military offensive, see: ‘Four Tamils Killed,’ The Hindu, 22 January 1987: 9; ‘20 Killed in Fighting in Batticaloa Dt.,’ The Hindu, 29 January 1987: 1; ‘200 Civilians Killed, Says LTTE,’ The Hindu, 31 January 1987: 1; ‘Sri Lankan Planes Strafe Village in Mannar,’ The Hindu, 7 February 1987: 9; ‘Sri Lanka Armed Forces Launch Major Attacks,’ The Hindu, 7 February 1987: 9; ‘Massive Army Operations in Mannar,’ The Hindu, 8 February 1987: 2; ‘22 Militants Die as Battle Continues,’ The Hindu, 9 February 1987: 9; ‘Massacre in Mannar,’ The Hindu, 11 February 1987: 1; ‘50 Civilians Killed, Tamil Village Wiped Out,’ The Hindu, 12 February 1987: 9; ‘Tamils Facing Starvation, Says LTTE,’ The Hindu, 14 February 1987: 1;

Notes 235

17. 18.

19.

20.

21.

22. 23.

24.

25. 26.

‘Unacceptable Response from Colombo’ (editorial), The Hindu, 24 February 1987: 8; ‘Jaffna Operation Leaves 30 Dead,’ The Hindu, 9 March 1987: 1; ‘No Let Up in Sri Lankan Army Attack on Civilians,’ The Hindu, 11 March 1987: 1; ‘Gunships Attack School,’ The Hindu, 13 March 1987: 1; ‘Jaffna Civilians to Get Some Relief,’ The Hindu, 13 March 1987: 1; ‘20 Die in Jaffna Shelling,’ The Hindu, 30 March 1987: 1; ‘11 Patients Killed in Jaffna Shelling,’ The Hindu, 31 March 1987: 1; ‘Sri Lanka Planes Bomb Jaffna,’ The Hindu, 23 April 1987: 1; ‘Over 150 Tamils Killed as Bombing Continues,’ The Hindu, 24 April 1987: 1; ‘MPs Want Jaffna Bombing Condemned,’ The Hindu, 24 April 1987: 9; ‘Jaffna Bombed for Third Day,’ The Hindu, 25 April 1987: 1. Government Forces and the IPKF were also accused of using civilians as shields from enemy fire (Rubin 1987). Media reports at the time put the toll at 107 people (The Hindu, 18 April 1987: 1). O’Ballance (1989: 76) provides the figure of 127 deaths including 31 servicemen. He adds that another 60 were injured, ‘the majority being Sinhalese.’ According to de Silva (1993: 116), 130 people were ‘mowed down by automatic weapons.’ Parliamentarians were apparently unaware that their government was already actually involved in the conflict, training and assisting the Tamil Paramilitaries. If Pakistan provided similar aid in this fashion to the suffering Muslim residents of Kashmir, Government of India objections would have been heard around the world. This was the unanimous assessment of all military and paramilitary actors interviewed by the author in Sri Lanka. It was also the view of Lalith Athulathmudali, who was Minister of National Security at the time and a close advisor of Jayewardene. Personal interviews, Sri Lanka, May–July 1992. For a useful overview of the movement for an independent Tamil state of Dravidistan, see Gunewardene (1986: 125–144) and Irshick (1969). There had been gun fights in Tamil Nadu among rival Tamil paramilitary organizations throughout the 1980s. In late 1986, a rash of bombings in Tamil Nadu fuelled the fear that Indian Government-sanctioned materials and training to Sri Lankan Tamil militants might be acquired by South Indian separatists: General Post Office explosion, Madurai (11 December 1986); train explosion, Tiruchi (23 December 1986); railway explosion, Madras (29 December 1986); and the Wallajah Road explosion, Madras (1 January, 1987). Athulathmudali, the Minister of National Security at the time of the Agreement, claimed that India intervened precisely because Sri Lanka was ‘winning’ in the ‘Tamil War’ and that Indian pressure increased after the Army’s victory at Vadammarachichi (north-east of Jaffna). Athulathmudal claimed that at this point, the Indian High Commission in Colombo said to him: ‘Under no circumstances will we allow you to take Jaffna.’ Eelam People’s Democratic Party is a Tamil paramilitary group which splintered off from the EPRLF in 1987. ‘As the Indians waited patiently for the weapons, the Jaffna-based LTTE attacked two Eastern Province militant groups, [PLOTE] and [EPRLF], as

236 Notes

27.

28.

29.

30. 31.

32.

33.

leaders of the two groups were on their way to a ‘peace conference’ with the LTTE. A round of attacks and counter-attacks followed, leaving more than 150 dead. The conflict demonstrated the durability of a long-standing political rift between Jaffna Tamils and their Eastern Province counterparts who have long complained of the Jaffna penchant for trying to dominate Jaffna political affairs’ (Pfaffenberger 1988: 143). Sustained attacks on members of PLOTE and EPRLF caused an unspecified number of casualties. Similarly in Vavuniya, on 22 September, the LTTE killed about 18 PLOTE members in an attack on the latter’s camp (Rubin 1987: 45–6). O’Ballance (1989: 97) reports that 66 people died in Tamil paramilitary feuding ‘and over 70 Tamil separatists of various groups surrendered to the police for their own safety. In the face of such LTTE deadly hostility, many other members of the EPRLF, TELO and PLOTE either surrendered to the authorities with their arms, or disappeared deeper underground.’ On completion of basic paramilitary training all LTTE forces are ceremoniously presented with a cyanide capsule which is worn on a thong around their necks. They pledge to fight until the last bullet and then swallow the cyanide rather than be captured by the enemy. There have been questions raised about how the 17 LTTE captives retained their cyanide after their capture. Rubin (1987: 50) provides details: ‘In late August, in response to the release of several thousand Tamil detainees by the government, the LTTE freed the three policemen. One soldier, perhaps a Muslim, was subsequently let go, leaving eight Sinhalese soldiers in [LTTE] custody. The militants had offered to arrange visits by the family of the prisoners, had taken journalists to see them, and had promised to treat them well. Within hours of the suicides … however, the LTTE executed the prisoners and dumped their bodies at the Central Bus Stand in Jaffna.’ ‘… at nearly seven percent of the men who fought, the rate was almost twice as high as in the wars against Pakistan. … Sources in South Block confirm that the ratio of officers to men of other ranks killed in Sri Lanka is at an all time high’ (India Today, 31 January 1988, cited in Austin and Gupta 1988: 16). ‘It should be a short, sharp exercise and our boys should be back soon’ (Rajiv Gandhi, quoted in Far Eastern Economic Review, 31 August 1987). Initially, the IPKF successfully performed its peacekeeping task: ‘to halt and reduce the manifest violence of a conflict through the intervention of military forces in an interpository role’ (Harbottle 1979). With the commencement of offensive military confrontation to disarm the militants, its role ceased to be one of peacekeeping. As noted in Chapter Two, according to the 1981 Census, the ethnic composition of the Eastern Province was Sinhalese 24.8 per cent, Tamil (all subgroups) 41.57 per cent, and Muslim 31.5 per cent. Clause 2.2: ‘During the period which shall be considered an interim period (i.e.) from the date of elections to the Provincial Council, as specified in para. 2.8 to the date of the referendum as specified in para 2.3, the Northern and Eastern Provinces as now constituted, will form one administrative unit, having one elected Provincial Council. Such a unit will have one Governor, one Chief Minister and one Board of Ministers.’

Notes 237 34. Such tactics continue to be used by the LTTE. As in the Balkans, civilians are not simply caught in the crossfire, they have become targets. This is a view shared by UTHR – Jaffna (1991: 22) and USCR (1991: 2). 35. In Batticaloa, for example, it was reported that 79 per cent of the voters risked Tiger reprisals and went to the polls (The Economist, t 26 November 1988: 32). 36. The LTTE was not always the most powerful paramilitary. Until the mid1980s, no particular paramilitary was dominant. At that time, PLOTE lost the favor of the Indian government for publishing anti-Indian pamphlets and for attempting to smuggle a cargo of weapons into India without tacit government permission. It thus began to atrophy due to diminished Indian Government patronage. TELO was weakened by the murder of its leaders in the Welikade prison massacre in 1983 and further hampered by subsequent assassinations by the LTTE. Continued LTTE assassinations of paramilitary and political opponents ultimately established it as the foremost Tamil paramilitary organization. 37. For example, during the SAARC Conference in Bangalore in 1986, the Tiger leader, Prabhakaran, was flown from Madras in an Indian airforce plane for meetings with Indian officials who were simultaneously meeting with Jayewardene. It is also well-known that the late Chief Minister of Tamil Nadu, M.G. Ramachandran, enjoyed a close relationship with Prabhakaran. 38. ‘Though India is not formally backing this group, it is believed to be supported by the Indian intelligence agency known as RAW (the Research and Analysis Wing). RAW is keen not only to protect the Tamils who cooperated with India, but also to stop the Tigers from lording it over the northeast again … . According to the Sri Lankan government’s intelligence agency, the Tamil National Army’s men arrived at the camps [vacated by the IPKF] in Indian lorries and helicopters. They had been trained by Indians and carried Indian weapons’ (The Economist, t 11 November 1989: 71). For detailed discussions, see Gunaratna (1993) and Kadian (1990). 39. This is a view that was candidly shared by both military and paramilitary leaders. Personal interviews, Sri Lanka, May–July 1992. 40. For details of the LTTE–UNP discussions, see: The Economist, t 31 March 1990: 65; FEER, 25 May 1989: 28 and 29 June 1989: 27–8; and India Today, 15 May 1989: 88–9. 41. Following the IPKF pull-out, the LTTE attacked all Tamil competitors and the Sri Lankan security forces which had been confined to barracks as part of an agreement it had signed with the Government. 42. For details on the arms transfers, see A.K. Menon. ‘The Other Battle Field,’ India Today, 15 October 1991: 99; The Hindu, 6 September 1991: 1; Frontline, reprinted in Christian Worker, 2nd & 3rd Quarter 1991: xvi. 43. Lalith Athulathmudali, who led the impeachment attack against President Premadasa in Autumn 1991 alleged that the list of UNP supplies to the Tigers included: 4000 T56 automatic rifles (the Chinese equivalent to AK47s), 38 vehicles (Pajero Jeeps which had formerly been used by the IPKF), cement (though Athulathmudali admitted that he did not yet have documentary proof of this), communications equipment, and Rs90 million (of which the vouchers for Rs25 million had so far been traced). Personal interviews, Colombo, May–June 1992.

238 Notes 44. The first incident is reported by the UTHR – Jaffna (1991). It was also recounted to the author in the course of discussions with residents of Batticaloa. The murder of the policemen is reported in Amnesty International (1991). 45. The Agreement received strong support from the parties on the left – the United Socialist Alliance formed in 1987 by the Lanka Sama Samaja Party (LSSP), the CP, the Nawa Sama Samaj Party (NSSP) and the Sri Lanka Mahajana Pakshaya (SLMP). However, these parties were electorally marginal and had very limited popular support. Two small but prominent groups of intellectuals supported the Agreement as well – the MIRJE and the CRD. 46. As a means of ensuring full Party support, Jayewardene had each UNP MP sign an undated letter of resignation with the understanding that it could be accepted at any time at the President’s discretion. For a discussion of this and other measures employed to ensure obedience, see Matthews (1992). 47. See, for example, Gunaratna (1987: 18–26); O’Ballance (1989: 31, 61–63, 67, 77); and Hoole et al. (1992: 75–98). 48. For example, the April 1985 formation of the ENLF (Eelam National Liberation Front) which included Eelam Research Organizations of Students (EROS), TELO, and the EPRLF; and in May 1987 the formation of the ENDLF Eelam National Democratic Liberation Front (EWDLF) which joined the EPRLF and PLOTE. 49. From the early 1980s to the assassination of Rajiv Gandhi, there had been direct, though covert, military assistance to Tamil paramilitary organizations from the governments of Tamil Nadu and the Central Government in Delhi (personal interviews, Sri Lanka, May–July 1992; Hoole et al. 1992: 345–347; Rao 1989). This fact, and the ‘natural linkages’ across international boundaries (cultural, economic, historic), underscores the continuous international elements of ‘domestic’ politics in Sri Lanka.

9 2002 Ceasefire Agreement 1. This chapter draws on the field work and subsequent report prepared for DFID Sri Lanka in early 2002 (Bush 2002). 2. Perhaps most interestingly, in the process of conducting field work for this chapter, it was possible to encourage successfully the adoption of multileveled perspectives into the discussions and decisions of the international community as they sought to nurture and support peace before and after the signing of Ceasefire Agreement – suggesting the policy relevance of a two-level approach. (Bush 2002; Bush 2001a; Bush 1999) 3. Thus, for example, immediately prior to, and following, the signing of the Ceasefire Agreement, G.L. Pieris (Minister for Constitutional Affairs, and chief architect of the peace process) held a detailed briefing with the Press, the donor community, and the Mahanayaka, while the Prime Minister flew straight from the signing ceremony in the Vanni to Batticaloa to address the troops. Comments supportive of the Peace Process are reported regularly in the local media. In other words, the support of potential spoilers of the peace process was very specifically cultivated.

Notes 239 4. The bombing appeared to be in protest against the Sarvodaya Movement’s effort to organize a peace meditation in Anuradhapura (14 March 2002). 5. In two waves, LTTE strike teams penetrated the high-security complex at Katunayake at 3.30 a.m. and destroyed a total of 11 aircraft and damaged three. Three passenger aircraft – two A 330s and an A340 – of Sri Lankan Airlines and eight Air Force military aircraft - two Israeli built Kfirs, a Ukrainian MiG-27, two Mi-17 helicopter gunships and three-Chinese K8 advanced training aircraft – were destroyed. Two other passenger aircraft – an A320 and an A340 – were badly damaged. See ‘A Wake Up Call,’ Frontline, Vol 18: Issue 17, 18–31 August 2001. 6. According to an Apparel Industry Executive quoted in the USAID supported study by Price Waterhouse Coopers, Sri Lanka Competitiveness Study, September 1998. 7. Kelegama (1999) offers the following estimate for pro-government forces in 1996: Army – 129,000, Air Force – 17,000, Navy – 21,000, and Police – 68,000 for a total armed forces strength of 235,000. However, this figure does not include the tens of thousands of village militias (or ‘Home Guards’); nor the pro-government Tamil Paramilitary Organizations; nor the estimated 30,000–40,000 army deserters who may, or may not, have kept their weapons. Further, any calculation of island-wide armed labour force must include the LTTE, estimated to be 13,200 (40–60 per cent of whom are estimated to be under the age of eighteen, Sentinel South Asia 2000). Related to the size of the armed forces, is the bloated size of the military budget. In a recent and rigorous study on the costs of the war in Sri Lanka, it has been estimated that from 1982 to 1996 defence expenditure as a percentage of total government expenditure has increased from 3.1 to 21.6 per cent. As a percentage of the GDP, the defence budget has increased from 1.1 to 6.0 during the same period of time (Arunatilake, Jayasuriya, and Kelegama 2000; National Peace Council/Marga 2001). Thus, military expenditures dwarf combined expenditure on education and health, the consequences of which affect availability and access to basic social services by all Sri Lankans (especially children). See Bush 2000a for an overview of the impact of militarized violence on children in Sri Lanka – inside and outside of the conventionally defined ‘war zones.’ 8. The details in this paragraph are based on an interview with a member of the foreign diplomatic community who was present at G.L. Peiris’s briefing to the diplomatic community on 24 February 2002. 9. The details in this paragraph are based on an interview with a member of the foreign diplomatic community who was present at G.L. Peiris’s briefing to the diplomatic community on 24 February 2002. 10. Peace Support Group Press Release, 19 July 2002, ‘PSG Calls on all Parties to Strictly Abide by Ceasefire,’ Centre for Policy Alternatives, http://www.cpalanka.org 11. Though as discussed elsewhere, relations with Tamil populations in the South – particularly the Hill Country and Colombo – were also characterized by oppressive, sometimes abused, security measures. 12. Field Interviews, Batticaloa District, February 2002. 13. The details of this assault were widely reported in the media. See for example: Sambandan 2000.

240 Notes 14. ‘Taxes in the north and east: Tigers openly flout truce agreement,’ ColomboPage News Desk, Sri Lanka. 27 June 2002. http://www.colombopage.com/#Saturday64004 15. Field Interviews, Colombo, Kandy, Batticaloa, February and December 2002. 16. In a particularly heart-rending case, the father of two scholastically oriented teenaged boys in the East was approached by the LTTE to donate Rs 5million – despite years of donations and moderate support to the Eelam cause. When he pointed out that he was unable to pay even if he sold everything (as would have been clear from the bank statements the LTTE had in their hands), he was told to choose which son he would donate to the cause. When I left him, he was still unsure about what he would do. Field Interviews. Batticaloa, February 2002. 17. Father Harry Miller. Batticaloa. February 2002. 18. http://www.colombopage.com 9 July 2002. 19. Field Interviews, Colombo, February and December 2002. 20. ‘Diplomats based in Colombo are preparing to visit Jaffna,’ [ColomboPage News Desk, Sri Lanka] 2002 March 25: Colombo: Reports said a group of around 30 foreign diplomats based in Colombo is preparing to visit Jaffna on Thursday. The diplomats intend to have firsthand experience of the situation in the peninsula after the implementation of the ceasefire. The group is also expected to look into measures that needed to be taken to improve common amenities in Jaffna. These diplomats are from countries including the United States, the United Kingdom, Norway, France, Italy, India and Canada. 21. Most publications by University Teachers for Human Rights (Jaffna) include extensive details of LTTE killings. See UTHR (J) 1998. 22. According to the terms of the February Ceasefire, all of these groups were to be disarmed and disbanded. Art. 1.8: ‘Tamil paramilitary groups shall be disarmed by the GOSL by D-day 130 at the latest. The GOSL shall offer to integrate individuals in these units under the command and disciplinary structure of the GOSL armed forces for service away from the Northern and Eastern Province.’ This process was initiated within the time frame of the Agreement. 23. These specific villages were identified by families in the East as well as local and international development and humanitarian NGOs 24. While conducting interviews in the Eastern Province, I heard the stories of two boys who were pulled out of a van of a recruiter who was in the process of transporting them into an uncleared area; and the story of a young boy who was in hiding after having escaped from training camp and walked through the jungle back to his village. Both stories came from very credible sources. February 2002. The indicator that this is public knowledge is when it is reported on the front page of the New York Times, see 6 January 2003 A1, A6. 25. Discussions in this section rely on intensive discussions with Muslim community leaders in the Eastern Province in February 2002. 26. This includes the Muslim Home Guard armed by the Government of Sri Lanka ostensibly to protect Muslim villages from attacks by the LTTE. However, being poorly trained and equipped, they were easy targets for experienced fighters of the LTTE, and there are stories of their fleeing

Notes 241

27. 28. 29. 30. 31.

32.

attacks on Muslim civilians only to launch reprisal attacks on unarmed Tamil civilians. Another armed faction within the Muslim community is Jihad. This section draws directly on the excellent fact-finding mission report by the Centre for Policy Analysis (2002). http://www.colombopage.com 09 July 2002. http://www.colombopage.com 10 July 2002 Patricia Lawrence. Telephone interview. 1 July 2002 The JVP’s strength in the Parliament climbed from one to 16 MPs in the preceding seven years. In October 2000, it polled 5.99 per cent of the total valid votes – a figure which climbed to 9.10 for the 2001 Election. Whereas, the JVP polled 518,774 votes countrywide in 1999 elections, it received 815,353 in December 2001. The JVP achieved this by exploiting the PA’s political crises and internal divisions Field Interviews, Colombo, Kandy, Batticaloa. February and December 2002.

10 Fitting the Pieces Together 1. Some analysts employ the term ‘conflict settlement’ to ‘indicate the formal ending of armed hostilities and the renunciation of the use of force. They believe the objective of conflict resolution to be unattainable, on the grounds that conflicts over fundamental values and needs will never be resolved to the complete satisfaction of the parties involved. In these circumstances, the best one can hope for is a settlement which ends the violence and in some small measure takes those conflicting values and priorities into account’ (Hampson and Mandell 1990: 193). 2. Often nudging states towards becoming terroristic themselves. As the military strategist, Martin Van Crenfeld put it: ‘he who fights terrorists for any period of time is likely to become one himself’ (1990). 3. The term ‘protracted social conflict’ was coined by Edward Azar to refer to ‘situations which arise out of attempts to combat conditions of perceived victimization stemming from: (1) a denial of separate identities of parties involved in the political process; (2) an absence of security of culture and valued relationships; and (3) an absence of effective political participation through which victimization can be remedied’ (Azar 1986: 30). The conflict in Sri Lanka meet these criteria. 4. The term ‘conflict mediating’ is used self-consciously to refer to mechanisms that manage, resolve, or settle conflicts and disputes non-violently as they arise. These are distinct from ‘conflict dampening’ mechanisms which have as their primary objective the limitation of the level of conflict using whatever means are deemed most effective, whether violent or non-violent. For example, in South Africa the apartheid system was certainly successful in dampening anti-apartheid challenges, but it did so through direct and indirect mechanisms of violence – using everything from security forces, to urban planners, to structural violence of poverty, illness, and illiteracy. 5. N.B. These would be indicators of prevalence, not efficacy. 6. One of the strongest sentiments that came out of a number of interviews with Athulathmudali was his adamant belief that any ‘solution’ must begin

242 Notes

7.

8.

9.

10. 11.

12.

13. 14.

15. 16.

17.

with the ‘total military annihilation of the LTTE’ (personal interviews, Colombo, May–July 1992). As Foreign Minister in 1988, Ranjan Wijeratne told Western correspondents: ‘We have to deal with terrorists in the most ruthless manner … we have taken a hard line. There is no question about that.’ Acting in his capacity as deputy defence minister, Wijeratne announced a series of measures to deal with the growing JVP violence and control in the south: ‘We have given orders to shoot at sight, arrest, detain, or deal with inciters including trade union officials, strikers and all troublemakers attempting to disrupt normal life’ (Far Eastern Economic Review, 20 July 1989: 13). Just prior to the signing of the 1987 Indo-Sri Lanka Agreement, the rhetoric of Sinhalese political leaders became increasingly antagonistic, as illustrated in the speech in parliament by Prime Minister Premadasa on 24 April 1987: ‘When the lives of our people are in danger, we are not prepared to go in for a political solution … any friend who tells us to find a political solution will be considered as the biggest enemy’ (The Hindu, 25 April 1987: 1). On 7 October 2001, American and British aircraft began bombing Afghanistan. While the Bush regime may define the world in ‘post-9–11’ terms, the central point of reference for much of the rest of the world is ‘10–7’. The uncritical appropriation of ‘9–11’ terminology, is a mark of the success of the Bush regime’s domination of political discourse. This was further enshrined in the Constitutions of 1972 and 1978. Article Nine of the 1978 Constitution states: ‘The Republic of Sri Lanka shall give to Buddhism the foremost place and accordingly it shall be the duty of the state to protect and foster the Buddha Sasana.’ ‘Ethnicization’ is used to refer to the process by which the potential autonomy of state actors is increasingly harnessed in the pursuit of the particularistic interests of a specific ethnic sub-group. Warnapala and Woodsworth specifically address this in their 1987 monograph, ‘Welfare as Politics in Sri Lanka.’ In interviews in December 2002 in Batticaloa with individuals who were being intimidated and extorted by the LTTE, I was particularly struck by the aggression shown towards one Tamil development worker because he had married a foreign woman, as well as the frequent challenges to a victim’s ‘Tamil-ness’ if they did not support the LTTE – i.e. cough up contributions, in cash or in kind. A violation of what is otherwise known as the Law of Equifinality. Although the number of people displaced by violence in Sri Lanka increased dramatically after the 1983 riots, the intimidation of Sinhalese out of the North began in the late 1970s. Some Sinhalese ethnic mobilizers have attempted to justify the 1983 riots as retaliation for the dislocation of the Sinhalese from the North. For example: Centre for Policy Alternatives; the Consortium of Humanitarian Agencies; Law and Society Trust; Sarvodaya; the National Peace Council; Peace and Community Action; Family Rehabilitation Centre; Satyodaya, INFORM, Lawyers for Human Rights and Development; Organizations for the Parents and Families of the Disappeared (OPFMD); Centre for the Study of Human Rights; the International Centre for Ethnic Studies; INPAC; the Batticaloa Peace Committee, the Mothers Front, the

Notes 243

18.

19. 20.

21. 22.

23.

24.

Movement for Inter-Racial Justice and Equality, the Civil Rights Movement, and the Centre for Society and Religion. These NGOs include workers from all ethnic communities. As happened to the Christian Children’s Fund of Canada (CCF) which was accused of attempting to send supplies to the LTTE in the North. Accusations in the government-run newspapers persisted even after the shipment in question was found to contain nothing suspicious or incriminating. Death threats and intimidation were directed to the head of CCF in Colombo. See The Ottawa Citizen, 24 November 1991: B6. The definition of an NGO by the government was so vague that it could include any actor or agency not formally associated with the government. Médecins Sans Frontieres (MSF) and the International Committee of the Red Cross (ICRC) have consistently raised concern over the shortage of medicine in the Vanni due to blocking by the Sri Lanka army, and in early May 2000, the surgery of the Mallavi hospital and two branches of the Mullaithivu hospital were particularly hampered by a severe shortage of medical supplies in the Vanni. Field Interviews. Colombo, Batticaloa, Kandy, 17–28 February 2002. Examples of the former are listed in Jayasuriya (2000) and Warnapala and Woodsworth (1987). Examples of the latter are found in any Human Rights report which includes a discussion of conditions in LTTE controlled areas – for example publications and reports by University Teachers for Human Rights (Jaffna). A recent but classic example is offered in local media reports of 17 March 2002 which noted the following: ‘LTTE leader V. Prabhakaran has issued a strict warning to his followers that any violators of the truce signed with the government would be punished severely. He has told political leader Thamil Selvan that he would not hesitate to punish any cadres violating the ceasefire, even with death. Selvan told journalists who met him in Wanni over the weekend that the leader was very clear on his instructions.’ www.lankapage.com In the current context, I am referring to the militarization of relations between government actors and Tamil sub-groups. During the 1987–90 JVP resurgence, relations between the government and Sinhalese sub-groups were similarly militarized. A Sri Lanka development worker in Colombo reported that the experience of government and JVP violence by Sinhalese civilians in Sinhalese majority areas has softened some attitudes towards the plight of innocent Tamil civilians caught in the cross-fire in the North and East (Personal interviews, Colombo, May–July 1992). The British Refugee Council reports that ‘police units regularly extort up to Rs 30,000 in bribes to release suspects. [Those] too poor to pay, stay in jail’ (Sri Lanka Monitor, February 1993: 3). The same system is employed by the LTTE. While conducting interviews on the East Coast, I met with a businessman who had been released from LTTE custody in order to raise the rest of his ransom (personal interviews, May–July 1992). From Fall 2001 to early 2002, there was a considerable increase in the scale and scope kidnappings: scale in terms of the excessively exorbitant demands for ransom money and scope in terms of the number and status of those being kidnapped – some of the most prominent businessmen and community leaders have been specifically targeted and affected. According to a human rights worker who

244 Notes

25.

26.

27.

28. 29. 30. 31.

32.

approached the military authorities concerning the kidnappings s/he was told that there was nothing they could do, because they feared destabilizing the security environment during the precarious ceasefire. One community leader in the East suggested that many of these extortionist activities might have been the actions of LTTE individuals rather than the organization as a whole. The motive behind them might may have been to collect the resources necessary to allow them to leave the country before the next phase of the military-political game. A kind of pension scheme that fits the modus operandi of the organizational culture of the LTTE and paramilitary organizations generally. There is a danger that this may well contribute to producing the opposite outcome than was intended: increased support for groups which use terror in the pursuit of political ends. The British Refugee Council offers a more conservative estimate. In a report published at the time, it estimated that ‘in the last eight months of 1993, Sri Lankan authorities arrested 8,000 Tamils in Colombo, of whom an estimated 1,000 to 2,000 remained in detention in December [1993] … [Y]oung men without jobs or relatives in the capital and who do not possess National Identity cards, which Tamils find difficult to obtain, are particularly vulnerable … [A]ccording to sources in Colombo, police there ‘run flourishing extortion rackets,’ demanding large sums of money from relatives seeking to secure detainees release’ (USCR 1994: 6). See Bush (1993), Abeysekere (1992), and Pieris and Marecek (1992) for examples of the Muslim Home Guard involvement in retaliatory massacres of Tamil villagers following attacks by the LTTE. http://www.hrw.org/wr2k/Asia-08.htm It should be noted that, in any given episode, a particular mediating factor or set of factors may be more or less salient. For example, the rise of S.W.R.D. Bandaranaike, and the mobilizational tactics of the JVP in both 1971 and 1987. For example: the material benefits acquired by the LTTE through its imposition of a ‘taxation’ and ‘tariff’ scheme in the Northern Province of Sri Lanka and the grey and markets inside and outside the war zones as a result of restricted access and the prohibition of goods. For discussion of ‘war weariness’ and ‘hurting stalemate,’ Zartman (1985); on collective learning, see Nye (1987); and on ‘re-perceptualization,’ see Keashly and Fisher (1990).

Bibliography Abeysekera, Sunila (1992). ‘Report on the Massacres in Karapola, Muthugala and Alanchipothana,’ unpublished report, INFORM, Colombo, 15 May 1992. Ali, Ameer (1992). ‘Sri Lanka’s Ethnic War: The Muslim Dimension,’ Pravada, 1: 11, 5–8. Allardt, Erik (1979). ‘Implications of the Ethnic Revival in Modern Society,’ unpublished paper, Woodrow Wilson International Centre for Scholars, 22 May 1979. Alles, A.C. (1977). Insurgency 1971. Colombo: The Apothecaries Company. Allison, Graham (1969). ‘Conceptual Models and the Cuban Missile Crisis,’ American Political Science Review, LXIII: 3, 689–718. Allison, Graham (1971). The Essence of Decision. Boston: Little, Brown. Amnesty International (1983). ‘Report of an Amnesty International Mission to Sri Lanka: 31 January to 9 February 1982,’ AI Index: ASA 37/01/83, Amnesty International, London. Amnesty International (1986). ‘Sri Lanka “Disappearences” Campaign,’ AI Index: ASA 37/14/86, Amnesty International, London. Amnesty International (1991). ‘Sri Lanka – The North-East: Human Rights Violations in a Context of Armed Conflict,’ AI Index: ASA 37/14/91, Amnesty International, London. Amnesty International (1992). ‘Sri Lanka: Deliberate Killings of Muslim and Tamil Villagers in Polonnaruwa,’ AI Index: ASA 37/10/92, Amnesty International, London. Amnesty International (1993). ‘Sri Lanka: An Assessment of the Human Rights Situation,’ AI Index: ASA 37/1/93, Amnesty International, London. Amnesty International (1994). ‘Sri Lanka: Summary of Human Rights Concerns,’ AI Index: ASA 37/09/94, Amnesty International, London. Amnesty International (1994a). ‘Sri Lanka, Balancing Human Rights and Security: Abuse of Arrest and Detention Powers in Colombo,’ AI Index: ASA 37/10/94, Amnesty International, London. Amnesty International (1999). Sri Lanka. Torture in Custody. Amnesty International Report – ASA 37/10/99. June 1999. http://www.amnesty.org/ ailib/aipub/1999/ASA/33701099.htm Anderson, Bendict (1983). Imagined Communities. London: Verso. Anderson, Mary (1999). Do No Harm: How Aid Can Support Peace – or War. Boulder and London: Lynne Rienner Publishers. Anderson, Scott (1994). ‘Making a Killing: The High Costs of Peace in Northern Ireland,’ Harper’s, 288: 1725 (February), pp. 45–55. Arunatilake, Nisha, Sisira Jayasuriya, and Saman Kelegama (2000). The Economic Cost of the War in Sri Lanka, Research Studies: Macroeconomic Policy and Planning Series No. 13. Colombo: Institute of Policy Studies, January. Asia Foundation (1999). ‘Effective Approaches to Law and Human Rights,’ Emerging Issues in Asia. Asia Foundation Office of External Relations. 245

246 Bibliography Austin, Dennis and Anirudha Gupta (1988). ‘Lions and Tigers: The Crisis in Sri Lanka,’ Conflict Studies # 211, London: Centre for Security and Conflict Studies. Bandara, Jayatilleke S. (1999). ‘The Impact of the Civil War on Tourism and the Regional Economy,’ in Siri Gamage and I.B. Watson (eds), Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop, pp. 241–52. Banks, Michael (1971). ‘Caste in Jaffna,’ in E.R. Leach (ed.), Aspects of Caste in South India, Ceylon, and North-West Pakistan. Cambridge: Cambridge University Press, 61–77. Bastian, Sunil (1993). ‘Political Economy of Ethnic Violence in Sri Lanka: The July 1983 Riots,’ Pravada, 2: 6, 18–23. Also printed in Veena Das (ed.), Mirrors of Violence: Communities, Riots and Survivors in South Asia. Delhi: Oxford University Press, 1990, 286–304. Bechert, Heinz (1978). ‘S.W.R.D. Bandaranaike and the Legitimation of Power through Buddhist Ideals,’ in Bardwell Smith (ed.), Religion and the Legitimation of Power in Sri Lanka. Chambersberg, PA.: ANIMA Books, 199–211. Bechert, Heinz (1978a). ‘The Beginnings of Buddhist Historiography: Mahavamsa and Political Thinking,’ in Bardwell Smith (ed.), Religion and the Legitimation of Power in Sri Lanka. Chambersberg, PA.: ANIMA Books, 1–12. Berdal, Mats and David M. Malone (2000) (eds). Greed and Grievance: Economic Agendas in Civil Wars. Boulder and London: Lynne Rienner. Brass, Paul (1991). Ethnicity and Nationalism: Theory and Comparison. New Delhi: Sage. Brass, Paul (1984). (ed.), Ethnic Groups and the State. London: Croom Helm. Brown, David (1989a). ‘Ethnic Revival: Perspectives on State and Society,’ Third World Quarterly, 11: 4. Brown, David (1996). (ed). The International Dimensions of Internal Conflict. Cambridge and London: MIT Press. Brown, David (1989b). ‘The State of Ethnicity and the Ethnicity of the State: Ethnic Politics in Southeast Asia,’ Ethnic and Racial Studies, 12: 1. Bush, Kenneth (2003). ‘The Commodification, Compartmentalization, and Militarization of Peacebuilding,’ in Andy Knight and Tom Keeting (eds), Peacebuilding in Post-Conflict Societies, Edmonton: University of Alberta Press. Bush (2002). From Putty to Stone: A Report of a Mission Investigating Human Rights Programming in Sri Lanka. Colombo: DFID Sri Lanka. Bush (2002a). ‘Participatory Assessment and Analysis of Democracy and Governance in Ethiopia,’ Concept Paper prepared for the Ethiopia Bilateral Country Programme of the Canadian International Development Agency with (Calvin Piggott), January. Bush (2001) ‘Peace and Conflict Impact Assessment (PCIA) Five Years On: The Commodification of an Idea,’ The Berghof Handbook for Conflict Transformation (Berlin: Berghof Research Center for Constructive Conflict Management). html www.berghof-center.org/handbook/bush/ Bush (2001a) ‘Peace and Conflict Impact Assessement (PCIA) of Swedish Development Cooperation With Sri Lanka,’ (Stockholm & Colombo: SIDA, March). Bush (2000). ‘Polio, War and Peace’ Bulletin of the World Health Organization: The International Journal of Public Health. 78 (3). http://www.who.int/bulletin/

Bibliography 247 Bush (2000a). ‘Stolen Childhood: The Impact of Militarized Violence on Children in Sri Lanka,’ Paper presented at the International Conference on War-Affected Children, Winnipeg, Canada, 10–17 September. Bush (2000b). ‘The Peace and Conflict Impact of Overseas Development Assistance in Sri Lanka,’ Sandra Maclea, Fahimul Quadir & Timothy M. Shaw (eds), Crises of Governance in Asia and Africa. (Aldershot, England: Ashgate, 2000), pp. 261–78. Bush (1999). The Limits and Scope for the Use of Development Assistance Incentives and Disincentives for Influencing Conflict Situations: Case Study – Sri Lanka, OECD, Paris: DAC Informal Task Force on Conflict, Peace and Development Cooperation (August). Bush (1998). A Measure of Peace: Peace and Conflict Impact Assessment (PCIA) of Development Projects in Conflict Zones, Ottawa: International Development Research Centre. http://www.idrc.ca/peace/p1/working_paper1.html Bush (1997) ‘Identity-Based Conflict: Rethinking Security in a Post-Cold War World,’ (w. Fuat Keyman) Global Governance, 3: 3 ( autumn ). Bush (1996). ‘Beyond Bungee Cord Humanitarianism: Towards a Developmental Approach to Peacebuilding’, Canadian Journal of Development Studies, Special Issue, December. Bush (1996). ‘Rocks and Hard Places, Human Rights Abuse, Bad Governance, and Population Displacement’, Canadian Foreign Policy, 4 (1), pp. 49–83. Bush, (1993). ‘Reading Between the Lines: Intra-Group Heterogeneity and Conflict in Sri Lanka,’ Refuge, 13: 3, pp. 15–22. Bush, Kenneth and Diana Saltarelli (2000) (eds). The Two Faces of Education in Ethnic Conflicts: Towards a Peacebuilding Education for Children. Florence, Italy: UNICEF Innocenti Research Centre. www.unicef-icdc.org/pdf/twofaces.pdf Butterfly Garden (1999). Blood of the Mango and Other Tales from the Butterfly Garden of Batticaloa, 3rd reprint (Sri Lanka: The Butterfly Garden). By the Children and Animators of the Butterfly Garden with illustrations by Thuraisamy Naguleshwaren, as collected and re-told in English by Paul Hogan. Canadian Human Rights Mission to Sri Lanka (1992). ‘Report,’ Toronto. CARE / Sri Lanka / Ravano, J. (2001). ‘The Economy of the Wanni: An Analysis of the Impact of Restrictions’. Colombo: CARE International – Sri Lanka, Draft Only, 2001. CARE / Sri Lanka, S. / Meyer, J. (2000b). ‘Household Livelihood Security Assessment in LTTE-controlled Areas of the Wanni’. Colombo: CARE Sri Lanka, Assessment Report, October 2000. CARE / UK, I. (2000c). ‘Vulnerable Groups Organized in Conflict Areas (VOICE)–Project Proposal’. London: CARE International UK. Casparsz, Paul (2002). ‘Human Rights in the Plantation Sector,’ Draft msc. CHA /Consortium of Humanitarian Agencies (2000). CHA Newletter, Vol. IV, Issue II, (March/April). CHA (2000a). ‘The Framework for Relief, Rehabilitation and Reconciliation – Sectoral Workshop Reports’. Colombo: Consortium of Humanitarian Agencies, National Peace Council, assisted by the WB, UNDP, UNHCR, the Dutch Government, CIDA on behalf of the Sri Lankan Government, Final. CHA (2000b). CHA Newletter, Vol. IV, Issue I, (January/February). CHA (1999). Consortium of Humanitarian Agencies Yearbook 1998 (Colombo: CHA).

248 Bibliography CHA (1998). Yearbook 1997: Sri Lanka, Consortium of Humanitarian Agencies, Colombo. CHA (1998a). ‘Important Commitments on the Protection of Children Made by Government and LTTE in Sri Lanka Arising from the Visit of Mr. Olara Otunnu’, CHA Newsletter, II: 8, May–June 1998. Chalk, Peter (2000). ‘Liberation Tigers of Tamil Eelam’s (LTTE) International Organization And Operations – A Preliminary Analysis,’ COMMENTARY No. 77, a Canadian Security Intelligence Service (CSIS) publication. Chandraprema, C.A. (1991). Sri Lanka: The Years of Terror. The JVP Insurrection, 1987–1989. Colombo: Lake House Bookshop. Chandraprema C.A. (1989). ‘Putschism, Ethnic Chauvinism and Social Revolution: Reflections on the Sri Lankan Experience,’ Colombo: Independent Students Union, (mimeo). Chase, Robbie (2000). The Butterfly Garden. Batticaloa, Sri Lanka. Colombo: Vishva Lekha press. Chase, A. Doney, S. Sivayogan, V. Ariyaratne, p. Satkunanayagam and A. Swaminathan (1999). ‘Mental Health Initiatives as Peace Initiatives in Sri Lankan Schoolchildren Affected by Armed Conflict,’ Medicine, Conflict and Survival, Vol. 15, pp. 379–90. Children as Zones of Peace (1998). ‘A Call for Action: Promoting and Protecting the Rights of Children Affected by Armed Conflict in Sri Lanka,’ A Summary. (n.p.) May 1998. Clare, Paul K. (1990). ‘Subcultural Obstacles to Control of Racketeering in Northern Ireland,’ Conflict Quarterly, X: 4, 25–50. CPA/Centre for Policy Analysis (2002). Report of the Fact-Finding Mission to Valaichchenai and Batticaloa. 24 July 2002. http://www.cpalanka.org CPA, CHA, LST (2001). ‘Internally Displaced Persons and the Human Rights Commission,’ A study undertaken by the Centre for Policy Alternatives (CPA), the Consortium of Humanitarian Agencies (CHA), and the Law and Society Trust on Behalf of the Human Rights Commission of Sri Lanka. Supported and Facilitated by the Asia Foundation. Colombo, August. CRD/Committee for Rational Development (1984). Sri Lanka, The Ethnic Conflict: Myths, Realities and Perspectives. New Delhi: NAVRANG. Cuny, Fred and Christopher Cuny, (1992). ‘The Return of Tamil Refugees from Sri Lanka, 1983–1989’, in Fredrick Cuny, Barry N. Stein and Pat Reid (eds), Repatriation During Conflict in Africa and Asia, Centre for the Study of Societies in Crisis, Texas. de Silva, C.R (1977). ‘Education,’ in K.M. De Silva (ed.), Sri Lanka – A Survey. Honolulu: University of Hawaii Press. de Silva, C.R. (1979). ‘The impact of Nationalism on Education,’ in M. Roberts (ed.), Collective Identities, Nationalisms and Protest in Modern Sri Lanka. Colombo: Marga Institute. De Silva, Chandra R. (1999). ‘The Role of Education in Ameliorating Political Violence in Sri Lanka,’ in R.I. Rotberg (ed.), Creating Peace in Sri Lanka. Washington, DC.: Brookings Institution Press, 109–30. de Silva, K.M (1988). ‘Sri Lanka’s Muslim Minority’ in de Silva, Pensri Duke, Ellen Goldberg and Nathan Katz (eds), Ethnic Conflict in Buddhist Societies: Sri Lanka, Thailand and Burma. Boulder, Colorado: Westview Press, 202–14.

Bibliography 249 de Silva, K.M (1982). ‘The Model Colony: Reflections on the Transfer of Power in Sri Lanka,’ in A.J. Wilson and Dennis Dalton (eds), The States of South Asia: Problems of National Integration. London: C. Hurst & Company, pp. 77–88. de Silva, K.M (1986). Managing Ethnic Tensions in Multi-Ethnic Societies: Sri Lanka, 1880–1985. Lanham, MD: University Press of America. de Silva, K.M and R.J. May (1991) (eds). The Internationalization of Ethnic Conflict. London: Pinter Publishers. de Silva, K.M and S.W.R. de A. Samarasinghe (1993) (eds). Peace Accords and Ethnic Conflicts. London and New York: Pinter Publishers. de Silva, K.M. (1993). ‘The Making of the Indo-Sri Lanka Accord: The Final Phase – June–July 1987,’ in K.M. de Silva and S.W.R. de A. Samarasinghe (eds), Peace Accords and Ethnic Conflicts. London and New York: Pinter Publishers, 112–55. Demusz, K. (2000). ‘Listening to the Displaced – Action Research in the Conflict Zones of Sri Lanka.’ Oxford: Oxfam GB, Oxfam Working Paper. Dewaraja, L.S. (1971). The Kandyan Kingdom, 1707–1760. Colombo: Lakehouse Publishers. DFID/Department for International Development (1999): Sri Lanka Country Strategy Papers, London. Dharmadasa, K.N.O. (1992). Language, Religion, and Ethnic Assertiveness: The Growth of Sinhalese Nationalism in Sri Lanka. Ann Arbor, Michigan: University of Michigan Press. Dharmadasa, K.N.O. (1990). ‘The People of the Lion: Ethnic Identity, Ideology and Historical Revisionism in Contemporary Sri Lanka,’ in Spencer (ed.), Sri Lanka: History and Roots of the Conflict. London and New York: Routledge. Dhume, Sadanand (2000). ‘The Cost of War,’ Far Eastern Economic Review, 1 June. Druckman, Daniel (1978). ‘Boundary-Role Conflict: Negotiation as Dual Responsiveness,’ in I. William Zartman (ed.), The Negotiation Process: Theories and Applications. Beverly Hills: Sage. Dubeau, Michilynn (1993). ‘Political Representation of Estate Workers in Sri Lanka: Social Mobilization as a Challenge to the Leadership of S. Thondaman of the CWC,’ in Louis E. French et al. (eds), South Asian Symposium 1992: A Reader in South Asian Studies. Toronto: Centre for South Asian Studies Graduate Students’ Union, 13–24. Dunham, David and Sisira Jayasuriya (1998). ‘Is All so Well With the Economy and With the Rural Poor?’ Pravada, Vol. 5, Nos. 10 &11. Edirisuriya, P. (1999). ‘Determinants of Military Expenditure in Sri Lanka,’ in Siri Gamage and I.B. Watson (eds), Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop, 229–39. Esman, Milton (1990). ‘Political and Psychological Factors in Ethnic Conflict,’ in Joseph V. Montville (ed.), Conflict and Peacemaking in Multiethnic Societies. Lexington and Toronto: Lexington Press, 53–64. Esman, Milton (1994). Ethnic Politics, Ithaca and London: Cornell University Press. Evans, Peter, Dietrich Rueschemeyer and Theda Skocpol (1985) (eds). Bringing the State Back In. Cambridge: Cambridge University Press. FEER /Far Eastern Economic Review, (1985). 21 February 1985.

250 Bibliography Feldman, Allen (1991). Formations of Violence: The Narrative of the Body and Political Terror in Northern Ireland. Chicago and London: University of Chicago Press. Frerks, G./Van Leeuwen, M. (2000): Conflict Policy Research Project (CPRP): The Netherlands and Sri Lanka – Dutch Policies and Interventions with regard to the Conflict in Sri Lanka, The Hague, Netherlands Institute for International Relations Clingendael, First Draft, March 2000. Frerks, Georg and Mathijs van Leeuwen (2000). The Netherlands and Sri Lanka. Dutch Policies and Interventions with Regard to Conflict in Sri Lanka. Conflict Policy Research Project, Conflict Research Unit, Netherlands Institute of International Relations ‘Clinendael,’ the Hague, Netherlands. Gamage Siri and I.B. Watson (1999) (eds). Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop. Geertz, Clifford (1973). ‘Thick Description: Towards an Interpretive Theory of Culture,’ in his The Interpretation of Culture. New York: Basic Books, 3–30. Gillies, David, (1992). ‘Principled Intervention: Canadian Aid, Human Rights and the Sri Lankan Conflict’ in Robert Miller (ed.), Aid as Peacemaker: Canadian Development Assistance and Third World Conflict, Carleton University Press, Ottawa: pp. 33–50. Goodhand, J. (2000). ‘Aid, Conflict and Peacebuilding in Sri Lanka.’ London, DFID, Final, November 2000. Goonaratne, Y. (1986). Relative Merits: A Personal Memoire of the Bandaranaike Family of Sri Lanka. London: C. Hurst & Company. Goonetileke, Ian (1984). ‘July 1983 and the National Question of Sri Lanka – A Bibliography,’ in the Committee for Rational Development, Sri Lanka, The Ethnic Conflict: Myths, Realities and Perspectives. New Delhi: NAVRANG, 243–77. Greenwald, Alice (1978). ‘The Relic on the Spear: Historiography and the Saga of Duttagamani,” in Bardwell Smith (ed.), Religion and the Legitimation of Power in Sri Lanka. Chambersberg, PA.: ANIMA Books, 13–35. GSL/Government of Sri Lanka (2001). National Framework for Relief, Rehabilitation and Reconciliation in Sri Lanka. Draft. Colombo. November. GTZ and RRAN (1999). Rebuilding Lives: The Jaffna Rehabilitation Project, Sri Lanka Colombo: Resettlement and Rehabilitation Authority and GTZ. Gunaratna, Rohan (2000). ‘Suicide Terrorism: A Global Threat,’ Janes on-line, 20 October 2000. Gunaratna, Rohan (1998). ‘Childhood – A continuous Casualty of the Conflict in Sri Lanka,’ Jane’s Intelligence Review, July. Gunaratna (1998a). ‘LTTE Child Combatants,’ Jane’s Intelligence Review, July. Gunaratna, Rohan (1993). Indian Intervention in Sri Lanka, The Role of India’s Intelligence Agencies. Colombo: South Asia Network on Conflict Research. Gunaratna, Rohan (1990). Sri Lanka, A Lost Revolution: The Inside Story of the JVP. Kandy: Institute of Fundamental Studies. Gunaratna, Rohan (1990a). War and Peace in Sri Lanka. Kandy: Institute of Fundamental Studies. Gunasekera, Rohan (1988). ‘Sri Lankan Armed Forces Switch Focus,’ International Defense Review, 21: 5, 509–12.

Bibliography 251 Gunawardena, R.A.L.H. (1985). ‘The People of the Lion: Sinhala Consciousness in History and Historiography,’ in Ethnicity and Social Change in Sri Lanka. Colombo: Navamaga. Gunawardene, Chandra (1975). ‘Ethnic Representation, Regional Imbalance, and University Admissions in Sri Lanka,’ Comparative Education, 11: 1. Gunawardene, Victor (1986). ‘Impact of Internal Ethnic Conflicts on the Region,’ in The Indian Ocean as a Zone of Peace, International Peace Academy Report #25, International Peace Academy, New York, 125–44. Gurr, Robert Ted (1993). Minorities at Risk: A Global View of Ethno-political Conflict Washington, DC: United States Institute of Peace. Gutman, Roy, and David Rieff (1999) (eds). Crimes of War: What the Public Should Know. New York and London: W.W. Norton & Co. Hampson, Fen Osler and Brian Mandell (1990). ‘Managing Regional Conflict: Security Cooperation and Third Party Mediators,’ International Journal, XLV: 2. Harbottle, M. (1979). ‘The Strategy of Third Party Intervention in Conflict Resolution,’ International Journal, 25: 1. Hechter, Michael (1986). ‘Theories of Ethnic Relations,’ in John F. Stack (ed). The Primordial Challenge: Ethnicity in the Contemporary World. Westport, Conn.: Greenwood Press. Heisler, Martin O. (1990). ‘Ethnicity and Ethnic Relations in the Modern West,’ in Joseph V. Montville (ed). Conflict and Peacemaking in Multiethnic Societies. Lexington and Toronto: Lexington Press, 21–52. Hellman-Rajanayagam (1989). ‘The Tamil Militants – Before the Accord and After,’ Pacific Affairs, 61: 4. Hellman-Rajanayagam, Dagmar (1994). The Tamil Tigers: Armed Stuggle for Identity. Stuttgart: Franz Steiner Verlag. Herring Ronald J. (1998). ‘Carrots, Sticks and Ethnic Conflict: Explaining Civil War in Sri Lanka,’ Draft, Cornell University, November. Hogan, Paul (2000). Play, Art, Ritual and Healing at the Butterfly Garden of Batticaloa: Presence in Poesia as the Practice of Peace. Unpublished manuscript. Hollup, Oddvar (1992). ‘Ethnic Identity, Violence and the Estate Tamil Minority in Sri Lanka,’ Round Table, 323, pp. 315-38. Homer-Dixon, Thomas (1991). ‘On the Threshold: Environmental Changes as Causes of Acute Conflict,’ International Security, (autumn), 16: 2, 76–116. Hoole, Rajan D. Somasundaram, K. Sritharan, Rajani Thiranagama (1992). The Broken Palmyra, The Crisis in Sri Lanka: An Inside Account, t 2nd reprint. Claremont, CA: The Sri Lanka Studies Institute. Horowitz, Donald L. (1991). ‘Ethnic and Nationalist Conflict,’ in Michael T. Klare and Daniel C. Thomas (eds), World Security: Trends and Challenges at Century’s End. New York: St. Martin’s Press, 225–36. Horowitz (1990). ‘Ethnic Conflict Management for Policy Makers,’ in Joseph V. Montville (ed.), Conflict and Peacemaking in Multiethnic Societies. Lexington and Toronto: Lexington Press, 115–30. Horowitz (1990). ‘Making Moderation Pay: The Comparative Politics of Ethnic Conflict Management’ in Joseph V. Montville (ed.), Conflict and Peacemaking in Multiethnic Societies. Lexington and Toronto: Lexington Press, 451–76. Horowitz (1985). Ethnic Groups in Conflict. Berkeley, Los Angeles, London: University of California Press.

252 Bibliography Horowitz, Donald L (1980). Coup Theories and Officers’ Motives: Sri Lanka in Comparative Perspective. Princeton: Princeton University Press. Horowitz, Donald L (1985). Ethnic Groups in Conflict. Berkeley, Los Angeles, London: University of California Press. Horowitz, Donald L (1973). ‘Direct, Displaced and Cumulative Ethnic Aggression,’ Comparative Politics, 6. Hull, Harry F. (1999). ‘Fighting Stops for Polio Immunization,’ Updated, 14 December 1999. http://www.who.int/inf/polio.html Human Rights Watch World Report 2000, ‘Human Rights Developments in Sri Lanka,’ http://www.hrw.org/wr2k/Asia-08.htm Hyndman, Patricia (1988). Sri Lanka: Serendipity Under Siege. Nottingham: Spokesman. Hyndman, Patricia (1987). ‘The 1951 Convention Definition of Refugee: An Appraisal with Particular Reference to the Case of Sri Lankan Tamil Applicants,’ Human Rights Quarterly, 9: 1, 49–74. Hyndman, Patricia (1985). Democracy in Peril, Sri Lanka: A Country in Crisis. Report to the Lawasia Human Rights Standing Committee, unpublished manuscript. ICES/International Centre for Ethnic Studies (2000). ‘Electoral Corruption in Sri Lanka, Interim Report,’ Studies on Governance and Democracy in South Asia, International Centre for Ethnic Studies, Kandy, Sri Lanka. September. INFORM various years. Sri Lanka Information Monitor Situation Report, t INFORM Colombo. International Alert (1989). Political Killings in Southern Sri Lanka: On the Brink of a Civil War. Prepared by Eduardo Marino, London: International Alert. IRB (Immigration and Refugee Board) (1992). ‘Sri Lanka: Internal Flight Alternatives,’ Research Directorate, Documentation, Information and Research Branch, Immigration and Refugee Board Documentation Centre, Ottawa. Irshick, Eugene (1969). Politics and Social Conflict in South India: The NonBrahman Movement and the Politics of Tamil Separatism. New Delhi: Oxford University Press. Islam, Nasir (1987). ‘Devolutionary Strategies in Sri Lanka: Political Decentralization and Ethnic Conflict,’ University of Ottawa Working Paper 87–46, Ottawa. Ivan, Victor (1989). Sri Lanka in Crisis: Road to Conflict. Ratmalana, Sri Lanka: Sarvodaya Book Publishing Services. Ivan, Victor (1993). ‘Caste Oppression and Youth Unrest,’ Counterpoint. (April), pp. 26–7. JAA 1998, Sri Lanka Competitive Study, Final Report by J.E. Austin Associates and SRI International for USAID, September 1998. Jayasuriya, Laksiri (2000). Welfarism and Politics in Sri Lanka: Experience of a Third World State. Perth: University of Western Australia. Jayawardena, Kumari (2000). Nobodies to Somebodies: The Rise of the Colonial Bourgeoisie in Sri Lanka. Colombo: Social Scientists’ Association and Sanjiva Books. Jayawardena, Kumari (1990). Ethnic and Class Conflict in Sri Lanka. Colombo: Sanjiva Books. JICA 1998, JICA in Sri Lanka 1998, Colombo: Japan International Cooperation Agency Sri Lanka Office.

Bibliography 253 Jiggins, Janice (1979). Caste and Family in the Politics of the Sinhalese, 1947–76. Cambridge: Cambridge University Press. Joshi, Chani Lata (2000). ‘Sri Lanka: Suicide Bombers – Ultimate Sacrifice,’ Far Eastern Economic Review, 1 June. http://www.feer.com/_0006_01/p64currents.html Joshi, Manoj (1987). ‘Mission Eagle Accomplished,’ Frontline, (13–26 June), pp. 10–13. Kadian, Rajesh (1990). India’s Sri Lanka Fiasco: Peacekeepers at War. New Delhi and Bombay: Vision Books. Kanapathipillai, Valli (1990). ‘July 1983: The Survivor’s Experience,’ in Veena Das (ed.), Mirrors of Violence: Communities, Riots and Survivors in South Asia. Delhi: Oxford University Press. Kapferer, Judith (1984).’The Demand for Education in Sri Lanka: Antecedents and Consequences,’ in Kenneth Orr (ed.), Appetite for Education in Contemporary Asia. Canberra: Australian National University. Karunatilleke, Waruna (1993). ‘Concentration of LTTE Forces in Jaffna: Changing of the Guard,’ Counterpoint, t (April), 30–1. Kearney, Robert (1978). ‘Language and the Rise of Tamil Separatism in Sri Lanka,’ Asia Survey, 18, pp. 521-34. Kearney (1975). ‘Educational Expansion and Political Volatility in Sri Lanka: The 1971 Insurrection,’ Asian Survey, 15: 9. Kearney, Robert (1967). Communalism and Language in the Politics of Ceylon. Durham, NC: Duke Universtiy Press. Kelegama, S. (1999). ‘Economic Cost of Conflict in Sri Lanka,’ in R.I. Rotberg (ed.), Creating Peace in Sri Lanka. Washington, DC.: Brookings Institution Press. Kemper, Steven (1991). The Presence of the Past: Chronicles, Politics and Culture in Sinhala Life. Ithaca and London: Cornell University Press. Kodikara, Sheldon U. (1989). (ed.), Indo-Sri Lanka Agreement of July 1987. Colombo: International Relations Programme, University of Colombo. Kodikara, Sheldon U (1965). Indo-Ceylon Relations Since Independence. Colombo: Sanjiva Books. Kriesi, Hanspeter (1988). ‘The Interdependence of Structure and Action: Some Relfections of the State of the Art,’ in Bert Klandermans, Hanspeter Kriesi, and Sidney Tarrow (eds), From Structure to Action; Comparing Social Movements Across Cultures. Greenwich, Connecticut and London, England: JAI Press, 349–68. Kurian, Rachel (1989). State, Capital and Labour in the Plantation Industry in Sri Lanka, 1834–1984. geboren te Tirchur, India: Academisch Proefschrift. Laitin, David (1985). ‘Hegemony and Religious Conflict: British Imperial Control and Political Cleavage in Yorubaland,’ in Peter Evans, Dietrich Rueschemeyer and Theda Skocpol (eds), Cambridge: Cambridge University Press, 285–316. Laitin, David (1986). Hegemony and Culture. Chicago and London: Chicago University Press. Lawrence, Patricia (1999). ‘The Changing Amman: Notes on the Injury of War in Eastern Sri Lanka,’ in Siri Gamage and I.B. Watson (eds), Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop, pp. 197–216.

254 Bibliography Leary, Virginia (1981). Ethnic Conflict and Violence in Sri Lanka, A Report of a Mission to Sri Lanka in July–August 1981 on Behalf of the International Commission of Jurists, International Commission of Jurists, Geneva. Lebow, Richard Ned (1973). ‘Ireland,’ in Gregory Henderson, Lebow, and John G. Stoessinger, Divided Nations in a Divided World. New York: David McKay Company, 197–265. Little, Angela (1999). Labouring to Learn: Towards a Political Economy of Plantations, People and Education in Sri Lanka (London: Macmillan – new Palgrave Macmillan). LST/Law and Society Trust 2001. Sri Lanka: State of Human Rights 2001. Colombo: Law and Society Trust. Malalgoda, Kitsiri (1970). ‘Millenialism in Relation to Buddhism,’ Comparative Studies in Society and Buddhism, 12. Manogaran, Chelvadurai. (1994). ‘Colonization as Politics: Political Use of Space in Sri Lanka’s Ethnic Conflict,’ in Manogaran and Bryan Pfaffenberger (eds), The Sri Lankan Tamils: Ethnicity and Identity. Westview Press: Boulder, CO, 84–125. Manogaran, C. (1987). Ethnic Conflict and Reconciliation in Sri Lanka, University of Hawaii Press, Honolulu. Manor, James (1989). The Expedient Utopian: Bandaranaike and Ceylon. Cambridge: Cambridge University Press. Manor, James (1988). ‘Collective Conflict in India,’ Conflict Studies # 212. London Centre for Security and Conflict Studies. Manor, James (1984) (ed). Sri Lanka in Change and Crisis. London and Sydney: Croom Helm. Manor, James 1979. ‘The Failure of Political Integration in Sri Lanka (Ceylon),’ Journal of Commonwealth and Comparative Politics, 17: 1, 21–44. Manor, James and Gerald Segal (1985). ‘Causes of Conflict, Sri Lanka and Indian Ocean Strategy,’ Asian Survey, XXV: 12, 1165–85. Marino, Eduardo (1987). ‘The Armed Component of Ethnic Conflict in Sri Lanka: From Discrimination to War,’ PRIO Report 23/86, International Peace Research Institute, Oslo, Norway. Matthews, Bruce (1993). ‘Devolution of Power in Sri Lanka,’ paper prepared for the International Centre for Ethnic Studies, Colombo. Matthews (1993a). ‘Sri Lanka’s Crushing Civil War at a Turning Point,’ unpublished paper. Matthews (1992). ‘Trouble in Sri Kotha: Strains and Perils of Democracy in Sri Lanka,’ The Round Table, 322, 215–27. Matthews, Bruce (1989). ‘The Janatha Vimukthi Peramuna and the Politics of the Underground in Sri Lanka,’ Round Table, 312, pp. 425–39. Matthews, Bruce (1988–89). ‘Sinhala Cultural and Buddhist Patriotic Organizations in Contemporary Sri Lanka,’ Pacific Affairs, 61: 4, 620–32. Matthews (1987). ‘Devolution of Power in Sri Lanka: The Problems of Implementation,’ Round Table, 301, 74–92. Matthews (1985). ‘The Sri Lankan Buddhist Philosophy of History and its Relationship to the National Question,’ Ethnic Studies Report, t III: 2, 81–7. McCarthy, John D. and Mayer N. Zald (1977). ‘Resource Mobilization and Social Movements: A Partial Theory,’ American Journal of Sociology, 82, pp. 1212–41.

Bibliography 255 McKay, James (1982). ‘An Exploratory Synthesis of Primordial and Mobilizationist Approaches to Ethnic Phenomena,’ Ethnic and Racial Studies, 5: 4, pp. 395–420. Meade, Peter C. and James M. Mirocha, ‘Civilian Landmine Injuries in Sri Lanka,’ Medecins Sams Frontieres Medical News, Vol. 9, No. 1 (November 1999), pp. 37–48. Meyer, Eric (1984). ‘Seeking the Roots of the Tragedy,’ in James Manor (ed.), Sri Lanka in Change and Crisis. London and Sydney: Croom Helm, pp. 137–52. Migdal, Joel S. (1987). ‘Strong States, Weak States: Power and Accommodation,’ in Myron Weiner and Samuel P. Huntington (ed.), Understanding Political Development. Boston and Toronto: Little, Brown, 391–435. Migdal, Joel S. (1988). Strong Societies and Weak States: State–Society Relations and State Capabilities in the Third World. Princeton: Princeton University Press. Miller, Robert (1992). ‘Introduction,’ in his (ed.), Aid as Peacemaker: Canadian Development and Third World Conflict. Ottawa: Carleton University Press, 1–17. MIRJE/Movement for Inter-Racial Justice and Equality (1983). 1981: The Year of Racial Violence. Colombo: MIRJE. MIRJE (1981). What Happened in Jaffna: Days of Terror. Colombo: MIRJE. MIRJE: (1980). Emergency ‘79. Colombo: MIRJE. Moore, Mick (1990). ‘Economic Liberalization versus Political Pluralism in Sri Lanka,’, Modern Asian Studies, 24 (2). pp. 341–83. Moore, Mick (1990a). ‘Economic Liberalization, Growth and Poverty: Sri Lanka in a Long Run Perspective,’ Institute of Development Studies Discussion Paper 274, International Development Studies, University of Sussex, Brighton, England. Moore (1985). The State and Peasant Politics in Sri Lanka. Cambridge: Cambridge University Press. Nagel, Joane (1986). ‘The Political Construction of Ethnicity,’ in Susan Olzak and Nagel (eds), Competitive Ethnic Relations. London: Academic Press, 93–112. Nash, Manning (1989). The Cauldron of Ethnicity in the Modern World. Chicago and London: University of Chicago Press. Newman, Saul (1991). ‘Does Modernization Breed Ethnic Political Conflict?’ World Politics, 43: 3, 451–78. Nissan and R.L. Stirrat (1990). ‘The Generation of Communal Identities’, in Jonathan Spencer (ed.). Sri Lanka: History and Roots of Conflict. London and New York :Routledge, pp.19–44. Nissan, Elizabeth (1984). ‘Some Thoughts on Sinhalese Justification for the Violence,’ in James Manor (ed.), Sri Lanka in Change and Crisis. London and Sydney: Croom Helm, 175–86. Nordstrom, Carolyn (1992). ‘The Backyard Front,’ in Carolyn Nordstrom and JoAnn Martin (eds), The Paths to Domination, Resistance and Terror. Berkeley, Los Angeles, London: University of California Press, 260–76. NPC/National Peace Council and Marga (2001). The Cost of the War: Economic, Social and Human Cost of the War … Sri Lanka. Colombo: National Peace Council of Sri Lanka. O’Ballance, Edgar (1989). The Cyanide War: Tamil Insurrection in Sri Lanka, 1973–88. London: Brassey’s (UK). Oberst, Robert (1985). ‘Democracy and the Persistence of Westernized Elite Dominance in Sri Lanka,’ Asian Survey. 15: 7, pp. 760–72.

256 Bibliography Obeyesekere, Gananath (1984). ‘The Origins and institutionalization of Political Violence,’ in James Manor (ed.), Sri Lanka in Crisis and Change. London and Sydney: Croom Helm, pp. 153–74. Obeyesekere, Gananath (1979). ‘The Vicissitudes of the Sinhala–Buddhist Identity Through Time and Change,’ in Michael Roberts (ed.), Collective Identities, Nationalisms and Protest in Modern Sri Lanka. Colombo: Marga Institute. Obeyesekere, Gananath (1974). ‘Some Comments of the Social Backgrounds of the April 1971 Insurrgency in Sri Lanka (Ceylon),’ Journal of Asian Studies, 33: 3, pp. 367–84. OXFAM/Lewer, N. (1998): ‘Peace building and conflict prevention, Bradford’. OXFAM – SRI LANKA, final version, December 1998. OXFAM/Lewer, N. (2000): ‘Conflict Reduction and Transforming Community Dynamics – Integrating a ‘Conflict Sensitive Approach’ to Oxfam’s work in Sri Lanka by working with others to explore new ways of reducing conflict through local community development initiatives’. Bradford: Centre for Conflict Resolution, Department of Peace Studies, University of Bradford, July 2000. OXFAM-UK and Save the Children-UK 1998. Listening to the Displaced and Listening to the Returned: A Community Study, Summary Report, photocopy. Perera, Jehan (1999). ‘The Cost of War’, in Consortium of Humanitarian Agencies (CHA) Annual Report 1998. Colombo: CHA, 122–32. Perera, Melloshini (1999). ‘Children’s Rights,’ Sri Lanka: State of Human Rights 1999. Colombo: Law and Society Trust, 91–110. Perera, Sasanka (1993). ‘Messengers of Peace or Creators of Chaos: The Role of Teachers in the Sri Lanka Ethnic Conflict,’ Pravada, 1: 12, 4–11. Pfaffenberger, Bryan (1990). ‘The Political Construction of Defensive Nationalism: The 1968 Temple-Entry Crisis in Northern Sri Lanka,’ Journal of Asian Studies, 49: 1, pp. 78–96. Pfaffenberger, Bryan (1988). ‘Sri Lanka in 1987: Indian Intervention and the Resurgence of the JVP,’ Asian Survey, XXVIII: 2, pp. 137-47. Philip, Anton (1991). ‘Sri Lanka: Chronology of Events,’ Sri Lanka Resource Centre, Oslo, Norway. Philipson, L. (1999) ‘Breaking recurring themes in the cycles of war and peace in Sri Lanka,’ Research Paper 3, Centre for the Study of Global Governance, London. Pieris, Shantha J.R. and Jeanne Marecek (1992). ‘Report on the Welikanda Massacre,’ unpublished report commissioned by the International Centre for Ethnic Studies, Colombo, 11 May. Piyadasa, L. (1984). Sri Lanka: The Holocaust and After. London: Marram Books. Ponnambalam, Satchi (1983). Sri Lanka: The National Question and the Tamil Liberation Struggle. London: Zed Books. Premachandra (1991). Sri Lanka: The Years of Terror. The JVP Insurrection, 1987–1989. Colombo: Lake House Bookshop. Premdas, Ralph and S.W.R. de A. Samarasinghe (1988). ‘Sri Lanka’s Ethnic Conflict: The Indo-Lanka Accord,’ Asian Survey, XVIII: 6, pp. 676–90. Rahula, Walpola (1974). The Heritage of the Bhikkhu: A Short History of the Bhikkhu in the Educational, Cultural, Social, and Political Life. New York: Grove Press.

Bibliography 257 Rahula, Walpola (1956). The History of Buddhism in Ceylon. Colombo: Gunasena. Rao, P. Venkateshwar (1989). ‘Foreign Involvement in Sri Lanka,’ Round Table, 309, pp. 88–100. Refuge (1993). Special issue on Sri Lanka, 13 (3). June. Roberts, Michael (1982). Caste, Conflict and Elite Formation. Cambridge: Cambridge University Press. Rogers, John (1987). Crime, Justice and Society in Colonial Sri Lanka. London: Curzon Press. Rogers, John (1987a). ‘Social Mobility, Popular Ideology, and Collective Violence in Modern Sri Lanka,’ The Journal of Asian Studies, 46: 3, pp. 583–602. Roosens, Eugeen E. (1989). Creating Ethnicity: The Process of Ethnogenesis. California: Sage Publications. Rotberg R.I. (1999) (ed). Creating Peace in Sri Lank. Washington, DC.: Brookings Institution Press. Rothschild, Donald (1981). Ethnonationalism. New York: Columbia University Press. RRAN (1998a). ‘Resettlement and Rehabilitation Authority of the North’, The Dawn of a New Era … Programme Framework for Resettlement, Rehabilitation, Reconstruction and Development: Jaffna Peninsula (n.p., July). RRAN (1998b). ‘Resettlement and Rehabilitation Authority of the North’, The Dawn of a New Era … Programme Framework for Resettlement, Rehabilitation, Reconstruction and Development: Mannar Plan (n.p., November). RRAN (1998c). ‘Resettlement and Rehabilitation Authority of the North’, The Dawn of a New Era … Programme Framework for Resettlement, Rehabilitation, Reconstruction and Development: Vavuniya Plan (n.p., November). Rubin, Barnett R. (1987). ‘Cycles of Violence: Human Rights in Sri Lanka Since the Indo-Sri Lanka Agreement,’ Asia Watch Report, Washington. Rupesinghe, Kumar (1988). ‘Ethnic Conflicts in South Asia: The Case of Sri Lanka and the Indian Peace Keeping Force,’ Journal of Peace Research, 25: 4. Rupesinghe, Kumar and Berth Verstappen (1989). Ethnic Conflict and Human Rights in Sri Lanka: An Annotated Bibliography. London: Hans Zell Publishers. Rupesinghe, Kumar, Berth Verstappen, and Anton S. Philip (1993). Ethnic Conflict and Human Rights in Sri Lanka: An Annotated Bibliography, Vol. II, 1989–92. London: Hans Zell Publishers. Samarakone, Priya (1984) (Pseud.). ‘The Conduct of the Referendum,’ in James Manor (ed.), Sri Lanka in Change and Crisis. London and Sydney: Croom Helm, 84–117. Samaranayake, Gamini (1987). ‘The Changing Attitude Towards the Tamil Problem Within the Janatha Vimukthi Peramuna (JVP),’ in Charles Abeysekera and Newton Gunasinghe (eds), Facets of Ethnicity in Sri Lanka. Colombo: Social Scientists Association, 272–97. Samarasinghe, S.W.R. de A. and Kamal Liyanage (1993). ‘Friends and Foes of the Indo-Sri Lanka Accord,’ in K.M. de Silva and S.W.R. de A. Samarasinghe (eds), Peace Accords and Ethnic Conflicts. London and New York: Pinter Publishers, 156–72. Samaraweera, V. (1981). ‘The Evolution of a Plural Society,’ in K.M. de Silva (ed.), Sri Lanka: A Survey. London: C. Hurst. Sambandan, VS (2000). ‘A National on the Brink,’ Frontline, Vol. 17, Issue11. 27 May–9 June 2000.

258 Bibliography Schafer, Teresita C. (1999). ‘Peacemaking in Sri Lanka: The Kumaratunga Initiative,’ in R.I. Rotberg (ed.), Creating Peace in Sri Lanka. Washington, DC.: Brookings Institution Press, 131–41. Scott, George M. (1990). ‘A Resynthesis of the Primordial and Circumstantial Approaches to Ethnic Group Solidarity: Towards an Explanitory Model,’ Ethnic and Racial Studies, 13: 3, pp. 147–71. Scott, Nicholas (1984). (Pseud.), ‘A Report from Sri Lanka,’ America, 14 January, pp. 11–12. Sentinel South Asia, November 1999 – April 2000. www.janes.com Shanmugaratnam, N. (1993). ‘The National Question and Obstacles to Peace in Sri Lanka: Some Reflections,’ Pravada, 2: 5, pp. 4–13. Shastri, Amita (1990). ‘The Material Basis for Separatism: the Tamil Eelam Movement in Sri Lanka,’ The Journal of Asian Studies, 49: 1, pp. 56–77. Silva, Kalinga Tudor (1986). ‘Segregation as an emerging aspect of Ethnic Relations in Sri Lanka,’ unpublished paper, Department of Sociology, Peradeniya University. Singer, Marshall A. (1964). The Emerging Elite: A Study of Political Leadership in Ceylon. Cambridge, Mass.: MIT Press. SIPRI (1987). SIPRI 1987 Yearbook. Oxford: Oxford University Press. SIPRI (1992). SIPRI 1992 Yearbook. Oxford: Oxford University Press. Siriwardena, Reggie (1984). ‘National Identity in Sri Lanka: Problems in Communication and Education,’ in the Committee for Rational Development, Sri Lanka, the Ethnic Conflict: Myths, Realities and Perspectives. New Delhi: NAVRANG, pp. 215–29. Skocpol, Theda (1985). ‘Bringing the State Back in: Strategies of Analysis in Current Research,’ in Peter Evans, Dietrich Rueschemeyer and Theda Skocpol (eds), Bringing the State Back In. Cambridge: Cambridge University Press. Smith, Bardwell L. (1978). (ed). Religion and the Legitimation of Power in Sri Lanka. Chambersberg, PA: ANIMA Books. Somaratne, G.P.V. (1989). ‘Renewal of Ties Between Sri Lanka and Israel,’ Jerusalem Journal of International Relations, 11: 1, pp. 74–86. Somasundaram, Daya (1998). Scarred Minds: The Psychological Impact of War Sri Lankan Tamils. Colombo: Vijitha Yapa Bookshop. Spencer, Jonathan (1990). (ed). Sri Lanka: History and Roots of Conflict. London and New York: Routledge. Spencer (1990a). ‘Writing Within: Anthropology, Nationalism, and Culture in Sri Lanka’, Current Anthropology, 31: 3. pp. 283–300. Spencer, Jonathan (1984). ‘Popular Perceptions of the Violence: A Provincial View,’ in James Manor (ed.), Sri Lanka in Change and Crisis. London and Sydney: Croom Helm, 187–95. Sri Lanka Resource Centre (1981). ‘Sri Lanka: Reign of Terror in Jaffna,’ Sri Lanka Resource Centre, Oslo, Norway. Tambiah Stanley J. (1996). Leveling Crowds: Ethnonationalist Conflicts and Collective Violence in South Asia. Berkeley, Los Angeles, London: University of California Press. Tambiah Stanley J (1993). ‘Bridges Across the Ethnic Divide,’ Pravada, 2: 6, pp. 4–11. Tambiah Stanley J (1992). Buddhism Betrayed: Religion, Politics and Violence in Sri Lanka. Chicago and London: University of Chicago Press.

Bibliography 259 Tambiah Stanley J (1986). Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy. Chicago and London: University of Chicago Press. Tambiah Stanley J (1983a). ‘Civil War and Women of Jaffna’ Pravada, 1: 11, pp. 8–12. Tarrow, Sidney (1989). ‘Struggle, Politics, and Reform: Collective Action, Social Movements, and Cycles of Protest,’ Western Societies Program Occasional Paper #21, Center for International Studies, Cornell University, Ithaca, New York. Tarrow, Sidney (1989a). Democracy and Disorder: Social Conflict, Politics and Protest in Italy, 1965–1975. New York: Oxford University Press. Thambayah, Jennifer S. (1992) ‘The Displaced in Sri Lanka: The Relevance of International Humanitarian Law,’ unpublished paper, Law and Society Trust, Colombo. Tilly, Charles (1984). Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Foundation. Tilly, Charles (1978). From Mobilization to Revolution. Englewood Cliffs NJ: Prentice-Hall. Tilly, C., L. Tilly and R. Tilly (1975). The Rebellious Century, 1830–1930. Cambridge, Mass.: Harvard University Press. Tiruchelvam, Neelan (1999). ‘Devolution and the Elusive Quest for Peace in Sri Lanka,’ in Siri Gamage and I.B. Watson (eds), Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop, 189–201. Trawick, Margaret (1999). ‘Reasons for the Violence: A Preliminary Ethnographic Account of the LTTE’ in Siri Gamage and I.B. Watson (eds), Conflict and Community in Contemporary Sri Lanka: Pearl of the East or the Island of Tears. Colombo: Vijitha Yapa Bookshop, 139–64. UN Secretary General (1999). ‘Protection of children affected by armed conflict’, Note by the Secretary-General, United Nations A/53/482. Fifty-third Session Agenda Item 106. Promotion and protection of the rights of children, 12 October 1998. UN Working Group on Enforced and Involuntary Disappearances (1992). ‘Report on Sri Lanka’ (February). UNICEF/SCF (2000). ‘Report on the Workshop, The Challenge of Child Soldiers,’ Colombo, Sri Lanka, 24 March 2000. Uphoff, Norman (1992). Learning from Gal Oya: Possibilities for Participatory Development and Post-Newtonian Social Science. Ithaca and London Cornell University Press. Uphoff, (1992a). ‘Monitoring and Evaluating Popular Participation in World Bank-Assisted Projects’ in Bhuvan Bhatnagar and Aubrey C. Williams (eds), Participatory Development and the World Bank: Potential Directions for Change, World Bank Discussion Paper 183. Washington: World Bank, 135–53. Uphoff, Norman (1990). ‘Distinguishing Power, Authority and Legitimacy: Taking Max Weber at his Word by Using Resource-Exchange Analysis,’ Polity, 22: 2, pp. 295–322. Uphoff, Norman and Warren Ilchman (1972). The Political Economy of Development. Berkeley, Los Angeles and, London: University of California Press. US State Department (2001). Sri Lanka Country Report on Human Rights Practices for 2000, released by the Bureau of Democracy, Human Rights, and Labor, February 2001.

260 Bibliography USCR/US Committee for Refugees (1991). ‘Sri Lanka: Island of Refugees,’ prepared by Court Robinson, US Committee for Refugees, Washington, DC, October. USCR (1994). ‘People Want Peace: Repatriation and Reintegration in War-Torn Sri Lanka,’ Issue Paper, prepared by Hiram Ruiz, US Committee for Refugees, Washington, DC, January. UTHR(J)/University Teachers for Human rights – Jaffna (2000). The Scent of Danger. Sir Lanka Information Bulletin No. 23 (30 January). http://www.spur.asn.au/UTHR_bulletin_no_23_dated_30_january_2000.htm UTHR(J) (1993). ‘Land, Human Rights and the Eastern Predicament,’ Report #11, Thirunelvely, Jaffna, 15 April. UTHR(J) (1991). ‘The Debasement of Law and of Humanity and the Drift Towards Total War,’ Report #8, University of Jaffna, 28 August. UTHR(J) (1991a). ‘The Politics of Destruction and Human Tragedy,’ Report #6, University of Jaffna, February. UTHR(J)(1998). Living Through Jaffna’s Sultry Sunset – Special Report #10. (Jaffna: University Teachers for Human Rights). Uyangoda, Jayadeva (1989). ‘Cinema in Cultural and Political Debates in Sri Lanka,’ Framework, 37, pp. 37–43. Van Crenfeld, Martin (1990). The Transformation of War. New York: Free Press. Vittachi, Tarzie (1958). Emergency ‘58: The Story of Ceylon Race Riots. London: Andre Deutsch. W.A. Wiswa Warnapala and D.E. Woodsworth (1987). Welfare as Politics in Sri Lanka, Monograph Series No. 22, Centre for Developing Area Studies, McGill University, Montreal Canada. Weaver, Mary Anne 1988. ‘The Gods and the Stars,’ The New Yorker, 21 March. Weber, Max (1947). The Theory of Social and Economic Organization. New York: Oxford University Press. Wilson, A. Jayaratnam (2000). Sri Lankan Tamil Nationalism. London: Penguin. Wilson, A. Jayaratnam 1988. The Break-up of Sri Lanka: The Sinhalese and Tamil Conflict. Honolulu: University of Hawaii Press. Wilson, A. Jayaratnam (1980). The Gaullist System in Asia: The Constitution in Sri Lanka (1978). London: Macmillan – now Palgrave Macmillan. Wilson, A. Jayaratnam (1974). Politics in Sri Lanka, 1947–1973. London: Macmillan – now Palgrave Macmillan. World Bank (1998). The World Bank’s Experience with Post-Conflict Reconstruction, Volume V V, Operations Evaluation Department, Report No. 17769, May 4. Wriggins, Howard (1960). Ceylon: Dilemmas of a New Nation. Princeton: Princeton University Press. Zartman, I. William (1985). Ripe For Resolution. Oxford: Oxford University Press. Zunzer, Wolfram 2002. Annotated Bibliography on the Nexus of Relief/Development Aid and Violent Conflict in Sri Lanka, Working Paper No. 1, Berghof Foundation for Conflict Studies, Sri Lanka Office (March).

Index Allison, Graham, 12–13, 16, 206n Amirthalingam, A., 30, 60, 63, 121, 229n, 233n Annuradhnapura Massacre (1985), 139 Ariyaratne, Dr A.T., 157 Aryan race, 213n assassination, 29–30, 32, 56, 57, 58, 60, 63, 64, 67, 71, 85, 94, 119, 120, 123, 139, 145, 157, 169, 185, 202, 214n, 215n, 217n, 225n, 233n, 237n, 238n, Athulatmudali, Lalith, 43, 58, 59, 117, 152, 153, 185, 210n, 215n, 227n, 231n, 232n, 235n, 237n, 241n Azar, Edward, 8, 34 Balasingham, Anton, 169, 171 Balkans, xviii, 237n Bandaranaike, Anura, 57, 102 Bandaranaike, Sirimavo, 57, 63–4, 103, 104, 117, 152, 214n, 227n, 233n Bandaranaike, S.W.R.D., 55, 56, 84–98, 99, 101, 117, 180, 186, 196, 200, 202, 214n, 220n, 221n, 244n Bandaranaike Katanayake Airport Attack (2001), 63, 158, 168 Batticaloa Peace Committee, 168 B–C Pact / Bandaranaike– Chelvanayakam Pact, 56, 94, 214n B-Specials (Northern Ireland), 193 Betrayal of Buddhism, 96 Bhikkhus (Buddhist monks), 64, 86, 89, 94, 96, 105, 142, 152, 216n, 220n, 222n, 224n Black Cats (vigilante group), 54, 65, 106 Bopage, Lionel, 110, 111 Bosnia, 3, 6, 7, 8

Brass, Paul, 10, 12, 22, 33 Buddha Jayanthi, 95 Buddhist clergy, see Sangha Buddhist Revival Movement, 86 Burma, 220n Canada, 66, 67, 165, 228n, 240n Casparsz, Paul, 213n caste, 48–53, 101, 102, 110, 111, 202, 212n, 213n, 214n and Tamil paramilitaries, 51–3 causality, 10–11, 205n Ceasefire Agreement (CFA) of 2002, 42, 44, 59, 66, 70, 156–76, 181, 186, 190, 192, 205n, 217n, 238–44n Ceylon Citizenship Bill, 41, 45, 217n, 218n Ceylon Tamil Congress, 81–2, 136 Ceylon–UK Defence Agreement (1948), 70 Ceylon Workers’ Congress, 77, 146, 218n Chandrasekeran, P., 53, 54 Chelvanayakema, S.J.V., 214n child soldiers, 30, 67, 165, 168, 209n China, 71, 144 Citizenship Bills, 76–7, 189 class, 35, 39, 50, 58, 110, 114, 126, 134, 181, 202, 218n, 226n, 231n conflict management vs conflict resolution, 204n, 241n costs of war, 31 coups d’état, 61 CTC/ Ceylon Tamil Congress, 75 demobilization, 16, 194 demography, 35, 40, 42, 79, 101, within Armed Forces, 60 deserters, 30, 62, 157 Dharmapala, Anagarika, 46 disenfranchisement, 73–83, 180 Dissanayake, Gamini, 58, 120, 229n

261

262 Index Donoughmore Commission, 78–9 double minority complex, 36 Druckman, Daniel, 201 DUNF/Democratic United National Front, 58, 59 Dunham, David and Sisira Jayasuriya, 158 Dutta Gamani (King), 47

colonial rule, 29, 38, 45 and myth, 45–8, 127, 213n Hogan, Paul, xiv Hameed, A.C.S., 227 Hollingworth, Steve, 61 Home Guard, 30, 43, 153, 193, 209n, 240n, 244n Horowitz, Donald, 3, 20, 92, 203

EBP/ Eksath Bhikkhu Peramuna (United Front of Monks), 89, 220n education, 30 ,38, 41, 43, 44, 45, 46, 47, 48, 49, 53, 101, 110, 186, 228n emergency regulations, 30, 118, 124, 128, 137, 140, 162, 189, 190, 192–4, 205n, 208n, 216n, 228n, 229n, 230n, 231n, 233n environmental scarcity, 7 Eelam War III, 156, 205n ENDLF/ Eelam National Democratic Liberation Front, 149, 238n ENLF/ Eelam National Liberation Front, 238n EPDP/ Eelam people’s Democratic Party, 68, 145, 146, 154, 217n, 235n EPRLF/ Eelam People’s Revolutionary Liberation Front, 66, 68, 69, 145, 146, 149, 150, 151, 154, 168, 206n, 216n, 217n, 226n, 235n, 238n EROS/ Eelam Research Organization of Students, 154, 234n, 238n Esman, Milton, 207n ethnic cleansing, 36, 41, 42, 187, 188

identity, 3, 5, 39, 206n and conflict, 195–7 contingent identities, 20–3 formation, 4–7, 8, 10, 11, 155 mobilization and politicization, 17–23, 45–8, 55–6, 65, 182, 186, 194, 196 and legitimation, 9, 46, 187 India, 24, 51, 69, 70–71, 103, 113, 135–155, 161, 191, 206n, 208n, 234n India–Sri Lanka Agreement of 1987 (‘Indo-Lanka Accord’), 36, 54, 57, 64, 70, 71, 105, 115, 135–55, 157, 181, 191, 223n, 232–8n, 242n India–Sri Lanka Free Trade Agreement, 191 inter–intra-group relations, xvii–xix, 10, 12, 14, 18, 21, 24, 32, 33, 34, 53, 56, 61, 81–3, 91–4, 112, 114, 115, 125, 130–34, 142, 146, 177–203 IPKF/ Indian Peacekeeping Force, 24, 36, 55, 61, 66, 68, 71, 135, 144–55, 181, 191, 227n, 233n, 235n, 236n, 237n Israel, 67, 71

Federal Party, 80, 90, 94, 136, 187, 214n

Jaffna Public Library, 228n Jayrawardene, J.R., 47, 56, 57, 58, 64, 70, 87, 94, 103, 104, 115, 118, 122, 124, 126–7, 131, 132, 136, 142, 143, 144, 149, 152, 153, 155, 200, 227n, 229n, 230n, 235n Jiggins, Janice, 91, 102, 226n JVP/ Jathika Vimukti Peramuna (People’s Liberation Front), 12, 32–33, 53, 57, 62, 63–66, 68, 99–117, 133, 144, 146, 150, 151, 152, 153, 157, 175, 179, 180, 182,

Gandhi, Rajiv, 29, 50, 61, 71, 143, 146, 149, 153, 217n, 236n, 238n Gal Oya Water Project, 207n, 229 garment industry, 158 Hakeem, Rauf, 171–6 Hisbullah, M.L.A.M., 173 history anachronistic renderings, 33–5, 109

Index 263 213n, 216n, 222n–227n, 231n, 241n, 242n, 243n, 244n Karikalan, V, 174 Kataragama, 207n Kattankudi Mosque massacre of 1990, 43, 169 Kent and Dollar Farms Massacre (1984), 139, 212n kleptocracy, 7 Kotelawala, Sir John, 85, 87, 88, 92, 95 Kumaratunge, Chandrika Bandaranaike, 23, 51, 57, 63, 66, 156, 157–61, 175, 185, 205n, 227n

multidimensional approaches, 4 thick description, 16 two-level critical juncture approach, 15–17 Miller, Father Harry, 168, 240n militarization, 227n military expenditures, 31 military remittances, 30, 158 Mohamed, M.H., 173 monks, see bhikkhus Movement to Protect the Motherland, 152 Muslim, 36, 37, 39, 40–4, 68, 76, 129, 131, 147, 169–75, 181, 187, 188, 207n, 212n, 235n, 240n, 244n National Framework for Relief, Rehabilitation, and Reconciliation in Sri Lanka (3–R Framework), 190 Northern Ireland, 3, 6, 7, 13, 14, 22, 24, 77, 184, 188, 192–4, 196, 200, 206n, 208n Norway, Government of, 161, 191

landmines, 30, 63, 112, 113, 123, 130, 139, 158, 230n LTTE (Liberation Tigers of Tamil Eelam), 14, 29, 30, 31, 36, 37, 38, 40, 41, 42, 43, 52–5, 60, 63, 65, 66–70, 112, 119, 135, 137, 141, 142, 144–6, 147, 147–55, 156–76, 181, 182, 184–6, 190, 191, 194, 199, 205n, 207n, 210n, 216n, 217n, 233n, 234n, 235n, 236n, 237n, 239n, 240n, 243n, 244n

O’Neill, Terence, 200 opportunity structure, 17, 25–6, 94–8, 115, 182, 194

Mahavelli Water Project, 42 Manor, James, 20, 81, 85, 87, 90, 221n massacre, 43, 106, 107, 108, 123, 124, 139, 141, 142, 145, 147, 148, 151, 169, 170, 210n, 230n, 234n Mathew, Cyril, 50, 117, 120, 122, 128–9, 132, 185 Matthews, Bruce, 55, 57, 64, 105, 112, 115 media, 62, 68, 121, 123, 125–7, 128, 132–33, 139, 147, 152, 153, 166, 170, 181, 189, 230n, 231n, 232n MEP/Mahajana Eksath Peramuna, 84–98, 196 methodology, xvii, 12–28, 156, 182, 194–5, billiard ball models, 3–9, 10, 195 indicators, 15 large-N studies, 12

PA (Peoples’ Alliace), 57, 62, 66, 156, 174–5 Pakistan, 144, 235n, Palestine, 13, 67, 167 Palestinian Liberation Organization, 234 pass system, 159 Peace Support Group, 239n Perumal, A., 149 Pieris, G.L., 160–1, 238n, 239n Plantation Tamils, see Up-Country Tamils PLOTE, 53, 66, 68, 145, 146, 154, 216n, 226n, 234n, 235n, 237n Ponnambalam, G.G., 75, 90 Popular Front for the Liberation of Palestine, 234n population displacement, 30, 36, 42, 43, 51, 77, 117, 118, 125, 138, 148, 171, 188, 208n, 228n

264 Index Posen, Barry, 8 poverty alleviation programmes, 158, 215n Prabhakaran, V., 145, 151, 162, 165, 166, 167, 169, 171, 174, 210n, 237n, 243n Premadasa, Ranasinghe, 29, 57, 58, 59, 64, 68, 71, 85, 106, 117, 132, 142, 150–1, 152, 184, 185, 199, 206n, 217n, 232n, 242n proscription, 104, 111, 112, 115, 121, 127, 133, 137, 163, 181, 188, 223n Radzek, 168 Rajarata Rifles, 225n Ramachandran, M.G., 207n, 237n Rattwatte, Anuruddha, 62 realism, 4–9 referendum of 1982, 57, 83 repatriation, 77, 111 resource mobilization and competition, 17, 23–5, 85–6, 92–4, 108, 114–15, 155, 182, 194, riots of 1915 (Anti-Muslim), 41 riots of 1958, 94, 101, 221n riots of 1983, 38, 49–50, 51, 63–4, 104, 112, 115, 118–34, 136, 137, 149, 179, 181, 188, 197, 198, 232n, 242n riots of 1983 and economic liberalization, 130–31 Rothschild, Donald, 22 Rwanda, xviii, 3, 6, 7 Sangha, (Buddhist Clergy) 68, 86, 89, 95, 152, 186, 213n, 221n, 222n, 238n SAARC Confernece (1986), 237n security/ insecurity, 3, 4, 7–8, 196 segregation, 188–9 Senanayake, D.S., 55, 75, 78, 86, 92 Senanayake, Dudley, 55, 92, 221n Sinhala, 36, 86 Sinhalese Buddhism, 36, 45, 64, 84–98, 186 Sinhalese Nationalism, 36, 45, 117, 128, 186

Sinhala-Only language policy, 41, 44, 84, 87, 89, 90, 94, 97, 101, 117, 200, 221n Siripada, 207n SLFP/ Sri Lanka Freedom Party, 54, 55, 57, 58, 68, 84–98, 106, 116, 117, 118, 121, 146, 152, 153, 185, 200, 220n, 221n, 223n SLMC/ Sri Lanka Muslim Congress, 44, 146, 171–6 SLMM/ Sri Lanka Monitoring Mission, 161, 162, 165 South Africa, 14, 71, 208n Sri Lankan Security Forces, 30, 36, 52, 60–3, 103, 107, 147, 157, 158, 160, 162, 164–5, 182, 209n, 217n, 222n, 225n, 229–30n, 231n, 237n, 238n, 243n suicide bombers, 29, 30, 31, 67, 157, 164, 209n, 217n, Tamil diaspora, 66, 165, 216n, 228n Tamil Eelam, 36, 38, 55, 59, 115, 119 Tamil Madrasi Regiment, 148 Tamil Nadu, 39, 50, 51, 66, 70, 137, 138, 141, 143–4, 149, 159, 207n, 208n, 217n, 226n, 228n, 234n, 235n, 238n Tamil National Army, 237n Tamil paramilitaries, 30, 32, 33, 36, 42, 50, 53, 55, 60, 61, 66–70, 108, 109, 112, 113, 114, 116, 118, 119, 121, 122, 123, 125, 127, 135, 136, 137, 138–141, 144–6, 149–51, 154, 184, 187, 194, 205n, 207n, 225n, 234n, 235n, 237n, 240n, 243n Tamil paramilitary feuding , 66–9, 139–41, 145, 206n, 226n, 235n Tarrow, Sidney, xv, 25 tautology, 8, 26 TELO, 51, 61, 66, 68, 69, 145, 146, 154, 168, 216n, 235n, 237n Thamilchelvam, 171, 243 Thamil Selvan, see Thamilchelvam Thondaman, S., 218n Tilly, Charles, 13, 16, 24, 207n Tiruchelvum, Neelan, 30, 60, 223n TNA, 68, 69, 149, 151, 168 trade unions, 80, 122

Index 265 TULF/ Tamil United Liberation Front, 30, 55, 59–60, 119, 120, 121, 123, 127, 136, 137, 140, 146, 187, 214n, 228n Udagampola, P., 106, 107, 224–5n, UNP/ United National Party, 50, 54, 57, 58, 79, 81, 82, 85, 86, 87, 88, 90, 92, 94, 95, 96, 112, 116, 117, 118, 120, 123, 126, 128, 130, 131–4, 136, 142, 146, 148, 149–52, 153, 174, 175, 181, 185, 188, 200, 218n, 220n, 223n, 225n, 233n, 241n Up-Country Peoples’ Front, 54, 77 Up-Country Tamils, 35, 39, 44, 49, 50, 53–4, 65, 73, 83, 108, 109, 111, 112, 113, 118–20, 180, 182, 210n, 212n, 216n, 221n, 226n, 230n, 239n Uphoff, Norman, xiv, 17, 23, 205n, 207n, 229n UTHR (Jaffna)/ University Teachers for Human Rights (Jaffna), 66, 210n, 212n, 238n, 240n vigilantism (see also Black Cats), 65, 106, 107, 125, 210n,

Vikalpakandayama, 113, violence, 29, 43, violence, patterns, 11–12, 31, 33, 37, 66, 103, 118, 120–5, 128–34, 138–41, 145–6, 153, 157–8, 171, 183–4, 188, 210n, 224n, 229, 234–5n, 235n, 236n violence, post-election, 14, violence and shifting identities, 22 Walpola, Rahula 86, 220n War for Peace, 156, 175, 185, 205n, 227n War on Terror, 3, 7, 163–4, 185, 242n, 244 war weariness, 163, 202 Weerakoon, Batty, 225n Weerawansa, Wimal, 66 Welikade Prison massacre, 124, 230n Wickramasinghe, Ranil, 59, 156–62, 165, 174, 175, 181, 185 Wijeweera, Rohana, 100, 103, 104, 108, 110, 111, 112, 114, 222n, 223n, 226n Yogeswaran, V., 229n, 233n