JACKSON SCHOOL PUBLICATIONS

IN INTERNATIONAL STUDIES
 JACKSON SCHOOL PUBLICATIONS

IN INTERNATIONAL STUDIES

Senator Henry M. Jackson was convinced that the study of the history, cultures,
political systems, and languages of the world's major regions was an essential pre-
requisite for wise decision making in international relations. In recognition of
his deep commitment to higher education and advanced scholarship, this series
of publications has been established through the generous support of the Henry
M. Jackson Foundation, in cooperation with the Henry M. Jackson School of
International Studies. and the University of Washington Press.

The Crisis of Leninism and the Decline of the Left:

The Revolutions of 1989, edited by Daniel Chirot

Sino-Soviet Normalization and its International Implications.

1945-1990, by Lowell Dittmer

Contradictions: A.rtistic Life, the Socialist State, and the

Chinese Painter Li Huasheng, by Jerome Silbergeld with Gong Jisui

The Found Generation: Chinese Communists in Europe

during the Twenties, by Marilyn A. Levine

Rules and Rights in the Middle East: Democracy, Law, and Society,
edited by Ellis Goldberg, Re'iat Kasaba,. and Joel S. Migdal

Can Europe Work?: Germany and the Reconstruction of

Postcommunist Societies, edited by Stephen Hanson and Willfried Spohn

Marxist Intellectuals and the Chinese Labor Movement:

A Study of Deng Zhongxia (1894-1933), by Daniel Y. K. Kwan

Essential Outsiders: Chinese and Jews in the Modern Transformation

of Southeast Asia and Central Europe,
edited by Daniel Chirot and Anthony Reid

Days of Defeat and Victory, by Yegor Gaidar

 THE PRODUCTION OF

Hindu-Muslim Violence
IN CONTEMPORARY INDIA

PAUL R. BRASS

UNIVERSITY OF WASHINGTON PRESS

Seattle and London
 This publication was supported in part by the Jackson School
Publications Fund, established through the generous support of the
Henry M. Jackson Foundation and other donors, in cooperation with
the Henry M. Jackson School of International Studies
and the University of Washington Press.

Copyright © 2003 by the University of Washington Press

Designed by Pamela Canell
Printed in the United States of America

All rights reserved. No portion of this publication may be reproduced or

transmitted in any fonn or by any means, electronic or mechanical, including
photocopying, recording, or any information storage or retrieval system,
without permission in writing from the publisher.

Library of Congress Cataloging-in-Publication Data

Brass, Paul R.
The production of Hindu-Muslim Violence in contemporary India i
Paul R. Brass.
p. cm. - (Jackson SchooI publications in international studies)
Includes bibliographical references and index.
ISBN 0-295-98506-2 (alk. paper)
1. Communalism-India-Aligarh.
2. Riots-India-Aligarh.
3. Hindus--Illdia--Aligarh.
4. Muslims-India-Aligarh.
1. Title. II. Series.
DS422.c64873 2003 954'.2-DC21 2002027192

The paper used in this publication is acid-free and recycled from 10 percent
post-consumer and at least 50 percent pre-consumer waste. It meets the minimum
requirements of American National Standard for Information Sciences-
Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.@()\
In memory of my teacher,

MYRON WEINER,

ever a source of inspiration,

ever a friend
 CONTENTS

Abbreviations Used in This Book ix

Maps, Figures, and Tables xi
Preface and Acknowledgments xv

PART I I INTRODUCTION

1 1 Explaining Communal Violence 5

PART II I COMMUNAL RIOTS IN INDIA AND ALIGARH

2 I Aligarh: Politics, Population, and Social Organization 43

31 Hindu-Muslim Violence in India and Aligarh 60

41 The Great Aligarh Riots of December 1990 and January 1991 116

5 I The Control of Communal Conflict in Aligarh 132

PART III I DEMOGRAPHIC, SOCIAL, AND ECONOMIC

FACTORS IN THE PRODUCTION OF RIOTS

6 I The Geography and Demography of Riots 149

7 I The Economics of Riots:

Economic Competition and Victimization 199

vii
 viii I Contents

PART IV / RIOTS AND THE POLITICAL PROCESS

81 Riots and Elections 219

9 1 The Practice of Communal Politics 240

10 I Communalization and Polarization: Selected

Constituency-Wise Results for Aligarh Elections 262

11 I Communal Solidarity and Division at the Local Level 286

12 I The Decline of Communal Violence and the

Transformation of Electoral Competition 296

PART V / THE PROCESS OF BLAME DISPLACEMENT

131 Riot Interpretation, Blame Displacement,

and the Communal Discourse 305

141 Police Views of Hindu-Muslim Violence 328

15 / The Role of the Media 344

PART VI/CONCLUSION

16/ The Persistence of Hindu-Muslim Violence:

The Dynamics of Riot Production 355

Postscript: Aligarh and Gujarat 385

Appendices 393

Notes 413

Index 463

Index of Mohallas 475

 ABBREVIATIONS USED IN THIS BOOK

ADM additional district magistrate

AMU A1igarh Muslim University
AMUSU A1igarh Muslim University Students Union
BJP Bharatiya Janata Party
BJS Bharatiya Jan Sangh
BKD Bharatiya Kranti Dal
BLD Bharatiya Lok Dal
BSF Border Security Force
BSP Bahujan Samaj Party
CPI Communist Party of India
CRPF Central Reserve Police Force
DM district magistrate
DSP deputy superintendent of police
FIR First Information Report
INC Indian National Congress
lSI Inter Services Intelligence
LA Legislative Assembly
LS Lok Sabha (Indian Parliament)
MLA Member, Legislative Assembly
MP Member of Parliament

ix
 x / Abbreviations Used in This Book

PAC Provincial Armed Constabulary

PUCL People's Union for Civil Liberties
RPI Republican Party of India
RSS Rashtriya Swayamsevak Sangh
SO station officer
SP Samajwadi Party
SSP senior superintendent of police
SUA standard urban area
U.P. Uttar Pradesh (United Provinces before Independence)
VHP Vishwa Hindu Parishad
 MAPS, FIGURES, AND TABLES

India 4
MAP 1. Aligarh City 151

MAP 2. Upar Kot 154

MAP 3. Location and population composition of mohallas
officially designated as communally sensitive, riot-hit,
and/or crime-prone 164
MAP 4. Riot sites, 1990-91 174

FIG.2.1. Achal Talab, 1999 52

FIG. 3.1. Number of deaths in Hindu-Muslim riots in Aligarh
by five-year periods, 1946-95 68
FIG. 3.2. Shiv temple at akhara site 91
FIG. 3-3. Wrestlers 92
FIG. 3.4. Wrestling pit 93
FIG. 3.5. Police picket at Phul Chauraha 114
FIG. 5.1. Mufti shahar, Aligarh, Upar Kot, December 1984 140
FIG. 6.1. Posh new house in the Civil Lines area, March 1999 153
FIG. 6.2. Mosque, Upar Kot 155
FIG. 6.3. Mosque, Aligarh Muslim University campus, 1962 156
FIG. 6.4. Number of deaths and number of sites of riotous activity in
Aligarh riots in which there were five or more deaths 161
FIG. 6.5. Phul Chauraha 165
FIG. 6.6. Lock manufacturing, Sarai Suitani, Aligarh, 1999 187
FIG. 6.7. Victim of bomb blast in 1990 riots 195
FIG. 6.8. Riot damage, Sarai Sultani, 1999 196

xi
 xii / Maps, Figures, and Tables
FIG. 7.1. Muslim medical clinic in Sarai Hakim 202
FIG. 8.1. Number of deaths in riots by political period 224
FIG. 8.2. Vote shares for Congress and militant Hindu parties,
Legislative Assembly elections, 1952-96 226
FIG. 8.3. Valid votes turnout in Aligarh City Legislative Assembly
elections, 1952-96 228
FIG. 8.4. Percent interval between winning and runner-up candidates,
Aligarh City Legislative Assembly elections, 1952-96 230
FIG. 8.5. Correlations for Congress and militant Hindu candidate vote
shares with percentage of Hindus and others, 1957-91 238
FIG. 8.6. Correlations for militant Hindu party vote shares with
percentage of religious/caste groups, 1957-91 238
FIG. 8.7. Correlations for Congress vote shares with percentage
of religious/caste groups, 1957-91 239
FIG. 9.1. Krishna Kumar Navman, November 1997 244
FIG. 10.1. Vote shares for two leading parties, 1989 Legislative
Assembly election 277
FIG. 10.2. Vote shares for two leading parties, 1991 Legislative
Assembly election 281
FIG. 11.1. Militant Hindu vote share in Aligarh Legislative Assembly
constituency and Manik Chauk mahalia, 1957-91 288
FIG. 11.2. Congress vote in Aligarh Legislative Assembly constituency
and Manik Chauk mohalla, 1957-91 289
FIG. 11.3. Vote shares for all militant Hindu candidates and for the
Congress, Manik Chauk mohalla, 1957-91 290
FIG. 11.4. Militant Hindu party vote shares in Aligarh constituency and
Manik Chauk and Sarai Sultani mohallas, 1980-91 293
FIG. 11.5. Congress vote shares in Aligarh constituency and Manik
Chauk and Sarai Sultani mohallas, 1980-91 293
FIG. 11.6. Party vote shares in Sarai Sultani mohalla, 1980-91 294
FIG. 14.1. PAC encampment, Aligarh, 1999 332

TABLE 2.1. Population (in percentages) of Aligarh City by religion and

caste, 1951-91 47
TABLE 2.2. Mahallas in which particular castes/baradaris/sects are
predominant 54
 Maps, Figures, and Tables / xiii
TABLE 2.3. Caste/community of members of the Aligarh municipal
corporation, 1995 57
TABLE 2.4. Caste/community of members (in percentages) of the
Aligarh municipal corporation, 1995 58
TABLE 2.5. Political identification of members of the Aligarh municipal
corporation by caste/community, 1995 58
TABLE 3.1. Riots and riot deaths in Aligarh City, 1925-95 63
TABLE 6.1. Population of Aligarh City wards, 1951 158
TABLE 6.2. Correspondence between 1951 and 1995 wards and the
distribution of corporators by new ward number 159
TABLE 6.3. Major Aligarh mohallas not included in the 1951 census 175
TABLE 8.1. Winning party or independent candidate, Aligarh City
Legislative Assembly, Lok Sabha, and mayoral elections,
1951-98 222
TABLE 8.2. Winning party in 14 Legislative Assembly contests, Aligarh
constituency, 1952-96 225
TABLE 8.3. Comparison of turnout rates in Aligarh and Uttar Pradesh
Legislative Assembly elections, 1952-96 227
TABLE 8.4. Riots and elections 232
TABLE 8.5. Correlations between votes for militant Hindu candidates
and local population composition across Aligarh mohallas
in selected elections, 1957-91 236
TABLE 8.6. Correlations between votes for Congress and local popula-
tion composition across Aligarh mohallas in selected
elections, 1957-91 236
TAB LE 10.1. Election results for Aligarh City Legislative Assembly
constituency, 1962 263
TABLE 10.2. Vote shares for party candidates in their top and bottom
five polling stations, 1962 Legislative Assembly elections,
and demographic data for the mohallas included in them,
according to the 1951 census 265
TABLE 10.3. Election results for Aligarh City segment of Aligarh Lok
Sabha constituency, 1962 268
TAB LE 10.4. Correlation coefficients of party vote shares with percent
population Muslim, Hindus and others, and Scheduled
Castes, 1962 Legislative Assembly (LA) and Lok Sabha (LS)
elections 271
 xiv / Maps, Figures, and Tables

TABLE 10.5. Election results for Aligarh City Legislative Assembly

constituency, 1989 274
TABLE 10.6. Vote shares for party candidates in their top and bottom
five polling stations, 1989 Legislative Assembly elections,
and demographic data for the mohallas included in them,
according to the 1951 census 275
TAB LE 10.7. Correlation coefficients of party vote shares with percent
population Muslim, Hindus and others, and Scheduled
Castes, 1989 Legislative Assembly elections 278
TABLE 10.8. Election results for Aligarh Legislative Assembly
constituency, 1991 280
TABLE 10.9. Vote shares for party candidates in their top and bottom
polling stations, 1991 elections, and demographic data for
the mohallas included in them, according to the 1951 census
or 1995 voters' lists 282
TABLE 10.10. Election results for Aligarh City Legislative Assembly
constituency, 1993 284
TABLE 12.1. Election results for Aligarh City Legislative Assembly
constituency, 1996 299
TABLE A.I. Demographic data from 1951 census, by Muslim popula-
tion percentage, for sensitive, riot-hit, and crime-prone
mohallas 395
TABLE A.2. Mohallas and other sites identified as riot-hit in major
Aligarh riots, 1956-95 397
TABLE A.3. Vote shares for party candidates in their top and bottom
polling stations, 1957 Legislative Assembly elections, and
demographic data for the mohallas included in them
according to the 1951 census 402
TABLE A.4. The twentieth-century decline in Muslim representation
in the U.P. police 402
TABLE C.1. Community and caste composition of three mohalIas, 1951
census 408
TABLE C.2. Number and percentage of registered voters by community
in Ward 26, 1995 410
TABLE C.3. Party vote shares in Ward 26,1995 corporation elections
410
 PREFACE AND ACKNOWLEDGMENTS

This book follows upon my last two books on collective violence, Riots and
Pogroms and Theft of an Idol, published in 1996 and 1997, respectively.
Although temporally earlier than this book, many of the ideas contained in
them were developed first in my work on riots in Aligarh. It was here, dur-
ing my field work in 1983, that I first developed the notion of the "institu-
tionalized riot system" as a central factor in the produ<..iion of Hindu-Muslim
violence.
I had originally intended to include my work on Hindu-Muslim riots in
Aligarh in Theft of an Idol, but concluded that the material was too extensive
to go alongside the other case studies in that volume. My next thought was
to produce a book focusing specifically on Hindu-Muslim violence based on
my research in several districts of Uttar Pradesh (U .P.), including especially
Meerut and Kanpur, on which I have collected very considerable materials
over the years, as in Aligarh. However, after looking over my interview data
in Aligarh over thirty-eight years and digging into boxes of documentary mate-
rial and election data that I had collected in the same period, I decided finally
on a book in which the city of Aligarh, standing in for so many other cities
and towns in India, would form the center. That decision has allowed me to
do something that I believe is unprecedented in studies of collective violence,
namely, to carry out a diachronic study at a single site, keeping my analysis
sharply focused-so I hope the reader will agree-on the same set of ques-
tions and problems throughout. Although studies have been done of riot-
prone cities (such as, for example, Detroit) that analyze each riot in succession,
those I have looked at treat each riot as something new and different from
its predecessor. Here, on the contrary, I have discovered continuity, exten-
sion, and development of what I intuitively felt in 1983 was an institutional-
ized system of riot production. I now feel that I have established my case in

xv
 xvi / Preface and Acknowledgments

this book and that the findings herein can be generalized to other parts of
India and to other times and places in the world.
I first visited Aligarh in the winter of 1961-62 to carry out field research
for my Ph.D. dissertation on the Congress Party in Uttar Pradesh. That was
a different time in many respects. Aligarh then was a relatively small town
with a population around 185,000, now over half a million. The Congress was
the dominant party in the district. Many of the prominent politicians I inter-
viewed then are now gone. Although party politics then was not lacking in
volatility, bitter conflict, and some violence, it appears relatively genteel in
retrospect compared to the atmosphere of recent years. During the past twenty
years, a new generation of militant Hindu politicians has risen to prominence;
I have met most of the leading persons among them. I have also maintained
and extended my contact with politicians from an other political parties and
organizations in Aligarh, Hindu and Muslim alike, and with members of the
faculty of the Aligarh Muslim University. In most of my visits to Aligarh, I
have always also interviewed key members of the civilian administration and
police, and many subordinate civilian and police officials as well.
Aligarh was very different in 1961-62 in many other respects as well. It was
a relatively much quieter and more peaceful place in general, not only with
respect to incidents of violence. Persons of prominence from the pre-
Independence era were still present in those days, including not only most
senior Congressmen, but men like the Nawab of Chhatari, former leader of
the National Agriculturalist Party and later a member of the Muslim League,
and A. M. Khwaja, a leading so-called nationalist Muslim, and others of sim-
ilar aristocratic or landlord backgrounds. Upper-caste and upper-class per-
sons dominated in all spheres oflife, something that has changed considerably
since then with the rise to self-assertion of the middle and lower castes in
politics. Most of the senior politicians spoke good English then, fewer do so
now. One could breathe the air everywhere in the absence of the internal com-
bustion engine, which now pollutes the atmosphere even in this place far from
any major industrial conurbation.
The Aligarh Muslim University (AMU) in the Civil Lines area of the city,
like the whole area around it, was then a kind of oasis, a quiet, appealing, and
peaceful place, though the AMU simmered internally with conflicts between
so-called conservative/communal and progressive/Communist faculty. The
AMU now has the appearance more of a fortification, surrounded with high
walls in an effort to keep out rowdy, criminal, and other unfriendly elements
from the campus. It is at the same time a place of internal turmoil, where
confrontation and violence between groups of students, students and faculty,
 Preface and Acknowledgments / xvii

faculty against each other, and students and faculty against the vice-chancellor
have occurred repeatedly over the years.
I had selected Aligarh as one of five districts for my research in 1961-62
specifically for the purpose of analyzing how the Congress functioned in an
environment of Hindu-Muslim tension. As if to demonstrate the validity of
my selection of this district for that purpose, my visit, between December 25
and January 20, occurred between the riots of October 1961 and the General
Elections of 1962, held in February. I returned to Aligarh again in September
1962 to continue the research on that district in the aftermath of the elections
that were influenced decisively in the city by the riots that had occurred the
previous October. I did not visit Aligarh again for seventeen years. Since then,
I have visited the city and the district numerous times, for short trips during
elections when I toured U.P. in connection with several election studies
projects, for an extended research period in August 1983, and since then on
several occasions when I have returned to India for research, conferences,
and workshops. On more than a few occasions in those years, I arrived to
find that another riot had recently occurred, or, as in 1990-91, I arrived just
as the great riots of December 1991-January 1992 were coming to an end.
My experiences in this latter respect were mirrored in others of the dis-
tricts that I have visited repeatedly during the past thirty-eight years. So, dur-
ing these later years, I increasingly built in to my research visits to north India
more focused and increasingly systematic questions, interviews, and data on
the reasons for the recrudescence of Hindu-Muslim violence. I continued this
practice during the writing of this manuscript in my most recent visits to
Aligarh in November 1997 and March-April 1999.

I have presented earlier versions of aspects of my research on Hindu-

Muslim violence in Aligarh at many universities, conferences, and workshops
between 1987 and 2000, far too many to note here. It is more important that
I note and acknowledge with appreciation colleagues and others who have
assisted me in the final preparation of this rather complex manuscript. At
the top of the list are two persons who read the entire manuscript in earlier
versions. David Laitin read tlle first version when it was several hundred pages
longer and still in preliminary form. Kanchan Chandra read a complete, but
still imperfect, second draft. The comments of both were indispensable to
me in making the revisions that preceded my submission of the manuscript
for review by the University of Washington Press. Elizabeth Mann read sev-
eral chapters of the earliest version of the manuscript and her comments also
led me to make several changes. Walter Andersen and Richard Flathman gave
 xviii I Preface and Acknowledgments

me the benefit of their comments on particular chapters. The two anony-

mous reviewers for the press and Michael Duckworth, the acquisitions edi-
tor, will, I hope, also note that I have taken their criticisms and suggestions
seriously. Of course, I am alone responsible for the arguments and points of
view adopted as well as any errors that may be found herein.
N aresh Saxena facilitated my visits to Aligarh during the past twenty years.
Kanchan Chandra and Violette Graff provided me with valuable maps of
Aligarh that I had not been able to obtain. Iqbal A. Ansari and Asghar Ali
Engineer cleared up in correspondence with me a few details on which I needed
information. Several persons have accompanied me to Aligarh over the years
to assist me in moving about the city and interpreting when necessary; they
include Pallav Kumar, Gyan and Jayati Chaturvedi, Sumit Mehta, and Aftab
Ahmad. Arup Singh has been unfailingly helpful to me during all my recent
visits to India.
My past practice in citing interviews has been to provide simply the date
and place of the interview. I have modified that practice somewhat in this
manuscript. I have masked most of my sources for interviews. However, I
no longer invariably promise my respondents confidentiality, and carry out
the great majority of my interviews with a tape recorder plainly in view. Since
so much of my material comprises direct quotes that lose part of their signifi-
cance if the identity of the respondent is masked, I felt it important not to
do so in such cases where no confidentiality was promised.
I have been engaged more or less continuously in the research and writ-
ing of this manuscript for the past four years, that is, since my teaching respon-
sibilities at the University of Washington ended in June 1997. In that period,
others also have been extremely helpful to me. They include Irene Joshi, since
retired as the South Asia librarian at the University of Washington, and her
successor in that position, Alan Grosenheider. Michael Shapiro has helped
me from time to time in translating some lines from Hindi newspapers and
from my tape-recorded interviews. Jere Bacharach, Director of the Jackson
School of International Studies, University of Washington, made available a
small grant from Rockefeller funds, which provided partial funding for the
drawing of the maps included herein, prepared by Guirong Zhou. Fred Nick,
director of the Center for Social Science Computation and Research at the
University of Washington, and his staff, especially Dixielynn Gleason, have
been a tremendous help to me on countless occasions with computer and
software problems of all kinds, including recovery from a total and irretrievable
crash of my previous computer ten minutes before the Seattle earthquake of
February 28, 2001, that hit just as I was trying to plug my portable computer
 Preface and Acknowledgments / xix

with the backups for this book into the wall under my table. Thanks to the
portable computer, a few other backup disks, and Fred's help in deciding on
the purchase of a new computer and getting me through the process of reestab-
lishing my work on new software, this book is now presented here.
Susan Halon bore patiently my apparently unending absorption in the
details and complexities involved in the construction of this book. She trav-
elled with me to India and to Aligarh during my last two visits there and took
all but three of the photographs included herein.
While this book was being written, my teacher, Myron Weiner, passed away
on June 3, 1999. It was from him that I first learned the methods of field
research that I have practiced during the past four decades. It was under his
supervision that I carried out the first field research in Aligarh and the other
districts of U .P. in 1961-62. He was himself then also in India, carrying out
the research for his book on the Indian National Congress. He advised, helped,
and encouraged me then and remained a source of inspiration and a friend
to me for the rest of his life, even when my work went off in directions and
with methods and modes of analysis different from his own. I think he would
have liked this book and I dedicate it to him.

P.R.B.
Seattle, Washington
December 16, 2001
 THE PRODUCTION OF

Hindu-Muslim Violence
IN CONTEMPORARY INDIA
 PART I

Introduction
 China

I
" ESH I
,L.."YA
l.>'U1{..... I
I
sh -I
,Bunna(1

!
o

'" IND~ S;ATES ~D UNION TERRITORIES I

• Capitals I
• Riot sites mentioned in the text JI
--'---'
 1/ Explaining Communal Violence

A mong the many paradoxes and contradictions that must confront

observers of India is the competing imagery of violence and non-
violence, symbolized in two recurrent representations of that coun-
try. One is the image that has been flashed countless times during the past
half-century in the media and the cinema of the bloody riots that occurred
immediately before and after Independence as a consequence of the events
associated with the simultaneous partition of the country into two new, mutu-
ally hostile sovereign states that immediately fought their first war and have
since fought two more. The second image is that of the saintly Mahatma
Gandhi traversing the country for decades proclaiming the message of non-
violence and devising strategies of nonviolent opposition to British rule that
have since been adopted round the world by the weak to fight against
exploitation and discrimination by the strong and privileged. The two images
merge in Attenborough's film, Gandhi, when Gandhi appears in Noakhali in
the province of Bengal during the pre-partition riots there to end the killing.
He is shown lying on his bed, fasting to death against the violence, which is
brought to an end as the repentant, weeping murderers deposit their weapons
at his side.
Forty-five years after Independence, the world was presented with another
image of India, that of violent mobs of Hindus descending upon the old,
mainly Hindu religious town of Ayodhya to climb upon a five-hundred-year-
old mosque to destroy it. This image was then followed by the pictures flashed
round the world of Bombay in flames from the riots that followed after the
destruction of the mosque a thousand miles away. Few people outside India,
however, knew that similar riots also took place in cities and towns in large
parts of the country, in which Muslims, having seen one of their mosques
destroyed on BBe television or having otherwise learned of it, were now being

5
 6/ Explaining Communal Violence

slaughtered, allegedly because they came out into the streets in shock and out-
rage and engaged in riotous behavior.
Even fewer people-indeed, only specialists-know that Hindu-Muslim
riots and anti-Muslim pogroms have been endemic in India since Indepen-
dence. I They have occurred and recurred in many cities and towns through-
out the country, but especially in the northern and western parts. Their
frequency and intensity have fluctuated from time to time and place to place,
but hardly a month passes in India in which a Hindu-Muslim riot does not
occur that is large enough to be noted in the press. But there are also many
such events on a smaller scale that occur much more frequently. Indeed, it
is likely that not a day passes without many instances of quarrels, fights, and
fracases between Hindus and Muslims in different places in India, many of
which carry the potential for conversion into large-scale riots in which arson,
looting, and killing may take place.
Neither in December 1992, nor on most of the occasions between Inde-
pendence and 1992 in which so much destruction of people's lives, homes,
and property have occurred, have many saintly figures appeared to quell the
violence. In fact, both these images-of frenzied, murderous masses in India
and saintly figures moving about spreading their message of nonviolence as
a cure for their frenzy-are part of a grand discourse of violence that I hope
to undermine in this book. Riots are not explained by the spontaneous furies
of mad mobs nor are there any weeping murderers among them nor can they
normally be stopped by saints.
On the contrary, it is a principal argument of this book that the whole polit-
ical order in post-Independence north India and many, if not most of its lead-
ing as well as local actors-more markedly so since the death of Nehru-have
become implicated in the persistence of Hindu-Muslim riots. These riots have
had concrete benefits for particular political organizations as well as larger
political uses. Hindu-Muslim opposition, tensions, and violence have pro-
vided the principal justification and the primary source of strength for the
political existence of some local political organizations in many cities and
towns in north India linked to a family of militant Hindu nationalist organi-
zations whose core is an organization founded in 1925, known as the Rashtriya
Swayamsevak Sangh (RSS). Included in this family, generally called the Sangh
Parivar, are an array of organizations devoted to different tasks: mass mobi-
lization, political organization, recruitment of students, women, and work-
ers, and paramilitary training. The leading political organization in this family,
originally called the Jan Sangh, is now the Bharatiya Janata Party (BIP), cur-
rently (2001) the predominant party in India's governing coalition. All the
 Explaining Communal Violence 17
organizations in the RSS family of militant Hindu organizations adhere to a
broader ideology of Hindutva, of Hindu nationalism that theoretically exists
independently of Hindu-Muslim antagonisms, but in practice has thrived only
when that opposition is explicitly or implicitly present. 2
The benefits for the consolidation of Hindu communal sentiment behind
the organizations of militant Hindu nationalism in the RSS family of orga-
nizations and the Shiv Sena in the western state of Maharashtra (which also
adheres to an ideology of militant Hindu nationalism) have been great: they
have served to bring these parties to power in numerous states in India out-
side the south and have at last brought them to power at the Center,3 in New
Delhi, as well. These formations have launched numerous Hindu-oriented
campaigns since Independence in which the Muslims have been portrayed
directly as obstacles to the achievement of national aspirations or have been
clearly assumed to be the main obstacle. The two most massive such cam-
paigns were the cow protection movement of the mid-196os and the Ayodhya
movement of the mid-1980s and early 1990s. In the former movement, there
was no Muslim structure to stand as the centerpiece to be brought down;
but, obviously, it was the Muslims of India, who slaughtered and ate beef,
whose opposition was implied. The political context at that time, however,
did not provide the same potential benefits for the RSS family of organiza-
tions as the context that existed in the 1980s during the Ayodhya movement. 4
For, after a brief resurgence in its strength after the assassination of Indira
Gandhi in 1984 and the installation of her son, Rajiv Gandhi, as prime min-
ister, the Indian National Congress, India's dominant ruling party since
Independence, began a severe decline in its fortunes in U.P., the most
important state in the country, which has since spread to most of the other
Indian states as well.
The years 1988-89 were to mark the beginning of the end of Congress dom-
inance in the state and the country. In a bitterly contested national election
campaign centered around north Indian personalities, the Congress was
defeated by a coalition of non-Congress parties in 1989. When that coalition
itself disintegrated and new elections were held in 1991, the leader of the
Congress, Rajiv Gandhi, was assassinated in the midst of the campaign; the
Congress was nevertheless once again returned to power under Prime
Minister P. V. Narasimha Rao, whose government lasted its full parliamen-
tary term till 1996. While the Congress was weakening even during its last
term in power after losing its popular leadership from the Nehru-Gandhi fam-
ily, and while its rivals among the left parties were unable to consolidate their
strength and emerge as a stable alternative governing force, the militant Hindu
 8/ Explaining Communal Violence

organizations, including the BJP, were gathering strength from the mass mobi-
lizations associated with the Ayodhya movement.
The Ayodhya movement made explicit use of a Muslim structure, a
mosque, which stood for a religion that militant Hindus-and many non-
militant Hindus as well, for that matter-disdain as foreign, immoral, and
evil. But, even more important, just five years short of the celebration of India's
fiftieth year of independence from British rule, this mosque stood in the minds
of its enemies as the mark of an earlier "slavery," as they called it, of Hindu
subjection to Muslim rule. The movement also created martyrs, especially
after the killing of sixteen Hindus in Ayodhya in 1990, who were then por-
trayed in a grossly exaggerated and fantastic manner as the latest in a long
series of martyrs to the cause of the removal of the defiling mosque from its
place in the Hindu town of Ayodhya on the allegedly sacred ground of the
god, Ram. This movement was also accompanied throughout north India by
deliberate provocations directed against Muslims, whose effect was certain
to bring down violence upon them, as it did massively in many cities and towns
in northern and western India between 1989 and 1993. 5 The movement also
contributed significantly, along with other changes taking place in India at
the time-particularly the controversy over reservation of places in public
sector jobs for a large section of India's Hindus commonly referred to as
"backward castes"-to a major transformation in the sources of strength of
the BIP, to such an extent as to tum tlle BJP of the 1991 and 1996 elections in
its northern India stronghold state of Uttar Pradesh (U.P.) into a virtually
new and much stronger political formation than ever before. 6 It should be
clear enough by now, therefore, how valuable Hindu-Muslim opposition,
antagonism, and violence have been to the fortunes of the BJP.
But Hindu-Muslim riots in India obviously did not begin only with the
Ayodhya movement. They have been a recurring feature of modern Indian
politics for nearly a century. Moreover, there have been periods during which
Hindu -Muslim rioting has occurred in what are commonly referred to as great
"waves" or "chains." These periods include especially the years 1923-27, after
the collapse of the noncooperation/Khilafat movement against British rule
in India in which Hindu and Muslim political and religious organizations
and groups worked together; 1946-48, when massive waves of rioting and mas-
sacres preceded, accompanied, and followed the partition of India and the
consequent formation of the two successor states to the British Raj of India
and Pakistan; and the succession of riots that occurred between 1989 and 1993
during the Ayodhya movement.
But it would be a mistake to confine our gaze only to these great waves of
 Explaining Communal Violence /9

rioting, for there has never been a period in modem Indian history, most espe-
cially in the north, when Hindu-Muslim riots have not occurred. It is a cru-
cial part of the argument to be developed and demonstrated in this volume
that the maintenance of communal tensions, accompanied from time to time
by lethal rioting at specific sites, is essential for the maintenance of militant
Hindu nationalism, but also has uses for other political parties, organizations,
and even the state and central governments. It is necessary, therefore, for a
fuller understanding of the phenomenon of the persistence of Hindu-
Muslim rioting and its manifestation from time to time in great waves, to
examine as well its appearance in relatively quieter times and at sites where
it is endemic.
My first question, therefore, is: why do Hindu-Muslim communal riots
persist in India? Or, put another way, how have such riots become endemic
in that country? Consider in this connection the available data on Hindu-
Muslim riots for the period 1960-93.7 Using a restrictive definition of riot-
proneness, Varshney and Wilkinson have pointed out that the incidence of
Hindu-Muslim communal riots in India is skewed towards urban India in
general and towards 24 cities in particular. There are, therefore, only certain
sites in which riots may be considered endemic. At the same time, while the
incidence and timing of Hindu-Muslim riots vary from region to region
and city to city, it is not incorrect to consider India as a whole a country in
which Hindu-Muslim riots persist and are endemic. First of all, the number
of such riots in the worst-hit cities account for only half the total incidents
in the country. Second, even the worst-hit cities are scattered throughout a
very large part of the country, in ten of its states. Third, from time to time,
new sites that have either never before experienced large-scale Hindu-
Muslim riots, or have not experienced them in several decades, have entered
or reentered the lists.
Given the situation just described, a satisfactory explanation of the phe-
nomenon of persistence must account both for the dispersion of rioting in
India in time and space and for its concentration in particular sites. The expla-
nation to be provided in this volume will attempt to encompass India as a
whole through an analysis of the discursive framework of communalism that
affects, however differentially, all parts of the country where Hindus and
Muslims abide side by side. However, I will focus the detailed analysis of riot
production on a single site, the town of Aligarh, where riots have persisted
since Independence, and which stands as a choice exemplar of riot persis-
tence for other reasons as well, especially because of the presence there of the
Aligarh Muslim University (AMU). The Pakistan movement itself grew out
 10 / Explaining Communal Violence

of conditions in U.P., and the AMU was one of its storm centers. Conse-
quently, in the minds of many Hindus, the AMU stands in for the Muslims
of India, for Partition and the creation of Pakistan, and for so many of the
ills that afflict Indian society. Without the presence of the AMU, the Jan Sangh,
the BIP, and other local communal groups would have had greater difficulty
in establishing a strong presence in this city.
A second question is: why do Hindu-Muslim riots ebb and flow, appear-
ing now here, now there at different times? As noted above, Hindu-Muslim
riots sometimes occur in what appear to be great waves that spread from town
to town and region to region of the country, affecting a great number of sites
either simultaneously or one after the other. Those great waves are usually
associated with large-scale political movements that precede them. On the
other hand, there has never been an extended period of time in India since
Independence when Hindu-Muslim riots have not occurred in some town
or other, apparently unrelated to any broader movement in a region or in
the country as a whole. So, there is persistence as well as variation in the coun-
try as a whole and at particular sites.
A third question concerns how it happens that large-scale violent events,
in which mostly Muslims are killed, mostly by the police, get classified in the
press, by the authorities, and by the public as riots rather than pogroms. I have
argued elsewhere that there are two types of struggles that take place when
riots occur. The first is the violent conflict between riotous groups and between
rioters and the police. The second is the rhetorical struggle that takes place
afterwards to control the interpretation of the riot, determine its meaning,
explain the violence. s It is at this stage that classification occurs. However,
such classification is often automatic: often one explanation emerges dom-
inantly, and sometimes a hegemonic consensus arises that lasts for a long
time in the form of a master narrative that requires no knowledge of facts
on the ground for its immediate acceptance. 9 Such a master narrative exists
in India, comprising two key elements. First, riots in general are perceived
as spontaneous occurrences that arise out of petty quarrels that become con-
verted into mass frenzies through the spread of rumors that exaggerate the
precipitating incident. Second, Hindu-Muslim riots in particular are said to
arise from the prejudices and hostilities that exist between these two reli-
gious groups, such that there is a natural tendency to expand any quarrel
between a Hindu and a Muslim into a riot. In order to prevent such con-
versions of quarrels into riots, it is commonly urged, peace committees and
other forms of intercommunal cooperation need to be developed to com-
bat them. It is in this way, almost automatically and without reflection or
 Explaining Communal Violence / 11

challenge, that trivial incidents involving Hindus and Muslims that precede
large-scale riots are said to have been their cause. The automatic mechanism
that produces such an explanation is based upon the deep-seated belief that
popular passions are aroused as much by a preexisting history of commu-
nal antagonisms and a pervasive atmosphere of tension between Hindus and
Muslims as by the actual or perceived circumstances surrounding the pre-
cipitating incident. It is my purpose in this volume to demonstrate that nei-
ther the prior history of communalism nor the immediate circumstances
surrounding the so-called precipitating incident, nor the two alleged causes
combined, provide a satisfactory explanation for the outbreak of large-scale
Hindu-Muslim riots or anti-Muslim pogroms. On the contrary, the deci-
sive factor is the action that takes place before the precipitating incidents
and immediately thereafter, action that is often planned and organized and
that fills the intermediate space and time between past history and imme-
diate circumstance.
The fourth question asks: what interests are served and what power rela-
tions are maintained as a consequence of the wide acceptance of the real-
ity of popular communal antagonisms and the inevitability of communal
violence?
These four questions may be summed up in a nutshell as issues of per-
sistence, differential incidence/timing, classification/meaning, and power.
They are large questions that require diverse approaches. They raise issues
of causality, function, and discourse. Issues of persistence and incidence!
timing seem to require causal and/or functional analysis: why do riots per-
sist, occur here and not there, occur now and not then? Whose interests are
served by the occurrence, persistence, or disappearance of Hindu-Muslim
riots? The struggle over meaning, explanations, and power relations requires
attention to discourse. To what extent is there a communal discourse that
accounts for the persistence of communal rioting over time by providing a
framework of explanation and meaning, an ordering of relations between
Hindus and Muslims, and an ordering of the respective relations of these two
categorical groups to the state? Further, to what extent does such a discourse
itself contribute to the persistence of the violence that it claims to explain? I
will make use of methods appropriate to each type of analysis in an attempt
to arrive at as full a picture as possible of the mechanisms that lead to the
production of riots in India and of the dynamic processes that precede them,
produce them, and explain them after their occurrence. The framework that
unites the various themes and questions pursued herein is a theoretical for-
mulation, a kind of ideal type, of the manner in which large-scale riots are
 12/ Explaining Communal Violence

produced in sites where riots are endemic, whether in Aligarh, other parts of
India, or other parts of the world.

THE DYNAMICS OF RIOT PRODUCTION

The primary approach taken in this volume to the dynamic process of riot
production has some affinity with theories of collective action and social move-
ments, developed most notably in the works of Tilly, Tarrow, McAdam and
others of their colleagues. lO That affinity as well as the differences were noted
in part in Theft of an Idol," but will be restated and elaborated briefly here.
In agreement with this group of scholars, riots are conceived in this book as
a form of collective action, one among a number of repertoires of collective
action that developed in India primarily in the late nineteenth and twenti-
eth centuries. As much, therefore, as the great movements of noncoopera-
tion and civil disobedience and a whole host of other nonviolent forms of
agitation, demonstration, and protest, riots have become a common and even
an anticipated form of collective action.
The term "anticipated" is used here in two senses. People expect riots to
occur from time to time and in certain places in India without being able to
predict exactly when and where they will occur, but they also anticipate and
expect riots during particular types of mass mobilization. Euphemistic terms
have even been developed for the latter, such as "direct action." A different
rhetoric is also used during movements in which Hindu-Muslim violence is
anticipated, an inflammatory rhetoric of hostility with scarcely-veiled encour-
agement to listeners to act out against the other community. Most commonly,
the rhetoric is laced with words that encourage its members not to put up
any longer with the attacks of the other but to retaliate against their aggres-
sion. There are also specific forms of action that are designed to provoke the
other community into aggressive action, which is then met with a stronger
retaliatory response. These forms include, especially, processions through
neighborhoods inhabited primarily by persons from the other community,
and the insistence by processionists that shopkeepers "down their shutters"
and dose their shops to honor whatever demand is being made during a
demonstration. In the latter case, a demonstration ostensibly directed against
the state or local administration may turn into a communal riot.
Consistent with the Tilly-Tarrow conceptualization of the development
of social movements, every great wave of rioting in modern India has been
preceded by new mobilizing tactics that become integrated into the new reper-
toire and promote violence. For example, between 1923 and 1927, the rioting
 Explaining Communal Violence 113

was accompanied by competitive movements for the conversion and recon-

version of Hindus and Muslims in many localities in India-again especially
in the north and west-to the other religion. Also common at this time was
the interference of political and religious organizations in religious processions
organized by the other community, especially the opposition of Muslims to
the Hindu Ram Lila processions that marched through localities in which
Muslims were concentrated. In contrast, the great massacres of 1946 to 1948
were more directly linked to political actions and mobilizations around the
demand for Pakistan.
The various forms of religious and political mobilization that were devel-
oped in these earlier waves of violence have persisted into the present, but
they have been subordinated to more direct appeals to Hindu religious sen-
timent and solidarity that confront and directly offend Muslim religious sen-
timents. These have included yatras (journeys) from one Hindu religious site
to another or from one emotionally charged nationalist site to another, of
which the most famous was the ruth yatru12 of the BJP leader 1. K. Advani in
1990 from the Hindu temple of Somnath in the western state of Gujarat to
the Babri Mosque at Ayodhya. Although this yatra was aborted before it
reached U.P., it gathered a massive following of Hindus and was accompa-
nied by rioting in its wake at numerous sites through which it passed. Other
actions associated with this movement included the carrying of bricks from
all parts of India-and even from far-flung parts of the world where Hindus
reside-to be consecrated and carried to Ayodhya to be used in the con-
struction of the new temple to Ram at the site of the destroyed mosque; the
carrying of the ashes of "martyrs" killed by the police at Ayodhya during the
first assault on the mosque in 1990; and the travel of thousands of kar sevaks
(volunteers) to Ayodhya to participate in the work of construction itself. In
the localities of India, the passage of the ruth yatra or the movement of kar
sevaks was preceded or accompanied by "street corner meetings, the blow-
ing of shankh (conch shell), clanging of ghanta-gharial (ringing of prayer bells
and striking at a plate of alloyed metals), hoisting saffron flags in the day-
time and mashals (flaming torches) at night on terraces, and organizing mashal
jaloos (processions bearing lighted torches)."13 In the western state of
Maharashtra, the Shiv Sen a adopted the tactic of blocking the Bombay
streets with maha-aartis, that is, gatherings for mass Hindu worship, mim-
icking the similar street-blocking that had for long accompanied Muslim
namaz (worship) at prayer times.
Although, therefore, Hindu-Muslim and other forms of collective violence
in India are expected, anticipated, have a fairly stable set of forms of action
 14 / Explaining Communal Violence

that are fortified by new forms from time to time, and a distinct rhetorical
form as well, they differ from the nonviolent forms in critical ways aside from
the violence itself. However frequent and anticipated and however accom-
panied by new forms of mobilization that become integrated and legitimized
in a repertoire of mobilizing acts, the riots that follow from them are illegit-
imate. Their illegitimacy, moreover, is acknowledged both by those who
deplore the violence and by those who enact it. The authorities and the English-
language press invariably condemn the violence that ensues after Hindu-
Muslim riots, but the promoters of the violence also recognize its illegitimacy
by claiming that the aggressor community was not aggressing, but was acting
only in desperation in defense against the attacks of the other.
Another critical difference between these violent and nonviolent move-
ments follows from the illegitimacy of the former. 14 Their violent manifes-
tations appear spontaneous, undirected, unplanned-and even the most
carefully planned and well-organized assaults on the other community are
designed to appear so. Since such riotous violence is illegitimate and the ele-
ments of preplanning in it are disguised, the struggle that takes place after-
wards to explain it-that is, to control its interpretation-is crucial. The most
common explanation is that the violence was in fact an unplanned, sponta-
neous expression of the deep feelings of an aggrieved people, but there are
many others that will be illustrated in this volume. Here I want to note mostly
the multiple functions served by capturing the meaning of a Hindu-Muslim
or any other intercommunal, interreligious, interethnic riot in a particular
way. These include legitimizing illegitimate violence, concealing the extent
of preplanning and organi7.ation that preceded it, and maintaining intact the
persons, groups, and organizations most deeply implicated in the violence
by preventing punishment of the principal perpetrators.
The illegitimate and hidden aspects of riotous violence have posed almost
insurmountable obstacles to those who have set out to analyze them. Only
one author, not an academic, has dared to engage in actual participant obser-
vation of riotS.15 Most who care to ascertain the "facts on the ground" appear
after the riots to interview their victims. Within a few days, however, often
even the victims, coached by lawyers, have prepared their accounts, been
advised what to emphasize, who to name and blame, what to conceal. Most
social science studies of riots in the West have not even been based on any
direct or indirect observation of the events themselves, but on information
derived from census and other "datasets." In this study, a variety of instru-
ments and methods have been used to penetrate the dynamic processes of
riot production in contemporary India, which will be discussed below.
 Explaining Communal Violence 115

Before doing so, however, I want to outIine the specific structure and argu-
ment that will be used to organize the information and data collected.

Phases in the Production of Riots

The structure and argument to be developed in this volume depart from

nearly all previous studies of collective violence by insisting that, in sites where
riots of a particular type are endemic, they are a grisly form of dramatic pro-
duction in which there are three phases: preparation/rehearsal, activation/
enactment, and explanation/interpretation. In sites of endemic riot pro-
duction, preparation and rehearsal are continuous activities. Activation or
enactment of a large-scale riot takes place under particular circumstances,
most notably in the case of competitive political systems in a context of intense
political mobilization or electoral competition in which riots are precipitated
as a device to consolidate tile support of ethnic, religious, or other culturally
marked groups by emphasizing the need for solidarity in face of the rival com-
munal group. The third phase follows after the violence in a broader strug-
gle within, but also outside, the local community to control the explanation
or interpretation of the causes of the violence. In this phase, wider elements
in society become involved, including journalists, politicians, social scien-
tists, and public opinion generally. This third phase is marked by a process
of blame displacement in which the social scientists themselves become impli-
cated, a process that does not isolate effectively those most responsible for
the production of violence, but diffuses blame widely, blurring responsibil-
ity, and thereby contributing to the perpetuation of violent productions in
future.
In India, all this takes place within a discourse of Hindu-Muslim hostil-
ity. The second major theme of the book concerns the link between this dis-
course of Hindu-Muslim antagonism and the practice of violence enacted
in what are called Hindu-Muslim riots. The third theme is that the practice
of violence, especially in sites where riots are endemic, becomes embedded
in what I call institutionalized riot systems. These systems of riot production
are marked by two interrelated features: the existence of a multiplicity of roles,
in which there is a specialized division of labor.
It is an essential part of the argument herein that the third phase in riot
production, the interpretation phase-and the struggle for control over the
explanation of riots that occurs during this phase-is as important as the pro-
duction itself, as in any dramatic production. Finally, the keynote of this phase,
suffusing it, is a process of blame displacement in which all are involved and
 16/ Explaining Communal Violence

whose end result is the diffusion of responsibility in such a way as to free all
from blame and allow the principal perpetrators to go scot-free.
The explanation of riot production to be developed herein relegates all
spontaneity theories of the causes of riots to the realm of blame displace-
ment. It is the most common form of blame displacement, in fact. Such the-
ories are at worst utterly false, at best-and invariably so-misleading.
Although the primary focus of the book, therefore, is on the dynamic
process of riot production, that is, on how riots are produced rather than on
why they happen, several causal explanations of Hindu-Muslim violence in
India will nevertheless be examined. In the course of this examination, con-
ventional social science techniques of correlation and regression will be used
to uncover associations from which causal explanations are often inferred.
For the rest of my analysis both of the functional utility of persisting riots to
a multiplicity of social, economic, and political groups and of the struggle to
control the meaning of riots, the principal sources will be interviews and per-
sonal observations carried out in India during field trips over the thirty-eight
years from 1961 to 1999. Newspaper accounts of riots will also be used, though
my experience with such accounts is that they are invariably deficient, some-
times utterly false, and altogether inadequate for serious social science research
on the subject of collective violence. Public and confidential documents will
also be used, but they are sparse on this subject and some potentially useful
reports are inaccessible.
Before laying out the outline of this volume, I want to take up some
methodological issues concerning causal analysis of collective violence and
indicate how I will make use of causal explanations as well as alternative meth-
ods of functional analysis and the analysis of discursive formations in inter-
preting and understanding the incidence and dynamics of the production of
riots and pogroms in contemporary India.

EXPLANATION: CAUSE, FUNCTION, DISCOURSE

Problems of Causal Analysis in the Comparative Study

of Forms of Collective Violence
Virtually every scholar who has written about riots, pogroms, and other forms
of collective violence seeks their causes, and not a few scholarly articles fea-
ture the word in their titles.16 Many of those who seek causes also seek cures-
as if dealing with an illness-or solutions, as if dealing with a social problem.
Commissions of inquiry are often appointed after particularly serious riots
or after a wave of riots. Their charge is invariably to determine the causes of
 Explaining Communal Violence /17

the riot or riots in question. Their purpose is often different from the work
of scholars. It is to assign responsibility, especially upon the state authorities
and their agents, administrative officers, and police. In this case, riots are seen
as problems for the authorities, disturbances of law and order in which their
own competence and effectiveness in allowing riots to occur and in failing
to control them when they do occur are judged. Such an approach introduces
a secondary search for causes, namely, for the failure to control a riot in its
early stages and allowing it to get out of hand.
Numerous problems have arisen in the literature that seeks causes for riots.
One is that participants in riots often do not conform to expectations based
on the imputed causes. For example, many participants in riots whose causes
have been said to be poverty and unemployment turn out to be employed. 17
Such a finding raises another kind of causal issue, that of individual moti-
vation for participation. But the two sets of findings, one based on ecologi-
cal factors, the other on individual motivations, may conflict. In statistics,
this is, of course, the problem of the ecological fallacy.
It will be argued throughout this volume that it is essential to the under-
standing of the dynamics of riot production that we separate explanations for
riots seen as crowd behavior from explanations for individual behavior in riots.
Yet, this is not an easy task, for the former usually take the form of justifications
masquerading as explanations, while individual actions are normally hidden
from view in the hubbub and furor during large-scale rioting, after which indi-
vidual motivations are obscured by the public explanations, especially those
offered by political persons. 18 Three solutions to this analytical problem are
proposed here. The first i" to note how a discourse of public good and evil
becomes integrated into individual thought and behavior, thereby providing
an internal motivation for the instigation of and participation in acts of vio-
lence. Second, I note throughout the volume the multiplicity of types of vio-
lent action that occur during large-scale rioting under the cover of the discourse
of communalism, actions that cannot be explained or justified in terms of that
discourse, but can easily be fit into more parsimonious explanations of indi-
vidual pursuit of political advantage, profit, and vendetta. But it must at the
same time be recognized that there can be no perfect separation between pub-
lic explanations and private motivations, that the ecological fallacy is not merely
a problem of methodology but a deliberate and often impenetrable form of
political and public obfuscation. Third, all forms of deep psychological inter-
pretation of individual motivations for participation in riotous activity will be
eschewed as essentially futile; instead, the focus will be on the hidden face
itself, what can be made visible behind the mask of discourse.
 18/ Explaining Communal Violence

Another problem concerns the relationship between the immediate acts

that precipitate riots and the "underlying causes" that make it possible for
such acts to be followed by large-scale crowd violence. There is often a dis-
crepancy here, as well, between imputed underlying causes and the immedi-
ate acts that precipitate an event classed as a riot. It is not, say some scholars,
the pig in the mosque or some other insult to the religious beliefs of Muslims
or Hindus that is responsible for riots in India, but the economic issues that
lie behind these incidents; rivalry between Muslim and Hindu manufactur-
ers and wholesalers in urban areas, disputes between Hindu and Muslim vil-
lagers over land.'9
A critical problem in assessing the relationship "between underlying
causes and immediate precipitants of racial disturbances" in the United States
has been that the former are "relatively stable," whereas the latter "are ran-
dom occurrences, the kind of events which occur daily in most communi-
ties and usually are disposed of routinely. "20 Spilerman found little or no
connection or correlation between the standard list of underlying causes and
the occurrence of black ghetto riots in the 1960s. Instead, he found that the
strongest predictor variable was the size of the Negro population. He then
asked the question, which is of central importance in this study as well, how
one explains the escal.ation of random, routine occurrences that affected black-
white relations in American cities into large-scale riots. His answer is the exis-
tence of racial consciousness among the black population that allows
"bystanders to the conflict to interpret it in primarily racial terms," and to
respond accordingly, a consciousness stimulated not by differences in the
extent of unemployment, dilapidated housing conditions, and other economic
disadvantages experienced by blacks, but by racial solidarity produced uni-
formly across the black population in the United States, primarily by means
of television coverage of the problems confronting them throughout the coun-
try. In this study, it will be demonstrated that not television, but a pervasive
discourse that emphasizes Hindu-Muslim differences and hostilities in India,
provides the framework that allows the escalation of trivial incidents into
major riots. But this process also is neither automatic nor spontaneous; it
requires the presence of other factors that will be spelled out below, princi-
pally political-process factors that Spilerman does not discuss.
Social scientists and political scientists are divided among methodologi-
cal individualists, that is, rational choice and game theorists, devotees of the
case study method, and dataset enthusiasts. Between and within these alter-
native approaches, there are other divisions as welL In the search for causes,
is it not necessary, say some, to consider places where riotous violence has
 Explaining Communal Violence / 19

not occurred as well as those where it has occurred? The question here is:
why here and not there? What are the social, economic, and political differ-
ences between places where riots have occurred and places that have been
free from riots? One method adopted to answer this kind of question is "paired
comparison analysis," which can be done by using statistical data 21 or by his-
torical narrative comparison. 22 Either way, the method is designed to pair
two sites or two sets of sites that are as much alike as possible except in the
one respect under consideration (the dependent variable), namely whether
or not they have experienced incidents of collective violence.
Game theorists, notably Fearon and Laitin,23 have sought to develop com-
prehensive theories of contlict, including interethnic contlict. Like the pro-
ponents of paired comparison, they have also challenged the validity for the
development of a theory of interethnic violence of an exclusive focus on the
situations that lead to violence, arguing that, in fact, interethnic cooperation
rather than violence is the norm. What is required, they argue, is a theory
that explains both why violence is so infrequent, given the numbers of
potential conflicts in the world or any part of it, and why the norm some-
times breaks down. They have developed a model to explain both situations,
derived from a theory of games, in which the crucial issue concerns the con-
trol of individual opportunism that threatens the relations between ethnic
groups. They identify two mechanisms that are used for such control. The
first, which they call the spiral equilibrium, controls individual opportunism
through the fear based on knowledge that an infraction of interethnic rela-
tions will lead to mutual violence that will spiral out of control. The knowl-
edge is sufficient most of the time to prevent such dangerous infractions, but
when it is not, the violence that ensues is likely to be awful.
The second mechanism is "in-group policing" by which any member of
one ethnic group who commits an infraction against a member of another
ethnic group will be disciplined by his own group. Violence between ethnic
groups ensues only when the intraethnic policing mechanisms fail. The authors
are aware that their model leaves out many other critical factors, whose inclu-
sion would require a further extension of it, but have provided examples of
empirical situations to demonstrate its applicability. There will be occasion
to refer later to examples that relate to their hypotheses in the discussions of
riots in Aligarh. For the present, it is sufficient to note first that riots can fol-
low from a breakdown of either equilibrium situation, but that the frequency
of rioting and the type that has occurred in Aligarh since Independence falls
primarily in the category of the spiral equilibrium that repeatedly breaks down.
Second, as will be argued later, a different game is being played in Aligarh, a
 20 / Explaining Communal Violence
political game of brinkmanship whose purposes are not to maintain intereth-
nic cooperation and prevent violence, but to keep always in readiness the
mechanisms to bring group relations to the brink of conflict, and to let the
violence loose at times considered advantageous to one side or the other or
both.
Whatever the method in the literature on collective violence, the search
for causes remains primary. I believe it is overemphasized and often mis-
placed, for the following reasons. First, for aU the scientific pretensions of
causal analysts, the search for causes cannot be separated from the values of
the observer, whether politician, judicial enquiry commissioner, scholar, or
journalist. It is obvious in the case of the politicians,24 more subtle in the
case of social scientists and historians, but it is nevertheless present amongst
all, whether consciously or unconsciously, by design or in the implications
of our findings. Indeed, all categories of persons just mentioned merged to
produce probably the most famous riot inquiry of the twentieth century,
the Kerner Commission Report on the racial disturbances in the 1960s in
the United States: former politician Governor Kerner, acting as judicial com-
missioner, other politicians, and members of interest groups-racial, labor,
and business-a team of social scientists, and, of course, all making use of
newspaper reports as well as their own sources for their findings. 25
Second, the search for causes easily turns from an expression of the val-
ues of the observer and his identification with those perceived as the victims
of violence to the assignment ofblame.26 If the cause of riots is system strain,27
structural conditions, or any general condition prevailing in society, then
blame may be dispersed or dissolved or it may be directed towards the regime
considered responsible for the strains. If it is perceived as poverty, inequal-
ity, unemployment, or discrimination, again, depending upon whether these
in tum are considered aspects of a general social transformation or are seen
as themselves caused by state policies, blame may be dispersed or concen-
trated on the regime. In the case of interethnic violence, cause and blame may
be placed upon the prejudices of particular groups and their upbringing, their
"family values," or upon objective conditions of economic interaction, com-
petition, or perceived exploitation of one group by another. And so on.
All these forms of causal analysis are deficient, Keith argues, because of
their adherence to a Humean model of probability that makes use of state-
ments of the form "if x, then y," or, put more precisely, statements that most
probably the phenomenon y (in our case standing for riots) will follow when-
ever x (the cause or causes) recurs in a social situation. This kind of focus
ignores or downplays in particular three aspects of human action and strug-
 Explaining Communal Violence / 21

gle, namely, intentionality, process, and meaning. 28 In other words, first, it

ignores or underplays the self-fulfilling and self-denying prophecies, the pos-
sibility that riots may be willed actions, concerted productions of thinking,
acting people who may also decide to cancel a performance. Second, follow-
ing from the first, it ignores the dynamic processes of riot production, being
satisfied instead with explanations that focus on social, political, or economic
conditions or on spontaneous crowd responses to stimuli such as rumors or
atrocities perpetrated against ethnic compatriots by members of a rival eth-
nic group or police brutality against the former. Third, it ignores the con-
stant struggle to control the meaning of riots after they occur, to represent
them appropriately, which then feeds back into both common-sense and
social-science causal explanations converted into cures. The primary focus
in this volume will be on these three elements of riot production and repre-
sentation, from out of which I will also seek to generate causal statements
that are limited and cautious with regard to their truth claims. The causal
statements that will be made herein, however, will not be of the "if ... then"
variety. It will not be said, for example, that, if the possibility for enhancing
a party's electoral chances will be increased by a riot, there is a high proba-
bility that a riot will take place. It will be said only that large-scale riots are
associated with intensified party/electoral competition in which the causal
arrow points from riots to intensified competition rather than the reverse. It
will also be inferred that the instigators of such riots have in mind such a result
in which their side will be favored. 29
Permeating all the forms and varieties of causal analysis applied to the
study of collective violence-sometimes hidden, but often quite explicit-
is a foundational substratum of opposition between those who attribute col-
lective violence primarily to the prejudices, hostilities, aggressions, and
propensities to violence of particular peoples, segments of them, or the pop-
ulace in general, on the one hand, and those who argue to the contrary that
one must look elsewhere for explanations of violence, not only to other under-
lying causes, but to the mechanisms for crowd mobilization. In its simplest
form, the opposition is between those who perceive collective violence, espe-
cially that which takes the form of riots, as arising from the acting out spon-
taneously by mobs of feelings that pervade a population, on the one hand,
and those who perceive mass mobilization as a difficult task at best that requires
a degree of skill, planning, and organization on the part of specialists.
The problem with this dichotomization and indeed with most of the causal
arguments concerning riots and other forms of collective violence is the ne-
glect of the intermediate space in which riotous activity actually takes place,
 22 / Explaining Communal Violence
that space between popular sentiments and state action that is occupied not
just by the social science category of "riot participants" gleaned from police
reports but by those who organize and plan acts of violence, which include
a whole range of specialists carrying out a variety of tasks. Indeed, the fram-
ing of the causal debate in the terms described above has served no purpose
so much as obfuscating the operations of riot systems, the mechanisms of
pogroms, and the organization of collective violence. 30
There is a further question, moreover, that goes to the heart of the prob-
lem of causal reasoning in the contemporary social sciences and the uses to
which it is put in the explanations for riots as well as the solutions offered to
prevent, control, or suppress them. Caught up in the scientific pursuit of the
causes of riots, all observers who have participated in that pursuit have either
failed to see or paid insufficient attention to the fact that a fully satisfactory
explanation will always remain elusive, and, further, that the very pursuit of
causes is itself implicated in the political process. A full causal explanation
of a complex event such as a large-scale riot can never be arrived at. Every
attempt to do so must involve reduction and concentration on sets of man-
ageable, observable "variables." But riots involve often many thousands of
people who, despite the existence in today's world of widespread electronic
devices of surveillance, seek anonymity in crowd action and to a high degree
still succeed in doing so. A full explanation of a large-scale riot, pinpointing
all the true causes, would involve some kind of exercise in repetition of the
event in the manner of Jorge Luis Borges.J1 Social scientists seek to avoid this
by reduction or by modelling, both processes that cannot fail, by the very meth-
ods of reduction and abstraction, from feeding into the interests and pur-
poses of individuals, groups, governments, and societies that seek satisfying
explanations.
With regard to the second statement, that the pursuit of causal explana-
tions is implicated in the political process, I mean that this pursuit in itself
constitutes a political struggle that occurs invariably after every riot, for the
capturing of its meaning, for the establishment of a hegemonic consensus,
which in turn will influence, even determine power relations in society
thereafter-relations among groups, within groups, and between state and
society. It is not only a question of control of policy decisions taken in the
aftermath of riots to prevent their future occurrence, but of deciding what
is the social problem of which riots are the outcome, who are the persons
and groups to blame for their outbreak, and whether or not those persons
and groups deserve blame and punishment or are to be seen as victims deserv-
ing immediate succor and future amelioration of their condition. It may also
 Explaining Communal Violence / 23

lead in some societies, including India, to definitions of groups as either loyal

and patriotic citizens or antinational persons owing loyalties to a foreign
power.

Functional Analysis

The persistence of a system, a set of institutions, or practices such as collec-

tive violence may sometimes be better understood by a form of functional
analysis used by Robert Merton in his classic study and explanation of the
persistence of the political machine in American big-city politics. How, he
asked, could an institution universally associated with corrupt and criminal
practices for which no public figures had a good word to say and all respectable
citizens condemned persist for a century in the great American democracy?
He argued that there was no simple causal explanation, particularly no
explanation that singled out some "powerful subgroups in the society"
manipulating things from behind the scenes. On the contrary, he found that
the machine performed necessary functions for society and served the needs
of many groups.
Merton did not see his kind of functional analysis as a displacement of
"causal interpretation," but as a supplement to it. 32 He did not say that the
machine persisted because it fulfilled "x, y, and i' functions for particular indi-
viduals, groups, and society as a whole. He argued rather that it was simply
not in the interest of any substantial or powerful social force in society to dis-
place the machine, from which at the same time particular social, political,
and institutional forces benefit. In that case, it would not necessarily follow
that the machine would not persist if it failed to perform the stated functions,
since it would still be the case that no powerful force would gain from its
destruction.
I intend to make use of a kind of Mertonian functional analysis in my search
for an explanation for the persistence of Hindu-Muslim riots and anti-Muslim
pogroms in contemporary India. No societal functional necessity for the per-
sistence of riots will be assumed nor will it be argued that there is any kind
of "feedback loop," such that riots benefit certain groups, who then support
directly or indirectly the persistence of riots. It is sufficient for my purposes
simply to note that riots serve the interests of particular individuals, groups,
organizations, and even society as a whole in concrete, useful ways that are
beneficial to them. Further, I will adopt one of the more common uses of the
term junction, that of use or utility, and will speak of the functional utility
of the persistence of Hindu-Muslim riots in India for a wide variety of inter-
 24 / Explaining Communal Violence

ests, groups, institutions, and organizations, including ultimately the Indian

state. Under these circumstances, it is not possible to produce a broad
enough consensus in society to eliminate violent riots from Indian public life,
just as it was not possible for a century to eliminate the machine from American
public life.

Discourse

Much of my recent work and a good part of the present volume are concerned
with the question of the struggle for control of the meaning of riots in their
aftermath. Although I will take up causal issues below, especially with regard
to the association between riots and political competition, and I will also seek
to demonstrate the functional utility of riots in India, I am especially con-
cerned with the analysis of the explanations for the occurrence of riots given
by ordinary people, politicians, the media, the police, and the civil authori-
ties. These explanations will be treated "as representations," whose contruc-
tion and organization as well as "the types of function they serve"33 need to
be analyz.ed in and for themselves as part of the struggle of sodal and polit-
ical forces in contemporary Indian society.
I intend to show also that a hegemonic discourse exists in Indian society,
which I call the communal discourse, which provides a framework for explain-
ing riotous violence. That framework allows Indian citizens, particularly its
dominant castes and classes, to accept the persistence of such violence in their
society without seeing it as a fundamental flaw in their democra(.l', their essen-
tial nonviolence, their acceptance of Indian cultural diversity, in short, their
ideals. People claim to live according to ideals and become uncomfortable when
the discrepancy between their ideals and the prevalent practices in their soci-
ety is too stark. There are only a few ways of dealing with that kind of dis-
crepancy. One is denial. Another is recognition followed by some form of social
action in the direction of reform through political organization, social work,
writing, cultural protest, resistance, nonviolent demonstration, or the taking
up of arms. There is a third way that is the most prominent in all societies
with recognized severe social problems, including India, namely, blame dis-
placement. Blame displacement makes it possible to acknowledge the existence
of evils such as riotous communal violence and pogroms by attributing vio-
lent practices to others or to natural human propensities that must be accepted
by any realistic person as a part oflife.lt makes it possible, also, either to accept
the violence as inevitable or to direct rhetoric or action towards one's favorite
causes that mayor may not have anything to do with the violence.
 Explaining Communal Violence / 25

THE STUDY OF RIOTS IN INDIA

Argument among Historians

Historical analysis of Hindu-Muslim communal conflict, its causes and pre-
conditions, has been highly contentious in character. Contemporary histo-
rians of India do not even agree on whether or not there existed before the
nineteenth century anything that could be called Hindu or Muslim commu-
nal identities, and, a fortiori, on whether or not Hindu-Muslim conflict was
endemic. On one side are those who argue that Hindu-Muslim conscious-
ness and conflict are largely modern constructions, in which the British colo-
nial rulers played a major role, either through deliberate "divide and rule"
policies or through the ways in which they categorized, classified, and counted
the various peoples of India.34 Historians who accept this point of view also
tend to see Hindu and/or Muslim communal consciousness or communal-
ism as forms of ideology or discourse connected to class, group, and elite polit-
ical interests. In this perspective, the creation or development of communal
consciousness is an instrument of struggle, either against the British or between
Hindus and Muslims for political advantage or supremacy. In the course of
struggle, communal violence could be and often was the result of conflicts
framed within a communal discourse. 35
On the other side are historians who argue that there is more continuity
between past and present, extending backward at least to the early eighteenth
century and, in some arguments, into the earlier period of Mughal rule. In
this view, interreligious strife and riots that resemble contemporary Hindu-
Muslim conflict were present, even endemic, in premodern times. 36 Argu-
ments concerning the causes of Hindu-Muslim violence flow logically from
these distinctive views, with constructivists taking the position that com-
munalism is a cover that hides a multiplicity of mainly political and economic
causes, while those who emphasize continuities between past and present
modes of conflict place greater weight on their religious significance and on
the existence of strong communal identities that preceded them.
My argument in this volume and elsewhere has been on the constructivist
side. Whatever similarities, continuities, and persisting idioms may be found
before the nineteenth century, it would seem idle to overemphasize them.
The consolidation of the heterogeneous Hindu and Muslim groupings on
the subcontinent and the politicization of the differences between them are
overwhelmingly a modern phenomenon deeply connected with the striving
for control over the modern state apparatus, involving a claim to rightful inher-
itance on the part of Hindu and to self-determination on the part of Muslim
 26/ Explaining Communal Violence

leaders. In the course of the struggles for power that developed during British
rule, intensified in the late nineteenth century, and culminated in the divi-
sion of India in 1947, a discourse of Hindu-Muslim difference was created
that has struck deep roots in both communities and acquired a partly self-
sustaining momentum that at the same time continues to be fed by political
competition. In the construction of this discourse, competing historiogra-
phies and historians have themselves played and continue to play substan-
tial contesting roles. 37

Contemporary Social Science Studies

Most available social science methods have been applied to the study of riots
in South Asia. Contemporary and historical case studies are by far the major-
ity. Asghar Ali Engineer has been the most prolific writer on Hindu-Muslim
riots in India since Independence. 38 In countless articles and numerous edited
books, he has chronicled virtually every major riot and many smaller ones.
Deeply committed to secular, universal values, he has blamed neither Hindus
nor Muslims as communities for the carnage that has been inflicted upon the
victims of communal riots. He has instead consistently pointed the finger at
the politicians, on the one hand, and at forms of economic competition
between Hindus and Muslims, on the other hand.
The incidents that precipitate riots in Engineer's view are sometimes sec-
ondary, sometimes primary, but they either hide or reveal the hands of politi-
cians and political movements. Minor disputes are exploited by "petty-minded
politicians" who do not shrink from the sacrifices in human lives that follow
upon their exploitation of such disputes for ilieir political advantage. 39 At times,
political movements themselves are ilie cause of violence, as in the famous
"Ramshila pujan processions" of militant Hindus carrying bricks to Ayodhya
in the movement to bring down the Babri Mosque there and replace it with
a temple to the god Ram. Riots followed in the wal<e of these processions in
almost every part of the country.40 For Engineer, therefore, the primary cause
of communal riots in India is the pursuit of political advantage at any cost.
Beth Roy has produced an exceptional case study of a single Hindu-Muslim
riot in a village in East Pakistan, now Bangladesh, that occurred soon after
Independence and Partition. She has reconstructed the course of ilie riot from
its beginnings in a trivial conflict to the full-scale confrontation iliat ultimately
resulted. Although the way in which this riot developed and its relatively minor
consequences in casualties and injuries appear almost quaint in comparison
 Explaining Communal Violence I 27
with the seriousness and brutality that accompany urban riots in contem-
porary South Asia, her case study nevertheless illuminates many aspects of
the process of riot production that will be analyzed throughout this volume.
Of particular interest is her analysis of how a trivial conflict between two vil-
lagers, one Hindu, the other Muslim, over an incident of one villager's cow
eating the plants of the other developed into a politicized confrontation between
Hindus and Muslims in the surrounding area. She argues further that the vil-
lagers did not act spontaneously, but deliberately, choosing to riot. Moreover,
she has noted distinct stages in the development of thi" riot that, though differ-
ent from my conceptualization of the phases of riot production, neverthe-
less are consistent with the view that riots are indeed deliberate productions. 41
Paired comparison analysis has been done by Varshney. Varshney and
Wilkinson 4Z have pioneered in the creation of an extensive dataset for all riots
noted in one of India's leading English-language newspapers between 1960
and 1993. Both authors in their separate and joint works have insisted that
"To explain the causes of violence we must ... move beyond explanations
which ignore the many times and places where, though communal tension
may exist, riots do not take place. 43
Aside from the general problems that the method of paired comparison
shares with causal theorizing in the social sciences, discussed earlier in this
chapter, there are three additional problems specific to it. One is that it may
lead to downplaying the significance of collective violence in those places where
it does occur. This is clear enough in the case ofVarshney's paired compar-
isons of specific sites and in the Varshney and Wilkinson comparisons of the
incidence of riots by state and city in India. By localizing and regionalizing
the sites of violence in India, they do not refine their gaze, but instead dis-
perse it. In Varshney's case, it leads as well to a doubtful conclusion that civic
engagement between Hindus and Muslims can prevent violence, when it is
more likely that the creation of institionalized riot systems overrides and dis-
places whatever forms of civic engagement and interethnic cooperation exist
at specific sites. But neither alternative argument can be proved on the basis
of such comparisons. 44
The second problem is that such comparisons, by diverting one's gaze from
the specific sites of institutionalized violence, fail to reveal the dynamic
processes of riot production. They claim to tell us why riots occur here, not
there, but they do not tell us how they happen. Until we know how they hap-
pen, the former question cannot be answered.
Third, the method of paired comparison as it has been generally used in
 28/ Explaining Communal Violence

practice suffers from a fundamental flaw of misidentification or inadequate

specification of the actual sites of collective violence. The sites of riots are
too often referred to as the cities in which they occur rather than the specific
neighborhoods or street locations from which they originate or to which they
are even sometimes confined. When, as in the United States, it has often been
the case that the entire black ghetto has been affected by riot activity, there
is some justification for using the cities as the named sites when what is actu-
ally meant is the black ghetto within those cities. However, it is never the case
in either the United States or India that an entire city becomes a site of col-
lective violence. 45 It is also rare that all parts of a city that share similar demo-
graphic features are affected equally or even at all. In such circumstances, the
method of paired comparison by city or town is a kind of unnecessary and
wasteful methodological flourish, for such comparison would be better done
within the specific city or town, comparing instead the neighborhoods and
streets that are affected with those that are not. In this volume, I specify as
clearly and precisely as possible those localities that have been repeatedly, occa-
sionally, or never affected by communal rioting. I demonstrate also that, while
there are demographic, economic, and caste/communal differences that dis-
tinguish the populations in such localities, it is political activity, organiza-
tion, and leadership that demarcate most clearly the riot-prone or riot-affected
from the less affected localities.
Kakar has adopted a psychoanalytic, social-psychological approach to the
study of riots in India, finding their source in mass persecutory fantasies. 46
Although his approach falls squarely in a primordialist interpretation of inter-
communal relations and ethnic violence that I have consistently opposed in
all my writings on ethnicity, nationalism, and collective violence, there are nev-
ertheless some points of contact between hi" work and mine. Our approaches
coincide insofar as we both, along with the leftist historians in India, agree that
the "representations of collective pasts" and the way "collective memories are
transmitted through generations" are of considerable importance in con-
tributing to the persistence of communal violence in India. Kakar adds to this
perspective the argument tl1at these "representations of collective pasts" are
psychic ones, not just intellectual ones. 47 They do not merely justify collective
violence, they explain it. 48 I will have occasion to refer to several explanations
from my interviews that suggest a form of psychic delusion on the part of those
militant Hindus who speak from the vantage point of the communal discourse.
Their statements do indeed suggest that they harbor persecutory fantasies.
Moreover, these fantasies seem also to have become quite widespread among
middle- and upper-class, upper-caste Hindus in northern India.
 Explaining Communal Violence /29

However, Kakar carries his psychic argument too far, into a kind of psy-
chological essentialism. For example, he argues that "the Muslim butcher in
his blood-flecked undervest and lungi, wielding a huge carving knife was ...
a figure of awe and dread for the Hindu child and of a fear-tinged repulsion
for the adult. "49 But this fear that may conceivably afflict Hindu vegetarians
can hardly extend to all "Hindus," since it is the same butcher who slaugh-
ters the goats that nonvegetarian Hindus love to eat. More important, it is
difficult to accept Kakar's argument that it is "religious ultimacy," fed by "an
arsenal of ideational and ritual symbols" on both sides of the communal divide
that makes communal riots so much more "violent and ... difficult to con-
trol" than other types of collective violence. 50 This is a form of objectification
of religious difference that does not hold up in at least two respects. It does
not differentiate communal violence from, say, the slaughter of landless
Scheduled Castes by Hindu landlords in Bihar and South India, or even more
atrocious forms of genocide in places such as Rwanda and Burundi or
Cambodia, where religious conflict is not at issue. Second, neither the killers
interviewed by Kakar nor those I have interviewed appear to me to be moti-
vated by "an arsenal of ideational and ritual symbols.»
Further, Kakar repeatedly writes of Hindu perceptions in general, seeing
them as causes of the tension that leads to riots, as if riots simply follow from
particular tensions, such as a Bihar government move "to raise the official
status of Urdu" that, he argues, "was perceived by the Hindus as a step down
the road of Muslim separatism which led to the Ranchi riots in 1967. "51 Kakar
here not only presents "the Hindus" as an undifferentiated mass, but mis-
takes precipitants for causes. 52 Further, Kakar adopts the utterly misleading
social-psychological approach, going back to Le Bon, that treats riotous
crowds as an undifferentiated mass of individuals who adopt the identity of
the crowd, losing their own in the process.53
This view of the crowd is all made up, pure conjecture that has been shat-
tered by the personal participant observations of Bill Buford.54 It is one of
the principal arguments of this book that we cannot understand what hap-
pens in riots until we examine in detail the multiplicity of roles and persons
involved in them and the justifications presented concerning them by their
promoters and participants. The argument that individuals lose their iden-
tity in crowds belongs among those interpretations of riots that displace blame
onto entire collectivities, who cannot be held responsible for their actions
because they have lost any sense of what it is that they do as individuals. Kakar
tal(es the argument even further towards the primordialist perspective by argu-
ing that group identity in general "is inherently a carrier of aggression,"55 thus
 30/ Explaining Communal Violence

placing the ultimate responsibility for communal violence in the irrational

tendencies of the human psyche, in other words precisely nowhere.
Tambiah has included in his massive work on riots in South Asia case
studies of Sikh violence in the Punjab and the anti-Sikh massacres in New
Delhi in 1984, as well as the riots that accompanied and followed the destruc-
tion of the mosque at Ayodhya in 1992. His work and mine are in agreement,
especially, on three aspects of collective violence in India: (1) the tendency for
explanations of riots to be variously contextualized in a manner that suits the
interests of political actors and others involved in the struggle to control the
meaning of riots in their aftermath; (2) the routine rather than exceptional
character of violence, making it, in effect, an aspect of the political process
that is as predictable/unpredictable as other aspects; (3) the existence of a mul-
tiplicity of roles performed in the production of riots, a kind of division of
labor that makes arguments concerning the spontaneity of riots suspect.56

THE PRODUCTION OF HINDU-MUSLIM

COMMUNAL VIOLENCE

The focus of this volume is on the production of Hindu-Muslim commu-

nal violence in post-Independence India and its relationship to the con-
struction of the categories "Hindu" and "Muslim" in modern Indian history.
In the course of my research, I have arrived at three conclusions concerning
the persistence of Hindu-Muslim communal violence in the specific form
called "riots." The first, based on my analysis of the available evidence mar-
shalled in favor of various causal interpretations, i" that no single causal expla-
nation of Hindu-Muslim riots and anti-Muslim pogroms will suffice to
explain all or even most instances of such collective violence in India. Nor
could it be otherwise, given the vast differences from instance to instance in
what are classified as riots in India, ranging from fights between small groups
that lead to injuries with or without the death of one or more parties to vast
conflagrations that occur in many areas of a large town or city leading to
tens or even hundreds of deaths, large-scale arson, looting, and property
destruction.
It is, in fact, rather a loose form of scientific analysis that would lump the
latter with the former. It is also extremely dubious even to classify the larger
events as a single instance of the phenomenon. An event described as a riot
or pogrom that takes place more or less simultaneously in widely spread areas
of a large town mayor may not have been precipitated by a single incident,
have spread to other areas by means of news, rumor, or "contagion," have
 Explaining Communal Violence /31
been preplanned or spontaneous. In other words, such a large event as the
Bombay riots of December 1992 or the Aligarh riots of 1990-91 may be one
riot or many, each with its own precipitants, underlying enmities, and sets
of interests involved.
Although, given these difficulties, no single cause of riots can possibly be
adduced successfully, I have also found that there is at least a kernel of truth
in all but the most bizarre explanations of riots. There is sufficient evidence,
for example, that riots between Hindus and Muslims often occur where Hindu
and Muslim areas are in close juxtaposition to each other, in nasty slums where
landlords or businessmen of one or the other community seek to displace
persons from the other community in order to acquire valuable real estate,
and at times of intense political competition and mass mobilization when
different political parties seek to mobilize the votes of one or the other of the
two communities. But the mere presence of anyone of these factors or any
combination of them is neither necessary nor sufficient to produce large-scale
riots. At the same time, it is also the case that many or all of these factors and
others to be discussed in this book are present in all large-scale riots.
For many social scientists and historians, the obvious solution to the causal
dilemma here is simply to resort to a multicausal explanation, either through
the listing of aU the factors discovered to have operated in one or more cases
or through statistical methods of correlation and regression that will assign
percentages of explanatory power to each of the active variables. But, as I will
show in this volume, this also will not do. It will not do principally because,
when one examines the actual dynamics of riots, one discovers that there are
active, knowing subjects and organizations at work engaged in a continuous
tending of the fires of communal divisions and animosities, who exercise by
a combination of subtle means and confrontational tactics a form of control
over the incidence and timing of riots. But their control is not total. Some of
the pogroms that have occurred in India in recent history have been shown
to have had a high degree of organization and preplanning, notably the anti-
Sikh pogroms after the assassination of Indira Gandhi in 1984 and the Shiv
Sena-organized pogrom against Muslims in Bombay in January 1993. Even
in these cases, however, control is not total. So, while the production of many
large-scale riots can be seen as deriving from the political interests and cal-
culations of specific organizations, groups, and leaders, many other interests
also then come into play whose actions at the multiplicity of sites in which
violence occurs require secondary explanations.
It is the combination of "objective," underlying factors of demography,
economics, and electoral competition with intentionality and direct human
 321 Explaining Communal Violence

agency that makes causal explanation of riots in general so difficult. We can

surely say that Krystallnacht in Germany was a preplanned, coordinated attack
on carefully specified targets, namely, the Jews and their property and reli-
gious objects. But even this attack was designed to appear spontaneous, to
fool outside observers. What makes riots and pogroms in India or the United
States or nineteenth-century Russia so much more difficult to analyze and
comprehend is that they combine objective and intentional factors, spon-
taneity and planning, chaos and organization. They are best conceived as dra-
matic productions in which the directors are not in complete control, the cast
of characters varies-some of them being paid, some of them acting volun-
tarity for loot or fun-and many of the parts have been rehearsed, but others
have not.
This then brings me to the second argument of this book, that, where riots
are endemic, what I call "institutionalized riot systems" exist in which known
persons and groups occupy specific roles in the rehearsal for and the pro-
duction of communal riots. In such systems, a central role is played by what
I call "conversion specialists," those whose task it is to decide when a trivial,
everyday incident will be exaggerated and placed into the communal system
of talk, the communal discourse, and allowed to escalate into communal
violence.
The production of communal riots is very often a political one, frequently
associated with intense interparty competition and mass political mobiliza-
tion. This was true before Independence as well as after. This fact goes a long
way also towards understanding Gandhi's emphasis on nonviolence as the
basis for the mobilizations against Briti~h rule in India. Nonviolence may have
been a religious principle for him as well as a political tactic of the weak against
the strong, the relatively unarmed against the armed, but it also arose from
his profound knowledge of Indian social and political life, from his own under-
standing of the violent mechanisms that could so easily be brought into play
under the cover of the vast mass movements that he launched, which would
undercut their purposes and direct local political energies to other targets.
Moreover, there were other politicians already playing a different game of
direct instigation of communal violence for other political purposes.
The institutionalized riot systems or networks that exist in riot-prone cities
and towns comprise a multiplicity of roles that I have identified elsewhereY
A good part of this book will be devoted to showing how and when they come
into operation. For now, it need only be noted that they include informants
who carry messages to political group leaders of the occurrence of incidents
that may affect the relations between Hindus and Muslims; propagandists
 Explaining Communal Violence /33
who create messages to be conveyed to particular segments of society, to the
press, to the general public; vernacular journalists who publish these mes-
sages in the form of "news," poster plasterers who place them on walls, rumor-
mongers who transmit them by word of mouth; recruiters who collect
crowds from colleges and universities and goondas (thugs) to kill, loot, and
burn when the time is ripe.
But there are two roles that are crucial in the dynamics of riot produc-
tion, designated herein by the terms "fire tender" and "conversion special-
ist." The fire tender keeps the embers of communal animosities alive by
bringing to the notice of the politicians, the authorities, and the public situ-
ations that are known to be sensitive in the relations between Hindus and
Muslims. These situations may be genuine or bogus, they may refer to inci-
dents that actually happened naturally or that were created deliberately for
the purpose of stoking communal passions.
The second role, that of the conversion specialist, is the pivotal one of turn-
ing a mere local incident or a public issue affecting the two communities into
one with riot potential by inciting a crowd and giving a signal to the special-
ists in violence to let loose the violent action: stone throwing, stabbing, or
arson. The conversion specialist is generally, if not always, a political person.
He is part of the political group whose interests are to be served by the vio-
lence and may even be a leader in the group. He usually does not himself
engage in violence, but instigates others to do so when the political context
favors it.
The third and fourth arguments concern the issue of riot persistence. Since
much if not all of the information to be presented in this book concerning
the mechanisms of riot production is known to the leaders of the country,
the journalists, the local authorities where riots occur, and to the literate and
illiterate public, and is deplored by all, how is it that riots persist? The ques-
tion is especially puzzling since the politicians and the authorities know where
and when riots are likely, who the principal riot-mongers are, how to pre-
vent riots and to control them when they break out. How is it that they do
not do so always? In fact, how is it that they do so only irregularly?
The first answer to these questions refers primarily to large-scale riots and
those that occur in waves. There is a dear association between such riots and
waves of riots, on the one hand, and electoral competition and mass politi-
cal mobilization. Not only is there an association, but the evidence to be pre-
sented below demonstrates that there is a direct causal link between riots and
electoral/political competition, such that Hindu-Muslim riots are a product
of actions designed to consolidate one community or the other or both at
 34 / Explaining Communal Violence

the local, regional, and national levels into a cohesive political bloc. Riots do
in fact have that result. Riots precede elections and intensify political com-
petition. Riots accompany political mobilizations around religious symbols
and contribute to the strengthening of the movements, which in turn solid-
ify communal solidarity in subsequent elections.
The second answer is functionalist: riots persist because they are func-
tionally useful to a wide array of individuals, groups, parties, and the state
authorities. In other words, this particular form of endemic, recurrent vio-
lence is a function of the unwillingness, failure, or lack of desire of such bod-
ies and entities to take preventive measures, the effectiveness of which are
well known whatever the imputed causes of particular incidents, which vary
widely. Their functional utility is in turn heavily influenced by the political
benefits that derive from them.
The third answer is that there exists in India a discourse of Hindu-
Muslim communalism that has corrupted history, penetrated memory, and
contributes in the present to the production and perpetuation of communal
violence in the country. Large-scale riots involve considerable mass partici-
pation in their enactment as well as a commitment to the belief on the part
of many people that they could not have been prevented by the actions of
well-intentioned persons or governments. It involves a belief that they are
in fact endemic to and a consequence of fundamental hostilities, prejudices,
and passions in society. It is necessary here to say a few words about the ori-
gins of this discourse before proceeding further.

THE DISCOURSE OF COMMUNALISM:

HISTORY, MEMORY, AND COMMUNAL VIOLENCE
IN POST-INDEPENDENCE INDIA

India is a country whose peoples live today under the signs of several great
historical ruptures that are perceived as having disrupted the historial real-
ity of the pre-existence of an Indian nation and prevented its full realization
in political form: the prolonged rupture that is seen as the Muslim conquest of
the subcontinent, the establishment of British rule in the eighteenth century,
and the division of the country in 1947. Two of these events are associated in
the minds of most Hindus and in the ideology of militant Hindu national-
ism with the large Muslim population of the country. Despite the contrary
attempts of secular nationalist leaders and historians, a divisive history of India
has acquired a hegemonic place in the school textbooks and in the national
mythology of the country which defines the miUennially long arrival of both
 Explaining Communal Violence /3S
the religion of Islam and Muslim arms into the subcontinent as a foreign,
Muslim conquest. Despite the fact that probably 95 percent of the Muslim
population of the subcontinent is of indigenous origin, descendants of con-
verts to Islam, Islam is considered in the history of the Hindu nation "as a
foreign element."58
The millennium that saw the establishment of Islam and Muslim ruler-
ship is seen as one coherent period in Indian history, the period of the Muslim
conquest that followed upon the classical Hindu period. Both periods are
defined in communal-religious terms, the first as the "Hindu period"-
though it encompasses the rise of other world as well as specifically Indian
religions such as, respectively, Buddhism and Jainism-the second as the
Muslim period.59 The first period is described as the glorious age of imperial
Hindu achievement in politics and culture, the second -despite its own glo-
ries of art and architecture, which are acknowledged-as a period of con-
quest, destruction, and consequent decay of Hindu civilization. The Hindu
practices, customs, and superstitions that exist today are a consequence of
that decay and, as many Hindu reformers of the nineteenth century argued,
did not exist in the past. To revivify India and build a great, new, modern
nation-state, it is necessary to "revive the true ideals of the past."60
This process of historical rectification also has been accompanied by a
demonization of the Muslims as a separate people, a foreign body implanted
in the heart of Hindu India, perpetually "warlike," who "believe it is their
religious duty to kill infidels."61 Muslims are also held responsible for the par-
tition of the country because so many of their leaders remained aloof from
the nationalist movement and ultimately fought for the creation of the sep-
arate independent state of Pakistan. The memory of the partition and the
violence associated with it is ingrained in the minds of most Hindus and is
kept alive by the constant tension in the relations between India and Pakistan.
During the last decade, the "memory" of Muslim violence in Indian history
has been kept vivid also by the militant Hindu demand to recapture and restore
temples allegedly destroyed by Muslim conquerors and replaced by mosques,
a movement that led to the destruction of the mosque at Ayodhya on
December 6, 1992.
The periodization of Indian history that achieved ascendancy in Indian
nationalism-not just militant Hindu nationalism-in the late nineteenth cen-
turywas itselflargely a product of British history-writing of the mid-nineteenth
century that sought to place British rule within the long sweep of Indian his-
tory. Slightly refurbished by Indian nationalists, it emerged as the tripartite
division of Indian history as a movement from "classical glory" to "medieval
 36/ Explaining Communal Violence

decline" under Muslim rule to a "modern renaissance"62 of a new India which,

in its secular version, encompassed Muslims, but in its Hindu nationalist ver-
sion either did not or did so only grudgingly. In the latter version, as
Chatterjee has put it in his discussion of one Bengali example, "a stereotyp-
ical figure of 'the Muslim'" emerged, "endowed with a 'national character':
fanatical, bigoted, warlike, dissolute, and cruel. »63
Although Muslim history and Muslim character in the Indian subconti-
nent were blackened and demonized in these ways and often were used to
justify demands by Hindu nationalists in the nineteenth century for the British
to rectify matters and restore to Hindus their rightful place in the new India
emerging under colonial rule, ultimately the target of this historiography was
the British, then the West in general, and finally the great powers, particu-
larly the United States. India soon was perceived in the minds of Indian nation-
alists, when they gradually directed their aims towards the elimination of
British rule, as a potential great power, the equal of the great powers of the
West. It was a great, modern state that Indian nationalists, both secular and
Hindu, sought to create after Independence.
In the pursuit of that goal, the Muslims of India came to be seen, partic-
ularly by Hindu nationalists, as an obstruction, along with Pakistan, whose
very existence has, in the minds of such Hindu nationalists, been the princi-
pal post-Independence obstacle to India's achievement of its rightful place
in a world dominated by great nation-states. This historical consciousness
and teleology of Hindu nationalism has framed the modern discourse of
Hindu-Muslim communalism and violence. The demographic distribution
of the peoples of India and the landscape of the country have become pop-
ulated with lieux de memoire64 that signify the violence done by Muslims to
the Hindu body, the dangers of the Muslim populations that reside in the
midst of Hindus in cities and towns, and Muslim institutions that teach Mus-
lims to become traitors, all of which must be reformed, replaced, or extir-
pated before India can become whole, united, and powerful.
In the first category, signifiers of the violence done by Muslims to the
Hindu body, are the mosques, said to number three thousand, that are alleged
to have been built upon the ruins of Hindu temples destroyed by Muslim
conquerors. In the second, signifying the dangers to the Hindu body in the
present, are the concentrations of Muslim populations in cities and towns,
described by militant Hindus as "mini-Pakistans." The leading example in
the third category, signifying the traitors in the midst of the country, is the
Aligarh Muslim University (AMU) located in the town of Aligarh in west-
ern Uttar Pradesh, ninety miles south-southeast of Delhi. The AMU was in
 Explaining Communal Violence /37

fact one of the principal sites from which the ideology of Muslim separatism
and then the Pakistan movement developed and spread in the late nineteenth
and twentieth centuries. 65 It stands today in Aligarh on the outskirts of the
city as a symbolic presence that signifies to militant Hindus the persistence
into the present of Muslim separatist, communalist, and anti-Hindu designs,
and justifies, along with the existence of "mini-Pakistans" in the center of
the old city, violence against Muslims that is enacted in periodic outbursts
of large-scale rioting. Aligarh has become, in fact, one of the principal sites
of Hindu-Muslim violence in all of India since Independence up to the very
recent past.

PLAN OF THE VOLUME

I have selected Aligarh as the site of the present study and have chosen to
focus on a single site for several reasons. First, as just indicated, it has been
for more than a century, since the founding in 1875 of the Mohammadan
Anglo-Oriental College that later became the AMU, a principal site of Hindu-
Muslim tension. Second, numerous riots, large and small, have occurred in
Aligarh both before and since Independence. In fact, since Independence,
though there have been other cities and towns in India in which riots have
been endemic and where some much larger riots and pogroms have occurred,
it will be shown below that Aligarh stands close to the top, in some respects
absolutely, in others relatively. In absolute terms, it has been second in the
state of Uttar Pradesh (UP.) in the number of known riots and deaths asso-
ciated with them since Independence. Moreover, relative to its size, that is,
by a crude measure of deaths per thousand persons, it stands close to the top
in the country as well. Third, during thirty-eight years in which I have gone
back and forth to Aligarh for research, I have heard here all the explanations
for riots that are commonly given for riots everywhere else in the country:
demographic, economic, and political. Finally, the city, as will be demonstrated
later, is characterized by many of the features that characterize other riot-
prone cities: notably, the relative size of the Hindu and Muslim populations
and their juxtaposition in relation to each other, the existence of economic
competition between segments of the two communities in particular trades
and industries, and a history of intense electoral competition.
I have chosen to concentrate here on a single site rather than several sites
for methodological reasons. First, although I have gathered considerable mate-
rials from several other sites of the same sort to be presented herein, and have
published some of it elsewhere, I have concluded that studies from different
 38/ Explaining Communal Violence
sites of riots that have occurred in the same time period or at different times
cannot explain their occurrence at a particular site at a particular time. For
example, I have analyzed the post-Ayodhya riot in the metropolis of Kanpur
in 1992.66 An even worse riot occurred in Aligarh in 1990-91. Why did the
Kanpur riot not occur the previous year, and why did no Aligarh riot occur
in the following year? No form of ecological analysis can answer such a ques-
tion. To answer that kind of question, diachronic rather than synchronic analy-
sis i" required. Second, concentration on a single city makes it possible to focus
one's gaze more closely and intently at the specific localities in which riots
have occurred within it, to determine what especially characterizes them. In
other words, one achieves thereby a more fine-tuned look at riots than is pos-
sible when one classifies an entire city as riot-prone and seeks to compare it
with other equally large units. Third, whereas much work has been done in
attempting to explain "waves" of rioting, including many analyses that have
claimed to have found a "contagion" effect in the spread of riots from one
site to another in a short time period, I am not aware of any studies that have
looked at rioting in a different way, namely, as a phenomenon endemic to a
particular site or sites. While I remain alert ill this volume to what is hap-
pening elsewhere when rioting does and does not occur in Aligarh, I am seek-
ing an explanation for its persistence at this site, which can in turn be
generalized as an explanation for recurring riots elsewhere as well. Fourth,
during the thirty-eight years in which I have visited Aligarh, I have collected
all the available election data for the city since Independence, down to the
polling booth level, which I have converted into a large dataset that has made
it possible for me to analyze precisely the relationship between riots and elec-
toral competition.
Although Aligarh is a site of persistence, it is also a site of variation in two
respects. First, even here, where Hindu-Muslim tensions seem to be always
on the edge of violence in some parts of the city, there have been periods of
relative relaxation. Those periods of abatement in rioting need to be explained
as well as tlle numerous explosions. Of course, it is obvious that no society
can live with continuous rioting, so there must be some break. If there is not,
the situation is best described as a civil war. Since this is not the case in Aligarh,
the absence of rioting ill the town at certain times must be explained as well
as its presence at others. It will be shown that presence and absence are
explained by the same sets of factors.
Aligarh is a site of variation in another respect as well, namely, in the
differential spatial incidence of rioting within the city when it does occur.
Not all parts of the city are always affected by Hindu-Muslim riots and some
 Explaining Communal Violence 139
have never been touched by it. For these reasons, it will be possible to demon-
strate at this one urban site precisely in which localities rioting occurs, and
to point out the differences between them and localities where rioting is not
endemic, in ways that are not possible by the paired comparison method that
selects entire cities as sites.
The rest of this volume is organized into five parts. Part II, consisting of
four chapters, first presents a description of the general features of the demo-
graphic, caste, and communal composition of the city, then provides a
detailed history of riots in Aligarh since Independence, with a separate chap-
ter that focuses specifically on the great riots of December 1990 and January
1991, followed by a chapter that discusses the relative decline in riotous activ-
ity in the city during the following decade. Part III, consisting of two chap-
ters, focuses on those factors in the social composition, distribution, and
economic characteristics of the population that have commonly been con-
sidered conducive to Hindu-Muslim riots. Part lV, comprising five chap-
ters, marshals evidence primarily from interviews and electoral data to
demonstrate the close connection between riots and political competition.
Part V contains three chapters that focus on the discourse of communalism
and alternative contextualizations of riotous violence by politicians, police,
and press. The conclusions concerning the persistence of riots and pogroms
in Aligarh are provided in Part VI. A postscript has been added that brings
the situation in Aligarh up to date as of the February 2002 elections. I have
also included a discussion of the pogrom in the state of Gujarat that began
at the end of February 2002 and continued until the end of March, with spo-
radic incidents of violence for weeks thereafter.
 PART II

Communal Riots
in India and Aligarh
 2/ Aligarh
Politics, Population, and Social Organization

P arty pOlitiCS., elections, and riot.s are political dramas of contestation

involving simplification of complex realities. They reduce hetero-
geneity and complexity in order to mobilize large enough numbers
to form a party or movement, win an election, or confront a rival ethnic or
religious group in riotous violence. This is especially so for competitive polit-
ical systems in which election contests are decided by the first-past-the-post
system in single-member, plurality constituencies, that is, where there is only
one winner in a constituency in which the outcome is determined simply by
who gets the largest number of votes. In the society of unparalleled hetero-
geneity that is India, such simplification is both necessary for political suc-
cess and difficult to achieve. Consolidation and mobilization of large ethnic
groupings in violent conflict with each other is one of the methods that has
been used in many parts of India, whose effect has been to create solid vot-
ing blocs in succeeding elections.
Since the death of India's first prime minister, Jawaharlal Nehru, in 1964,
and the decline of India's former ruling party, the Indian National Congress,
along with the consensus that prevailed throughout Nehru's lifetime on the
maintenance of a secular, composite nationalism, politics in India have oscil-
lated between periods of fragmentation and consolidation, complexity and
simplification. At the national level, repeated attempts have been made to
articulate appeals and programs that would be effective in mobilizing large
categories of voters across regions to encompass the country as a whole. These
appeals have been both economic and cultural, including Indira Gandhi's pro-
grams and policies designed to capture the votes of all the poor and disad-
vantaged segments of society as well as the lowest castes (termed "Scheduled
Castes") and minorities. The center and left opposition parties in north India
countered with appeals to a huge segment of Indian society known as "back-

43
 44/ Aligarh: Politics, Population, and Social Organization

ward castes," comprised primarily of middle agricultural and artisan casts

encompassing a wide range of economic conditions and social status. In recent
years, also, a new political formation has become politically important in north
India, the Bahujan Samaj Party (BSP), claiming to speak for another "major-
ity" (bahujan), all the lowest and most deprived castes and classes of the coun-
try. Against both the Congress and the parties claiming to speak for caste
categories and minorities have stood the Bharatiya Janata Party (BJP) and its
associated organizations, claiming to spe-& on behalf of all the Hindus of India,
north and south. In contrast, the parties opposed to the BJP have sought to
amplify their own support bases among the backward and lower castes by
seeking Muslim support; the Samajwadi (Socialist) Party (SP), the principal
rival of the BJP in Uttar Pradesh (U.P.), has been particularly successful in
doing so during the past decade.
Each of the regions of India has its own complexities. Indeed, the par-
ticularities of language, religious distributions, and regional cultures have been
sufficient so far to prevent any reconsolidation of power at the Center behind
a single new dominant party to replace the former hegemony of the Congress
in the country as a whole. The B]P, presently (2001) the leading party in a
central government coalition, has been the principal claimant for such a role.
India is so heterogeneous a country that complexity, and the consequent
political need for simplication to overcome it, are evident not only in each region,
but in every district, town, and village as well. Insofar as the huge state ofU.P.
is concerned (see frontispiece map), there have been four primary political
formations in recent years, each of which has sought to consolidate large seg-
ments of the population. The formerly dominant Congress has declined to
such an extent that it cannot be said any longer to be an effective political
party in the state. The major political parties at present are the BJP, claiming
to speak for all Hindus; the SP, which has its greatest strength among certain
of the backward castes, particularly Yadavs; and the BSP, which has its great-
est strength among the largest of the Scheduled Castes, the Chamars and Jatavs.
None of tl1ese parties have been able to produce effective governing majori-
ties in this state in recent years, either alone or in coalition. The ruling coali-
tion in the state in 2001 was led by the BJP.
The same political formations have been important in the politics of Aligarh
City, but Aligarh has been different in one important respe(,:t that foreshad-
owed the state's future. In 1962, twenty-five years before the disintegration
of the Congress in the state as a whole, that party was defeated in the city in
a critical election from which it never recovered. That victory was achieved
by a consolidation and simplification of complexities through an alliance
 Aligarh: Politics, Population, and Social Organization /45

among the Muslims and &heduled Castes, made possible by Muslim anger
against the Congress after the communal riots of October 1962. As a conse-
quence of that anger, Muslims joined with Scheduled Castes under the ban-
ner of a party called the Republican Party of India, then the leading party
claiming to speak for the low castes. Since then, politics in Aligarh as well as
in the state as a whole have oscillated between periods of consolidation and
fragmentation. In Aligarh, in common with several other of the larger towns
and cities in the state, consolidation has been associated with interreligious
violence-indeed, as I will show, has followed upon communal rioting. When
communal rioting has declined, the natural heterogeneity of the population
has been reflected in increased fragmentation of voting and dispersion of votes
among a larger number of political parties.
The purpose of this chapter is to layout both the social heterogeneity of
Aligarh's urban population as well as the simplified categories that are used
in the census to encompass that heterogeneity. It has been noted by many
historians and in my own earlier work that British census definitions of the
population of India, imposed upon its diversity, were ultimately converted
into social realities. One can witness the same process at work in contempo-
rary census and political definitions of India's population, where political for-
mations seek, in effect, to match in politics the consolidated numbers reflected
in census categories. I will begin this discussion of complexity and sim-
plification in Aligarh with the simplified religious and caste categories used
in the census, after which I will show the very considerable heterogeneity that
lies beneath those categories and that continues in peaceful times to be reflected
in politics.

RELIGION AND CASTE IN THE CENSUS OF ALIGARH

Census figures for the total population of Aligarh are available from 1901 to
1991/ during which period the population multiplied by nearly seven times
from 72,084 to nearly half a million, or 480,520. The growth rate in both the
municipality of Aligarh and in the census-defined standard urban area (SUA)
surrounding it in the decade 1981-91 was huge, somewhat under 50 percent
in the city and over 50 percent for the entire SUA, which includes a large num-
ber of villages. The population of the city has certainly passed the half-mil-
lion mark during the last decade. However, some of this increase must be
attributed to changes in the boundaries of the city through incorporation of
outlying rural areas, including some previously unpopulated areas where new
settlements have been established as well as some populated villages. The area
 46/ Aligarh: Politics, Population, and Social Organization

incorporated within the city limits nearly tripled between 1941 and 1961, from
just over 11 to nearly 32 square kilometers. Since then, the city area has increased
by another two square kilometers, to 34.05 in 1991. It is believed by many
Muslim and Hindu politicians in the city that the larger number of persons
in the newly populated and incorporated areas comes from Muslims migrat-
ing from the villages of Aligarh District to the city, a matter that will be dis-
cussed further below (see Chapter 6).
For purposes of this book, the most important figures concern the religious
and caste composition of the population. The census provides information-
rather sporadically and incompletely-for three very broad caste/communal
categories: Hindus, Muslims, and Scheduled Castes. Scheduled Castes are
sometimes included in the Hindu population, sometimes not. Jains, Sikhs,
Buddhists, and Christians have at times been lumped together in a general
category designated "Hindus and others." Fortunately for our purposes, the
numbers of these religious groups are quite small, although some are promi-
nent in particular mohallas (neighborhoods) of the city. Muslims, because of
prejudices against them and political fears concerning their allegedly more rapid
increase than Hindus, are sometimes undercounted both in census enumer-
ations and in the preparation of voters' lists, another source of information
on the Muslim population. 2
Only four censuses provide a three-way division of the total population
into Hindus, Muslims, and Scheduled Castes, those for 1951, 1971, 1981, and
1991 (Table 2.1). In 1951, Hindus comprised 54.17 percent of the city's popu-
lation, Muslims 34.53 percent, and Scheduled Castes 9.99 percent. By 1991,
however, the relative proportions of the three categories had changed con-
siderably: they were 44.52 percent for Hindus, 37.41 percent for Muslims, and
16.51 percent for Scheduled Castes.
These figures are quite important for political reasons and calculations.
Militant Hindus have claimed that the Muslim population has been increas-
ing in the city through natural population growth and migration from the
rural areas, thereby reducing the Hindus to a minority. While the figures now
do show non-Scheduled Caste Hindus comprising less than a majority of the
city's population, the result has been obtained not so much by a rapid increase
in the Muslim population, which-despite a higher growth rate than that
for Hindus-has increased in the census by less than 3 percentage points
overall, but by a much greater increase in the Scheduled Caste population,
which increased by 6.52 percentage points, or by 61 percent from 1951 to 1991.
However, when the figures for Scheduled Castes are added to the non-
Scheduled Caste Hindu population, they can be read as showing a continu-
 Aligarh: Politics, Population, and Social Organization 147

TABLE 2.1 Population (in percentages) of Aligarh City

by religion and caste, 1951-91

Scheduled
Year" Hindus Muslims Castes Others

1951 54.17 34.53 9.99 1.31

1971 50.34 33.08 14.44 2.14
1981 46.75 34.47 17.04 1.75
1991 44.52 37.41 16.51 1.55

• No figures available in 1961 for Hindus and Muslims; the percent Scheduled Caste
in that year was 14.45.

ing Hindu majority in the city of above 60 percent. The political significance
of these numbers lies in the fact that, when elections have been held in Ali-
garh City at times of tension between Hindus and Muslims, that is, in an
atmosphere of communaliz.ation tending towards polarization, militant
Hindu candidates have been likely to prevail. If, however, Scheduled Castes
and Muslims combine, then no militant Hindu candidate can win the Ali-
garh Legislative Assembly seat. If a Scheduled Caste party or leader enters
the fray, then the vote will be split at least three ways. All three of these out-
comes-Hindu-Muslim polarization, Muslim-Scheduled Caste combina-
tion, and three-way division of the votes-have occurred at different times
in the course of the electoral history of Aligarh from 1951 to 1998, as will be
demonstrated in Part IV below.
Despite the fact that Muslims are sometimes undercounted, we get a clearer
picture of Muslim population change in the city for the longer period from
1931 to 1991. The Muslim population stood at 42.85 percent in 1931, increased
to 45.90 percent in 1941, then took a dramatic drop to 34.53 percent in 1951 in
the aftermath of the mass migrations to Pakistan at the time of Partition in
1947 and after, which affected Aligarh profoundly. It is likely, considering the
drop in the percentage of Muslims between 1941 and 1951, that somewhere
around 13,000 to 16,000 Muslims left the city. The figure given during my
first visit to Aligarh in 1961 was approximately 15,000 people, who left after
the 1946 and 1950 riots.3 After Prime Minister Nehru intervened to facilitate
the return of Muslim emigrants who wanted to come back to India after the
rioting ended, many did come back, but certainly the vast majority did not.
In some localities, the communal character of the population changed rad-
ically during this period. According to Elizabeth Mann, one of the most riot-
 48/ Aligarh: Politics, Population, and Social Organization

prone localities of the city, Manik Chauk, changed from a Muslim-majority

mohalla to one "almost entirely inhabited by Hindus."4
In the intervening years, as already indicated, the Muslim population
increased by less than three percentage points to reach 37.41 percent in 1991.
It is generally believed that the Muslim population has increased still further
during the past decade. As will be shown in Chapter 12, the dual increase in
the Scheduled Caste and the Muslim population in the city linIits has recently
undermined the political base of the militant Hindu party, the BJP, in the
politics of Aligarh, rendering its victory in the Aligarh City Legislative Assem-
bly contests problematic.

CASTE, COMMUNITY, BARADARI

Hindus and Others

Bania castes. Census data on the distribution of Hindus and others by caste
and community are not available. Among the Hindu castes, Vaishyas, also
commonly called Banias, are generally considered to be the largest broad
grouping, comprising around a fifth of the total population of the city and
nearly 40 percent of the Hindu population. The two most prominent castes
in this broad category in Aligarh are Barahsenis (also called Varshneys) and
Agarwals. In Aligarh City, persons from both castes are primarily business-
men, shopkeepers, bankers, and industrialists. 5
In Aligarh District and Aligarh subdivision (tahsil Koil), Agarwals have
comprised the largest of the Bania castes, although in Aligarh City the
Barahsenis are generally considered to be twice as numerous, accounting for
12 percent of the total population of the city compared to 6 percent for the
Agarwals. 6 They are also "the most powerful group in the city-politically
and economically," they "largely dominate trade and industry and have an
elaborate group organization."7 The mayor of the city, elected in 1995, was
Ashutosh Varshney, a manufacturer and exporter of brass building hardware,
handicrafts, and food products. The previous mayor, from 1989 to 1994, was
O. P. Agarwal.
Before Independence, Agarwal cotton traders identified with the 1930
Congress boycott of British doth, since they "had acquired much wealth by
using locally produced cotton in ginning and pressing mills," while both
Muslims and "the Baraseni [sic] wholesalers" did not participate in the boy-
cott. 8 For some time after Independence, also, it was said that the Agarwals
tended to be "pro-Congress,"9 while the Barahsenis identified with militant
Hindu nationalism and with the Jan Sangh and the BIP.
 Aligarh: Politics, Population, and Social Organization /49

The local colleges in the city, whose faculty and students have provided
leadership as well as mass support for political parties, rallies, and riotous activ-
ity, were also identified in 1962 with different Bania castes and different polit-
ical identifications. Thus, the Maheshwaris, another Bania caste in the city,
had their own college, the Maheshwari Inter-College, then managed by the
Jan Sangh candidate for the Legislative Assembly seat in 1962. The Barahsenis
had their own college also, the Barahseni Degree College, the manager of which
at that time was also said to be a Jan Sanghi, while the D.S. (Dharma Samaj)
College was "in the hands of the Congress."10
Politically, as late as 1980, Barahsenis and Agarwals were reported to have
been divided, with the Agarwals declining to support the BJP candidate for
the assembly election in that year. However, for the most part, as the com-
munal polarization in the city has intensified after Independence, Barahsenis
and Agarwals have been generally found on the same side politically, as mem-
bers of the RSS,ll as supporters of the BJP,12 the Vishwa Hindu Parishad, and
other RSS/BJP-affiliated organizations,13 and as antagonists of the AMU.14 In
recent Aligarh elections, marked by polarized contests between Krishna
Kumar Navman, a militant Hindu Barahseni politician from the mohalla of
Manik Chauk, and a Muslim candidate, all Bania castes have voted overwhelm-
ingly for N avman.15 Nevertheless, there remains some sentiment among per-
sons from the Agarwal community in favor of the Congress, one of whose
members, Vivek Bansal, contested for the Aligarh Legislative Assembly con-
stituency as the candidate of that party in 1993.16
Brahmans. After the Banias, the largest group of elite castes in the city is
Brahman. Their proportion in the total population of the city has been esti-
mated at approximately 8 percent. The Brahman population is widely dis-
tributed in the "predominantly Hindu localities," none in particular being
identified as primarily Brahman. 17 Before Independence and for many years
afterwards in U.P. and in Aligarh as well, Brahman castes provided both lead-
ership and a relatively stable basis of support for the Congress, while at the
same time contesting with other elite castes, particularly Thakurs,18 for lead-
ership predominance in both the district and city Congress organizations.
Brahmans, for the most part, remained Congress supporters in Aligarh City
as well as the rural areas of the district for many years after Independence. It
was only in the 19805, with the decline of the Congress in the state as a whole
and with the rise of Hindu communal mobilization during the Ayodhya move-
ment in the 1980s, that Brahmans, along with other high-caste Hindus such
as Thakurs and Kayasthas, turned away from the Congress to join with the
Banias in supporting the BJP. 19 However, Brahmans have never been promi-
 50 IAligarh: Politics, Population, and Social Organization

nent in Aligarh in the militant Hindu politics of the RSS and its affiliated orga-
nizations, including the Jan Sangh and the BJP.20
Maithel Brahmans. The Maithel Brahmans, who take the name of a high
Brahman caste category from the north Bihar region of India,21 are not
accepted locally as such. They are an artisan caste, among whom those with
the surname Sharma have been traditionally associated with carpenty.
Maithels are described by Haqqi as an artisan caste "mostly engaged in the
lock industry.» They are said to constitute "one of the most cohesive and highly
organized group [s l" in Aligarh, maintaining "an Intermediate College for the
benefit of the community." Haqqi estimated the voting strength of this "lower
rank" Brahman caste in 1971 at "between four to five thousand,"zz so they
would certainly number above five thousand today. They have large con-
centrations in some localities of Aligarh.
The Maithels do not always behave politically in the same manner as the
other Brahman castes. One local politician remarked that, ordinarily, the
Maithel and other Brahmans "are against each other," but in the case of a
political/electoral contest between Brahmans and Thakurs, "they will join
hands."23 Moreover, in contemporary Aligarh politics, they have tended to
vote, like upper-caste Hindus, for the BJP. All three of the Maithel Brahman
municipal corporators were elected on the ticket of the BJP. At the same time,
they identify themselves as backward castes and associate politically with other
backward castes in local politics.
Other Hindu castes. Very little information is available on the other Hindu
castes of Aligarh. Among the elite castes, the population of Kayasthas is esti-
mated at around 4 percent of the electorate.24 There are not many Rajputs
(Thakurs) living in the city, but, like Rajputs everywhere in U.P., there are
prominent persons from Rajput landowning families living in the city, who
have some influence in local politics. Among the backward castes, the most
important in the city numerically and politically are Yadavs, who support the
SP, and Lodhs, Sainis, and Nais, who support the BJP.

Scheduled Castes

Among the Scheduled Castes, the largest caste is Jatav, estimated at about 11
percent of the total population,25 but comprising the overwhelming major-
ity of the Scheduled Caste population in Aligarh City, as in the state as a whole.
Jatavs in Aligarh have been an upwardly mobile caste for over a century; the
caste's members having changed their preferred name from Chamars (tra-
ditionally leather workers) to Jatavs. Some converted to Sikhism and
 Aligarh: Politics, Population, and Social Organization /51

Christianity in the pre-Independence years. Many in Aligarh, following the

example of Dr. B. R. Ambedkar, who initiated this form of protest against
Hindu caste discrimination in Maharashtra years before, adopted Buddhism
in a mass conversion at the Achal Talab (see Figure 2.1), site of the central
place of worship for Hindus in Aligarh, in 1957.26 Although most Jatavs are
manual laborers, some are "engaged in various types of professions and occu-
pations including small-scale industries. "27
Politically, the Scheduled Caste population, particularly its Jatav/Chamar
segments, has played a critical role in several elections in Aligarh. Although
in the state as a whole, Muslims and Scheduled Castes were part of the broad
Congress coalition that underpinned that party's dominance in most of the
post-Independence period up to the 1980s, as noted above, Jatavs in Aligarh
broke early from the Congress to join with Muslims in administering the
party's first major defeat in the city and the district. Since that time, Scheduled
Castes have at times supported the Congress, at times opposition parties, and
lately have concentrated their votes behind the BSp.28 Despite fluctuations
in Jatav voting behavior, there has been a consistent underlying basis of antag-
onism to parties such as the Congress and the BIP that have been dominated
by upper castes, and a potential reservoir of support amongst the Jatavs for
leaders or parties that speal( on their behalf.
As in other parts ofU .P., Scheduled Castes are divided politically. The low-
est of the low castes, those whose traditional occupation is sweeping and scav-
enging, traditionally called Bhangis, but nowadays Balmikis, formerly
supported the Congress. 29 They did not join with the Jatavs against the
Congress in the 1960530 and in the 1990S have not supported the new
Scheduled Caste party, the BSP, which they are said to consider a party "for
the Jatavonly, not for any other."31 There are a number of other smaller groups
of Scheduled Castes in the city, notably Koris (traditionally weavers}, Dhobis
(washermen), Dhuniyas (cotton carders), Aheriyas,32 Kolis (foundry work-
ers},33 and Baghels. The bulk of the votes among these non-Iatav Scheduled
Castes probably now goes to the BJP.34
Although Muslims and Jatavs among the Scheduled Castes have sometimes
joined together electorally, it has also been reported, especially in recent years,
that persons from some Scheduled Castes have participated in riots against
Muslims. I was told that Scheduled Castes had "participated in a big way" in
the 1990-91 riots in A1igarh, destroying and looting Muslim property, and had
become "tools" in the hands of militant Hindus. 35 It was also charged that money
was provided to lower castes by Krishna Kumar Navman for the purpose. 36
Persons from the Koli and Bhangi castes, particularly, were mentioned as hav-
 52 / Aligarh: Politics, Population, and Social Organization

FIG. 2.1. Achal Talab, 1999

ing engaged in "arson, looting, and stabbing,"37 including working hand in

hand with the police in singling out and stabbing and killing Muslims.38

Muslim Baradaris

Although Islam does not recognize status or caste categories, Muslims in India
are generally divided in fact into status and caste-like categories, called bara-
daris (brotherhoods) in Aligarh as well as elsewhere in the subcontinent. These
baradaris are endogamous and are frequently associated "with a specific tra-
ditional occupation or lineage."39 The baradaris also sometimes act as cor-
porate groups, establishing mosques, shrines, and schools,40 cooperating in
economic activity, and taking political decisions on a group basis, influenced
by "formal or informal leaders" who "establish liaison" between their
baradaris and political parties. 41 Mann has argued for the importance of
baradaris, which she characterizes as "the core unit of social organisation in
the [old] City"42 and which she claims have persisted as boundary-defining
and identity-forming groups that are not necessarily all moving "towards a
high Islamic tradition,"43 nor, by implication, therefore, towards an over-
arching Muslim solidarity. Although she does not deny the "relevance of a
corporate Muslim identity," she stresses the importance of "context" in
 Aligarh: Politics, Population, and Social Organization /53

determining the influence of baradari, on the one hand, and Islam, on the
other.44 Although there are other ways in which the Muslim populations of
Aligarh and elsewhere have been classified,45 for political purposes it is
baradaris, on the one hand, and the Muslim community as a whole, on the
other hand, that are most important.
Mann has identified twenty-four baradaris in Aligarh and listed them in
hierarchical order based on her interviews between 1984 and 1986.46 She has
also identified some of the localities in which large concentrations of particu-
lar baradaris are found or where a particular baradari is prominent "in its inter-
nal affairs."47 Table 2.2 lists those localities as well as others that Haqqi and
others, including myself, have identified as being numerically dominated by
or otherwise associated with particular Hindu castes and/or Muslim baradaris.
There are no precise or reliable estimates of the size and relative numerical
ranking of the Muslim baradaris. The most "prominent," socially, econom-
ically, and politically, in Mann's view 4B as well as in my own observations are
the Qureshis, Ansaris, and Saifis. These three groups have undergone con-
siderable social change and upward mobility in Aligarh and elsewhere and,
like many Hindu castes, have adopted new names that do not carry the stigma
associated with the older names. 49
Qureshis are a quite numerous baradari that is commonly identified with
the name Qasai, which in turn was traditionally identified with the occupa-
tion of butcher. However, the name Qasai, as well as the occupation associ-
ated with it, is rejected by virtually all Qureshis in the city, whose members
have a multiplicity of occupations, including lock manufacturing, fruit and
vegetable selling, transport, and manuallabor. 50 The baradari as a whole has
been upwardly mobile for some time; it has been undergoing "rapid economic
growth," according to MannY Further, many of the most active and promi-
nent Muslim politicians in the city come from this baradari. There are also
Qureshis on the faculty of the AMU.
Haqqi characterizes the Qureshis as "the largest and most compact single
unit among the Muslims." In 1978, he described the baradari as socially cohe-
sive, but economically poor, with a low literacy rate and "negligible political
influence,» with a "voting strength of ... a little over 5,000"52 or approximately
5 percent of the electorate. Much has changed in the intervening years, as the
members of this baradari have become more prominent in Aligarh public
life and more politically influential.
Twelve mohallas have been identified in Table 2.2 in which Qureshis are
the predominant baradari, and two in which they share predominance with
Pathans. Pathans predominate in five mohallas and share dominance in three
 54IAligarh: Politics, Population, and Social Organization
TABLB 2.2. Mohallas in which particular castes/baradaris/sects are predominant
Predominant Riot-hit/
Ward Mohalla caste/baradarilsect Crime-prone

Kanwariganj Sunet Ansari

Kanwariganj Chandan Shaheed Ansari
Kanwariganj Tila Ansari
Kanwariganj Bani Israilan Ansari Riot-hit and
crime-prone
Kanwariganj Chira Ghachain Ansari
Achal Talab Surendranagar Bania/Thakur
Shahpara Sarai Ghosian Ghosi
Achal Talab Gambhirpura Maithe1 Brahman
New Ward 33 JwaJapuri Maithe1 Brahman
Achal Talab Vishnu Puri Maithe1 Brahman
(Sharmas/Misras)
Kanwariganj Gali Hajjarnan Nai
Jaiganj Turkman Gate Pathan
Kanwariganj Atishbazan Pathan Riot-hit and
crime-prone
Turkman Gate Turkman Gate Pathan Riot-hit and
crime-prone
Turkman Gate Usmanpara Pathan Sensitive and
crime-prone,
but not riot-hit
Turkman Gate Ataiyan Pathan
Jaiganj Sarai Pathanan PathanlQureshi
Turkman Gate Delhi Darwal,ll Pathan/Qureshi Riot-hit and
crime-prone
Kanwariganj Tantanpara Pathan/Saifi Sensitive
Mamubhanja Sarai Behram Beg Qureshi Riot-hit, but
not crime-prone
Raghubirpuri Sarai Rahman Qureshi Riot-hit, but
not crime-prone
Mamubhanja Phapala Qureshi
Shahpara Qanungoyan Qureshi
Shahpara Sarai Sultani Qureshi
Shahpara Qazipara Qureshi
Shahpara Sarai Bibi Qureshi
 Aligarh: Politics, Population, and Social Organization /55
TABLE 2.2. (continued)

Predominant Riot-hitl
Ward Mohalla castelbaradarilsect Crime-prone

Shahpara Khaidora Qureshi Riot-hit and

crime-prone
Jaiganj BabriMandi Qureshi
Jaiganj Sarai Kale Khan Qureshi
Kanwariganj Khaidora Qureshi Crime-prone
Turkman Gate SaraiMian Qureshi
Kanwariganj Sheikhan Sheikh/Saifi Sensitive
Jaiganj Syedwara Shi'a
Mamubhanja ManikChauk Varshney (Barahseni) Riot-hit

others. Three of the five predominantly Pathan mohaUas are considered com-
munally sensitive, riot-hit, and/or crime-prone.53 Four of the twelve Qureshi
mohallas are also so classified. Two others in this category are mohallas in
which Pathans share dominance with Qureshis and Saifis. Several of my inter-
view respondents and some documentary sources as well claim that Qureshis
are particularly active during riots, but this is a disputed matter that will be
taken up at several places later in this book. Strangely, though the associa-
tion in the table between riot-proneness and Pathan localities is even stronger
than that with the Qureshis, I have not myself heard the Pathans mentioned
as a baradari that is particularly active in riots.
The third numerous and important baradari is the Momin Ansars
(Ansaris), who predominate in five mohallas, only one of which is classed as
riot-hit and crime-prone. Ansaris in Aligarh today are mostly manufactur-
ers and workers, according to Mann, but there are also prominent Ansaris
in many walks of life, including at the AMU.S4
In the pre-Independence era, Ansaris, the largest Muslim baradari in the
whole of north India, were mostly pro-Congress, while the upper-status
baradaris tended towards the Muslim League and ultimately towards the
Pakistan movement. While members of some of the high-status baradaris,
such as the Shamsi, left Aligarh for Pakistan after partition,55 "the Ansaris of
Aligarh largely remained in the city." Their continuing support for the
Congress after Independence was rewarded by their recognition "as a back-
ward group by the government," entitling them to receive "favourable terms
for business loans and educational opportunities from state and government
 56lAligarh: Politics, Population, and Social Organization

agencies,» which in turn made it possible for many members of the baradari
to rise on the economic and status ladders.56
The question of baradari versus a broader Muslim identity is of consid-
erable political importance in Aligarh, as we will note throughout this vol-
ume, where it will be shown that baradari has sometimes undermined
Muslim political solidarity, while at other times Muslim identity has con-
solidated and brought together all or virtually all baradaris on a common polit-
ical platform. However, insofar as baradari members do recognize "a common
sense of being a Muslim," I agree fully with Mann that this sense of unity "is
not vested in a high Islamic tradition." There is, in her view, both an inter-
nalization of a sense of Islamic oneness (which, however, "does not stand up
very well to external pressures") as well as an externalization that arises
"through contrast, and at times even confrontation, with other religious com-
munities."57 As I will argue below, it arises in Aligarh primarily out of polit-
ical contestation against the rise of militant Hinduism. In the absence of such
direct contestation, the tendency among Muslims, as among Hindus, is to
identify with the contradictory but" empirical reality of baradari and elites,"
rather than with "solidarity under a banner of Islam."58
Mann has provided a dear example of inter-baradari conflict in Aligarh,
between the high-status Shamsi and the lower-status baradaris, led by the
Ansaris, for control over the dargah (Muslim shrine) of Shah Jamal, a strug-
gle over "both spiritual and financial" resources under the control of the dar-
gah, as well as an assertion of status on the part of the Ansaris.S9 Such examples
of inter-baradari conflict over the symbols and rituals of Islam challenge, in
her view, the role of the latter in building Muslim solidarity. My own inter-
views provide evidence, also, of political divergence among baradaris in elec-
toral contests, which will be noted in the analysis of electoral contests in Part
IV. These examples of inter-baradari conflict support the argument that will
be made below, that Muslim solidarity arises primarily as a consequence of
political struggle with militant Hindus, that even then it is difficult to achieve,
and that it is not at all religious sentiment that provides the catalyst for polit-
ical solidarity.

Caste, Community, and Political Influence

Varshney has noted that, if the figure of 21 percent is taken as an accurate

estimate of the population of the trading castes in the city, and if one adds
to this figure the known figures for Muslims and Scheduled Castes, then these
 Aligarh: Politics, Population, and Social Organization /57

TABLE 2.3 Caste/community of members

of the Aligarh municipal corporation, 1995

Total Hindus and others 29

Bania and other business castes 11
Agarwal (5)
Varshney (4)
Khatri (1)
Sindhi (1)
Brahman castes 7
High Brahman castes (4)
Maithel Brahmans (3)
Thakurs (Jadon) 2
Backward castes 8
Lodhi (4)
Saini (1)
Nais (2)
Other (1)
Jain
Muslims 19
Scheduled Castes 12
TOTAl. 60

three categories of the population comprise "'about 70 per cent of the town
population."60 Although there has been a tendency in Aligarh politics for these
three potential groups in the population to act as if they were politically solid
blocks, I have noted above that there are multiple divisions among all three
of these broad categories that also express themselves in politics from time
to time. Both tendencies will be analyzed in detail in succeeding chapters. It
is readily observable, however, from the distribution by caste and religion of
the members of the Aligarh municipal corporation elected in 1995, that
Varshney's calculations are precisely reflected therein (see Tables 2.3 and 2.4):
the Bania castes taken together account for 18 percent of the seats, the Mus-
lims for 32 percent, and the Scheduled Castes for 20 percent, amounting in
all to 70 percent of the total.
Table 2.5, which gives the distribution by political identification of the
castes and communities represented in the municipal corporation, also
 58 / Aligarh: Politics, Population, and Social Organization

TABLE 2-4. Caste/community of members (in percentages)

of the Aligarh municipal corporation, 1995

Hindus and others 48

Bania and other business castes 18
Brahman castes 12
Thakurs 3
Backward Castes 13
Jain 2
Muslims 32
Scheduled Castes 20
TOTAL 100

TAB LE 2.5. Political identification of members of the Aligarh

municipal corporation by caste/community, 1995

Political Identification
Castelcommunity BJP BSP SP Ind.

Hindus and others 27 0 0 2

Scheduled Castes 5 6 0
Muslims 0 5 8 6
TOTAL 32 11 8 9

gives a sense of the relative importance of caste and communal solidarityl

division as a basis for political influence and control in Aligarh. In the 1995
municipal corporation, all but two Hindus (a Varshney and an Agarwal)
affiliated with the BJP. The Scheduled Castes divided three ways: six with
the BSP, five with the SIP, and one Independent. The five Scheduled Caste
SIP members provided that party with majority control in the corporation.
Among the three main caste/communal categories in the city, the Muslims
were the most divided: six were independents, eight affiliated with the SP,
and five with the BSP. As of 1995, therefore, one could say that political
identifications in the municipal corporation were communalized to a degree,
but not polarized. Non-Scheduled Caste Hindus and Muslims shared no polit-
ical identification. "Hindus and others" (the latter including one Jain) were
virtually completely unified across all caste and religious boundaries, but
 Aligarh: Politics, Population, and Social Organization /59
Muslims were divided. Scheduled Castes constituted a divided swing force.
It will be shown in succeeding chapters how politics in Aligarh City have
fluctuated since Independence between poles of consolidation and frag-
mentation, communalization and internal division, polarization and depo-
larization, and how these fluctuations have been influenced by riotous
violence.
 31 Hindu-Muslim Violence in India and Aligarh

THE INCIDENCE OF HINDU-MUSLIM RIOTS

AND THEIR CONSEQUENCES FOR THE MEMBERS
OF THE TWO COMMUNITIES

Incidence and Basic Facts: 1950-95

T he annual reports of the Home Ministry of the government of India,

whose responsibility includes reporting on the state of law and order
in the country, including the incidence of communal riots, has failed
to produce its annual reports for the past fifteen years, the most recent one
available being for the year 1984-85. Those reports that were produced in
earlier years have many defects, including absence of a definition of what
constitutes a communal incidentl and a failure to specify for most years the
numbers of persons killed according to religious community as well as the
numbers killed in communal clashes and those killed by the police. The latter
types of figures certainly exist, but have been available even unofficially only
for the years 1968 to 1980.
Official figures compiled from several sources, including the Home
Ministry, calculate the total number of incidents of communal violence
between 1954 and 1982 as 6,933, but provide no other details between 1954 and
1967 and after 1980. Between 1968 and 1980, the Home Ministry reported that
there had been 3,949 communal incidents in which 530 Hindus, 1,598 Mus-
lims, and 159 "other" persons and police personnel were killed.2 The latter
figures confirm, at least in the period for which such a breakdown by com-
munity is available, the often-stated fact that a disproportionate number of
Muslims have been killed in communal riots. In some riots, the ratio of Mus-
Hms to Hindus killed has been very much higher. For example, during the

60
 Hindu-Muslim Violence in India and Aligarh /61

riots of September 1969 in the city of Ahmadabad, 512 persons were killed of
whom 24 were Hindus, 430 Muslims, 58 "others" and unidentified;3 the latter
category is a rather grisly one since it suggests either burning of the murdered
person's body beyond recognition, or mutilation. The official figures that are
available as well as media and other reports concerning police treatment of
Muslims during riots demonstrate clearly, also, that police arrest, fire upon,
and kill disproportionately more Muslims than Hindus. 4 Moreover, con-
cerning several major riots, commissions of inquiry have established that the
police arrest innocent Muslims, kill them inside their homes, and enter
mosques to shoot and kill Muslims as well.
Varshney and Wilkinson, using news reports from the Times ofIndia, have
compiled sets of figures on the numbers of riots-defined as communal inci-
dents in which there was at least one death-for India as a whole and by state
for the period 1960-93. Unfortunately, the figures have been published only
in chart form; consequently, exact numbers for the entire period or for any
part of it cannot be calculated. The trends, however, are quite clear from their
charts. The number of Hindu-Muslim riots rose during the 1960s, reaching
a peak in 1969, declined between 1971 and 1977, then began "an unambigu-
ous and alarming increase during the years from 1978-93."5
Rioting and killings in the years between 1990 and 1993 reached peaks not
seen since 1947. In these three years, there were two waves of riots across large
parts of northern and western India, associated with the mass mobilizations
and provocative and incendiary tactics used by two of the organizations in
the RSS family, the Vishwa Hindu Parishad (VHP) and the BJP, to mobilize
the Hindu community behind the demand to remove the Babri Mosque from
its site in Ayodhya. This movement was also associated with the electoral strat-
egy of the BJP to displace the Congress from power in the northern and west-
ern states and at the Center by consolidating the Hindu vote in its favor. Since
the last wave of riots occurring in December 1992 and January 1993 in the after-
math of the destruction of the mosque, there has been a marked decline in
the incidence of communal riots, but no exact figures are available at this
writing. 6
Among the fifteen largest states in the Indian Union, five ranked especially
high in the incidence of Hindu-Muslim clashes involving fatalities; in rank
order by number of such clashes, they were Gujarat, Maharashtra, Uttar
Pradesh, Bihar, and Madhya Pradesh.? In Uttar Pradesh (U.P.), which is the
focus of our inquiries here, Wilkinson identified 193 riots in the period between
1950 and 1993, in which 1,313 deaths occurred.8 Varshney's figures for riot deaths
 62/ Hindu-Muslim Violence in India and Aligarh

in D.P. in riots in which there was at least one death in the period 1950-1993
show a sharp drop in the years after Partition, between 1950 and 1980, when
the number of deaths in a single year never reached the level of 1950 (approx-
imately 75 deaths). In many years during this period, there were no riots with
fatalities. Although there were several years in which there were major riots
with large numbers of fatalities, particularly 1961, 1968, 1971, 1972, and 1978,
the peak year in this period was 1978, when fatalities rose to nearly 60. Between
1980 and 1992, however, the situation in the state changed dramatically.
Successive new peaks of riot-produced fatalities were reached in 1980 (over
160 deaths), 1990 (over 180 deaths), and 1992 (over 200 deaths). In 1987, also,
the number of deaths (above 140) surpassed the number in 1950. Moreover,
the general level of such violent riots remained quite high during this second
period, except in 1981 and between 1983 and 1985. 9
The violence in D.P. has also been spread widely among a number oflarge
and small cities and towns. Between 1960 and 1993, Wilkinson identified 24
cities and towns in India that he characterized as "riot-prone." Of the 24, six
(25 percent) were located in D.P., the largest state in India, containing 16 per-
cent of the population of the country as a whole. Fifty-nine of the 249 com-
munal incidents, or 24 percent, occurred in these six towns, and 579, or 14
percent, of the total of 4,005 deaths. Within D.P., figures provided by Wil-
kinson for the longer period from 1950 to 1993 placed the town of Aligarh at
the top; according to his dataset, there were 25 riots in this town in that period,
in which 388 persons were killed.lO However, his count of both Hindu-Muslim
riots and of reported deaths is not accurate (see Table 3.1). The count I have
compiled includes 18 riots and riotous periods, in 10 of which there were deaths
in the period 1950-93, giving an estimated total death toll of 176; adding the
pre-Independence and the 1994 and 1995 riots to the count gives a total of 23
riots, 14 with deaths totalling 195. If the other figures in the Wilkinsonl
Varshney dataset are accurate, Aligarh would stand second to the nearby dis-
trict of Meerut, for which their reported death toll, 1950-95, is 265. 11 Even
given the lower figures for Aligarh, it remains clear enough that the town is
among the most riot-prone in the country. Moreover, although Aligarh is
associated in the minds of most Hindus in north India with the name of Sayyid
Ahmad Khan, the founder of the AMD and the Muslim League, and with the
Muslim separatist movement that led to the creation of Pakistan, it is a mat-
ter of some surprise to note that in the entire period from 1900 to 1949, the
number of riots and riot deaths was substantially lower; there were 4 Hindu-
Muslim riots in Aligarh and 11 riot deaths in that period. Aligarh ranked eighth
in the state in that period in the number of riotsP Aligarh's rise to promi-
 Hindu-Muslim Violence in India and Aligarh /63

TABLE 3.1. Riots and riot deaths in Aligarh City, 1925-95

Official or
Date newspaper death toll

September 22, 1925 6

April 10-11, 1927 o
November 3, 1937
March 29, 1946 4
Pre-Independence Total 11
March 3-4, 1950a 5
June 6, 1954 o
September 14, 1956 o
October 1-3, 1961 b 15
1966 (date not specified)' o
1969 (date not specified) o
March 2, 1971 d 17
June 1971 e o
June 1972 f
October 3-5, 1974 o
October-December 1978g 28
May 1979 h 5
June 17, 1979i o
August-November 1980i 11
October 1988 k 2
November 10,1989 1 o
November 1990-January 1991lTI 92
December 1992 Some, but no figure
March 10, 1995" 8
Post-Independence Total 184

GRAND TOTAL 195

·Wilkinson's code sheet from the Times of Indja lists 4 dead in this riot; so does the Free
Press Journal, March 6, 1950. My figure comes from the Times of India (Bombay edition),
March 6, 1950; however, the headline of the article lists 7 dead while the text refers to 5 dead.
There is clearly no consensus in the press on this matter. I have simply chosen the Times of
India over the Free Pre.<s Journal to try to make the data source the same as that for Varshney-
Wilkinson wherever possible. However, the Varshney-Wilkinson dataset lists 68 dead in this
riot, which is certainly incorrect. Wilkinson has acknowledged in a personal communication
that their total may have included Moradabad (a town in a different district) as well as Ali-
garh. So, there is a small discrepancy between my and the Varshney-Wilkinson data on what
 64/ Hindu-Muslim Violence in India and Aligarh

TABLE 3.1 (continued)

the actual figure is in the Times of India as well as a very large one with regard to my figure
and their dataset.
hI have two unpublished documents prepared by a former district magistrate, one of
which, "Riots in Aligarh," gives the official figure of deaths in this riot as 13, while a second,
"Muslims and the Aligarh Muslim University," gives the official death toll as 15, all Muslims.
Since the latter document also notes that "unofficially the number [of deaths of Muslims1
was believed around 40," I am using the larger figure as doser to the truth.
'References to a riot this year and in 1969 from "Riots in Aligarh."
dSource for figure in column 2: Times of India (New Delhi edition), March 31, 1971.
·Centre for Research in Rural and Industrial Development (CRRID), Chandigarh,
"Communal Violence and its Impact on Development and National Integration," unpub-
lished, undated, p. 35. However, I have not been able to find any press reports on this riot.
fSource for this riot and the single death is from "Riots in India." However, the date given
for this riot in the report is August 1972, which is incorrect, so the reported death also may be
erroneous.
8Source for figure in wlurnn 2: Patriot, November 26, 1978. This, of course, is an extended
riotous period, not a single event or duster of events.
hSource for figure in column 2: Times of India, May 12, 1979.
iGraff, "Religious Identities and Indian Politics," p. 72.
j As indicated in the discussion below of this series of riots, there is ambiguity in the press
reports concerning whether or not the death toll, including two policemen, was actually 13.
kSource of figure in column 2: Times of India, October 9, 1988.
IThe Wilkinson dataset lists 3 deaths in this riot, but I found no reported deaths in either
the Times of India (New Delhi edition) or in a Hindi newspaper of which I have only copies
of the relevant artides with no source or date.
mThe Wilkinson dataset gives a figure of 240 deaths in this riot, but is certainly incorrect
and is probably derived from a lumping together of riot deaths in other towns at the same
time with those from Aligarh. The figure of 92 is the official death toll, which neither Muslim
groups nor the People's Union for Civil Liberties ("Communal Riots in Aligarh, Dec.
1990-Jan. 1991," PUeL Bulletin [March, 1991)) accept. Muslim groups daim a documented
list of 100 deaths for Muslims alone, whereas the PUCL estimates the total death toll for both
Hindus and Muslims as between 125 and 150.
nThe figure of 8 deaths is the official figure reported by the District Magistrate to Asghar
Ali Engineer, "Aligarh Riots-An Unplanned Outburst," in Towards Secular India 1, no. 2
(April-June, 1995): 92. The last report that mentioned the munber of deaths in the 1'imes
of India (New DeIhi), March 12, 1995, gave a figure of 6 till that date. The discrepancy again
reveals the dangers of relying on newspaper reports in such matters.

nence as one of the most riot-prone towns in the state and country, there-
fore, is a phenomenon of its post-Independence history.
From the point of view of persistence, there can be no doubt that Aligarh
ranks dose to the top in India in the number and seriousness of Hindu-Muslim
riots since Independence. In Varshney's count, using four different criteria
for riot-proneness in Indian cities and towns, the same five cities appear at
the top by all four counts as well as by the total number of deaths in riots
 Hindu-Muslim Violence in India and Aligarh /65

between 1950 and 1995: Bombay, Ahmadabad, Hyderabad, Meerut, and Ali-
garh. The four criteria are: minimum of 15 deaths in 3 riots over two five-
year periods; minimum of 20 deaths in 4 riots over three five-year periods,
minimum of 25 deaths in 5 riots over four five-year periods, and minimum
of 50 deaths in 10 riots over five five-year periods. i3

Some Other Important Features of Hindu-Muslim Riots

Deaths in Hindu-Muslim riots have three sources: "mob action," police

killings, and "isolated incidents." "Mob action" may take the form of con-
frontations between gangs or crowds from different communities or segments
of them, armed with sticks, knives, swords, occassionally bombs and small
weapons, and kerosene. It often also involves armed gangs from one com-
munity seeking out defenseless persons or whole families in their homes, slash-
ing and cutting up the male members and sometimes the female members,
raping the latter, and burning all alive. A second source is police killings, which
account for a large percentage of deaths in several major riots for which figures
have been provided in inquiry commission reports, and which cannot be
justifi.ed in terms of "crowd control." These killings are disproportionately
of Muslims. Third, a good part of the killing takes the form of cold-blooded
murder of individuals in "isolated incidents" rather than in killings arising
from "mob frenzy." Such killings are often the precipitant or perhaps the start-
ing signal for the production of a riot.
One further feature of riots deserves note. The government of India and
the state governments do virtually nothing after a riot to prosecute and con-
vict persons suspe(.ied of promoting or participating in riots. Occasionally,
but less frequently in recent years, commissions of inquiry are appointed. If
the final reports are not too damaging to the government of the day or to the
political supporters of that government in the Hindu or Muslim communi-
ties, the report may be published. More often than not, there is a significant
delay before publication. Some reports are never made public.14 The con-
sciences, if any, of the authorities are fully served by the payment to the sur-
vivors of riot victims a fixed sum of money-fixed in advance by law and in
a published schedule-for the lives of the persons killed, thereby transforming
riots into a ghastly, crude form of public relief, which in turn is abused by
some members of the public, whose sons and brothers conveniently disap-
pear from the slums after riots while other family members claim-and some-
times quarrel over-the death benefit.
 66/ Hindu-Muslim Violence in India and Aligarh

HISTORY OF RIOTS
AND OTHER COMMUNAL INCIDENTS IN ALIGARH

Problems in Riot Enumeration

Both riot counting and the counting of deaths in riots are precarious exer-
cises. They also carry an odor of callousness about them that must be avoided
as far as possible, for they resemble nothing so much as body counts in con-
temporary insurrectionary warfare. Like such body counts, the numbers of
riot deaths are suspect and contested by opposing sides.
There is first the matter of defining a riot. In Indian criminal law, a riot
is "an assembly of five or more persons" engaged in unlawful activities directed
against government institutions, the laws, persons, or property for the pur-
pose of commiting "mischief or criminal trespass, or other offence."15 Such
a sweeping definition allows for much disputation in law concerning whether,
when, and for what purposes an assembly is or is not "unlawful" and whether
or not the persons assembled committed any of the prohibited acts. Further,
although the case law on the subject refers to various incidents of rioting
between Hindus and Muslims, there is no separate law defining communal
riots. The definition of a riot as a communal or Hindu-Muslim riot, there-
fore, is a matter of public interpretation and labelling that takes place on the
streets, in the media, in the legislatures, in cabinet meetings, and in the civil-
ian and police administration.
There is, therefore, a double ambiguity built into riot counting that stems
from the definition of riots in general and the absence of a definition of com-
munal riots. This ambiguity is compounded when the variable of adminis-
trative action is considered. The latter enters the picture before the law when
the local or state administration does or does not decide to be vigilant in pre-
venting and/or controlling riots. The paradoxical result in this case is that
riot figures may go up-and have in fact gone up-when the administration
is most vigilant in enforcing the law against riots, including communal riots.
Still another complication arises when the numbers of persons involved
in riots increases exponentially beyond the minimal five persons and when
such activities extend to numerous and sometimes widely separated areas of
a city or town. In such cases, the police mayor may not count 50 riots, while
the press and the politicians lump all or most incidents that occur within a
particular time frame as one riot, communal or otherwise. In the Aligarh case,
as will be shown below, this problem becomes extreme when riotous activ-
ity extends over weeks and months. It is a statistical artifice to count, for exam-
ple, periods of riotous activity extending over three or four months, such as
 Hindu-Muslim Violence in India and Aligarh / 67

occurred in Aligarh in 1978 and 1980 (see Table 3.1), as one riot or two or three,
when the number of incidents-some communal in nature, some not, some
ambiguous-may be very high.
One might think that there would be no similar ambiguity in the count-
ing of persons killed during a period of rioting. A death is not usually a mat-
ter of dispute, although the circumstances of the death may be. Nevertheless,
every large-scale riot in India occasions dispute over the numbers of persons
killed or injured. Members of the community targeted in riots always say
that the number of deaths was higher than that officially noted, while per-
sons from the other community say the figures were accurate. The local
administration obviously has an interest in minimizing its failure to main-
tain law and order by undercounting. Rival parties in and out of office exag-
gerate or minimize the death counts to blame and embarrass each other or
to decrease responsibility.
But, how can the counts be so disputable? There are grisly as well as some
less grisly reasons for overcounting or undercounting. Bodies are burnt to ashes,
thrown in canals, dumped in wells and sewers. As previously indicated, men
may disappear during riots so that their families can make a claim for death
payments from government. For all these and other reasons, both the count-
ing of riots and the numbers killed and injured must be treated cautiously.
Although Aligarh ranks very high in the country in riot-proneness pro-
portionately to population and in the number of deaths over the 50 years from
1946 to 1995, there has been, even here, considerable variation in the num-
ber of riots and the number of deaths that have occurred during them. Of
the twenty riots classifiable as Hindu-Muslim, eleven led to deaths, distrib-
uted as follows: 1946-50, two riots with 9 deaths; 1951-55, no riots with deaths;
1956-60, no riots with deaths; 1961-65, one riot with 15 deaths; 1966-70, no
riots with deaths; 1971-75, two riots with 18 deaths; 1976-80, numerous riots
with 44 deaths; 1981-85, no riots with deaths; 1986-90 (terminating in January
1991), two riots with 94 deaths; 1991-95, two riots with more than 8 deaths.
Since the number of communal riot~ in two of these periods cannot be counted
accurately, extending as they did over several months in 1978, 1980, and 1990-
91, they cannot be presented graphically. Therefore, the number of deaths in
each period will have to stand-and reasonably so, for that matter-for vari-
ation in the intensity of riot activity in Aligarh (see Figure 3.1).
There are three aspects of riot variation to note from these figures and
the accompanying chart. First, there was an initial decline in riot activity in
Aligarh after the rioting that occurred before and after Partition, as elsewhere
in the country. Second, there are four periods in which there were no riots
 68/ Hindu-Muslim Violence in India and Aligarh

lOO~------------------------------------------,

80

60

40

20

Deaths
o+-__~__~~__~__~__~~__~__~____~~
1946-50 J956-60 1966-70 1976-80 1986-90
1951-55 1961-65 1971-75 1981-85 1991·95

Five-Year Period
FIG. 3.1. Number of deaths in Hindu-Muslim riots
in Aligarh by five-year periods, 1946-95

in which deaths occurred. Third, however, over the entire period there has
been a marked increase in the peaks of violence, with the number of deaths
increasing sharply with each successive time of intense and murderous riot
activity.
Intense rioting in Aligarh has at times deviated from the pattern in the
rest of the country, at times mirrored it!6 For example, there were no riots
with deaths in Aligarh in 1966-70, when riot deaths in the country reached
a peak not to be encountered again until the post-Ayodhya period. Similarly,
although an unknown number of deaths occurred in December 1992 and an
additional 8 deaths occurred in March 1995, the number was certainly sman
in comparison to other parts of the state and country, which were experi-
encing between 1991 and 1995 the worst five-year period in the history of post-
Independence India, as wen as in the state of U.P. On the other hand, the
two five-year intervals from 1971 to 1980 were ones of intense riotous activ-
ity in Aligarh, partly mirrored in the state as a whole, but much less so in the
country as a whole. Finally, there is one period marked by intensive riotous
activity in Aligarh, in the state of U.P., and in the country as a whole, namely
that of 1986-90, in which the movement to destroy the mosque in Ayodhya
waxed. It is the case, therefore, that intense riotous activity in Aligarh must
 Hindu-Muslim Violence in India and Aligarh /69

be explained partly in terms particular to the city and partly in relation to

activities taking place simultaneously in the state and in the country as well
as in Aligarh.
There is a further aspect of variation in Aligarh's history of riot produc-
tion. If we count all riots in the post-Independence period, then it would
appear to be the case that Aligarh is a site of persistence, of recurring riotous
activities. That fact would be further confirmed by visits to particular riot-
prone mohallas where, as the Centre for Research in Rural and Industrial
Development study has noted, and I have personally experienced many times,
there is a palpable sense of tension "in the air," evidenced by the existence of
simmering local issues involving Hindus and Muslims on opposite sides and
by the reactions to my very presence in such mohallas, especially when I have
attempted to make inquiries on these local issues. Suffice it to say tl1at I and
my companions were always eager to leave such places without pursuing ques-
tions very far, as crowds, almost always Hindu crowds, began to form and
persons began to appear wanting to know my business and eager to set me
straight in no uncertain terms about my questions.
Yet, even given the situation just described in the most riot-prone areas
of the city, there is variation in the very fact that some riots lead to killings,
others do not; that some riots are prevented or controlled quickly by firm
administrative action, others are not; and that, at times, riots occur when least
expected and do not occur when most expected. In all these respects, the his-
tory of flux in riotous activity in Aligarh provides ample materials to con-
sider the dynamics of riot production from all relevant angles. The remainder
of this chapter will provide a detailed history, from information provided in
official records, newspaper reports (mostly from the Times of India), a few
secondary sources, and my own interviews, of all riots in which either deaths
occurred or there are reports of violence and injuries from five or more sites
in the city (see Table 3.1 and Appendix Table A.2) between 1925 and 1995, a
seventy-year period.
In this overview, especially with regard to newspaper reports and gov-
ernment documents, the information that comes from these sources is pre-
dominantly of two types. The first consists of factual accounts of specific
actions said to have precipitated the riots, occasionally including some back-
ground describing "rising tensions" that preceded the precipitating incidents,
and accounts from day to day of particularly dramatic incidents involving
clashes between Hindus and Muslims or between the police and rioters, all
of this accompanied by daily body counts and counts of the injured. In the
case of the larger riots, there also generally follows more or less extensive com-
 70/ Hindu-Muslim Violence in India and Aligarh

mentary from a variety of individuals, groups, political parties, the media itself,
and other sources on their causes, which are predominantly exercises in blame
displacement. In the largest riots, which occasion not only press reports but
reports from visiting teams of human and civil rights organizations, we get
glimpses of the dynamics of riot production that are otherwise hidden from
view. My own interviews here and in the succeeding chapters of this volume
are designed to enlarge to the extent possible the concealed and systematic
aspects of riot production. Sometimes the reporting on riots also notes a con-
nection between them and local and extralocal political circumstances. Some
attention will also be paid to those circumstances in this chapter, but they
win be given more detailed treatment in later chapters.

Pre-Independence Riots

It is noteworthy, given the nationalist historiography concerning British poli-

cies of divide and rule and their responsibility for communal antagonisms
and violence, that the number of riots in this town-and for that matter in
the twenty towns in U .P. classified by Wilkinson as "riot-prone"-were greater
in the forty-three years from 1950 to 1993 than in the 49 years from 1900 to
1949. The earlier period includes not only all the years of British rule in the
twentieth century, but the years of pre- and post-Independence, post-
Partition rioting in 1946-47.
Between 1923 and 1927. there were 88 riots classified as communal in the
United Provinces (as U.P. was then known) "in which thirty-nine Hindus
and forty-two Muhammadans were killed and one thousand five hundred
and sixty-six Hindus and seven hundred and thirty-five Muhammadans were
wounded. "17 Serious rioting was reported from several of the large cities and
towns in U.P. in each of the five years between 1923 and 1927. These riots
occurred during an intense period of competitive Hindu-Muslim political
and religious mobilization accompanied by a great wave of rioting following
upon the collapse of Hindu-Muslim cooperation during the combined non-
cooperation/Khilafat movement of 1921-22. A riot occurred in Aligarh dur-
ing this period on September 22,1925. The Aligarh riot of 1925 was the major
riot noted in that year in U,P, and was among the worst of "sixteen com-
munal riots reported in 1925" in the country as a whole,1s Indeed, it even drew
the attention of the London Times in two notices on September 25 and 26,19
Although the London Times referred to this riot as having occurred in "the
Moslem university city of Aligarh in connexion with a Hindu procession,"
it appears that the riot had nothing to do with the university itself, having
 Hindu-Muslim Violence in India and A.ligarh /71

occurred in the old city in the Madar Darwaza (Madar Gate) area (referred
to as "Madho Dharwaza" in the report) and having in fact begun during a
Ram Lila procession. The official death toll in this riot was 6, of whom 4 were
Muslims, 2 Hindus. 20
Information is also sparse concerning the second riot, which occurred on
AprillO-ll, 1927. The press referred to its origins in a "fracas" arising out of
"a private quarrel at the carriage stand" between two castes or baradaris over
a contract for the stand awarded by the District Board. News of the fracas
"spread in the city," and fighting was "free and fierce," but did not involve
the university. Both "the official version" and the press version took the view
that religious antagonism was not the "cause" of the riots.21
The third pre-Independence Aligarh riot occurred on November 3, 1937.
It took place in an atmosphere of intensifying political competition in the
province as a whole and in Aligarh between the Congress and the Muslim
League and among Muslim groups as well, some of whom supported the for-
mer, others the latter political organization. Although another considerable
wave of rioting occurred during the Hindu festival of Holi in U.P. in March
1938 at several sites, Aligarh was not affected at this time.l l
In 1946-47, the Congress government in power on the eve of Independence
paid special attention to "law and order in Aligarh district with particular ref-
erence to the Aligarh University." They specifically were concerned to pre-
vent an "invasion of the City by the University students," and devised a "riot
scheme" that provided "for guarding six main points on the railway line"
(dividing the Civil Lines area where the university is located from the city;
see Map 1) "through which the students are likely to enter the city." An entire
"military police company" was to be posted "at three of these points. "23
The last riot in this pre-Independence period occurred in Aligarh on March
29, 1946. At the end of March and in eady April 1946 there were several riots
in other western districts of U.P. as well, which were attributed in press reports
directly to political conflicts between militant pro- and anti- Pakistan forces,
namely, the Muslim League on one side and the RSS on the other. Although
the origin of this series of riots was said to have been an incident in Aligarh
on March 29, which then "had repercussions" in other western districts, the
press reports of the time did not attribute the Aligarh incident itself to conflict
between RSS and Muslim League activists. On the contrary, it was attributed
to "student indiscipline," in the Times of India as well as in the Governor's
Report found in confidential records since made available. According to both
accounts, the riot began with an altercation between AMU students and the
proprietor of a Hindu cloth shop, in which the students beat up the shop-
 72/ Hindu-Muslim Violence in India and Aligarh

keeper, after which "a full-blown riot ensued during which the whole grain
market got burnt down."
The British governor dearly blamed the students for the precipitating inci-
dent, referring to them as "a bobbery lot," for fear of whom the "Hindus in
the town always live in a kind of half panic." This riot occurred after several
months of electioneering in the province and on the eve of the transfer of
power in the province to the Congress, which had won a decisive victory. The
governor noted that "the local authorities" had been charged with doing "noth-
ing at all to control the students" and anticipated "vindictive" action on the
part of the incoming chief minister, Pandit Govind Ballabh Pant, against both
the local administration and the students, but hoped to have "forestalled him
by ordering an immediate joint enquiry by the Commissioner and the local
D.I.G. of Police."24 The chief secretary's report dated four days later also
blamed "student indiscipline" for the outbreak and provided a few additional
details on the manner in which it spread. 25
Although the precipitating incident appeared to have nothing to do with
the preceding ele(,:tions and political activity, the action of the AMU students
immediately became an interparty and intercommunal issue. The chief sec-
retary reported that "the condemnation of the Aligarh students for their indis-
cipline and hooliganism is being interpreted as a biased attempt to victimize
Muslims in general and the League in particular."26 Chief Minister Pant, in
fact, wished to take action against the AMU by imposing "some sort of penalty
on the University and to justify himself with his public, "27 a desire that caused
an internal dispute in the government between him and the British governor
Wylie. 28
Precise figures were provided for the numbers of persons killed and
injured as well as for the duration of this riot. It was a one-day affair, but 4
persons were killed and l6 injured. These latter figures are important for com-
parison with later events in the town, for they constitute a kind of bench-
mark. The years between 1946 and 1948 are when the vast carnage associated
with the partition of India occurred, when political leaders and organizations
on both sides openly called for violence, revenge, and retaliation over the par-
tition decision and over violent incidents attributed by one side to the other.
The Pakistan movement itself grew out of conditions in U.P., and the AMU
was one of its storm centers. Yet, in Aligarh City, the violence was contained
to a day, the deaths to 4, and the injured to 16. All riotous events in Aligarh
that have occurred since 1946 should be compared with this one that occurred
in the midst of an extremely hostile and violent political and social environ-
ment when India was still under British rule.
 Hindu-Muslim Violence in India and Aligarh /73

The British governor remained concerned throughout the remainder of

the year 1946 about the spread of communal violence in the province and
about the partiality of the Congress government in dealing with it. Although
Aligarh was not the worst site of violence in U.P. in this period,29 the British
were concerned with the intensification of political conflict between Congress
and the Muslim League over the Pakistan demand; the formation of and
recruitment to volunteer paramilitary forces established by the Muslim
League, the RSS, and other militant groups; and, after August, the possibil-
ity that the Muslim League's "Direct Action" movement for the achievement
of Pakistan, which was associated throughout north India with considerable
violence-most famously in Calcutta in August 1946-would spread to the
AMU campus. It was reported that, in June, the RSS was very actively orga-
nizing camps, offices, and branches in several districts, including Aligarh
(though in another town than Aligarh City) and that the Muslim League was
particularly active in three districts in the state, including Aligarh, in "the train-
ing of Muslim Guards," the Muslim League's paramilitary force. 30 It was, in
fact, reported that the staff at the AMU was "in favour of 'direct action'" and
that "it would be difficult to control the students if religious aspect is given
to the 'direct action."'31
For its part, the Congress government in the United Provinces set out in
October to establish "Home Guards" in nine towns where violence had
occurred between Hindus and Muslims. These Home Guards, the British gov-
ernor thought-most likely correctly-were designed as a government-sup-
ported volunteer force to deal with any Muslim League-sponsored "Direct
Action" in those towns. Aligarh was among the nine towns, with a force of
300 to be establishedY In November, it was reported "from Aligarh ... that
the University hopes to send a batch of 100 students for relief work and the
students are also collecting subscriptions."33 This reference suggests that the
orientation of the AMU students was outward to the broader areas of the prov-
ince and the country where Hindu-Muslim riots were taking place. The British
were concerned with the spread of what they perceived as a chain of riots
that had begun with the Great Calcutta Killings of August to the riots in
N oakhali where Hindus were the principal victims to the riots in Biliar where
mostly Muslims were killed. Indeed, this same report noted that a secret
meeting was held at the AMU campus, "attended by over 2,000 persons," at
which one ''A. T. Mustafa, Vice-President of the Muslim University Union
in appealing to all students to take revenge for events in Bihar, stated that he
had killed 10 persons with his own hand during the Calcutta riots." Further,
it was noted that "at another meeting on November 10 the necessity for revenge
 74/ Hindu-Muslim Violence in India and Aligarh

was again stressed. Members of the staff and ex-members are taking part in
these activities, the Khaksars under the leadership of Mr. Shamin, a lecturer
in the Chemistry Department, being particularly active." The Khaksars were
a Muslim paramilitary force whose purported aim was to defend Muslims
from Hindu attacks in riots.
In the second half of November, the chief secretary's fortnightly report
noted again the political activity at the AMU, where "there have been several
meetings both of the students and the staff, at which very strong speeches
were delivered against the Congress and the Hindus. "34 On the other side,
the RSS was recruiting intensively in the district of Aligarh, where the chief
secretary remarked that an attendance of 2,000 at the training camp in the
district was the highest among 8 districts where such camps were held on the
first of January 1947. 35
It is evident, therefore, that the AMU, at this critical moment in the mod-
ern historyofU.P. and the country, was seen by the incoming Congress gov-
ernment as a primary source of Muslim political organization and communal
activity directed against the Congress and Hindus in general. That the AMU
students were feared by the Hindu population in the town was also explic-
itly stated by the British governor. It is also clear from the disputes between
the governor and the chief minister that the incoming Congress government
was bent on action, considered vindictive by the governor, against Muslim
organizations and institutions, including the AMU. Furthermore, Pandit Pant
accused the police in Aligarh of partiality-obviously, in favor of Muslims-
in the March riot in Aligarh, at a time when the police force in the state was
47 percent Muslim. It is also known that Pant took immediate steps to change
this situation by various actions designed to reduce the Muslim component
in the police force, which, by the end of his tenure, had fallen to approxi-
mately 5 percent. That position has not changed significantly since these early
years of Congress rule in U.P.3 6
Several aspects of the riots in the pre-Independence period deserve par-
ticular notice in relation to the general issues raised earlier and to be discussed
throughout the remainder of this volume. First, the waves of rioting as well as
the particular incidents in Aligarh were placed by the authorities and the press
in a dual context: one religious-communal, the other communal-political.
The first form of contextualization referred to the riots as precipitated by con-
troversies associated with the manifestation of religious zeal or playfulness
that irritated the other community, in these cases particularly Muslim objec-
tions to Hindu Ram Lila processions or Holi pranksterism. The second form
of contextualization was communal-political, relating Hindu-Muslim riots
 Hindu-Muslim Violence in India and Aligarh /75

to the communal competition for power in the state and the country between
the Congress and the Muslim League. Sometimes, both contextualizations
were present. Thus, the first wave of riots between 1923 and 1927 followed
upon the collapse of Hindu-Muslim political cooperation, but it also occurred
simultaneously with a revival of Hindu-Muslim competitive religious pros-
elytization. The second wave of rioting in 1937-38 occurred during a very crit-
ical period in pre-Independence political history, when the Congress came
to power in U.P. and negotiations to share power with the Muslim League
broke down, which many persons consider to have been a turning point on
the road to partition of the country and the creation of Pakistan. The two
years of Congress rule were marked by intense political competition between
the Congress and the Muslim League, including the use of "direct action"
methods that often led to riots, but riots also occurred sometimes during reli-
gious festivals such as Holi.
The simultaneity of religious and political activities at times of rioting has
left room for differing interpretations of their sources: religious or political
differences, spontaneous resentments, or politically inspired actions. The
opposition between these two types of contextualization is, however, spuri-
ous. Religious observances in public spaces were highly politicized in the pre-
Independence years and religious proselytization was politically as well as
religiously motivated.
When we come to 1946, the political aspect of riot production is fullyevi-
dent. It is now a manifestation of the communal conflict over the future of
the subcontinent. Moreover, Aligarh provides at this moment a miniature
crystallization of the subcontinental divide. The AMU is at the center of polit-
ical activity, seen as the originating source of Muslim separatism, many of
whose faculty and students became supporters, members, and leaders of the
Muslim League. Moreover, the very division of the country as well as the prob-
lems that were to follow it were already symbolically present in Aligarh itself.
The city was divided by the railway line between the Civil Lines and the old
city. Forces massed on either side of the line, threatening to cross it, to invade
the other side. Militant nationalist and separatist organizations and para-
military formations were present and ready for confrontations. But so, too,
were complications present, for on the city side of the line there remained a
substantial Muslim population, the vast majority of whom would be left behind
in Aligarh bereft of political leadership and organization, living side by side
with the majority Hindu population, among whom a small militant Hindu
segment was to build political support around the resentments of Hindus at
the partition of the country for which the AMU stood as the symbol.
 76/ Hindu-Muslim Violence in India and Aligarh

From 1946 to 1961

Between 1946 and 1961, when I first visited Aligarh, there were three other
riots, in one of which deaths occurred. The latter riot-much more serious
than its predecessor in 1946-occurred over two days on March 3 and 4,1950.
The origin of this riot, like so many in the pre-Independence period as well,
was attributed to a trivial incident during the Hindu festival of Holi, "when
a Hindu boy threw coloured water on a passer-by," who, though the report
does not say so, was obviously a Muslim. In the communal disturbances that
followed, "five persons were killed and 40 injured" and "there were many cases
of looting and arson."37 It was reported, as in March 1938, that similar inci-
dents followed by communal "disturbances" occurred during the Holi festi-
val days in several other towns in western and central U.P., namely Moradabad,
Pilibhit, Bareilly, and Shahjahanpur.38
Pars Ram remarked somewhat ambiguously in his UNESCO Study ofSocial
Tensions in Aligarh in 1950-51 that "the Hindu students' record in inflaming
the Hindu-Muslim conflict to make it appear more violent" was "impres-
sive." His study also referred to "the stopping of a passenger train," "the way-
laying and the assaulting of stray Muslim pedestrians," "the assaulting of
Muslims in their residential quarters," and "assaulting of a number of Mus-
lim rickshaw pullers by Hindu students."39 Despite the ambiguity concerning
the extent of involvement of Hindu students in the actual violence perpe-
trated in this riot, it is dear that Pars Ram attributed some role in the actual
assaults to Hindu students and that he saw Muslims assaulted at random as
the victims in this riot.
A brief, quickly controlled riot occurred in Aligarh on June 6, 1954. The
Times of India account attributed its origins to "a dispute between a hawker
and customer" that "sparked communal disturbances," what Wilkinson
codes as a "private quarrel."40 No one was killed in this riot, though 8 people
were injured.
The third reported riot in the 1950S occurred on September 14, 1956.41 The
sequence of events leading up to this riot illustrates the process by which exter-
nal and local factors interact to produce communal confrontation and vio-
lence at particular sites. The symbolic pretext for rioting in Aligarh and
elsewhere was provided by the publication by the Bharatiya Vidhya Bha-
van 42 of a book called Living Biographies of Religious Leaders, written by an
American, that contained some references to the Prophet Mohammad that
were considered blasphemous by Muslims. The head of the Bharatiya Vidhya
Bhavan, K. M. Munshi, was a figurehead president only, who happened also
 Hindu-Muslim Violence in India and Aligarh /77

to be the governor of the state of U .P. Demonstrations took place in Aligarh

the last week of August,43 which were discussed in the U.P. Vidhan Sabha
(Legislative Assembly) and reported in the press some days later. The stu-
dents of AMU had taken out "a procession carrying an effigy of the U.P.
Governor, beat the effigy with shoes [a great insult in India] and ultimately
burnt it." In discussion in the Vidhan Sabha, it was alleged that "the proces-
sionists [had] shouted slogans like 'Hindustan Murdabad' [Death to India]
and 'Pakistan Zindabad' [Long Live Pakistan)" and that "Pakistani agents were
also behind the demonstrations. "44 The first allegation appears to have been
generally accepted, the second was never proved. The attention given to these
particular slogans immediately transformed the local incidents into an issue
affecting Hindus and Muslims throughout the subcontinent as well as India-
Pakistan relations. The charge that Pakistani agents were behind the events
has been since Independence to the present day a recurring theme in speeches
by Hindu politicians from both the Congress and the militant Hindu parties
and in news reports whenever Hindu-Muslim confrontations and violence
occur. These allegations were repeated in the press on numerous occasions
during the month and were made again in the Lok Sabha by militant Hindus
within the Congress. 45
No violence was reported from Aligarh in this first week of September,
but significant rioting, including deaths in some places, occurred in several
other towns of U.P. during the month, as well as in Bhopal, Jabalpur, and
N agpur in Madhya Pradesh, and in the East Pakistan capital of Dhaka. 46 The
riots in all these cities and towns appeared to follow a common pattern of
processions and counterdemonstrations: Muslim processions protesting the
Bharatiya Vidhya Bhavan book-including in some places ritual burning of
the Hindu religious text, the Bhagavad Gita, and attacks on idols of the Hindu
god, Ganesh, whose festival was being celebrated in various parts of north
India during this period 47-followed by Hindu counterprocessions or demon-
strations, often ending in arson and riotous violence. One Muslim religious
organization normally allied with the Indian National Congress, the Jamaat-
ul-Ulama, was reported to have "telegraphically urged its district units to take
part in the protest meetings and demonstrations." The Urdu vernacular press
also contributed to the inflammatory situation "by publishing the offending
contents of the impugned book in screaming headlines."48 The English-
language press, however, acting with restraint, never published the passages
from the book that had perturbed Muslims.
Insofar as Aligarh is concerned, the principal riotous incidents occurred
in connection with militant Hindu counterprocessions and demonstrations
 781 Hindu-Muslim Violence in India and Aligarh

on September 14. The processions, which were organized within the various
predominantly Hindu- and RSS-dominated degree colleges of the town, were
"taken out from one college to another," and demands were made for '''stern
al..."tion' against the Aligarh University authorities and students" for having taken
out or allowed their protest processions. When the procession approached
the Moti Masjid (a mosque near Phul Chauraha), "it was reported to have
been attacked with stones and brickbats."49 A few cases of arson were reported
in several localities and there were some injuries as a result of the stone throw-
ing, but no deaths occurred. It is noteworthy that these processions followed
a week after the Muslim student processions, that they were well organized,
and that they set out deliberately through areas known to have been sites of
Hindu-Muslim confrontations in the past. It is obvious, therefore, that the
processions were deliberately provocative, that they were not spontaneous,
and that violence was a probable if not predictable and desired result.
Within the university itself, its retiring vice-chancellor, Zakir Hussain-
a "nationalist Muslim" who had been opposed to the Pakistan demand-
later to become president of India, made a farewell speech in which he
condemned the actions of the AMU students in very strong terms. He went
even to the extent of saying that the AMU deserved criticism for its "past his-
tory," and that students should not react against such criticism.50 However,
a few days later, he issued a statement criticizing press reports on his farewell
speech that suggested he was giving "the impression that the university was
a hot-bed of anti-national feelings and anti-social activities." On the contrary,
he insisted that generally the AMU students were well behaved and that,
although their behavior in connection with their protests against the offending
book and the governor was "reprehensible," there was nothing communal
about itY
The incidents in Aligarh and in other towns where processions, demon-
strations, and riots occurred were framed within a broad national context
that extended beyond either the precipitating issue of a religious insult to the
Muslims ofIndia or the local contexts in which rioting occurred. Pandit Pant,
now home minister of the government of India, referred to "the old tendencies
of the Muslim League" reappearing among Muslims that could only "dis-
turb the communal amity prevailing in the country," balanced by criticisms
of those Hindus who wished to establish a "Hindu Raj" in the country. Pandit
Pant reminded the country that a major reorganization of the internal
boundaries of India was about to take place in the form of the linguistic reor-
ganization of states and remarked that, in this context, every citizen should
remember that he was an Indian first rather than a member of a particular
 Hindu-Muslim Violence in India and Aligarh /79

region, state, or community; otherwise, the Indian state itself might be weak-
ened. A disruption of "communal amity" was said to have the even more grave
implications of harming the "prestige and respect" that India was believed
to hold in the world. Finally, Pandit Pant used this occasion to claim that
India was a better protector of Muslim interests, not only in South Asia, but
in the broader Muslim world, where, at this very time, India had come to the
support of Egypt in the Suez Canal crisis, whilst Pakistan, then an American
ally, had not.52
The president of the militant Hindu organization, the All-India Hindu
Mahasabha, also used this occasion to condemn the "agitation against the
book" published by the Bharatiya Vidhya Bhavan and called for "firm action"
against the "trouble-makers." He described the Muslim protests as a "planned
attack by anti-national elements in the country" and, in a refrain that was to
be repeated and to gather force decades later, called for an end to the gov-
ernment's "policy of appeasement" of Muslims that could threaten even "the
security of India. "53
Several features of this sequence of events and rioting in Aligarh and else-
where are notable. The initial protests and demonstrations were begun by
Muslim organizations, and Muslim university students comprised the main,
if not the only participants in them. The protests were initially against an
offense to Islamic religious faith. Further, the Muslim protests provided an
occasion for responses from Hindu students and militant Hindu organiza-
tions that led to violence. It is essential to understanding the dynamics of riot
production, however, to note that the counterdemonstrations in Aligarh were
separately generated from within the city, that they constituted provocative
behavior in city mohallas with heavy concentrations of Muslims. That is, it
was not a case of a direct confrontation between two protesting groups, but
of separately organized incidents. Further, the Hindu counterdemonstrations
and attacks on Muslims produced symbolic countermoves in the allegation
that Muslims had insulted a Hindu religious text, the Gita, and in the shift-
ing of focus to the AMU as a grand symbol of antinational sentiments. Yet a
further broadening of the context occurred in the statements made by polit-
icalleaders that both Muslim and Hindu communalist groups were threat-
ening the communal harmony of the country and, thereby, the country's future
and its prestige in the world.
But what, then, was behind this wave of demonstrations, counter-
demonstrations, and rioting that broadened out from a few sentences in a
book published by a centrally funded educational foundation that certainly
had no intention of insulting Indian Muslims? There are particular, local
 80 I Hindu-Muslim Violence in India and Aligarh

answers to this question as well as more general ones. We cannot know with-
out detailed knowledge of each case in which rioting occurred why it occurred
in some places and not others. Many cities and towns in U.P. and Madhya
Pradesh were affected, but the demonstrations and riotous violence occurred
at specific sites in each such city. We know that there are two areas from which
politically charged demonstrations in Aligarh are generated: the AMU and
the mohallas in the city where the local degree colleges are located. They stand
in polar opposition to each other, but the actions in this case were separate.
In later riots, we will see militant Hindu activists and rioters moving directly
to the university area to attack Muslims, Muslim property, and Muslim insti-
tutions, but even later it is not always the case. What is constant is the exis-
tence of the AMU as a symbolic presence, providing a fund of Hindu enmity
for political mobilization and violent acting out that takes place at sites of
Muslim concentration, though tlle Muslim population at those sites has noth-
ing to do with the AMU or its history.
Finally, there is the question of the relationship between these riots in 1956
and the general elections that were to follow in five months. This was a time
of Congress dominance in Aligarh, U.P., and nearly all of India. The mili-
tant Hindu parties were not yet a major threat to Congress supremacy. Yet,
any consolidation of either Muslim or Hindu sentiment that could be
directed against the Congress could threaten its ascendance. The general secre-
tary of the Congress drew attention to the possible electoral purpose behind
the demonstrations and rioting in a speech in which he called upon party
workers to combat "communal forces." He remarked, "Both Hindu and Mus-
lim communal parties and organisations are trying to fan the flames and create
riots mainly for sordid political motives in connection with the forthcoming
general elections. "54 The Congress at this time was presenting itself as a sec-
ular force maintaining a fragile peace between dangerous Hindu and Muslim
communal forces, whose actions allegedly threatened the unity of the coun-
try, but more certainly threatened Congress preeminence. Insofar as Aligarh
is concerned, Congress dominance remained undisturbed in the 1957 elec-
tions. It was to take the far more serious and deadly riots of October 1961 to
overthrow it.

The Riots of October 1961

The riots of October 1961 began with a scuffle among Hindu and Muslim stu-
dents at the AMU in the aftermath of Students Union elections, which had
been fought wholly on communallines}5 Rumors spread to the town that a
 Hindu-Muslim Violence in India and A.ligarh /81

Hindu student had been killed in the university. Large mobs from the town,
mostly students from the Hindu colleges, massed, burned some shops in the
University market areas, and attacked some university employees. Rioting
then spread to the town. In the notorious Manik Chauk mohalla-which will
feature again and again in our accounts-a Hindu businessman used a mob
of students to raid the houses of a Muslim with whom he had a property dis-
pute, as well as other Muslim houses. The police were present, but stood by
without interfering, even though this incident alone led to the deaths of 7
Muslims and the injury of 21. Altogether, by official reckoning, 15 Muslims
were killed, but the true number is believed to be around 40; no Hindus were
killed. Police allegedly colluded with the Hindu students, deliberately denied
protection to Muslims, and, it is confirmed, entered the AMU hostels, beat
up Muslim students, and looted their belongings. Curfew was maintained
in the town for two weeks. Muslim conservatives on the faculty of the uni-
versity, where there was an ongoing struggle for primacy between so-called
communalists and so-called Communists, entered the 1962 general elections
thereafter, joined forces with the opposition to the ruling Congress, and helped
administer a major defeat to the Congress in the district as a whole in a com-
munalized election atmosphere. There is, therefore, a political cum commu-
nal rioting cum political sequence, from university politics to rival student
communal confrontations to rioting in the city to the general election.
Reports on this riot by the press, and the commentary upon it by others
afterwards that was reported in the press, are of great interest for what they
do and do not reveal about the riot itself, the kinds of explanations that were
generated about it, and the ways in which the riots were used in the process
of blame displacement. With regard to the reporting in the Times of India,
emphasis was placed on the conflict among the students on the AMU cam-
pus as the precipitant of the riot, from which the action in the city was seen
almost as part of an inevitable and understandable sequence. Yet, there were
two kinds of reactions from the city. One was the mobilization, massing, and
movement of a crowd said to be "about 8,000 including some students," who
"came from the city side and made an attempt to cross the railway line and
rush towards Aligarh university," but were stopped by tlle police.56 Here we
have in this major Aligarh riot-the most serious in its history to this time-
a situation exactly the reverse of that for which the U.P. government pre-
pared in 1946-47, namely, the attempted crossing of the railway line by Hindu
students seeking to reach and presumably to attack the AMU, rather than the
reverse. The second response to the incidents on the AMU campus was the
action within the city itself, notably in Manik Chauk, mentioned above. Yet,
 82/ Hindu-Muslim Violence in India and Aligarh

the names of none of the localities in which rioting occurred in the city, includ-
ing Manik Chauk, were given in the press.
The meager factual reporting in the Times of India on the riot in Aligarh
was confined to a single day, datelined October 4. However, commentary on
the riot by the paper's editor and reports of the statements of politicians on
it continued in the press almost daily for the rest of the month. In the com-
mentaries, there were two frames: one placed it in the context of Hindu-
Muslim relations, the unity of the country, and the place of the AMU in
relation to both; the second placed it in the context of interparty conflicts.
Within both these contexts, blame was partly focused, partly dispersed, but
largely failed to identify the principal sources of riotous activity in the city.
Blame was to a considerable extent focused on the AMU and its author-
ities for failing "to postpone elections to the Students Union even after it
became clear that the contestants were canvassing openly on communal
lines. "57 To a considerable extent, the blame directed at the AMU also focused
on the institution itself, its past and present. A Hindu journalist, Prem Bhatia,
wrote an analytical article in the Times of India which began by stating that
"Aligarh [referring to the riots] has posed the first serious challenge to the
efforts for national integration." However, he averred, the riots that began at
AMU and spread to other parts of the state, while they certainly reflected seri-
ous problems at the university, extended more broadly than that, reflecting
as well a "disease which afflicts a large number of Muslims in this country. ";8
Atal Bihar Vajpayee, then secretary of the Jan Sangh and leader of its parlia-
mentary group, urged the government of India to appoint a committee "to
investigate the state of affairs in Aligarh University" and to shed its belief "That
the University can be allowed to retain its 'communal character' without jeop-
ardising national interests and the principles of secularism."59
With regard to the second context, of interparty conflicts, the state's home
minister, Charan Singh, "attributed the communal disturbances in the west-
ern districts of the State, to an 'organized' attempt by some political parties,"
whose intentions were deliberately to "disrupt communal harmony" for polit-
ical advantage just before the general elections. His remarks were echoed by
the chief minister of the state as well, Mr. C.B. Gupta. 60 Two days later, Charan
Singh was more explicit in his charges, saying that the riots in Aligarh and
those that followed it elsewhere were a deliberate attempt "to discredit the
Congress among the Hindu and Muslim masses with a view to capturing polit-
ical power."61
Blame was also partly dispersed as commentary upon it extended up to
the highest levels of leadership in the country, including the president and
 Hindu-Muslim Violence in India and Aligarh /83

prime minister. The president, Dr. Radhakrishnan, said that the events in
Aligarh occurred just at the time when "a conference on national integration
had been held" and that they "showed 'how distant the goal is.' "62 Prime
Minister Nehru adopted his usual schoolmaster's stance, maintaining an
equable balance in his remarks, in which he was reported to have described
"the communal disturbances in the State" as "unfortunate, but a happy sign
was that the masses did not take any interest in such ugly happenings." He
was reported to have strongly denounced the "communal incidents" and
remarked that "it was most painful that students instead of pursuing their
studies participated in disturbances." He placed no blame on Hindus or
Muslims as such; on the contrary, he "said that Hindus and Muslims were
sons of the same soil and there should be no hatred."63
It is noteworthy in all the commentary on these riots that the vast major-
ity of statements took the riots out of their local context of a tussle between
undisciplined students from the two communities at a university, almost com-
pletely ignored the very considerable militant Hindu mobilization that fol-
lowed in the city itself and that was followed later in other towns in many
districts in the state, and instead placed it within the framework of Hindu-
Muslim relations in general, the unity of the country, and the loyalty to India
of Muslims in particular. Further, in the face of the wide dispersion of charges
in which virtually the only factual materials provided referred to the student
disturbances, the authorities at the state and national levels decided against
appointing an inquiry commission to fix responsibility for the criminal acts
committed during these riots. 64 In so avoiding an impartial judicial inquiry,
charges made by three ministers of the central government, which were utterly
lost in the extralocal contextualization of the violence, could not be investi-
gated. One of the ministers, after a visit to Aligarh on October 6, "expressed
the opinion that 'most of the deaths and injuries' caused in the incidents had
'no direct connection with the communal trouble.''' On the contrary, he said,
"advantage had been taken by certain elements of the 'tense situation in the
town,' following trouble in the university, 'for wreaking private revenge' and
in the course of these incidents eight persons had lost their lives and a large
number received injuries." In contrast, the incidents in the university area
led to "only three" deaths. 65
We have here, in short, a clear example of the distorting effects of post hoc
commentary on communal collective violence. The leading English-language
newspaper of the country devoted many of its pages for an entire month to
commentary by its own editors and journalists as well as the political lead-
ers whom it quoted extensively, without providing more than a fragment of
 84/ Hindu-Muslim Violence in India and Aligarh

reporting on the incidents themselves. Further, it raised no serious questions

concerning how such extensive mass mobilizations, most of them emanat-
ing from Hindu students at colleges and universities in several districts of the
state, could have been organized so quickly. Finally, it ignored completely
the main scene of riotous activities perpetrated in the center of Aligarh City
itself. Instead, virtually all commentary followed an agenda, a text that
emanated from the minds of the commentators and from a discourse of
national unity that emphasized the threat to it posed by communal conflicts
between Hindus and Muslims, which may not have yet infected the entire
populations of the two communities, but had done so before and might yet
do so once again.

1971 Riots

An even more deadly riot occurred in the midst of the polling for parlia-
mentary elections on March 1, 1971. 66 The election in A1igarh was of great
importance at this time because the years immediately preceding it had been
marked by an upsurge in Muslim political activity centered on a list of griev-
ances that included opposition to the central government's efforts to alter the
character of the AMU and, as it was seen by Muslim political organizations,
to undermine its character as a minority institution. 67
The importance of the Aligarh parliamentary seat to Mrs. Gandhi's
Congress, then known as Congress (R), was reflected in the choice of a can-
didate from this constituency, a Muslim member of the central government,
Deputy Minister for Railways Mohammad Yunus Saleem. His opponent, Shiv
Kumar Shastri, contested on the ticket of the BKD. Shastri was a militant
Hindu supported by, though not a member of, the Jan Sangh/RSS, which was
a component of a four-party alliance against the Congress (R). The election
campaign had been marked by agitations at the AMU in favor of restoring
its minority character and counterdemonstrations by students at the D.S.
College organized by the RSS.
Violence broke out on March 2 that lasted for four hours, during which
Congress (R) candidate Saleem alleged that "shops of Congress (R) supporters
had been 'looted' and set on fire by an unruly mob of RSS workers." He also
said that the Jan Sangh president, Shiv Hari Singhal, had been arrested in con-
nection with these incidents. Saleem's opposing candidate, Shiv Kumar
Shastri, gave a different explanation for the incidents, namely, that they fol-
lowed upon a protest by student demonstrators against an attempt by a young
 Hindu-Muslim Violence in India and Aligarh /85

supporter of his rival candidate, clad in a woman's burqa, to cast an illegal

vote, which precipitated "an altercation between two groupS."68 Whatever
the actual precipitant, considerable violence followed in several riot-prone
localities of the old city, during which 17 persons were said to have been killed. 69
The district magistrate and the senior superintendent of police were suspended
for failing to prevent and control the violencer the chief minister of the state
visited the town along with a former chief minister (both part of the four-
party alliance), and an inquiry commission was appointed.
There was complete disagreement that has never been resolved among
observers concerning every aspect of this riot. The Congress (R) candidate,
Saleem, insisted that the incidents were political, not communal, whereas his
opponent, Shastri, said they appeared "to be taking a communal colour."71
The former chief minister, C. B. Gupta, who visited the scene after the riots,
went further and declared that the riots were preplanned and that it was "the
majority community," that is, Hindus, who had suffered from most of the vio-
lence that occurred. He also implied that the Congress (R) candidate himself
was responsible for the violence,72 The state government appointed Justice
Mathur as a one-man commission of inquiry; his report conformed to the
views of the government that appointed him, blaming the riots on the AMU
and relieving the RSS and the Jan Sangh and several RSS/Jan Sangh leaders
who had been accused of fomenting the riots of any responsibility for them,
despite the report that students from the D.S. College and tile Hiralal Barahseni
Degree College had massed, taken out processions, and engaged in arson,73
The following features of this riot are especially relevant to the analyses
that will be presented in the later chapters of this volume. First, complete dis-
agreement on the origins and character of the incidents, not resolved even
by the findings of an inquiry commission. Second, the direct relationship
between the violent incidents and an election campaign in progress. Third,
the existence of sharp political divisions not only in the local election con-
test, but between the central and state governments, which were rdlected in
their opposed explanations of the causes of the riots. Fourth, the salience at
the time of the election of issues concerning the grievances of Muslims in
Indian society and politics, including the status of the Al\lfU. Fifth, the occur-
rence of agitations at the AMU on these issues followed by the mobilization
of students from the local Hindu- and RSS-dominated degree colleges pre-
ceding the outbreak of violence. Finally, the identification and arrest by the
local administration of a prominent leader of the Jan Sangh, Shiv Hari Singhal,
for contributing to the outbreak of violence.
 86/ Hindu-Muslim Violence in India and Aligarh

1972

During the general elections of 1962 in Aligarh, held in the aftermath of the
October 1961 riots, leaders of the so-called communalist group in AMU pol-
itics joined with opposition political party leaders to administer a defeat to
the Congress in the city from which it never recovered. Ten years later, the
internal politics of the university once again spilled outside into the broader
political arena, but this time embracing the entire country and merging its
ideological differences with the national issues concerning the identity of the
Indian nation, the place of secularism and Hindu values within that identity,
and the place of Muslims in Indian society as a whole. At the center of this
merged local/national debate was the former leader of the Communist/pro-
gressive group in AMU politics, Nurul Hasan, past head of the history
department, a lifelong member of the Communist Party of India (CPI), at
this time minister of education in the government of India led by Prime
Minister Indira Gandhi.
N urul Hasan undertook to restructure the internal governance of the AMU
and to reorient its mission as well. The latter was to be redefined from "devot-
ing special attention to the promotion of oriental and Islamic studies and the
teaching of Muslim theology" to promoting "the study of the religions, civi-
lisation and culture of India. "74 To ensure this reorientation of mission, the
structure of the AMU's internal governance was also to be changed and it
was to come more closely under the direct supervision of the central gov-
ernment, which provided most of its 11nances.75 The vice-chancellor was now
to be appointed, in effect, by the central government and was to be given
enhanced powers.
The Aligarh Muslim University (Amendment) Bill was moved in the Lok
Sabha on May 31, 1972, and passed after a day-long debate the next day. It was
then quickly moved in the Rajya Sabha as well, where it passed the next day
on a voice vote.76 Three sharply opposed points of view were expressed in the
debates on this long day, each reflecting one of the three main streams of Indian
nationalist ideology. Members of the Muslim League 77 and some other Mus-
lim MPs protested on the grounds that the minority character of the AMU
would be destroyed by this legislation. They argued, in effect, that the Muslims
of India were a minority with a distinct religion, culture, and civilization, that
the AMU had been founded by the Muslim community, and that its separate
existence as a minority-run institution ought to be maintained.
Minister of Education Nurul Hasan responded by saying that "the uni-
versity should have [aJ national and not minority and theocratic character
 Hindu-Muslim Violence in India and Aligarh 187

and in the name of tradition progress should not be hampered. "78 Indeed, he
averred that "the Bill would help the university to shed the 'dirt and filth of
obscurantism.' "79 In effect, the minister was presenting the secular progres-
sive view of Indian nationalism and the proper place of Muslims within it.
The third position, of militant Hinduism, was articulated by both the Jan
Sangh and by Shiv Kumar Shastri, the candidate elected to Parliament from
Aligarh in 1971 by the anti-Congress (R) coalition, which had included the
Jan Sangh. The Jan Sangh members took the occasion of the debate to attack
the AMU, in one case criticizing the bill as inadequate for its failure to order
an enquiry into "the snake pits and snakes"8o in the university. Their posi-
tion clearly was that the AMU was a source of anti-Indian sentiment that
needed to be utterly extinguished before the Muslims of India could be inte-
grated into the nation. Both the Jan Sangh members and Shiv Kumar Shastri
also criticized the bill for its failure to affiliate the two Hindu- and RSS-
dominated local degree colleges to the university, a measure which would more
surely finish any pretension on the part of AMU to being a center for Islamic
education and a minority institution. It would also have satisfied the long-
standing demand of the administrators, staff, and students of these degree
colleges to have access to the superior facilities, salaries, and emoluments avail-
able to the AMU faculty and students.
In response to the enactment of the AMU legislation, demonstrations were
staged in Aligarh by persons from among both communities opposed to the
bill for the reasons articulated by their representatives in Parliament. These
demonstrations were not confined to the AMU campus, but involved pro-
cessions in the communally sensitive Upar Kot area (see Map 2) of the old
city by local Muslim leaders on one side and militant Hindu leaders and orga-
nizations on the other side. The local administration acted promptly to arrest
some 63 processionists. 8• On the next day, a fortuitous incident occurred in
which a child was hit by a scooter driver in Sabzi Mandi (the wholesale veg-
etable market )82 just as a procession of Muslim activists reached the spot Some
stone throwing, looting, and arson followed. The administration again acted
promptly, dispersed the crowd by firing in tlle air, and arrested an additional
20 persons. 83 It also imposed curfew in all affected areas or areas likely to be
affected, in some cases for 65 hours at a stretch, and arrested other persons,
bringing the reported total to 334.84
On June 9, it was reported that Aligarh was "quiet" and no further inci-
dents had been reported. 85 No deaths were reported by the press in this riot. 86
A follow-up article on this riot in the Times of India raised another, local
explanation for the agitations that preceded it. It was suggested that they were
 88/ Hindu-Muslim Violence in India and Aligarh

as much concerned with local as with national politics, specifically the like-
lihood that municipal elections, which had not been held in the town for many
years, would soon be scheduled in Aligarh. In this interpretation, local forces,
particularly the Muslim MajIis and the Jan Sangh, were staking out positions
in anticipation of those elections to win support, respectively, from Muslims
and Hindus. 87
The Aligarh demonstrations against the AMU Act were part of a broader
movement throughout the state led by the Muslim Majlis-e-Mushawarat and
several other Muslim organizations, including the Aligarh Old Boys' [Alumni]
Association. In two other U .P. towns-Firozabad and Varanasi -the demon-
strations were followed by extensive rioting, destruction, and deaths. 88 So, in
comparison, on this occasion, rioting in Aligarh was far less severe than in
other towns in the state and, apparently, much more effectively controlled.
Nevertheless, the focus of the demonstrations and counterdemonstra-
tions as well as the reporting on these riots was on the AMU and the Act
that had changed its governance. Further, the press and the Congress (R)-
the ruling party at the Center and in the state capital in Lucknow-attacked
the Muslim organizations in ways that implied that they were antinational.
One report in the Times of India devoted several paragraphs to blaming the
partition ofIndia upon the AMU and its graduates, claiming that Mohammad
Ali Jinnah, the leader of the Pakistan movement, had also been "closely asso-
ciated with this university."s9 In an editorial, the Times of India character-
ized the leaders of the AMU agitation as "obscurantist" Muslims who were
"playing with fire" since their actions would only lead to provoking "a sim-
ilar ganging up by Hindu extremist bodies."9 0
Feeding militant Hindu concerns itself, the Times of India presented as
its front page lead article the very next day a report on the results of the decen-
nial Indian census by religion, showing that the minorities in India, includ-
ing the Muslims, were proliferating at a faster rate than the Hindus, thus
vindicating "the secular atmosphere in the country in which all communi-
ties can thrive." Far from demonstrating anything of the sort, such figures
only provide further impetus to militant Hindu fears that, having vivisected
India in 1947, the Muslims along with otl1er minorities would one day take
over India as well.
It would seem, therefore, that this riot belongs, in effect, to a series of inci-
dents that began in 1966, of which this and the previous riot in March 1971
are the most prominent. Both are notable for the intermixing of local and
national politics and issues, including issues concerning the very identity of
 Hindu-Muslim Violence in India and Aligarh /89

the Indian nation-state. Both involved also an intermingling of university,

local, and national politics. Both may have involved also some relationship
to electoral politics, evident in the March 1971 incidents, suggested in the 1972
incidents.
There are similarities also that pertain to the problem of explanation, par-
ticularly concerning the issues of precipitants (a scooter accident or the pro-
cessions) and of containment, that is, the effectiveness of administrative action.
With regard to the latter, the administration appears to have acted promptly
and effectively and to have learned from its failure to control the previous
riot, but we cannot be sure. It is quite possible that the protagonists them-
selves were less interested in provoking a full-scale riot on this occasion, though
there can be no doubt that some kind of confrontation was desired of the
type that so frequently does result in murderous rioting; it is obvious that
there was extensive organization of pre-riot activities. But was the riotous
action that ensued after the scooter accident-even if it was not a result of
it-spontaneous or deliberate, motivated by communal "frenzy" or by otl1er,
unknown motives? No matter how detailed the information available on such
events as even this relatively small riot, the realm of the unknown that is open
to competing interpretations remains always extensive. Those interpretations,
as we have just seen, extend from the trivial-a scooter accident-to the delib-
erate designs of communalist leaders, particularly "obscurantist" Muslim lead-
ers, to undermine the unity of the country.

October-November-December 1978

This series of riots 91 began with a scuffle between the supporters of two
wrestling akharas (one Hindu, the other Muslim) in Aligarh town. Attacks,
including stabbing incidents between the two groups, then occurred over suc-
ceeding weeks. 9z One of the Hindu wrestlers, called Bhura (or Bhure Lal),
who had been assaulted and wounded in a fight, finally died in the Civil
Hospital on October 5, after which a large crowd carried his body through
the Muslim localities demanding revenge. Rioting broke out "independently
but simultaneously" both in Chauraha Abdul Karim (an important four-way
crossing) when the procession arrived at that crossing and in the nearby
Hindu-majority mohalla of Manik Chauk,93 both in the old city (see Map 2).
Altercations occurred between processionists and shopkeepers when the for-
mer demanded that the latter close their shops. It was reported that 12 per-
sons were killed in this first phase of rioting. 94
 90 I Hindu-Muslim Violence in India and Aligarh

The wrestling akhara to which Bhure Lal belonged is located some dis-
tance from the city limits on the Khair road. It is situated within the precincts
of a Shiva temple. (See Figures 3.2, 3.3, and 3.4.) According to Alter, this asso-
ciation between Hindu religiosity and gymnastics is embedded deeply in the
minds of the wrestlers.95 However, it is also the case that wrestlers in Aligarh
and in places such as Hyderabad are often involved in criminal activities,
including Hindu-Muslim riots. 96
These riots occurred at a time of intense political division within the Janata
governments in both New Delhi and Lucknow, which had replaced the Con-
gress Emergency regime in the 1977 General Elections. The situation in AIi-
garh was immediately politicized at multiple levels by direct and contradictory
comments from all political sides, that is, from the Congress (1), now in oppo-
sition, as well as from the disparate elements that had formed the Janata Party,
which also included the former Jan Sangh. As a consequence, a series of indi-
vidual and collective teams of state and national politicians and religious lead-
ers descended on the city while the riots were still in progress and made partisan
pronouncements concerning them. 97
On October 10, a three-person team deputed by the national president of
the Janata Party, Mr. Chandra Shekar, visited Aligarh. 98 The team called for
a judicial inquiry into the riots, but the contents of its own report were dis-
puted. According to Chandra Shekar, it absolved the RSS of all blame for insti-
gating the communal riots. When asked specifically about the Aligarh leader,
Krishna Kuman Navman, he acknowledged that allegations had been made
against him, but claimed that "nobody had said that he had engineered the
riots" and that the team deputed had "also found no evidence of his pres-
ence at any of the troubled spots."99 Chandra Shekar's remarks were criti-
cized the next day by the presidents of the AMU Staff Association and the
Students Union, respectively, noting that they contradicted the views of most
"prominent leaders of different parties" that the RSS had, in fact, "engineered
the Aligarh riots." Members of the Congress (1) in Parliament, including its
Congress Socialist Forum group, also declared their belief that the RSS was
behind the Aligarh riots.loo
Two RSS ministers in the U.P. government, including Kalyan Singh, also
prepared a report for the state chief minister in which, predictably, they absolved
the RSS of any blame. Further, they specifically absolved from blame Krishna
Kumar Navman, despite the admitted fact that he had "an altercation with a
policeman" on October 7 while "touring the affected areas with a valid pass."
All blame was placed by state RSS leaders on the Congress (I) and the CPI for
snatching the body of Bhure Lal and taking out the provocative procession. IOI
 Hindu-Muslim Violence in India and Aligarh /91

FIG. 3.2. Shiv temple at akhara site

It needs to be noted, however, that, as on most occasions of rioting in the old

city, Mr. Krishna Kumar Navrnan's name is prominent and that it is dear that
he and other RSS workers somehow manage to roam about freely in affected
areas during riots, including, in the case of Navrnan, during curfew hours.
Several politicians blamed police inaction for the events leading up to the
riots, particularly the snatching of Bhure Lal's body.102 The state president of
the Congress (I) blamed the state government for failing "to maintain law
and order" and demanded the "dismissal" of the millistry.l03 There was also
a noticeable tendency on the part of politicians and the press to disperse blame
to no one in particular. One of the members of the Janata team, for exam-
ple, bemoaned the fact that" even after 31 years of independence we have not
been able to integrate ourselves and cultivate communal harmony." 104
A report prepared on behalf of the Peoples Union for Civil Liberties
(PUCL), however, was quite specific in assigning responsibility to particular
persons, groups, and state agencies. It commented on the outbreak of the
 92/ Hindu-Muslim Violence in India and Aligarh

FIG. 3.3. Wrestlers

riots on October 5 as follows: "Hindu communalists ran berserk against poor

Muslims. The Provincial Armed Constabulary of Uttar Pradesh echoed the
same Hindu communal sentiments by shooting down and killing innocent
and unarmed Muslims."lQ5 As for the slain wrestler, Bhure Lal, the PUeL
report remarked that "everyone with whom we talked agreed that he was a
notorious anti-social element." Further, both sides involved in the fight that
led to Bhure's death were, they were told, "comprised of mere gangsters and
rough-necks drawn from both Hindus and Muslims." Moreover, the pueL
report emphasized that the wrestler's death was misreported in the press as
"a sequel to the fight over a wrestling competition on September 12," but was
in fali nothing but the "result of [an] inter-gang fight" and had nothing what-
ever to do with" communal feeling. "106 In fact, the stabbing of Bhure Lal had
been preceded by weeks of intergang fighting in which several wrestlers had
been attacked and stabbed, including one other person, also a Hindu, who
was killed on September 18.107
The PUeL report argued that the actual buildup to the riot of October 5
was an agitation that had been led for several months previous to the killing
of Bhure Lal by known Hindu communal persons on the issue of the AMU
Bill concerning the status of the university, which was once again before
 Hindu-Muslim Violence in India and Aligarh 193

FIG. 3.4. Wrestling pit

Parliament at the time. The killing of Bhure Lal was, in effect, nothing but a
pretext on the part of these Hindu communal persons in the town to insti-
gate a riot for political reasons. The Hindu leaders mentioned were persons
whose names will recur throughout this book: B. D. Gupta, lecturer in the
psychology department of AMU; Manga Ram, lecturer in the Barahseni
Degree College in the town; Shiv Hari Singhal, advocate and former president
of the Jan Sangh; and Krishna Kumar Navrnan, businessman. Several if not
all of these persons were also involved, according to the PUCL report, in the
incident that precipitated the riot, in which Navrnan allegedly played a lead-
ing role, namely, the "snatching" of the dead body of Bhure Lal, which was
then taken out in a procession through the predominantly Muslim areas of
the old city accompanied by shouts from the processionists demanding "blood
for blood" and "ten for one," that is, ten Muslims for one Hindu killed. lOs
Allegations were made that rich Varshney Hindu businessmen provided
money for the criminal attacks on Muslims, that the RSS was behind it, and
that Navrnan played a very important role. The PUCL report alleged that "RSS
elements" in tlle Manik Chauk mohalla where Navman resided also used the
riot in an attempt to force Muslims occupying ten or fifteen houses in the
mohalla "to leave the houses or face serious consequences. ''109 The PUCL team
visited some houses in Manik Chauk and interviewed persons in the locality
 94/ Hindu-Muslim Violence in India and Aligarh

on the streets and in their homes. Witnesses provided them with names of
persons who had murdered others in the riots, including one of Krishna
Kumar Navrnan's sons, Satya, who was said to have been "seen with a gun
supervising the murder" of two persons by others, who used a long spear to
impale their Muslim victims. These and several other attacks upon Muslim
houses in Manik Chauk were carried out during curfew hours between 5:30
P.M. and midnight, a short distance away from the Madar Gate police
station.Ho In addition to Navrnan's son, Professor B. D. Gupta was also named
"among the prominent personalities who led the attack" on Muslim houses
in Manik Chauk. m The actual work of killing, burning, and looting, however,
was said to have been carried out by "a handful of notorious characters-
known hoodlums and criminals," none of whom were arrested before or after
"the carnage of October 5."112
Political divisions at the national, state, and local levels also clearly played
their part in the production of Hindu-Muslim-police violence in this riot.
As noted above, the 1978 riots occurred at a time when the former Jan Sangh
members had been incorporated within the Janata Party that came to power
in the central government and in U.P. after the landmark 1977 General
Elections. H) In Aligarh, the former Jan Sangh members constituted the core
of the Janata Party, whose candidate, a Muslim, won the Legislative Assembly
seat.U4 Divisions within the Janata Party between the former Jan Sangh/RSS
members and the other principal northern Indian party, the Bharatiya Lok
Dal, existed at all levels. It was reported that internal divisions in the state
cabinet were likely to produce a "cabinet crisis" and that "at least two cabi-
net ministers [had] threatened to resign, while a minister of state [had] openly
blamed a Janata Party constituent [that is, the Jan Sangh) for fomenting com-
munal trouble.""5 Consequently, neither the state government nor the cen-
tral government were prepared to take decisive action to prevent rioting in
Aligarh or to control it after it had broken out,u6
These divisions in turn made it difficult for the district administration to
act decisively to head off tlle riots and to control them after tlley had broken
out. Their efforts to "arrest all the wrestlers of the town" on October 4 and
their actual arrest of "five Hindu [s Jand two Muslims the same evening" were
vigorously opposed by Navrnan, whose supporters surrounded the central
police station that night "demanding the release" of a friend of Bhure Lal's,
"who had been arrested earlier." Navrnan himself was provided with police
protection by two Provincial Armed Constabulary (PAC) jawans (police
officers) posted at his house.1l7
AMU students did not get involved in these riots, but did help with relief
 Hindu-Muslim Violence in India and Aligarh 195

activities. lIS However, the university was closed down on the grounds that
the police could not ensure the safety of students attending classes.
The 1978 riots in Aligarh constitute a benchmark in the development of
Aligarh's institutionalized system of riot production, containing all the major
elements that will recur in many riots thereafter. These elements include
intense inner party and interparty division. Not only was the Janata Party
divided in Aligarh City at this time, but there was fierce interparty competi-
tion as well between that party and the Congress. These divisions in turn made
forceful administrative and police action impossible. Not only that, the PAC
revealed itself as a harsh anti-Muslim force. These prevailing conditions in
turn made possible the staging of a riot for which the killing of Bhure Lal
provided the pretext, though the buildup to the riot had been a several-
months-long political campaign over the continuing issue of the status of the
AMU. In the actual riots themselves, we witness the evident involvement also
of several types of persons performing different roles: known political, busi-
ness, professional, and university/college persons playing leadership roles in
the organization of actions designed to provoke a confrontation between
Hindus and Muslims, and known criminal and hoodlum elements recruited
for the commission of the actual acts of violence. Little information is pro-
vided about the composition of the crowds, though it is likely that, as in the
past, students were recruited from the local Hindu- and RSS-dominated degree
colleges. There is also a specific form for riot-provoking activities, namely,
the procession, in this case of the dead body of a Hindu wrestler, deliberately
carried through Muslim-populated and communally sensitive areas with the
accompaniment of inflammatory slogans designed to provoke Muslims to
respond with brickbats, to justify the slaughter that is to take place thereaft.er.
Finally, under the cover of the communal conflict and the political issue of
the status of the AMU, local economic factors are also revealed in the efforts
to displace Muslims from their homes in a Hindu-dominated locality in order
to gain possession of their property.
Political parties and partisan newspapers continued to make political cap-
ital out of tlle Aligarh riots into November. Opposition parties accused the
government of inept handling of the situation1l9 and the CPI newspaper, the
Patriot, continued with the theme that "the organised killing of Muslims at
Aligarh was due to the 'conspiracy of RSS and police and PAC."'120 While
charges, countercharges, and accusations were flying, further large-scale riot-
ing broke out in the city on November 6. 121 By November 9, 1978, the Times
of India reported that another 15 persons had been killed. i l l
The worst-hit localities were Sarai Suitani, Sarai Kaba, Turkman Gate, Jai-
 96/ Hindu-Muslim Violence in India and Aligarh

ganj, and Babri Mandi (see Map 2). 123 It was also reported that most of the
people killed in this new phase of rioting were shot by the police. Further vis-
its by government and political party fact-finding teams and by the Shahi Imam
of the Delhi Jama Masjid followed and the same charges and countercharges
were made concerning party, police, administrative, and RSS culpability.124
Specific names of RSS persons from the Hindu community were again men-
tioned in the press as bearing direct culpability for encouraging the rioting,
namely, Professors B. D. Gupta and Manga Ram.l2S By November 15, the official
death and injury tolls had increased to 16 and 54, respectively.l26 On November
20, it was reported that a warrant of arrest had been issued against NaVlllan
for involvement in the riots, but that he had absconded. I27 In the meantime,
the district magistrate and the senior superintendent of police were replaced. u.8
By November 26, it was reported that the official death toll from the begin-
ning of the first phase of the riots on October 5 had risen to 28.129
Curfew continued in the town throughout the month of December, ter-
minating on the last day of the month. Under the strict control of the new
district civilian and police administration, no further deaths occurred.
However, conditions in the old city remained disturbed and there were fur-
ther outbreaks of violence during the month, including bomb explosions and
stabbings.13o The administration not only retained curfew throughout the
month, but engaged in continuous house searches for arms, arrested "more
than 1,000 persons," and engaged local persons in the mohallas to participate
in peace committees.I31
By the time the curfew was lifted, large parts of the city of Aligarh, par-
ticularly the old center, had been under varying curfew hours for 88 days con-
tinuously, that is, since its imposition after the second phase of rioting on
October 6. The prolonged character of the disturbances in Aligarh, from the
beginning of the first incident on September 12, that is, the brawl between
the two wrestling groups, and the length of time the city was under curfew
in both the first and second phases of rioting, suggest that the term riot is a
misnomer for what occurred during these months. These events were of a
different order from the riots that occurred before in Aligarh, which, like most
events classed as riots, lasted only a few days. But it is not at all clear how to
label it, what specific events should be included as part of the communal
attacks involving Hindus and Muslims, and what should be considered mere
crimes. We cannot label it urban civil war, for that would minimize the activ-
ities of known criminal elements for whom the disturbances provided an occa-
sion to loot and plunder. It was certainly not an insurrection, since the
violence was not directed against authority except insofar as the police inter-
 Hindu-Muslim Violence in India and Aligarh /97

vened between Hindu and Muslim groups or attacked one side or the other.
All that can be said with a measure of accuracy is that large parts of the city
underwent disorder, anarchy, violence, and that state of potential war of all
against all in which no person could feel safe from harm and sudden death
and no one's property was secure, a state in which most people suffered while
some also gained. The gains for some were loot, harming of one's rivals, steal-
ing of the property of the weak and helpless, displacing of unwanted ele-
ments from one's neighborhood, and not least of all the political advantages
gained by political persons in the city, state, and country through stentorian
statements exclaiming sympathy for the victims while blaming others for the
violence.
Indeed, the process of blame displacement continued throughout these
long months of violence in Aligarh, in the press and in a full-scale debate in
the Lok Sabha on December 4-5, 1978.132 The string of outside visitors to
Aligarh also continued.
The Minorities Commission heard an interpretation of the riots as aris-
ing out of economic rivalries. One view was that rivalries between Hindu and
Muslim businessmen in the lock industry were "behind the tension." A sec-
ond was that Hindu businessmen who coveted Muslim property contrived
riots for the purpose. Similar reports appeared also in the press. One such
report argued that, in this riot, as in the previous riots of 1961 and 1971, "ris-
ing property values in some poorer sections of the town could be a motive
behind the destruction of the houses in the present riots as in previous ones....
In both these riots, separated by a decade, miserable hovels of poor people
were obliterated and in their place has come up a prosperous commercial
centre. "133 It was also reported that Muslim owners of houses that had not
been destroyed were "receiving ... offers by potential buyers" at prices
"lower than ever before."I34 Other reports referred to "business rivalries
between Muslims and upper caste Hindus" as having "played their role in
engendering feelings of hatred."135
Although the sequence of events in connection with the incidents involv-
ing the wrestling competition is generally accepted, it is not universally
accepted that the wrestling competition was itself tlle "cause" of the riot. On
the contrary, at least four interpretations have been put forth to explain the
transformation of the scuffle, the stabbings, and the funeral procession of
Bhure Lal into a major riot in which the passions aroused by the death of the
Hindu wrestler are considered mere pretexts. Those interpretations were
laid out for me in an interview with one respondent, a Muslim, who was presi-
dent of the Aligarh Muslim University Students Union, a supporter of the
 981 Hindu-Muslim Violence in India and Aligarh

Janata Party, and a person actively involved in efforts to restore peace in the
city at the time.

PRB: So, the '78 riots ...

JH: Yes, actually, it was, uh, the theory was that, uh, some local persons, they want to
purchase a particular piece of land, or a particular market. ... Therefore, there
were two theories. One theory was that it was an unplanned riot started just after
a wrestling competition. That was one version.
PRB: That fellow Bhure ...

JH: Yes, Bhure Lal, Bhure Lal and maybe any.... They said that it was just a sponta-
neous thing, but it had become a communal riot and because the mind of the people
is polluted always, therefore, it-and due to the lack of some policemen who were
on duty that, uh, the dead body of the pahalwan, that Bhure Lalor ... pahalwan,
it was allowed to go in a procession like that. That was one version. The second
version was that ... Mr. Navrnan was behind the riot, and that he planned a riot
to emerge as the leader of the Hindus, and that was the second. The third and very
famous theory was that ... some persons want to purchase a particular area of land
for their commercial and economic purposes. Therefore, they [original empha-
sis1created a situation of the riot in order to get the Muslim [persons in the 1local-
ity there migrated from that [area J.... These were three main theories. And there
were many local things and all. ... But, after three theories, the fourth element
was added by Congress (1) at that time. And they said that it is RSS and Janata
Party and Muslims must beware of that because Navrnan is ... part of Jan Sangh
and man of RSS, therefore, Janata Party means Jan Sangh and means RSS is respon-
sible for that.136

Four theories, each one of a type that is repeated in virtually every account
of the origins of Hindu-Muslim riots in India. Moreover, the realm of expla-
nation has remained wide open concerning this riot, for a judicial inquiry
commission appointed after the riot never completed its work and was ulti-
mately discontinued by the succeeding state Congress government in August
1980, while the state was experiencing yet another round of rioting (see below).
The same charges and countercharges were made in the U.P. Legislative
Assembly when the chief minister, Mr. Vishwanath Pratap Singh, announced
the termination of the inquiry committee. Opposition leaders who had been
in power during the 1978 riots charged that the inquiry was being withdrawn
because it had discovered the involvement of "some Congress (I) men" in it.
BJP leaders implied that it was being withdrawn for the same reasons as well
as because the commission found no RSS involvement in those riots, while
 Hindu-Muslim Violence in India and Aligarh 199

a CPI member bemoaned the withdrawal of the inquiry because it would not
reveal the truth "about the involvement of the RSS and a then cabinet min-
ister in the Aligarh riot." He remarked further that, "if inquiries were
scrapped like this, the people would never come to know who were the real
culpritS."137

May 1979

In this riot, the issue of the status of the AMU was again at the forefront. The
sequence of events began with the response of AMU students to the passage
under the Janata Party government of a bill in the Lok Sabha meant to restore
the status, powers, and governing institutions of the university to the posi-
tion prevailing before the passage by the previous Congress governments in
1965 and 1971 of laws that were held to have undermined them. The bill was
passed on May 3,1979, but was criticized by many Muslim leaders, organi-
zations, and the conservative group at the AMU for not explicitly restoring
the "minority character" of the university by designating it as such under the
terms of Article 30(1) of the Constitution of India. 138
A protest meeting was arranged by AMU students in Delhi for August 9.
Delhi being only a few hours' train journey from Aligarh, probably several
hundred students boarded the train on the morning of the day of the rally,
taking up four bogeys.139 On the way. for a distance of some 24 kilometers
between tlle Dadri Railway Station in Meerut District and Delhi, the students,
according to initial reports, were soundly thrashed and many of them beaten
bloody by groups of local men from the Gujjar caste, a predominantly rural
agricultural caste. The thorough thrashing of the students immediately gave
rise to conflicting explanations of the circumstances that led to it. The press
cited prominently "sources" who attributed the incident to the misbehavior
of the students, who sought to take up all the seats on the train, displacing
or making them inaccessible to the Gujjars, including their women and
children. The Gujjars responded by beating the students and calling for rein-
forcements from members of their own caste as the train passed tluough var-
ious stations en route. The students, however, many of whom arrived at the
rally in Delhi with their clothes torn and some of them "soaked in blood,"
claimed that they had been peacefully travelling to Delhi, that they were
"beaten up and looted by miscreants,» and that the RSS was responsible for
the attack. 140
This incident proved quite useful to several political organization and their
leaders. Dissidents in the Janata Party, who were then fighting an internal
 100 / Hindu-Muslim Violence in India and A.ligarh

battle to remove RSS members from party membership, used it as an exam-

pIe to support the need to rid the party and the country as well of the RSS
entirely.141 Congress (I) members of Parliament, now in opposition, also
found the incident useful to belabor the RSS and the Janata government
together. The former Jan Sangh and RSS members in Parliament, of course,
found wholly satisfying the explanation of the train beatings as having arisen
out of an altercation among passengers. The authorities, however, includ-
ing the home minister of the government of India and the local police and
administration, insisted that they did not have adequate information to reach
a conclusion, but that they had information that contradicted the student
accounts.142
Upon their return to Aligarh the next day, many AMU students took out
their resentment by first attempting to set fire to a state transport authority
bus and by setting fire to a couple of dozen shops owned by non-Muslims in
the Shamshad Market adjoining the campus. The authorities responded by
imposing an indefinite curfew on the city and the AMU and by posting PAC
and Central Reserve Police Force (CRPF) forces on the campus itself. Students
protested against the police emplacement on campus and police-student
clashes ensued, in which 5 persons were killed, whose identities-that is,
whether they were students or not-were not given in the press reports. 143
The campus was closed down and the students evicted from their hostels and
escorted by the police to departing trains to prevent further incidents.
In the meantime, the state government also took up the issue. The chief
minister, also a member of the Janata Party, though not of the former Jan
Sangh, gave an interpretation of the origins of the violence at Dadri differing
from the original reports that it was a brawl between Gujjars and the students.
He said that his intelligence reports indicated that it was a conflict "between
two groups of student commuters." He also said that "intelligence reports
showed that the AMU students going to New Delhi had molested a newly-
married bride in the train."144 He criticized those who were trying to give a
"communal colour" to the incidents, but then proceeded to lay the blame for
doing so upon "Muslim communalists," who had been "responsible for the
partition of India."145 Thus, once again do we see how a brawl of uncertain
origin is immediately subject to a multiplicity of explanations that provide
political benefits to political parties, groups, and leaders, while in the mean-
time expanding to encompass not only the relations between Hindus and
Muslims, Muslims and police, but the very origins and identity of India.
It was pointed out in the press that the incidents of May 1979 were entirely
different from those that had occurred the previous October, in that they had
 Hindu-Muslim Violence in India and Aligarh /101

"engulfed the campus and students," whereas, in the previous riots, the cam-
pus had been unaffected and the AMU students had provided relief to riot
victims in the city.146 On this occasion, however, it was the city that was
unaffected except by the imposition of preventive curfew. Yet there was some
effort to involve the city in the disturbances. The two leading traders' organi-
zations called for an "Aligarh bandh," that is, a closure of shops "to protest
against the violent incidents on AMU campus on May 10.» The Vyapar
MandaI, the larger of the two, and the one dominated by militant Hindu
Varshney businessmen and RSS members, also adopted a resolution demand-
ing "that the AMU should not be allowed to open unless sufficient security
arrangements were made for non-Muslim students and teachers on the cam-
pUS."147 One can only wonder why a traders' association in the city should
have taken up the issue of the protection of Hindu students and teachers on
the campus. The other organization, also dominated by Hindu businessmen,
was said to have "issued a pamphlet leveling false charges against Muslims."148
But why, then, did the AMU agitation not spread to the town on this occa-
sion, as it had in March 1971 and in the 1972 demonstrations and counter-
demonstrations over the passage of the AMU (Amendments) Act? In May 1979,
there simply was no connection to electoral politics, the last election having
been held in 1977 and no other election being in the offing at that time. Second,
administrative failings were considered to have been a factor in the failure to
prevent and contain the riots in 1971, though not in 1972. In May 1979, it was
generally acknowledged that the district and university authorities moved
immediately after the Dadri incident to prevent any incidents in the town, by
imposing curfew, and to contain the violent student demonstrations on the
AMU campus as well, through the use of force and by sending the students
home. Once again, therefore, political and administrative factors appear on
the face of it to make a critical difference in the development, prevention, and
control of potentially riotous communal confrontations.

Incidents of August to November 1980

In August 1980, one of the worst incidents of collective violence in the his-
tory of U.P. occurred in Moradabad City and District, near Aligarh, leading
to a massacre of upwards of 115 persons, of whom nearly all were Muslims.
Riots followed in several other U.P. towns across the state and elsewhere in
India. The first violent incidents in Aligarh were reported on August 16; it
was said that a "mob was protesting against the Moradabad incidents." Police
had fired, 2 persons had been killed, and indefinite curfew had been imposed
 102/ Hindu-Muslim Violence in India and Aligarh

on the entire city, excepting the Civil Lines. 149 Two days later, it was reported
that, "in all, two police constables" had been killed and "23 others including
19 policemen" had been injured. '50 It is not clear whether the two policemen
killed were the same as the two persons reported killed earlier. A clash of this
sort, in which policemen are killed and injured in the circumstances obtain-
ing in Aligarh at the time, could only mean large-scale confrontation between
police and Muslim crowds. Reports from Aligarh on this and succeeding days
referred to relaxation of curfew, arrests of a large number of people,151 and
formation of peace committees to visit "the affected areas in a bid to reduce
tension and restore communal harmony."'52
Nevertheless, violence broke out again on August 24 in the old city dur-
ing a period of curfew relaxation, though no further deaths were reported.
The district administration was reported to be "handling the Aligarh situa-
tion with firmness" and "the authorities" were said to "have posted four
columns of troops there besides units of the BSP, the PAC and the local
police."'53 There then followed several more days of reports in the press from
Aligarh that there had been no further incidents, curfew was being relaxed
gradually in terms of the numbers of hours of the day as well as the areas
affected, additional persons had been arrested to prevent further violence,
and peace committees had again begun visiting the affected neighborhoods. '54
Meanwhile, the situation in Moradabad had been brought under control
sufficiently to allow the beginning of army withdrawal from the town on
September 2.155
As usual, a multiplicity of explanations for the outbreak of violence in
Moradabad, Aligarh, and elsewhere in the state were offered instantly by politi-
cians and the press. In contrast to the situation in 1978, when Janata govern-
ments were in power in Lucknow and New Delhi while the Congress (I) was
in opposition, the situation was now reversed. A state Congress (I) minister,
himself a Muslim, visited Aligarh and declared at a meeting at the AMU that
"the current disturbances were not communal but seemed to be the work of
certain lawless elements," but a central government minister of state for home
affairs, Yogendra Makwana, blamed the RSS, Jan Sangh, and BJP.156 Indira
Gandhi, however, consistent with her views in general concerning the sources
of the country's problems, suggested "that foreign forces could be behind the
recent communal incidents in U.P. and other parts of the country."157 The
formulation that ultimately emerged as the central government's position on
these riots was that "a neighbouring country," which, of course, meant Pald-
stan, as well as "some communal parties" were involved in promoting them.
Union Minister Makwana noted, in support of this formulation, that "for
 Hindu-Muslim Violence in India and Aligarh /103

the first time riots had broken out between two minority communities-
Muslims and Harijans." Since these two "groups had always sided with the
Congress," the aim was said to be "to weaken the ruling party."158 Although
the formulation contained considerable vagueness, especially on the partic-
ular communal parties involved, that is, whether they were Hindu or Muslim
or both, one implication was that Pakistan, in collusion with Muslim com-
munal parties, was responsible, and that, therefore, it was the Muslim minor-
ity and the Harijans, but not the upper-caste Hindus, whose members were
doing most of the rioting.
Girilal Jain, editor of the Times of India, added his support for the plau-
sibility of the central government's interpretations. He blamed "anti-social
elements" among the Muslims for being "at least partly responsible for the
riots in Moradabad and other U.P. towns" and Muslim leaders for not admit-
ting the facts themselves, but instead blaming the RSS as usual. He absolved
the RSS from any responsibility whatsoever and lent support to Mrs. Gandhi's
contention that "foreign interference" was involved. 159 In a move that we have
seen practiced earlier during the 1978 riots, the Times of India followed the
next day with a news item listing the number of persons from Pakistan who
had been visiting U.P. in the past few years and noted that there had been a
considerable increase, particularly in towns such as Sareilly, Aligarh, and
Moradabad, where extensive rioting was now taking place. 16o
There occurred at this time, in fact, a convergence of views among the
Congress (I) leadership, the leading English-language newspaper, and tile SIP
leadership that must have appeared menacing to many Indian Muslims. For,
at the same time, the SJP leader, L. K. Advani, while denying any attempt on
his part to blame the Muslim community as a whole for this latest round of
rioting in northern India, nevertheless thought that "the manner in which
the riots spread to Meerut, Delhi, and other towns almost simultaneously and
in an identical pattern did suggest to him some kind of planning." He was
then quoted directly as follows: "It appears that some Muslim elements are
trying to pit the community against the law and order machinery. In Srinagar
[Kashmir] they clashed with the army; in Moradabad and some other towns,
they attacked the police." He tllen directed his accusation specifically at Muslim
organizations and leaders, namely, the Muslim League, the Jamaat-i-Islami,
and the imam of the Delhi Jama Masjid. 161
In the midst of reports that the rioting in Moradabad and Aligarh was being
brought under control and the post-mortem speculations, furdler violence
broke out in Moradabad and Aligarh. In this latest outburst, the killing was
greater in Aligarh than in Moradabad. In Al.igarh, nine people were reported
 104/ Hindu-Muslim Violence in India and Aligarh

to have been killed on September 8 in "clashes between groups of people,"

but none in police firing. 162 The reference to "groups of people" means that,
on this occasion, the confrontation was not between Muslims and the police,
but between Hindus and Muslims. The next day, it was announced that the
district magistrate and the senior superintendent of police had been replaced.
The new DM was Rajiv Ratan Shah, the SSP was Mr. B. P. Singh. B. P. Singh,
especially, was said to have dealt with the Aligarh "underworld" quite
effectively, that is, according to my sources, by gathering up more than forty
known and alleged Muslim goondas and simply taking them outside the city
limits and killing them. 163 From this point until 1990, Aligarh had no major
riots. 164
Although sporadic rioting continued in Moradabad over the next two
months, it was contained in Aligarh by the new administration.165
During these months and for some time thereafter, there was a multiplicity
of incidents that have not been given a clear focus in public documents, press
reports, or my own interviews. 166 Responses to my questions concerning the
riots during this period were quite diffuse. They suggest a general state of anar-
chy and local warfare that, as noted in other sources, included "pitched bat-
tles between Muslims and the Provincial Armed Constabulary. "167
In 1983, I asked respondents, both Hindu and Muslim, to tell me about
these riots. Some respondents, and others who have written about the riots,
mostly without direct knowledge of them and not having heard of any par-
ticular reason for riots to have broken out in Aligarh in 1980, attributed the
outbreaks to a spontaneous response on the part of Muslims to the police
firing against Muslims in Moradabad at that time, which involved much loss
of life amongst Muslims there. Others with more direct knowledge of the inci-
dents in 1980 and in the preceding years gave more elaborate responses.
According to one Muslim politician, these incidents were not riots, but a sys-
tematic attempt on the part of the Hindu Banias associated with the Jan Sangh
in the locality of Manik Chauk to harass and intimidate Muslims in order to
get them to flee so that they could obtain their valuable property cheaply.
Eleven persons were killed in consequence, but the police, this respondent
said, acted promptly and effectively to contain the effects to the mohalla of
Manik Chauk. 168

Communal Incidents between 1980 and 1990

Among the various incidents that occurred during this-for Aligarh-

relatively peaceful period, those that occurred in October 1988 are of con-
 Hindu-Muslim Violence in India and Aligarh /105

siderable interest because of the extent to which they reveal the dynamic
processes of riot production and riot control as well as their close connec-
tion to political events.
The Riots of October 1988. The riots of October 1988 were underreported
in the Times of India because they were overshadowed by the much larger
disturbances in the nearby district of Muzaffarnagar, which had never before
experienced communal rioting. Rioting occurred later in the month in other
districts in U.P. as well. All the rioting in the state in that month was associ-
ated in press reports with the opposed Hindu and Muslim agitations in con-
nection with the militant Hindu movement to bring down the Babri Masjid
in Ayodhya, to replace it with a temple to Ram. The agitation and counter-
agitation began with the announcement of a planned march to Ayodhya by
five hundred Muslim leaders sponsored by the Babri Masjid Action Committee
to offer prayers at the contested site. The march was scheduled for March 14.
VHP leaders promptly announced a countermove in opposition to the
march, calling for a bandh (closure of shops) throughout the state on March
8 in protest against it. The call for a bandh in a context in which Hindus and
Muslims have opposed interests is a highly provocative tactic that has many
times in modern Indian history precipitated violence, both during the nation-
alist movement and since Independence. In most cases, the violence has fol-
lowed attempts by Hindus to compel resisting Muslim shop owners to pull
down their storefront shutters. Press reports on the events in October in the
districts that experienced violence during and after the bandh call indicate
clearly that this is precisely what happened.
However, riots were reported from only two of the 54 districts in the state:
Aligarh and Muzaffarnagar. Riots also occurred later in the month in Faizabad
(adjacent to Ayodhya), Hardoi, and Bahraich. In Aligarh and Muzaffarna-
gar, rioting was clearly associated with the March 8 bandh call.
Although the precipitating incidents were said to be similiar in Aligarh
and Muzaffarnagar, the political context in the two districts was entirely
different, and so were the results. Rioting was reported in the first Times of
India dispatch from Aligarh and Muzaffamagar on October 8, in which it was
stated that two persons had been "stabbed to death in Aligarh, and eight seri-
ously injured in Muzaffamagar and Khatau}i" (a town neighboring Muzaffar-
nagar).169 The rioting was sufficiently serious in both places to warrant the
sending of CRPF personnel to both Aligarh and Muzaffarnagar towns. The
home secretary of the state government was given as the source for the infor-
mation concerning the events in Aligarh, where "two groups clashed ...
around noon," "exchanged brickbats and resorted to arson resulting in the
 106/ Hindu-Muslim Violence in India and Aligarh
destruction of several shops. Stabbing incidents were also reported from some
parts of the city." In these fracases, 27 people were said to have been injured,
of whom two died of their wounds. In Muzaffarnagar, the outbreal( of vio-
lence occurred "while members of a particular community [Hindu] were tak-
ing out a procession over the Ram Janmabhoomi issue," during which "some
participants ... asked shopkeepers to close their shops and this led to a clash."
Although curfew was imposed in both Aligarh and Muzaffarnagar through-
out most of the rest of the month, it is noteworthy that the situation in
Aligarh-despite its more violent history-was kept under control while that
in Muzaffarnagar was not. No further deaths were reported from Aligarh after
the initial stabbings, but the Times of India reported 13 killed in Muzaffarna-
gar and the neighboring town of Khatauli by October 11; the last report on
the death toll there gave a figure of 22 killed. 170
The contrast between the situations in Aligarh and Muzaffarnagar on this
occasion throws additional light on the factors that promote and sustain com-
munal violence in northern India. The Times of India itself raised the ques-
tion of why the violence was contained in Aligarh while it spiraled out of
control in MuzaffarnagarYl Its answers, based on its own reports from the
district and state headquarters, were that the administrative and political sit-
uations in the two districts differed. In Muzaffarnagar, the district adminis-
trative officials were accused of outright negligence on the first day of the
incidents, in contrast to Aligarh, where prompt and effective action was taken
from the first outbreak of violence. The second reason was that the political
situation in Muzaffarnagar was even more volatile at the time than in Aligarh
because two ministers in the state Congress government, one a Hindu, the
other a rising Muslim politician, both from Muzaffarnagar town, were hos-
tile to each other and contesting for local political control. A prominent Hindu
supporter of the Hindu minister, one Harish Chabra-not a BJP man, but
a Congress man-went around "mobilising support for the October 8 bandh"
that provoked the initial violence. Following upon the arrest of "scores of
persons ... in connection with the violence," this same Chabra led "a mob"
that "surrounded the Kotwali police station" demanding their release.172 Then,
on October 11, there was a further clash between a procession led by Chabra
and an interparty peace march, leading to further violence. The Times ofIndia
editorial suggested further that the administration's ability to control the riot-
ing was hamstrung by the reported involvement of two powerful ministers
in the state government and their supporters.
The differences between Aligarh and Muzaffarnagar at this time suggest
the limitations of both the method of paired comparisons and the game the-
 Hindu-Muslim Violence in India and Aligarh /107

ory approaches to explaining communal violence/peace. Muzaffarnagar had

experienced no communal violence since 1947, but it suddenly faced a mas-
sive outburst comparable to the worst in Aligarh in previous years. Yet
Muzaffarnagar, because of its demographic and economic similarities with
the two most riot-prone districts in western U .P., had been included by the
Centre for Research in Industrial Development for comparison with Aligarh
and Meerut.
The study, though it selected mostly on the dependent variable of riot-
proneness, identified several differences between Muzaffarnagar, on the one
hand, and Aligarh and Meerut, on the other hand, that appeared to explain
the former's freedom from violence. The differences included, especially, the
sharp separation of predominantly Hindu and predominantly Muslim local-
ities, the generally subordinate, noncompetitive economic and political posi-
tion of the Muslims in relation to Hindus in tlle town of Muzaffarnagar, and
a history of intercommunal relations in the surrounding area that included
even some intermarriage and joint decisions on political action, as well as a
secular attitude on the part of traders from both communities. Indeed, their
study reported that many Muslims supported the BIP and that RSS volun-
teers were engaged in promoting Hindu-Muslim friendship. Finally, they
reported that their survey research showed, in contrast to Aligarh and Meerut,
that the two communities shared an appreciation for both Hindu and Muslim
historical figures who are associated in Indian historical hagiography with sec-
ular attitudes. 173 The only ominous sign they saw in the future was the exis-
tence of "increasing competition" in a context of "locational confrontation"
between two subgroups from the Muslim and Hindu communities, Qureshi
Muslims and Scheduled Castes, in competition in the beef and meat business. 174
However, the evidence available from the newspaper reports suggests that
it was political competition between Hindu and Muslim leaders, rather than
any change in the existing economic and social patterns, that produced the
massive rioting in 1988. Further, it is evident that what occurred was a spi-
ralling of conflict that is inconsistent witll the Fearon/Laitin assumptions.
Indeed, as noted above, there was even a peace march "taken out by various
sections of the people" (meaning both intercommunal and interparty) dur-
ing the riots, which was attacked by rioters.175 Insofar as Aligarh is concerned,
the limited evidence available on the riots there suggests just the opposite of
the situation in Muzaffarnagar: a highly politicized, communalized political
environment known for both the intensity of its interparty competition and
the tendency for intercommunal incidents to spiral out of control was con-
tained by the presence of firm, prompt administrative action and the ab-
 108/ Hindu-Muslim Violence in India and Aligarh

sence of an immediate local political advantage. One of the variables, the politi-
cal context, was to change within the year, while the administrative context
remained the same.
The riots of November 1989. The context in which the reported incidents
of November 1989 took place was framed by the dramatic events preceding
the 1989 elections and the election itself. The announcement of elections and
the scheduling of polls was made by Prime Minister Rajiv Gandhi on October
17. During the period before the elections, the VHP intensified its campaign
to replace the Babri Masjid in Ayodhya with a newly constructed temple to
the god Ram, through the famous shilanyas march of militant Hindu vol-
unteers to Ayodhya, bringing bricks to be consecrated for the foundation of
the new temple to be constructed there. The procession reached Ayodhya and
laid the foundation for the Ram temple at a site adjacent to the mosque on
November 9. On October 26, one of the worst communal riots in post-
Independence India, including some incidents of a particularly atrocious char-
acter, began in the district of Bhagalpur in Bihar; the reports on these riots,
which extended to other parts of Bihar and continued for weeks, with
mounting death tolls, reverberated among Muslims throughout north India.
Polling for the Lok Sabha and Vidhan Sabha elections was held between
November 22 and 26 in different parts of the country.
A further issue that became prominent at this time concerned the grant-
ing to Vrdu the status of a second official language in the western districts of
V.P. Severe riots occurred between September 28 and October 2 in the dis-
trict of Budaun, adjacent to Aligarh District, attributed to agitation over this
issue. 176
Aligarh City and the AMV both were affected by these events in the month
of November. On November 7, a broadside was distributed in the city and
on the AMV campus announcing a meeting of intellectuals to oppose the
declaration of Vrdu as the second state language of v.p.m Three Hindu stu-
dents at the Engineering College of the AMV were noticed by two Muslim
students with the broadside and possibly some other material in their pos-
session, presumably to be distributed on the campus. The Muslim students
apparently reported the matter to the university authorities, who immedi-
ately took the strong action of suspending the three students from the uni-
versity and ordering them to vacate their hostel rooms for having "brought
some objectionable/provocative material" to the Engineering College on
November 7. On the following day, one of the students wrote to the dean of
the engineering faculty proclaiming his innocence and stating that he was given
the "printed cards" in question by another Hindu student. He wrote again
 Hindu-Muslim Violence in India and Aligarh /109

to the District Magistrate, Aligarh, to protest his innocence. It appears from

this letter that the grounds for his suspension included the allegation that
"some pamphlets inciting communal hatred" were found in his room, of
which he denied any knowledge.
In the midst of this argument over the distribution of the broadside call-
ing for the anti-Urdu meeting, it appears that there was a much more seri-
ous matter involved, namely, the plastering on the walls of buildings on and
near the university campus of scurrilous and indecent posters and banners
attacking the Congress (I) and Rajiv Gandhi personally. It has never been
revealed whether or not the three Hindu students had anything to do with
these posters, which instead were attributed to Muslim students of the AMU
angry over the shilanyas and the riots in Bhagalpur. These posters would be
plastered at night, then removed the next day by the authorities, only to reap-
pear the next night. The slogans plastered on the walls were considered so
uncouth and indecent that their substance was not reported in the English-
language press and only one or two of the least offensive of them were men-
tioned in the Hindi press. There were also reports of similarly offensive effigies
of Rajiv Gandhi. 178
In the midst of these provocative calls for meetings, plastering of slogans
and placing of effigies, and the election campaign itself, the Times of India
reported from Aligarh with a November 10 dateline on mob violence in "some
Muslim localities in the old city." The citing of Muslim localities and the
grounds given for the violence-anger over the news that the shilanyas in
Ayodhya had been allowed to take place the previous day-implies that the
rioting was done by Muslims. The story also referred to an attack by "an angry
mob of slogan shouting students on the Vice-Chancellor's residence," the
hoisting of a black flag there, shouting of "anti -Rajiv Gandhi slogans," a can
by a Muslim student leader for a "war against the ruling party," and the burn-
ing of "half a dozen effigies ofMr Rajiv Gandhi. "179 A Hindi newspaper, date-
lined November 13, referred to the violence of November 10 as having
involved firing, presumably police firing, and brickbatting in U par Kot. There
were no reports of deaths. Another report in a Hindi newspaper, datelined
November 11, referred to the same violent incidents and named a number of
mohallas in which they had occurred: four of them in the central Muslim areas
and one, Shamshad Market, adjacent to the university. ISO The report went
on to say that, although the city had gone back to normal, an atmosphere of
communalism persisted at the AMU and in some other parts of the city
because of the posters. There were also reports of processions and meetings
on the campus and in the city: by the nonteaching staff of the university accus-
 110/ Hindu-Muslim Violence in India and A.ligarh

ing the government of bowing down and giving support to Hindu commu-
nalism; meetings of people of "a religious community" (that is, Muslims) in
some parts of the city to oppose the Congress (0; and tearing up of the
offensive posters and banners around the university campus.
Both the district and the university administration were extremely alert
to all these manifestations of anger and protest and took firm actions, some
of which, however, backfired. For example, the district administration closed
down all educational institutions in Aligarh City, both the AMU and the local
degree colleges, from November 13 until November 27, that is, until after the
elections. However, the closing of the university had the opposite of the effect
intended. It was assumed, wrongly, that the students would take advantage
of the vacation time to return to their homes, but instead, according to the
Hindi press, the student organizations announced various actions such as the
declaration of November 15 as a "black day" on which students would court
arrest in protest against the shilanyas. Meanwhile, the poster campaign
against the Congress (I) continued on the AMU campus. The Hindu students
were also active, calling for the revocation of the suspension of the three Hindu
students and naming the two Muslim students who reported their activities
as "the real culprits." For their part, the Muslim AMU students announced
their support for the Janata Dal candidate for Parliament, Mr. Satyapal Malik,
a Hindu known for his secular attitudes, but said they would not necessarily
support the Janata Dal in other constituencies because of its arrangement for
adjustment of seats with the BJP in the Lok Sabha elections. Mainly, how-
ever, they announced that they would work for the defeat of the Congress,
including the defeat of Congress Muslim candidates.
The last newspaper report from Aligarh, from the Times of India, date-
lined November 15. reported on the basis of a large number of interviews con-
ducted on the AMU campus that there had been "a sharp change in the mood
of the Muslim voters ... in the last ten days," such that the Congress could
not expect to get more than 25 percent of the Muslim vote in the state. Muslim
disaffection was attributed particularly to Rajiv Gandhi's decision to launch
his election campaign from Ayodhya and "the fallout of the shilanyas cere-
mony at Ayodhya."181
It appears from the newspaper reports of the violence in the inner city that,
despite the fact that there were no deaths, the incidents constituted a riot under
existing laws. At the same time, the violence was also evidently contained
quickly and effectively. The news reports cited above all referred to quick police
action, institution of day-and-night street patrolling, and the closure of all
educational institutions. Further, in the midst of all the hubbub surround-
 Hindu-Muslim Violence in India and Aligarh /1ll

ing the shilanyas, election campaigning, poster plastering, and the anti-Urdu
agitation, the district administration was faced with the movement of large
crowds of Hindu devotees through the town on the way to the Ganga River
to observe the Hindu bathing festival of Kartik Purnima (eighth-month full
moon). Newspapers reported that, because of this festival, "the administra-
tion has kept special watch over persons, vehicles, and trains passing by or
near the [AMU] campus."182 It would appear, therefore, that the district
administration had multiple reasons for acting effectively to contain this riot,
and that they did so.
At the same time, we can learn from the course of events in this riot much
about how riots begin in Aligarh and elsewhere and how they spread. This
riot was marked by a sequence of provocative actions, of which the call for
the anti-Urdu meeting was the least provocative and was certainly within the
range of normal, legal, democratic action. The other two major events were
also within the range of democratic participation in some societies, but pass
the margin of democratic acceptability in India because of their association
with or deliberate instigation of communal enmity, a violation of Indian laws.
The shilanyas, while technically a legal act, was everywhere in India associ-
ated with provocative actions that aroused Muslim anger: processions through
their neighborhoods, ringing of bells in the night, and many other demon-
strative acts, ending in the ultimate provocation of laying the foundation stones
for a temple designed to be built over an existing mosque in Ayodhya.
The indecent posters were clearly both provocative and illegal. They were
attributed to Muslim students at AMU. If this is true/8;; then signals were
being sent from elements in both communities for the start of riotous or poten-
tially riotous activities. Further, these signals involved elements both in the
city and at the AMU. Whereas Muslims were angry over the shilanyas and
the implication of the Congress (I) and Prime Minister Rajiv Gandhi in allow-
ing it to take place, Hindus were opposed to the imposition of Urdu as a sec-
ond language. But, more important, militant Hindus in Aligarh remain always
ready to pounce upon the AMU as the source of Muslim communalism, sep-
aratism, and anti-Hindu sentiments. Whenever communally sensitive or
provocative events in the city and the university occur or are made to appear
to occur simultaneously, the danger of large-scale Hindu-Muslim violence
is extreme in Aligarh, as events in the following year (discussed in the next
chapter) were to demonstrate.
There is, finally, the association of these events with the election campaign,
another aspect of the situation that promoted simultaneous feelings of anger
among Muslims in both the city and the AMU who shared resentment against
 112 / Hindu-Muslim Violence in India and Aligarh

the Congress for its behavior in relation to the shilanyas at Ayodhya. The results
of this election, in fact, depended heavily on Muslim reactions throughout
the state and did indeed contribute significantly to the defeat of the Congress
in the state as well as in the country, leading to the installation of Janata Dal
governments in New Delhi and Lucknow. The association of communal riots
with election campaigns has been noted above and will be discussed further
in detail in succeeding chapters.

Summary

Since the 1990 riots took place after a long gap from their most violent pre-
cursors in 1980, and stand apart in other ways from their predecessors
between 1956 and 1989, the main features of this long series of riots will be
summarized here before discussing the new series of incidents and rioting
that began in 1990. First, it should be noted that there is a shift, from the early
to the later incidents in the decades between 1961 and 1989, in spatial origins.
Among the earlier riots, several involved actions and reactions between town
and university, with students playing major roles on both sides.184 These
include the 1956 disturbances involving the Muslim student protest on the
AMU campus over the republication of a book alleged to contain"disrespectful
references to the prophet," which was followed by the false reports that the
AMU students had burned a copy of the Bhagavad Gita, providing a pretext
for a student protest strike in tlle city's degree colleges. Also included is the
1961 riot, in which the precipitating incident was the scuffle over the AMU
Students Union elections, which was then followed by action on the AMU
campus, with news and rumors concerning it sent to the city, precipitating
an action by the Hindu students in the city colleges. The third, in 1971, aris-
ing out of the agitation on the university campus over the issue of restoring
the minority character of the university, also began on the campus and was
followed by riot in the city. The fourth, however, the riot in the aftermath of
the train beating at Dadri, did not move to the city.
Second, although most of the Aligarh riots of this type do not arise directly
out of quarrels between rivals from organized political parties, there is usu-
ally a political connection of some type. The 1961 riot led to Muslim politi-
cal action against and defeat of the Congress in the broader arena of district
politics. The 1971 and 1989 riots were associated with the elections of those
years. Other riots have been precipitated by the politicized issue of the sta-
tus of AMU, which is highly charged symbolically and "represents" Hindu-
Muslim relations in significant ways.
 Hindu-Muslim Violence in India and Aligarh /113

Many of the later riots, however, are dearly marked by incidents within
the town and by the involvement of local politicians in them. There is also
an allegation of profit motivations by greedy Hindu businessmen.
When we come to 1989, however, and then to 1990, there is a kind of merg-
ing of the two types of riots. The 1990 riots start in town, as we shall see, but
move to AMU as well. In this respect, the 1989 riots constituted a kind of
rehearsal for what was to take place the next year.
When we consider that several of the incidents that took place in the above
years spread over weeks, months, and even years and that each of them
involved the imposition of curfews, one could almost say that there was a
kind of semiperpetual curfew in Aligarh marked also by the presence of armed
forces at strategic crossings. At such times, they are mostly standing and sit-
ting idly and uselessly by crossings and other sites where incidents have
occurred in the past (see Figure 3.5). Similar situations exist in other towns
in western U.P., such as Meerut.
Some of the riots display a kind of ritual of provocation of a type noted
by Gaborieau, involving "codified procedures" that include the "selection of
key symbols representing each community" and the "selection of the means
by which such symbols may be most effectively desecrated," followed by the
implementation of acts of violence against those symbols-mosques and
temples-and extending, finally, to destruction of property belonging to per-
sons from the other community and to killing. 185 That patterning of Hindu-
Muslim riots has been present at times in this series of riots, for example, in
those involving the throwing of colored water on Muslims during the Hindu
festival of Holi, but it does not seem to be prominent in most of them, espe-
cially the later ones. Instead, there is a different kind of sequence or perhaps
two different types of sequences, even more deadly and provocative. In one,
people from the two communities are already massed in one way or another
in either peaceful contestation (student union elections, a wrestling match,
or a political demonstration) and a scuffle breaks out between individuals,
after which the rumors spread, there may be further provocative actions, and
larger mobs are brought into play. In the second type, an incident occurs
between two individuals, one Hindu, the other Muslim, and conscious
efforts are made to mobilize members of one community to avenge the harm
caused to one of its members. Sometimes, the second sequence is required
to bring into full play the first; the second sequence, however, is one that is
in operation all the time. I will return later to the significance of the perpet-
ual operation of this second sequence.
Insofar as the change in the dynamics of riot provocation and action are
 114/ Hindu-Muslim Violence in India and Aligarh

FIG. 3.5. Police picket at Phul Chauraha

concerned, the Minorities Commission itself took note of it in its first report,
quoting the prime minister at the time, Morarji Desai, who noted that "petty
quarrels between two individuals are transformed into riots between two com-
munities," in contrast to the traditional forms of provocation involving "music
before mosques" or "taking out processions during festivals, etc."186 Figures
were given for the country as a whole for the year 1977 regarding how many
of the 188 communal incidents of the year arose in each way: it was found
that the large majority (113) arose out of "petty quarrels." The Commission
expressed its inability to understand why this should be so, but offered three
possible alternative explanations.
The first explanation proffered by the Commission was "that the feelings
between the two communities are so strained that an attack on an individual
is construed as an attack on the whole community." This explanation would
seem to be consistent with the Gaborieau model. It also fits the spiral equi-
librium pattern. The second construction offered by the Minorities Commis-
sion was "that there is a particular section in each community which has a
design to take advantage of every opportunity to spark off a communal riot."
A third explanation is that "communal riots are engineered by persons who
have an avaricious eye on the property of the minority community and who
succeed in acquiring such property on their own terms after each riot. "187
 Hindu-Muslim Violence in India and A.ligarh /115

The first two explanations could be consistent with each other, but it
depends on how much organization and effort is implied in the second and
what kinds of people are involved in the riots according to the second expla-
nation. In fact, as will be shown later, none of these explanations can account
satisfactorily for the dynamics of riots: their timing, the triggering incidents
that allegedly set them off, and the course they follow. The first is wholly unsat-
isfactory. The second is unsatisfa(.iory, among other reasons, because it does
not define what constitutes an "opportunity" to set off a riot. The third, the
property explanation, can be fit into a more general one, but the idea that
large-scale riotous events are "engineered" by avaricious businessmen will
not stand.
Communal riots are preeminently political events in which many other
forces come into play once they are fomented. None of the explanations offered
by public, political men present the preeminently political aspects of these
events, because to do so would expose not only themselves, their political
organizations, and their inability to control riots when they occur, but the
overall functioning of a polity in which riot production is as much a routine
aspect of politics as interest articulation, mass mobilization, and electoral com-
petition. Indeed, it is embedded in these latter processes. Further, the very
act of explanation is itself a part of these processes that contributes to the
production and persistence of riots framed within a communal context.
 4 / The Great Aligarh Riots
of December 1990 and January 1991

I n the minds of most commentators, reporters, and my own interview

respondents, the 1990-91 riots in Aligarh were dramatically different from
all that preceded them in several respects. First, the scale surpassed all
previous riots in the city, including those just before Partition in 1946, both
in the extent of the areas touched by them and in the numbers of persons
killed. The official figure for deaths in these riots given out by the district mag-
istrate was 92, of whom "about 2/3 were Muslims." However, Muslim groups
provided a figure of 100 for Muslims alone, with a documentary list of the
victims. On the basis of their own inquiries, the Peoples Union for Civil
Liberties (PUCL) placed "the number of persons killed at 125 to 150."1
Pertinent also to this matter of scale is the allegation made by Muslims
and confirmed by some Hindu observers as well that the riots broke out simul-
taneously in a large number of places in the city. This pattern contrasts with
previous riots, which began with an incident in a particular site and then spread
to other parts of the city. This allegation also carries the presumption that
these attacks were preplanned and coordinated. As one Muslim respondent,
an AMU professor, put it, "they [the rioters] attacked the Muslim commu-
nity from every corner.":/.
Second, this riot, in contrast to all the others discussed above, was asso-
ciated with a mass movement that engulfed the entire state as well as other
parts of northern and western India. This movement was inaugurated with
the rath yatra of L. K. Advani, which began in Somnath in western Gujarat
on September 25,1990, and proceeded across central India to Bihar, where it
was stopped by the arrest of L. K. Advani on October 23. The journey was
originally planned to end at Ayodhya, where kar seva (voluntary work) was
to begin for the construction of a new temple on part of the disputed site.
Although the rath yatra itself was prevented from entering Uttar Pradesh on

116
 Riots of December 1990 and January 1991 /117

its way to Ayodhya, the movement of tens of thousands of Hindus from all
over northern and western India and their convergence upon Ayodhya was
not stopped. These crowds passed through many of the important cities and
towns of U.P. in processions, raising slogans along the way, many of them
deliberately directing insulting remarks towards Muslims, and bringing
along with them additional persons from the sites through which they passed.
During the movement, in most towns in U.P., including Aligarh, Hindu sup-
porters of the movement "would start a cacophony" every evening at a pre-
arranged time, involving ringing bells, beating thalis (stainless steel dinner
plates) and blowing conch shells."3
In the town of Ayodhya, on October 30, the police, under orders author-
ized by the chief minister of the state himself, Mulayam Singh Yadav, stopped
an assault upon the mosque with bullets that resulted in the deaths of 16
Hindus in the crowd. News of the killings in Ayodhya naturally spread quickly
through the media, but also by word of mouth as devotees returned from
Ayodhya and dispersed back into their home places. Processions of devotees
also moved out of Ayodhya carrying the ashes of victims of the police firing,
to be spread across the state, including in Aligarh. Serious rioting and killing
began in Aligarh City on December 7 and continued with great intensity for
several days thereafter. The rioting included an attack on December 8 by a
large Hindu mob that stopped a passenger train, the well-known Gomti
Express, killing as many Muslims as could be identified, in a manner that
recalled the horrors of Partition and the great train massacres associated with
that catastrophic event. Muslims in Aligarh expressed the view that the spread-
ing of the violence in the aftermath of the October 30 police firing to places
where there had been no rioting since Partition, and its intensity in Aligarh,
were attributable to the" communal feeling which was being created by Vishwa
Hindu Parishad, RSS, Bajrang Dal, and other elements of the same feeling."
That feeling, it was said, included "taking revenge of the death of kar sevaks,
who were shot dead in Ayodhya when kar seva started on 30th October and
to November 2."4
It is common to attribute rioting that follows in the aftermath of mass
movements such as the one that developed in connection with L. K. Advani's
rath yatra to the spontaneous outbreak of communal hostilities in areas that
have a history of Hindu-Muslim tension and violence. When the rioting
spreads to new areas or areas not well known for such a history, it is said that
communal passions were inflamed or provoked by the mobilization and, in
the case of Ayodhya, particularly by the police firing on the Hindu crowd.
Such attributions are quite misleading insofar as they focus on alleged com-
 118 / Riots of December 1990 and January 1991
munal passions that break out spontaneously. On the contrary, these pas-
sions are fostered and kept alive in particular places through the actions of
communal activists and deliberately stoked to inflammatory proportions in
conjunction with such broad movements. Such was certainly the case in
Aligarh in the months before the beginning of the 1990 riots on December 7.
The Peoples Union for Civil Liberties report comments on the way in which
such intense sentiments were stimulated in U.P. as a whole and in Aligarh
in particular.

With the Ramjanmabhumi/Babri Masjid controversy, and particularly the

Rathyatra of BIP leader L K. Advani, the divide between Hindus and Muslims
had deepened. The divide became near complete after certain events took
place-the karseva programme at Ayodhya on 30-10-90 [October 30, 1990 J and
2-11-90 [November 2, 1990) accompanied by violence and police firing, play-
ing of audio/video cassettes of provocative speeches by BJP/VHP/Bajrang Dal
members against Muslims, holding of so called religious 'melas', circulation
of highly objectionable leaflets by some Hindu organisations, kites with
provocative slogans against Muslims released over Muslim areas of the dty,
almost total communalisation of the Hindu press in UP-all these had already
built up a tense atmosphere}

All these actions require a considerable degree of organization that varies from
town to town in north India. In Aligarh, that organization is extensive, active,
and ever ready for violence. 6

THE PATTERN OF RIOTING IN ALIGARH

It is also evident from the pattern of rioting in the city that considerable
advance planning was undertaken. The rioting did not follow the usual
sequence of a disturbance, a fracas, a killing, or some other violent act in one
area mimicked in a so-called retaliatory act in another area, then spreading
to the usual riot-prone areas. On the contrary, in this riot, there were virtu-
ally simultaneous outbreaks extending to many parts of the city and into the
Civil Lines area as well, including especially the Aligarh Muslim University
Medical College.7 As we have seen, the AMU is not usually affected by riot-
ing in the city, though it stands as a symbol for militant Hindus as the source
of all communal problems and Muslim disloyalty not only in Aligarh, but in
India as a whole. II Extending the attacks to the university campus added fur-
ther emphasis to the fact that the Ayodhya movement itself was not only about
 Riots of December 1990 and January 1991 /119

a mosque allegedly destroyed by Muslim conquerors centuries ago, but was

about the relative place of Hindus and Muslims in India.
Although, therefore, the extent of the rioting in Aligarh compared to other
cities and towns in U .P. cannot be understood without an appreciation of
the higher degree of mobilization and organization among militant Hindus,
it is also generally accepted by most observers that there was much more mass
support for this riot than for any other riot in Aligarh since 1946. Nor was
this mass support confined to the lower classes. On the contrary, if anything,
militant Hindu sentiment and the feeling that it was past time to give the
Muslims their due in Aligarh as elsewhere in north India spread more widely
among the middle and upper-middle classes.9 The lower classes, particularly
the Scheduled Castes, participated significantly in the violence, fighting with
Muslims in various localities, not because they shared the militant Hindu
nationalist ideology, but for purposes of "looting."10
Also noteworthy was the role played by the Hindi vernacular press in this
riot, particularly by the newspaper Aaj, which had since the days of the nation-
alist movement a reputation as the leading Hindi newspaper of north India.
This newspaper was responsible for the spreading of a vicious and false rumor
on December 10 in "a full-page banner headline on the front page that 74
people including 28 patients in Jawaharlal Nehru Medical College Hospital
of Aligarh Muslim University were killed." 11 It is generally accepted that riot-
ing, particularly attacks by Hindu mobs on Muslim persons and property,
intensified after tllis date. However, Aaj was by no means an exception. The
Hindu populace of U.P. and Aligarh was fed with fantastic stories and lies
from many of the prominent Hindi newspapers in north India throughout
the rath yatra. These newspapers included Amar Ujala and Swatantra Bharat.
The stories, obviously designed to instill "hatred for the minorities, » included
"highly exaggerated" accounts of the death toll in Ayodhya and elsewhere in
the state, ranging in the thousands and tens of thousands. They also included
mythological fabrications designed to imply divine intervention on behalf
of Hindu activists and against the police who attacked the Hindu mob at
Ayodhya, featured as front page news in the Swatantra Bharat, published in
Lucknow.12
It was also alleged by some persons that the role of the PAC in these riots
was different, much more extreme in its anti-Muslim behavior than in pre-
vious riots. One Muslim AMU professor whom I interviewed on January 3,
1991, commented in this respect that, in contrast to previous riots in Aligarh,
in this one "the philosophy and the naked aggression of the PAC was very
much marked."13
 120/ Riots of December 1990 and January 1991
Although much is known about this riot, much is also in dispute. As is
almost invariably the case, especially in large-scale riots such as this one, there
are different versions of how the rioting began. Even the PUCL team that vis-
ited the city within a few weeks of the rioting, between January 4 and 6 and
again on January 12,1991, remarked in its report, "It is futile to try to find out
the exact starting point of the riots in view of the conflicting versions
given ... both by Hindus and Muslims." I visited Aligarh myself on January
1-3, 1991, and heard similarly conflicting versions.
The PUCL report takes the position that the atmosphere had been so "pol-
luted" by the mobilizing activities and techniques used in Aligarh, as else-
where, to mobilize Hindus in connection with the rath yatra and the kar seva
to be done at Ayodhya that it is not only "futile," but "pointless to ask the
question: who started the riots and to find out the starting point." In other
words, given the polluted atmosphere, a riot was bound to happen and it hardly
matters exactly how it started. However, it matters greatly to those on oppo-
site sides of the communal divide, who seek to convince themselves, their
followers, and especially the rest of public opinion outside Aligarh that it was
the other side that began the riots, which would not have occurred or would
not have been so uncontrollable had it not been started in a particular way
by the other side.
Despite the caveats expressed by the PUCL about the contradictory nature
of different versions of the origins of the riots, there is some convergence in
their summary of the two leading ones, which are quoted below.

One version is that after Friday prayer at a Mosque at Upperkot on 7.12.90

[December 7, 1990I when provocative speeches (by way of sermons) were deliv-
ered, some Muslim.. attacked two PAC jawans near the Police Station. In retali-
ation PAC kiUed a number of Muslims and in some other areas Hindus attacked
Muslims.
The other version is that a bomb was hurled at a Mosque in Sarai Sultani
on 7.12.90, and as a result fighting started between Muslims and Hindus at Sarai
Sultani. The sound of the bomb blast was easily audible at Upperkot Mosque
(Jama Masjid). Muslims gathered in front of Kotwali [police stationl to protest,
and there a section of the mob tried to snatch rifles from some PAC jawans.
PAC then opened fire kiUing 3 Muslims and injuring many.

In the first, Hindu version, all blame is placed on Muslims: their religious
leaders delivered provocative sermons, Muslims attacked the PAC, the PAC
merely retaliated. In the second, Muslim version, Hindus are presumed to
 Riots of December 1990 and January 1991 /121
have started the riots by throwing a bomb at a mosque not far from the Jama
Masjid itself, Muslims protested peacefully in front of the police station, some
persons tried to grab the rifles from some PAC men, and the PAC then
"opened fire." Although the precipitating event differs in these two accounts-
provocative sermons in the mosque at Upar Kot or a bomb thrown at a
mosque in Sarai Sultani-they converge in agreement that the critical event
was the confrontation between a Muslim crowd and the PAC at the police
station just opposite the principal mosque of Aligarh, where the PAC opened
fire and killed several Muslims!4 They also converge in agreeing that pro-
vocative action on the Muslim side-the attempt to grab the rifles of PAC
jawans-precipitated the first fatalities.
From this point on, there is no convergence between Hindu and Muslim
versions. For Muslims, the riots then became a series of attacks launched by
the PAC and Hindu mobs, sometimes in collusion, on Muslims and their prop-
erty in many areas of the old city, extending out to the Civil Lines area near
the university as well. For Hindus, the riots constituted spontaneous outrage
over the news, generally believed by the Hindus of the town, that Hindu
patients at the AMU hospital had been massacred, in effect a Hindu response
to a classic blood libel charge against Muslims, comparable to the alleged effect
in European history upon Christians who attacked Jews in countless pogroms,
allegedly in response to blood libel charges against Jews.
The PUCL team, whose members included four Hindus and one Muslim,
placed the weight of its report upon the aggressive behavior of both the PAC
and the Hindu mobs, giving the following examples. In Sarai Sultani, a Muslim
locality "surrounded by Hindu localities," as Muslims "gathered in the
Mosque for Friday prayer," they "were alarmed when they saw hundreds of
people on roof tops of Hindu houses throwing bombs at Muslims and their
houses and at the Mosque." Although a PAC force was on the scene, its jawans
not only failed "to control the rioters," they "provided a cover" for them, allow-
ing them to continue their attacks while firing on any Muslims who dared to
emerge from their houses. Some Muslims who, the PUCL report says, "here
were fairly well organized," "threw bombs or similar articles at the PAC who
then opened fire to kill."
In the adjacent Muslim locality of Jogipara (not on maps), the PUCL report
states that "17 Muslims were killed" on December 8, 1990. An FIR (First
Information Report) was filed by some person or persons concerning these
deaths, reporting that the victims were burned to death. These 17 persons were
not included in the official death toll because the police did not recover any
dead bodies there. Six of these 17 persons, said to have been witnesses to the
 122 / Riots of December 1990 and January 1991
killing of the other 11, were thrown into a factory furnace "used for melting
metaL"
The PUCL team visited several other Muslim riot-affected areas, from
which "more or less the same picture" emerged, "namely PAC resorting to
firing to kill indiscriminately, Hindus being helped by PAC cover, Mosques
attacked, houses destroyed." All the colonies so affected were "surrounded
by Hindu areas." The team was also "given graphic accounts of how PAC
entered some houses and shot people dead." Although the report places the
heaviest responsibility upon the PAC and Hindu mobs, it also notes that, in
some localities, Muslims attacked Hindus, notably in one with a large
Scheduled Caste population, where Muslims attacked, burnt, and looted Hari-
jan houses. As Hindu mobs, some of whom came "from nearby villages,"
extended the rioting to the Civil Lines localities adjacent to the university with
large Muslim populations, "a Muslim mob" also attacked a number of "estab-
lishments ... belonging to Hindus."
All the above events occurred on December 7 and 8, before the Gomti
Express killings and the false Aaj report on the massacre of Hindus in the
AMU hospital. After the latter two events, the PUCL report notes that riot-
ing spread to many other areas of Aligarh, but provides no further details.
Nor has any official report ever been published to document the events
described by the PUCL and those that it could not cover. Despite the huge
scale of this riot, no enquiry commission was appointed.
Insofar as the Gomti Express killings are concerned, a judicial enquiry was
ordered, but its findings were never published. We must rely again, there-
fore, upon the PUCL account. According to the latter, "a Hindu mob of about
1000" at first tried but failed "to stop the Kalka Mail" in the morning, but
"succeeded in stopping the Gomti Express (very near the Station) and killed
several passengers." The PUeL report found definite evidence of "human fail-
ure," as well as grounds for suspecting the collusion of Hindu railway officials
and workers at the scene in allowing the mob to stop the train and gain access
to the passenger compartments to do their work of killing.
Insofar as the rioting in the city is concerned, the PUCL report was unequiv-
ocal in its condemnation of the actions of the PAC.

There is no doubt that the PAC killed a large number of Muslims. On some
occasions Muslims may have acted in a provocative manner by way of trying
to snatch riiles, going out during curfew hours, etc. but the PAC retaliated with
disproportionate brutality.
The part played by the PAC in the Aligarh riots is indeed reprehensible.
 Riots of December 1990 and January 1991 /123
Almost every where [sic] Muslims told us that at least as many persons of their
community were killed by the PAC as by Hindus. In several places we were
shown the high terraces from which PAC personnel fired at fleeing Muslims.
Even young Muslim children were thus killed by the PAC. At many places PAC
fired at Muslims when neither they (the PAC) nor Hindus were attacked. The
PAC acted as a highly communalised force.
It may be noted that PAC's bonhomie with Hindus made Muslims angry.
For example Hindus fraternised with PAC personnel, offered them sweets, gar-
landed them; and if any of them declined to accept their offers Hindus raised
slogans and exhorted them to listen to the call of Hindutva and Ram Bhakt
[devotion to the god Ram). It would indeed require tremendous self-control
on the part of an average Muslim to tolerate such Hindu fraternisation with
the police. 15

Such charges have been made by Muslims against the PAC in numerous riots
allover the state of U.P. for the past several decades. When official person-
nel are asked to comment on such accusations, they rarely support the charges
against the PAC fully. Such was the case when the PUCL team spoke to the
new district magistrate (DM) who arrived on the scene towards the end of
the riots to replace the previous DM who failed to anticipate or control them.
Such has been the case also in all my interviews with officials over the past
decades. The PAC, it is said, is a tough, well-trained force that is brought out
to handle the most difficult situations and often faces provocative behavior
from crowds, such as the alleged rifle-snatching incident. Its responses are,
therefore, either justifiable or understandable.
There are, however, at least two strong reasons for accepting the Muslim
point of view that the PAC is, in effect, an anti-Muslim force as essentially
accurate. The first is that these charges have been repeated so often in so many
situations in so many different parts of the state, including countless eyewitness
testimonies and reports of impartial teams such as that of the PUCL, that the
weight of the evidence is too strong to discount. Second, the Muslim charges
are not always directed against other armed forces, that is, the civil police,
the CRPF, and the army. Muslims sometimes, and more frequently in recent
years, also accuse the local police of partiality, but they never blame either
the CRPF or the army. That was the case in Aligarh in 1990-91 as well. In fact,
Muslim leaders beg for the intervention of the army to displace the PAC when-
ever and wherever the latter force is deployed in their localities to control major
civil disturbances.
It is noteworthy also that the PUCL report found few credible witnesses
 124 / Riots of December 1990 and January 1991
among Hindus they interviewed "from various walks of life." They found that
the only matter that concerned Hindus in Aligarh during and after the riots
was "the alleged killing in the Medical College Hospital as reported in AAJ
[sic] and other Hindi Dailies." When Hindus were asked to provide instances
to support their claims that Hindu "properties had been "destroyed and burnt"
and "that Muslims had acquired a lot of arms and weapons," nothing con-
crete or substantial emerged. The team found no "evidence of Hindu tem-
ples having been damaged or attacked except a slight touch on a small temple
in Jamalpur." However, they "found ample evidence of Hindus having
attacked Mosques. Also, [they] found ample evidence of Hindus using abu-
sive language about Muslims and Islam.... The most foul and obscene lan-
guage was used in slogans written on walls. We were sad to observe that
communal hatred against Muslims has assumed pathological dimension."
Finally, the team pursued several specific charges made by Hindus that patients
whose names were provided to them had been "killed in the Hospital." After
extensive inquiries, the team concluded that all the charges were false.

THE FALSE RUMORS

OF THE MEDICAL COLLEGE HOSPITAL MURDERS

Thanks to the work of the PUCL, there is ample documentation concerning

the origins and consequences of the false reporting of murders at the AMU
Medical College Hospital. False rumors are central in the spread of many riots
everywhere in the world. They are usually attributed, like riots themselves,
to the credulity of the masses, already inflamed by prejudices against another
group and ready to gather in large crowds to take revenge for the actions falsely
attributed to that group or for real, but trivial, actions enlarged upon in the
rumor process. In other words, they are part of the myth of spontaneity that
has for long provided the dominant explanation of riotous behavior. The ori-
gin and spread of this rumor in Aligarh suggests the opposite conclusion and
provides further evidence of the existence of institutionalized riot networks
in which rumormongering and rumormongers play important roles.
It deserves to be especially emphasized that these riot-provoking networks
are not confined to the locality in which the riots occur. The outside vernacular
press in particular, many of whose journalists and editors are themselves sym-
pathetic to militant Hindu ideology, often play significant roles in spreading
rumors and interpreting incidents in a manner whose effect is to inflame
Hindu passions.16
As in most other aspects of research on riots, it is necessary to carefully
 Riots of December 1990 and January 1991 /125

distinguish self-evident truths from speculative linkages, cause from conse-

quence. The first matter that has to be considered is the timing of the news
report. The accepted date for the start of the rioting is December 7,1990. On
December 8, a mob of Hindus attacked the Gomti Express and butchered at
least 4 and possibly 10 or 15 Muslim passengers. On December 10, the false
news reports of the Medical College Hospital massacres appeared, completely
overshadowing all other reporting in the Hindi press on the riots, especially
the true report of the deliberate massacre of uninvolved Muslims targeted
by a Hindu mob that set out deliberately for the train station to carry out its
murderous acts. This sequence presents us with a speculative possibility that
the false news reports were deliberately planted in the Hindi press to distract
attention from the premeditated, obviously organized actions of the Hindu
mob. We have no way of knowing that this was the case, but it was certainly
the effect.
It is also certain that the Hindi-language press, its journalists and editors,
took a deliberate decision to publish news that they either knew was false or
that they took no trouble to verify, and that they knowingly published false
or unverified reports with front-page headlines-in the case of Aaj, "a fuH-
page banner headline on the front page that 74 people including 28 patients"
had been killed, in the case of Amar Ujala, 50 persons killed, and in the case
of Dainik Jagaran, 124 persons proclaimed dead. These "reports were filed
on the basis of accounts of two witnesses, Mahesh and Satish Aggarwal [sic] "
who allegedly reported to the district magistrate that they had escaped with
their lives from the hospital with six other men. I7 If they did in fact go to the
district magistrate with their false story, he did not believe it, for he contra-
dicted the news reports after they were published.
The newspaper owners certainly knew the consequences of publishing such
false reports. Their actions, therefore, raise the suspicion that these news-
papers maliciously sought to throw wood on a burning fire. Moreover, the
three leading newspapers that published these malicious fabrications were
important newspapers. Aaj is a Hindi newspaper with a hallowed national-
ist reputation going back to pre-Independence days. Dainik Jagaran andAmar
Ujala. which also published the false news, are circulated widely throughout
northern India. It is evident, therefore, that these three newspapers and their
journalists and editors fed into the institutionalized riot systems that exist in
north India. In short, they bear a large measure of responsibility for the addi-
tional rioting and deaths that followed in the wake of their reports.
One should also note those who took responsible actions in response to
the newspaper reports. They included the district magistrate and senior super-
 126/ Riots of December 1990 and January 1991

intendent of police, Aligarh, several Hindu members of the faculty of the

Medical College, and the national media, including especially the English-
language press and the official Hindi TV network, Doordarshan. The author-
ities and persons from the national press and broadcasting networks
investigated the reports, found them baseless, and announced their falsity,
while Hindu faculty members had the courage to deny publicly lies spread
by and believed by many, if not most, other members of the Hindu com-
munity living in Aligarh. 18 Yet, the reports continued in Aaj and other papers
even afterwards. 19
It is also evident that the effect, if not the purpose, of the false news reports
was to place the AMU, which had nothing to do with these riots, at the cen-
ter of them. The RSS and other militant Hindu organizations have consis-
tently argued over the years that the AMU is a great center of conspiratorial
activity against the Hindu nation, a hotbed of Muslim communalist and anti-
national activity, a place where Muslim criminals abide and are protected by
the university authorities, and the like. This propaganda is widely believed
by Hindus in the city. It is, therefore, not farfetched to accept the argument
of the Medical College Hospital authorities, published in the midst of the tur-
moil occasioned by the riots and the false reports, as cited below.

To escape from that reckoning [concerning the killings on the Gomti Express J
the stories about the AMU Hospital were concocted. This was the plan to divert
the attention of the nation, to hide the real massacre that was allowed to hap-
pen during the curfew, and to force the AMU to be on the defensive-first,
aghast at the stupidity of the charges and, then, busy in giving denials and
explanations. In all the resultant hue and cry and denial about the Medical
College, nobody remembered the unfortunate Gomti Express. And, it was
intended to be 80. 20

Once again, whatever the intent, the effect was such. The PUCL team found
in their interviews with Hindus in the city that "almost all Hindus" whom
they met "were only concerned with and agitated over the alleged killing in
the Medical College Hospital." They blamed the AMU "and wanted action
to be taken against the University." 21
There is usually some kernel, if not of truth, at least of circumstance that
gives occasion for the spreading of false rumors. In this case, there were two
circumstances that fed the rumor. One was that injured riot victims were
brought to the hospital on the 7th when the riot began and on the 8th in the
 Riots of December 1990 and January 19911127
aftermath of the Gomti Express killings. Most of the injured brought to the
hospital were Muslims, but there were a few Hindus among them as well.
Second, Muslim mobs, presumably agitated over the killing of Muslim pas-
sengers on the Gomti Express, did collect outside the hospital on those two
days and riot; they engaged in stabbing, killing, and looting of Hindus and
Hindu-owned shops. One of the persons stabbed was a Hindu, who was
brought to the hospital for treatment and survived. 22
However, these facts notwithstanding, the PUCL team noted that several
Hindus associated with trade and professional associations in the city, teach-
ers at the degree colleges, and businessmen and industrialists either them-
selves visited the hospital and the few Hindu patients who were being
attended there or had access to reports from Hindus who had done so after
the false news reports were published and were, therefore, in a position to
know the truth. These included "representatives of Vyapar Mandal, faculty
members oflocal colleges, members of the Bar, and senior citizens." The team
also met with a respected "local industrialist," Promod [sic] Kumar, "who
was one of the first to visit the hospital after the news was published." 23
The first question that must be asked is why certain people belonging to
particular organizations and associations either rushed to the hospital upon
the publication of the false news reports or in other ways took a special inter-
est in the situation there. The Vyapar MandaI is an association of traders dom-
inated by persons of the Barahseni (Varshney) caste, among whom most are
RSS members or sympathizers.24 Why should a trade association in the city
be preoccupied with the alleged murders of Hindus at a hospital in the Civil
Lines area-across the tracks from their businesses in the city, where large-
scale riotous activity was in progress? Not only did the Vyapar MandaI take
an interest in the alleged happenings at the AMU Hospital, but it was their
leadership who produced the "witnesses," the brothers Mahesh and Satish
Agarwal, one a general merchant, the other a cloth merchant, and both
"important members ofVyapar Mandal."25 The fact that they were merchants
could explain the interest of the leaders of the trade association in their story,
but it does not explain why they materially assisted them in spreading the
false stories.
The presence of the Agarwal brothers on the scene is adequately explained
by the fact that their mother was admitted as a patient in the hospital on
December 7 and they were concerned for her welfare, especially at a time when
a riotous Muslim mob was active outside the hospital. After seeing to their
mother's admission to the hospital, according to the account they gave to the
 128/ Riots of December 1990 and January 1991

pueL team, they set out for home, were attacked by a Muslim mob, and
"returned to the Hospital," where "they saw injured people coming in the
Hospital and students shouting slogans." They came again to the Hospital
on December 8 and witnessed similar scenes. They also heard stories of people
having being stabbed in the vicinity. Again, on the 9th, they came to the hos-
pital, witnessed a "person being stabbed in front of the Emergency section,"
and "managed to get away from the Hospital ... along with 6 others." The
brothers now "spoke to leaders of the Vyapar MandaI" and went to meet the
"District Magistrate and Police officials" in the company of "Shri Navman,
SIP M.L.A. During their meetings with officials," according to the pueL
report, "there were angry exchanges between Shri Navman and Police
officials." After this meeting, the brothers were taken to the hospital; their
mother was removed from the hospital and "transported to their residence
in [the1District Magistrate's car" on December 11.
While the concern of the brothers Agarwal for their mother and their anger
over the riotous activity outside the hospital are understandable, especially
if it is true that one of them was attacked by a Muslim mob, they personally
observed that the situation inside the hospital was in no way a danger to the
patients. The pueL report sums up their account as follows.

This account of these two important members of Vyapar Mandai makers] it

quite clear that they had witnessed no untoward happenings in wards. They
saw none being killed in wards; they saw one person being stabbed outside the
Emergency section; they heard but did not see themselves that 12/13 persons
were stabbed outside the Hospital. [They say 1that their brother was attacked
by a Muslim mob but not injured; they have not reported [the matter J to the
police. [They observed) that Hospital authorities did not allow anyone from
the mob t.o enter the Hospital building.26

It is understandable that any Hindus of the city whose relatives were under
treatment in the hospital would be fearful for the safety of their loved ones,
given the circumstances outside the hospital, and would even want to have
the persons removed to a safer setting. It might not have been sufficiently
reassuring for such persons to have noted that many of the medical practi-
tioners in the hospital were themselves Hindus. Furthermore, Hindus visit-
ing their relatives in the hospital during those days endangered their own lives
by doing so. The wife of one Hindu patient was murdered when she left the
hospital to get some medicine for him. There were, however, no grounds for
believing that patients in the hospital were being killed and that the AMU
 Riots of December 1990 and January 1991 / 129

was somehow responsible for their deaths. It is evident that the PUCL team
did not believe the account of the brothers Agarwal. The entry of the Vyapar
MandaI onto the scene and, even more so, the presence of K. K. Navman,
brought the RSS and the BJP centrally into the picture, politicizing it and com-
munalizing it.
The brothers Agarwal were not the only "witnesses" to the hospital "mur-
ders.» Two Hindu women were admitted to the hospital on December 7, each
of whom died on the following two days. Their husbands were later produced
by"Vyapar [Mlandalleaders and faculty members ofD.S. College, Aligarh"
as false witnesses to the killing of their wives in the hospital. The names of
the "faculty members of local colleges" with whom the team discussed the
above situation and other circumstances surrounding the spreading of false
rumors about the situation in the hospital are not given, but many of the fac-
ulty of the local degree colleges in Aligarh are also RSS members and sym-
pathizers, among whom are several who are classed as riot-mongers by the
district administration.
All in all, after extensive prodding of interested Hindu leaders from such
organizations as the Vyapar MandaI, the local colleges, and others to provide
the names of persons allegedly killed in the hospital, the PUCL team was given
the names of four persons, two of whom were found to have died of "natu-
ral causes," while the other two were alive. Yet such people continued to talk
about 28 patients who had been "killed in Medical College Hospital." The
PUCL team concluded its report with the flat statement that, in fact, "there
was no killing of patients in the Hospital" [emphasis in original]. Further, it
provided testimony from Hindu patients that they were extremely well
treated, that, in the account of one of them, there was "no discrimination
between Muslims and Hindu patients," and that, during the riots, "the doc-
tors and other staff members worked throughout nights because of the large
number of patients coming to the Hospital from 7th December onwards."
In short, the PUCL team discovered what any conscientious journalists them-
selves could have discovered had they been at all interested in the truth of
the matter.
Two prominent Hindus, Promod Kumar, an "industrialist," and Ashok
Chauhan, who visited the hospital to assess the situation there, are named
and are characterized in the PUCL report as honest and truthful persons. Why,
then, did they not publicly contradict the reports that had been published in
the newspapers when they discovered their falsity, as indeed they did?
These two named persons came to the hospital certain that the news reports
were true. The PUCL team members were present and asked them to pro-
 130/ Riots of December 1990 and January 1991

vide the names of the persons they believed to have been killed. At first, they
refused, saying it was up to "the Hospital authorities to prove that the news
was baseless and that none was killed." After considerable persuasion, they
did produce some names and also visited several hospital wards where they
had been told that people had been killed. They found two of the allegedly
killed patients alive in the hospital. Promod Kumar was then asked by the
hospital doctors to immediately issue a statement contradicting the false sto-
ries of murder in the hospital. He replied that he "wanted to go about the
matter in a detailed and thorough manner so that he could issue a compre-
hensive statement in order to allay public misgivings caused by the mischie-
vous and baseless news." His refusal to issue a statement immediately led to
an altercation with the doctors. No statement was ever issued. The PUCL team
felt that the doctors had made a mistake, that if Promod Kumar and Ashok
Chauhan had been allowed to do things their way, they might have been able
to persuade "the press and Hindus of Aligarh ... to tone down their mis-
conceived reaction." However, it is equally plausible that they might never
have issued the desired statement or have issued it too late to influence the
course of the riots and save lives. We cannot know what might have happened
in a counterfactual situation. 27
What we can know is that Promod Kumar, by his own admission, felt con-
strained from issuing an immediate statement because, as he told the PUCL
team later, "he had to face such formidable forces like the RSS (the Vyapar
Mandai, as we could see, is controlled by RSS) and Bajrangdal." In other words,
his statement either was not meant for the general public, but for those organi-
zations in the city that keep Hindu animosities against the AMU alive and
that are active on the Hindu side during riots, or, if it was meant for the gen-
eral public, these same organizations would become angry with him. In a word,
these organizations and/or many members of them are intertwined with the
institutionalized riot system of Aligarh: they produce the rumors or spread
them, if they do not themselves produce them; they either want to believe
the rumors themselves or want the Hindu public to believe them; they want
the public mobilized and they want revenge-if not for the acts of murder
that never took place, then for the continued existence of the AMU and what
it represents to them. The PUCL team expressed a similar view.
Moreover, once the riots had gained momentum and the Hindu slaugh-
ter of innocent Muslims on the Gomti Express had taken place, it clearly also
suited the purposes of riot organizers to distract attention from those events
and to focus blame upon the Muslims and the AMU for the entire riot. These
 Riots of December 1990 and January 1991 /131
remarks constitute my own interpretation of these events based upon what
I have seen and heard over 38 years in Aligarh City. They are also consistent
with the PUCL team report, as expressed in the statement below.

Rumours spread, and baseless news items published in Hindu papers have dou-
ble significance-they tarnish the image of the Hospital and AMU, and at the
same time they are communal in character in that they are anti-Muslim. 28

They do not, however, constitute evidence that would stand up in a court of

law. On the basis of this interpretation, however, I do not believe that the
two gentlemen, Promod Kumar and Ashok Chauhan, would ever have been
allowed or been able to issue a mollifying public statement; they would have
been castigated by the riot-mongers in their community.
It is also quite likely that, even if a mollifying statement had been issued
by the two persons in question, it would not have been believed by the Hindu
public. More important, the riot activists would not have been persuaded and
would not have cared. It is, I believe, a mistake to think that a riot is pro-
duced by rumors and stopped by falsification of them. This is a fallacy that is
part of the myth that riots are produced by spontaneous mass action inflamed
by such things as false rumors. These rumors are merely precipitants that may
or may not lead to the activation of the riot network, depending upon other
factors, usually political ones, as in this very case.
Moreover, the false news was publicly contradicted by "newspapers from
Delhi" and by the government ofU.P., which published an advertisement in
one of the irresponsible newspapers, Amar Ujala, directly contradicting the
false news reports. The advertisement was published in the December 18, 1990
issue. 29 The rioting continued throughout the rest of the month.
 5 I The Control of Communal Conflict in Aligarh

WHY WERE THERE NO RIOTS IN ALIGARH IN THE

DECADE BETWEEN 1978-80 AND 1988-90?

F or a decade there was relative peace in the town of Aligarh. No inci-

dents classed as riots occurred during this period. Yet, during this same
decade, in the nearby district of Meerut, two of the most vicious com-
munal riots in this state since Independence occurred in 1982 and 1987, respec-
tively. Why were there no riots in Aligarh in this period? Answers given to
this question also lead naturally to answers to the obverse question: why did
the worst riots in Aligarh's history break out in Aligarh in December 1990,
after a decade of relative peace towards the end of which two much less severe
riots occurred, in October 1988 and on November 10, 1989?
When I visited Aligarh in 1983, I did not feel that a peace that would last
for several more years had returned to the town. In fact, I did not feel that
there was peace, only a kind of hiatus enforced by heightened police vigi-
lance, marked by the continuous presence of armed police forces at many
major four-way street crossings, or chaurahas, as they are called in Hindi (see
Figure 3.5). I passed through Aligarh several times between 1983 and 1990-91,
before and after parliamentary elections. I was present in Aligarh the day the
votes were counted in November 1984, when the Congress under Rajiv
Gandhi won its greatest landslide victory since Independence, returning the
Congress to a position of dominance in this district, as elsewhere, that it had
not enjoyed in many years.
As an afterthought, in some of my interviews with people in January 1991
concerning the riots that had just occurred-indeed, that were still in
progress-I asked some respondents how they explained the absence oflarge-
scale riots in Aligarh between those of 1978-80 and the present ones. I heard

132
 The Control of Communal Conflict in Aligarh /133

four explanations for the relative peace of those years: economic, political,
administrative, and relating to police action. I discussed these several expla-
nations with an AMU professor, a Hindu, whose political sympathies were
with the left and secular parties. Although he mentioned first the economic
explanation, he gave primary importance to the political one, and some impor-
tance to the administrative one, but discounted police action except insofar
as it came under the head of "'good administration.» 1
The economic explanation was quite simple, though its implications are
not: there had been no riots in the intervening years because the "business
community" did not want further riots after the long series that disrupted
the economic life of the city in 1978-79. I have heard similar statements made
in other contexts after major riots, in response to questions from me con-
cerning whether or not there would be further riots at a particular site; I have
been told that, no, there would not be for a long time because people in busi-
ness do not want them. What is not simple about this kind of statement is
the implication that must lie behind it, that businesspeople play critical roles
in riot production, either as instigators or as financers of them, and that riots
cannot, therefore, be produced if they do not at least finance them.
It was dear that, for this respondent, the business role was necessary, but
not primary. The primary factor in riot production was, for him, political.
In the intervening years, he noted, the Congress was dominant and the "BIP
was not a force here." Riots, he said, arose out of "political necessity." Only
if political party leaders feel that a riot is necessary for their advancement,
particularly for their electoral success, can a riot take place, "but if they think
that there is no necessity for [a) riot, then there will be no riot." In the course
of elaborating on this explanation, this respondent articulated a set of elec-
toral calculations for each of the main political parties in competition for
national, state, and local power at that time, which he used to illustrate why
it was in the particular interest of the BJP, but not of other parties, that there
be riots in Aligarh and elsewhere at this time.
The core of his arguments was that the Ayodhya movement had lifted
the BIP both in Aligarh and in U.P. to a position from which it could hope
to increase its gains in the elections anticipated after the resignation of the
Janata Dal government of Prime Minister Vishwanath Pratap Singh in
October 1991. The V. P. Singh government's resignation was precipitated by
the withdrawal of the BIP's support to it in Parliament after the arrest of
L. K. Advani in Bihar on October 23, 1990. The BJP, which had electoral
adjustments with the Janata Dal in the 1989 elections, was now determined
to fight the coming general elections on its own. The rath yatra ofL. K. Advani
 134 / The Control of Communal Conflict in Aligarh

was, clearly, a calculated move to begin the building of electoral support in

anticipation of the fall of the central government, and to counter the
advantage that the Janata Dal hoped to gain from the announcement in
August 1990 of its intention to implement the MandaI Commission rec-
ommendations for reservation of 27 percent of all posts in the Union gov-
ernment sector for backward castes. The B}p, in effect, sought to counter
the electoral advantage the Janata Dal hoped to gain through dividing
Hindus, by seeking instead to consolidate Hindu sentiment and voting
around the demand to remove the mosque at Ayodhya and build a new tem-
ple to Ram.
In Aligarh, the Congress had already lost support in the 1989 U.P. Legislative
Assembly elections, in which Krishna Navrnan had won the city seat on the
BJP ticket. (See Chapter 9.) Navrnan himself was accused of having played
a major role in the instigation of the riots in Aligarh, along with his son,
who was the chief of the Bajrang Dal in the city. Navman senior, himself a
big businessman, had the support of the vyaparis (traders) of the town, who,
my informant stated, were "united behind him." Because of Navman's sup-
port among the business community and also because, my informant
alleged, some Muslim businessmen had been "coming up" in the city, chal-
lenging the dominance of Hindus in local trade, the (Hindu) business com-
munity was no longer opposed to riots, and the relative peace of a decade
was broken.
After listening to this explanation, I said that I had heard an entirely dif-
ferent one, namely, that the senior superintendent of police who had brought
an end to the rioting in Aligarh ten years earlier had accomplished his task
by executing extrajudicially all the known Muslim criminals who had been
responsible for so much of the killing and property destruction in those riots.
This respondent acknowledged that such executions had taken place, but said
that the SSP had not discriminated between Hindu and Muslim criminals
and had killed both, though there may in fact have been more Muslim goon-
das in those days than Hindus. Nor did he deny the important role that goondas
played in communal riots. However, he reiterated his emphasis on the politi-
cal factors in relation to criminal participation even more than with business
participation. Without any "political force" behind riots, he argued, goondas
could not "make" a riot. The goondas might be in the forefront, they could
certainly be purchased, but if "there is no political force behind them, they
can't do [it]."
Having minimized the importance of criminals in producing riots, and
by the same token, minimizing the police action that led to the elimination
 The Control of Communal Conflict in Aligarh /135
of a large number of them, this respondent then proceeded to eliminate police
action entirely from his explanation of why there had been no riots in Aligarh
for a decade. Instead, he added an alternative third factor, which he called
"good administration.» Good administration to him implied two things. The
first was prompt and effective action in anticipation of riots at times of polit-
ical turmoil and agitation, such as had occurred in Aligarh during the mobi-
lizations and demonstrations that accompanied the rath yatra of L. K. Advani.
This respondent and many others criticized the district magistrate for fail-
ing to take such action in Aligarh between August and December 1990. The
second aspect of good administration, in his mind, was specific to Aligarh,
namely, the establishment of cordial relations of cooperation between the
city administration and the administration of the university. Such coopera-
tion had existed during the decade of relative peace, but had broken down
at the onset of the riots in December 1990.
The PUeL report, published after my inteview with this respondent,
confirmed his views on the failings of the district administration both in pre-
venting and in dealing effectively with the 1990-91 riots. That report com-
mented on the ineffectiveness of the district administration as follows.

Riots in Aligarh did not start all of sudden. The seeds of riots had already been
sown, not onJy in Aligarh, but throughout U.P.; only the District Administration
did not take notice. And significantly enough a blue-print to meet riot situa-
tions was also circulated quite sometime [sic] ago and it was in the District
Magistrate's office. It appears that the then District Magistrate was benignly
ignorant about it. ...
Given the background, the polluted and totally communalised atmosphere,
near break-down of the law and order machinery, and increasingly scant respect
for rule of law, communal riots were almost inevitable. The Administration
knew, or ought to have known, that riots were in the offing. And yet killings
went on unchecked during the period 7-10 Dec. 90. Hardly any administra-
tion existed in Aligarh during this period. The District Magistrate Mr. Verma
and some other officials were transferred (this was the only punishment for
erring officials); and a new District Magistrate Mr. Misra took over on 10-12-90
[December 10 J. Many people we met said that given the debris left by his pre-
decessor and the story that appeared in AAJ on the very day he took over charge,
resulting in increased violence, the new District Magistrate did a fairly good
job. Even though riots were widespread upto [sic] about 15-12-90 [December
151 and Muslims were at [the J receiving end, Mr. Misra was able to instil some
confidence in the victims.
 136/ The Control of Communal Conflict in Aligarh

Two assumptions behind the PUCL report deserve to be especially noted.

The first is that riots do not start spontaneously. "Seeds" must be sown. In
the months preceding the first major outbreaks, those seeds were sown in the
actions and demonstrations associated with the rath yatra and the kar seva
to Ayodhya. The nature of those actions in Aligarh was such as to "pollute"
and communalize the atmosphere. Second is the assumption that the local
administration has the capacity to prevent riots or, at the very least, to pre-
vent killings from going on "unchecked.» This assumption is strong enough
in the PUCL reports for its authors to imply that the officials transferred
because of their ineffectiveness should have been punished more severely than
by a mere transfer. The assumption is also evident in the mild commenda-
tion given to the new district magistrate, Mr. Misra, for having done his job
of checking the spread of the riots under difficult conditions "fairly" well.
Finally, the assumption is also evident in one of the final recommendations
in the report for the prevention of such riots in the future: "Local adminis-
tration must be instructed to act promptly and impartially at the very emer-
gence of a communal riot situation."
The last statement, however, introduces a further question, namely,
whether the district administration was ineffective because it, too, was com-
munalized and acted partially. This charge has been made repeatedly in U.P.
wherever riots have occurred. It has been made with increasing frequency in
recent years. The PUCL report does not suggest that the district administra-
tion was anti-Muslim, though it does imply that the previous district mag-
istrate, Mr. Varma, lacked the confidence of Muslims, since it commends the
new district magistrate for having instilled confidence among them.
Before the publication of the PUCL report, I also asked our AMU pro-
fessor about the role of the district administration in the Aligarh riots. In his
view, the previous district magistrate and SSP had been unrealistic. They had
known many days before the outbreak of rioting that "the situation [was)
very dangerous." However, the DM assured people like the professor, who
visited him and "presented a memorandum to him" to express their concerns,
that the administration knew about it and would control it. This respondent
said that the DM said there was no need for curfew, and that he held a news
conference on December 1St in which he announced that, should there be
any incidents in any mohalla, the miscreants would be immediately punished.
However, the riots did not start with isolated incidents in one or two mohal-
las; "they started in ten places simultaneously, in all the Civil Lines and [the]
whole city, [on the] 7th and 8th and 9th." However, this respondent did not
 The Control of Communal Conflict in Aligarh 1137

consider that Mr. Verma was "a communalist" for failing to prevent such a
large-scale outbreak, in which mostly Muslims were killed; he was only
"unimaginative" and "incompetent."
However, Muslims in Aligarh did not accept this limited view of the for-
mer DM's failings. For example, the mufti shahar of Aligarh, who is not given
to extreme statements, said in an interview that the "mentality of the [ex-l
DM was very bad." He thought that the "mentality" of the senior superin-
tendent of police was not so bad, he was merely incapable of handling the
situation. However, he faulted both the former DM and the additional dis-
trict magistrate (ADM), City, and gave the following example of their men-
tality as well as their incompetence. He said that rioting began in the Muslim
locality of Shah lamal (no. 69, Map 4) on the night of the very day on which
he had spoken on the telephone to the DM, who had informed him that he
was imposing curfew in Aligarh not because of "any danger," but as "a pre-
cautionary measure." Then, when Hindus attacked the Shah Jamal locality,
the ADM informed him that four Muslims had attacked Hindus in that local-
ity, which was false. The mufti also claimed that the DM and other officers
had gone round the localities of Aligarh warning Muslims that they would
"create another Maliana, another Hashimpura in Aligarh also, as the PAC
had done in Meerut district."2

THE CONTROL OF COMMUNAL CONFLICT

IN ALIGARH IN THE 19908

Control of Rioting in December 1992

Rioting was reported from Aligarh in the aftermath of the destru{.iion of the
mosque at Ayodhya on December 6, 1992, but no details were given in the
press reports, which focused more on the major riots in the state (Kanpur
and Varanasi) and in other parts of the country, especially in Bombay. On
this occasion, however, Aligarh remained for the most part under control.
Aligarh was listed among thirteen towns in U.P. that were placed under cur-
few on December 7, the day after the destruction of the mosque. 3 On the fol-
lowing day, Aligarh was listed among cities and towns in U.P. where deaths
had occurred, but no figures were given. It was also reported that more than
nine hundred people had been arrested in cities and towns throughout the
state, of which the largest number, "about 400," were in Aligarh. 4 It appears
that, on this occasion, the large number of arrests in Aligarh reflected as much
or more preventive measures, on the part of a district administration which
 138/ The Control of Communal Conflict in Aligarh

wanted no repetition of the previous year's catastrophic violence, rather than

the seriousness of the rioting in the city.
Why, then, were there no riots in Aligarh on this occasion, when numer-
ous other cities and towns in U.P. and other parts of India, most notably
Bombay, experienced severe rioting-indeed, a classic "wave" of riots com-
parable to those that have occurred at previous major junctures in modern
Indian history, notably in 1946-47? A Times of India journalist appeared to
be anticipating rioting in Aligarh in his report of December 12, in which he
noted that "tension in this city was fast reaching a flash point. " However, the
tension, which this reporter detected "in the Muslim-dominated areas of the
city," was "suddenly released" when the announcement came that the BJP
government had been dismissed, President's Rule had been imposed on the
state, and the prime minister had made an "impassioned appeal to the
nation." But the prime minister's appeal had little noticeable effect elsewhere,
where rioting continued through the month and into January as well.
Moreover, though the tension was said to have been relieved in the Muslim
areas of the city, there was considerable agitation on the AMU campus, where,
"around midnight," "several thousand slogan-shouting, emotionally charged
students had gathered ... and were trying to take out a procession on the main
streets. " This dangerous situation was also" defused" by the appeals of "some
student leaders and officials," but was repeated again "around noon the next
day," when the "highly charged" president of the AMU Students Union
remarked that "things have reached a boiling point" and that he did not know
how he could "control the situation." Once again, the students were "per-
suaded to give vent to their feelings by courting arrest," a time-honored
Gandhian political practice of expressing protest without provoking violence,
of which "more than 300 students" took advantage. Then "several thousand
students" gathered on the campus next day and passed a resolution, another
peaceful form of protest, declaring that their "faith in the country's Con-
stitution [had] been shattered."5 The AMU staff association also convened
to express peacefully "its deep sense of shock, indignation and anger" over
the destruction of the mosque and called for "the resignation of the Central
government," while the vice-chancellor wrote to the president of India to con-
vey the "sense of insecurity, uncertainty and anger" that prevailed among
Indian Muslims. 6
Such meetings, courting of arrest, and protests emanating from the AMU
campus continued throughout the month. The Students Union "demanded
that action should be taken against all those responsible for demolishing the
mosque under the terrorist act," failing which '''the culprits of the Ayodhya
 The Control of Communal Conflict in Aligarh /139

tragedy' will provoke and face the wrath of the Muslim community." Yet, at
the same time, the Students Union president claimed that the union had "been
working hard to ensure communal harmony in Aligarh and [had] seen to it
that Hindu students do not feel afraid or persecuted in the Campus."7 AMU
students also travelled to Delhi for a protest rally in defiance of an order ban-
ning such assemblies, courted arrest, and were released without any reported
violence. 8 In the midst of these mostly peaceful actions, a repetition of the
May 1979 beatings of AMU students on the train to Delhi occurred, in which
fifteen students of the university returning to their homes for vacation "were
thrown out of the running train 'by Kar sevaks' returning from Ayodhya."
Further, two students "had been hammered as well," one of whom died in
hospital from his head injuries, while a second had "lost part of his mem-
ory."9 Even in the face of this provocation, which had led to rioting by AMU
students in May 1979, the Students Union responded only by complaining
that the incident had "not been investigated."lo
What were the factors, then, that prevented rioting in Aligarh in December
199z? Timely and firm admini~trative action has already been noted. The above
summary of the actions on the AMU campus also suggests that great efforts
were made by students, staff, and the vice-chancellor to maintain calm and
confine actions to peaceful protests. A third explanation was provided by "Mr
Gyan Chand Varshney, general secretary of the Udyog Vyapaar MandaI," who
was reported to have said, "The people here were so traumatised by the 1990
riots that they were not willing to fall prey to senseless violence again. The
economy of this city was almost completely shattered in 1990 and it had taken
us nearly two years to limp back to normalcy." In other words, the people in
general did not want riots again aft.er the experience of 1990, and the busi-
ness community, the mainstay of the BJP and RSS in Aligarh, did not want
disruption of its business again. In the same article, a "local journalist"
remarked that the people in general were "simply fed up with communal riots
here." A further explanation for the lack of rioting in Aligarh at this time was
that the BJP chief minister of the state, Kalyan Singh, who himself had played
a critically important role in making possible the destruction of the mosque
at Ayodhya, did not want riots in Aligarh and "issued dear directives to pre-
vent communal riots in the district and they were implemented meticulously"
after the destruction of the mosque. The mufti shahar of the city (Figure 5.1)
also issued an appeal to the Muslims of the city for restraint. l l Finally, it was
also reported that "mohaUah peace committees" had "been activated at Aligarh
well in time."12
This welter of explanations for the absence of rioting provides a mirror
 140 / The Control of Communal Conflict in Aligarh

FIG. 5.1. Mufti shahar, Aligarh, Upar Kot, December 1984

image of those that are given to explain rioting when it does occur, and reveals
the same assumptions. Those who say that "the people" were fed up with riot-
ing must assume that "the people" are responsible when rioting does occur;
similarly with regard to the business community. Those who say that appeals
from political and religious leaders to the people to show restraint were
effective must assume both that, absent such appeals, the people will act spon-
taneously to express their rage violently and that, in the past, the politicians
and religious leaders have been remiss in not issuing them. Those who praise
administrative action and political decision for preventing rioting must
assume that rioting can be prevented when firm political directions from above
are issued and followed with firm administrative action in the localities.
But all the explanations given above have left out of account the factors
that I have argued have been critical to riot production in Aligarh in the past.
The local BJP and RSS leaders did not seek to produce a riot there in December
1992. For one thing, their leaders were in Ayodhya participating in the destruc-
tion of the mosque or in cheering the demolition gangs. Further, in contrast
to the situation in other cities and towns where Muslim protests turned vio-
 The Control of Communal Conflict in Aligarh /141

lent, providing a pretext for massacres of Muslims by Hindu mobs later-

notably in Bombay and in Kanpur 13-Muslims in the city and at the AMU either
were prevented from engaging in violent action by firm administrative action
or were restrained by the appeals of the mufti. The AMU students also were
similarly held back. Riots in Aligarh had been produced previously either by
police-Muslim confrontations or by confrontations between Hindu and
Muslim crowds; the latter have always been the more serious. Had Muslims
in Aligarh not shown restraint on this occasion, it i" likely-if not certain-
that the BJP and RSS network would have been activated and further killings
such as those that occurred in 1990-91 would have taken place.

The Riot of March 1995

On March 10, 1995, the Times of India reported from Lucknow that "group
clashes" (meaning Hindu-Muslim) had occurred in Aligarh during which 3
persons were killed and 13 injured. "Twenty companies of security forces" as
well as the new Rapid Action Force (RAF) had been deployed to control the
situation. 14 In another news item from Aligarh, datelined the same day, the
newspaper reported that curfew had been imposed in the city and quoted
"official sources" that the violence had broken out in the locality of Sarai
Sultani, but they were not able to provide a clear statement of the reasons
for, nor even the precipitants that led to, the outbreak:

There are two versions of why the dashes took place. According to some people,
the violence erupted following an incident of eve-teasing [harassment of a
female], while others put it down to a quarrel between a shopkeeper and a
customer.
However, within minutes there was a heavy exchange of fire and brick-
batting between members of the two communities.

This brief description suggests the working of the Fearon/Lahin spiral equi-
librium in this instance. However, other events followed thereafter that do
not conform to that pattern. One small but fatal incident occurred on the
next day, March 11, when a married couple were returning from Aligarh to
their home village and were attacked by a man with a knife. The wife later
died from the injuries sustained during the attack. The same report referred
to the recovery of three bodies of persons who were said to have been "killed
in separate incidents," but the localities in which these incidents occurred
were not mentioned. An additional twelve companies of "police and para-
 142 / The Control of Communal Conflict in Aligarh

military forces" were deployed. "Indefinite curfew" already imposed "in the
five police station areas of the old city would continue." The same report also
referred to another incident in which a major catastrophe was averted
through the courageous action of a lone woman. 1;

Senior police officers said a major incident was averted yesterday evening,
when a bus carrying a marriage party of the minority community was sur-
rounded by rioters in the Achal Tank area. The members of the party, includ-
ing women, were forced to come out from the bus and one of them was stabbed
to death. A woman, Mrs Mithilesh Yadav, who was watching the scene from
the window of her house, rushed to the rescue of the female members of the
marriage party. Mrs Yadav and her husband Surendra Mohan Yadav, a local
advocate, succeeded in turning away the rioters and gave shelter to the entire
marriage party in their house.
Other members of the locality maintained a night-long vigil and early this
morning the police escorted the marriage party to safety. Prominent members
belonging to both communities have praised the woman's courage.

There are two features of this report that especially deserve notice. The first,
of course, is that we have here a documented case, for which I have no sim-
ilar reports in the history of post-independence rioting in Aligarh, of civic
action at the local level that prevented imminent death and destruction. The
second is the reported site of this incident, Achal Tank (Achal Talab or Achal
Sarowar in Hindi), an area not listed as among the riot-prone localities in
the city, but one which is the center of Hindu religious activity in Aligarh as
well as of militant Hindu presence (see Figure 2.1 and Map 2).
The last news report on this riot came on March 15, when it was said that
"normalcy" was returning to the old city and day curfew was being removed.
In the meantime, the brave woman received effusive praise, nomination for
a presidential gallantry award, and for a state government award as well.
Moreover, praise for the woman came from both sides of the communal
divide: from the president of the AMU Students Union and from the state
secretary (Gyan Chand Varshney) of the Uttar Pradesh Vyapar Mandal (Uttar
Pradesh Traders' Association).16
In addition to the news reports on this riot, we have the firsthand obser-
vations in its aftermath of Asghar Ali Engineer. Engineer, as noted above, has
been a leading proponent of the view that communal riots in India are most
often planned, with the direct involvement of local politicians, and that eco-
nomic rivalries also come into play. However, in this case and in a few other
 The Control of Communal Conflict in Aligarh /143

riots that occurred in other parts of India at this time, Engineer argued that
they were not the same as previous riots. On the contrary, he argued that they
were not "a result of either offensive propaganda or planning by the Hindutva
forces," but were "a result of spontaneous outbursts of violence on the part
of the people of these towns." At the same time, he attributed the very spon-
taneity of these riots to the "sustained communal propaganda at a very high
pitch for close to a decade," which had "created a deep chasm between the
majority and minority communities" such that "any trivial incident can lead
to serious communal incidents. "17 Engineer was here referring, of course, to
the decade-long Ram Janmabhoomi movement that culminated in the
destruction of the Babri Masjid on December 6, 1992.
Engineer noted, as had the Times of India, that there were conflicting ver-
sions concerning the precipitating incident from which rioting ensued, but
he was able to track its source to "a scuffle between a Hindu and a Muslim
trader" in the Madar Gate area near Sarai Sultani (Map 2). Although he con-
sidered this riot to be spontaneous, he noted a different view held by Muslims
living in Sarai Sultani, namely, that "communal riots are engineered to force
them to sell their properties and go away." 18 Engineer also confirmed the news-
paper reports that, "as soon as the trouble began ... a big crowd from both
the sides gathered and resorted to stone throwing.... Private fire-arms were
used from both the sides." Although local people interviewed by him made
the common charge that the local police did nothing to stop the rioting and
instead allowed it to spread, all also said that "this riot was not planned nor
was it provoked by the BJP."
With regard to the near-disaster visited upon the Muslim marriage party,
Engineer reported that this incident took place near the D.S. College and that
the attackers were thought "likely to be students" of that college. 19 The attack-
ers used bicycle chains and iron rods, one of which was used to kill the bride-
groom's brother. zo
Except for the initial police failure to act promptly and decisively with the
first outbreak of violence, administrative and police action thereafter was firm,
decisive, and politically coordinated. The state chief minister, Mulayam Singh
Yadav, the Samajwadi Party leader and the principal opponent of the BJP in
this state, had adopted a policy to deal with communal violence that was
imposed immediately in Aligarh. The DM and SSP, having failed to control
the riots within 24 hours, were immediately replaced with a new team, sub-
stantial police, paramilitary, and military forces were deployed, and the riot-
ing was effectively contained. 21
I believe that Engineer is correct in distinguishing this riot from previous
 144/ The Control of Communal Conflict in Aligarh

riots in Aligarh, but the distinction that he draws between planned and
unplanned or spontaneous rioting is too sharp. Moreover, his explanation
for the spontaneous character of this riot introduces the political context as
a kind of background element, namely, the militant Hindu movement of the
previous decade that created a Hindu-Muslim chasm. My argument differs
from Engineer's in the following respects. First, the Hindu-Muslim chasm
was not created by the movement of the past decade, only intensified by it.
Second, the spontaneous aspects of this rioting were located in areas of the
city in which the Hindutva elements, as well as elements in the Muslim com-
munity, are always alert and active and are always ready to rush to the scene
of an incident to defend their brethren and to counter the actions of crowds
from the other community. The areas in which this violence was centered
are adjacent localities. One, Sarai Sultani, which will be discussed in detail
below, is predominantly Muslim, surrounded by Hindu localities in which
militant Hindu organizations are active. The second, Achal Talab, is located
in the midst of Hindu and Jain religious and educational institutions. It is
here that the local Hindu degree colleges are located and from which Hindu
students have been mobilized in so many of the disturbances and major riots
discussed in this and the preceding chapter, including the March 1995 inci-
dents. We are here, therefore, at the center of Aligarh's institutionalized riot
system.
Absent in this case was a political context into which riotous activity could
be placed and that would justifY its expansion. Rather than use the term spon-
taneous to describe this riot, it would be better to describe it as unanticipated.
It was not preceded by mobilizations of crowds for local electoral purposes,
or in mass meetings connected with statewide or national issues or on issues
connected with the AMU, or even a wrestling match or a cricket match. What
appear as spontaneity are, rather, conditions and practices that extend
throughout both communities that allow for the instantaneous reaction of
persons from both communities to an incident, a rumor, a provocation from
the other side, to rush to the scene and to mobilize others for attack and defense.
It is, therefore, appropriate to use the FearonfLaitin term spiral equilibrium
for this kind of riot, an unfortunately infelicitous term tl1at nevertheless sug-
gests not preplanning, but a predictable outcome in communities ever alert
for provocations from the other side and ever ready to be mobilized.
Such riots are more easily contained than others that we have examined
above because the political mechanisms for expansion that inhibit the local
administration from acting against political persons who have influence with
the state authorities are not brought into play. Absent also was divided polit-
 The Control of Communal Conflict in Aligarh /145

ical authority. Although there was a Congress government in New Delhi and
an SP government in the state, the central government had no reason or
authority to interfere in a situation where the state government had made it
clear that it would not tolerate communal violence and would hold every DM
and SSP accountable for any riotous activity that occurred on their watch.
Thus, once again, we see clearly the elements of riot containment and con-
trol in situations where riots expand and where they do not. In broad terms,
the elements are threefold: (1) presence or absence, activation or inactivation
of a preexisting institutionalized riot system; (2) presence or absence of a con-
text of electoral competition, political mobilization, or other context in which
symbols of communal opposition are present; (3) presence or absence of polit-
ical and administrative coordination of the forces available to the authori-
ties for immediate suppression of violence.

Riot Prevention in Dahi Wali Gali

During a visit to Aligarh in November 1997, I was informed that there was
an ongoing dispute between Hindus and Muslims over a religious site in a
locality in the old city known as Dahi Wali Gali. This locality, which derives
its name from the production of dahi, or curd, that takes place along the gali,
or alley, is as crowded a qasbah lane as one can possibly imagine. Here, as in
many lanes and alleys in the old city, population has grown to such an extent-
as well as the number of vehicles-that one cannot move on foot in a straight
line in any direction. One has to weave and dodge constantly to avoid jostling
others and to prevent oneself from being hit by vehicles of every description.
I was taken to this gali by an AMU professor, a Muslim, to visit the dis-
puted site, where a mandir exists adjacent to a Muslim mazar, or saint's tomb,
which Hindus claim is temple property and where they wish to install addi-
tional idols. There is also a mosque across the way from which the call to prayer
could be heard as we were touring the gali. This dispute had been simmer-
ing for seventeen years (as of 1997), while a civil case concerning it had been
in the courts. Throughout these seventeen years, the PAC had been posted
here to prevent an outbreak of violence in connection with the dispute. In
fact, I saw also posted here a huge Rapid Action Force truck. The Rapid A<.:tion
Force is a relatively new force that has been brought into use in such riot-
prone areas only in the 19908. It is a "state-of-the-art" riot-prevention force.
Its huge truck is said to be provided not just with men and arms but with
medical facilities to care for the injured in case of bloodshed. 2 2-
I had a very clear sense, in my interviews in Dahi WaH Gali and with the
 146/ The Control of Communal Conflict in Aligarh

district officials at the time, concerning two aspects of this situation. First,
obviously, this situation is a generic type in India. Its counterpart exists in
thousands of sites all over the subcontinent. It is a very dangerous situation
that, by its very nature, could lead to local strife and even violence. It is also
of the type that is easily susceptible to political manipulation. In the absence
of political manipulation, or of the work of other larger forces, however, such
potential conflict can be limited to the specific site. Second, my impression
was that the district administration in Aligarh at the time was very much aware
of the potential for an outbreak of violence at this site, was on top of the sit-
uation' and was determined to ensure that no significant violence did in fact
break out. My impressions, therefore, are consistent with the analysis given
above concerning the factors that promote and contain large-scale rioting.
In the case of Dahi Wali Gali, however, we confirm those conclusions from
an opposite direction. In the March 1995 riots, we saw a "spontaneous" or
unanticipated riot, precipitated by a trivial, everyday quarrel, brought under
control. In Dahi Wali Gali, we see a continuing local dispute in which the
feelings of members of both communities are involved, in which both sides
remain always alert for transgression by the other side, but in which the same
mechanisms operate to prevent its expansion.
Much more will be said in the four chapters in Part IV concerning the rela-
tionship between riots and political competition. Before turning to that sub-
ject, however, I want to consider as well the extent to which other factors that
have been cited in my interviews above playa part in the production of Hindu-
Muslim violence in Aligarh. I will consider first the demography and geog-
raphy of violence in Aligarh City, paying special attention to the spatial location
and communal composition of the various mohallas of the city where riotous
violence has often been centered. Second, I will consider the economic fac-
tors commonly cited as either primary or secondary causes in the produc-
tion of violence. I will then return in some detail in Part IV to an analysis of
the political and electoral context in which communal violence occurs in
Aligarh.
 PART III

Demographic, Social,
and Economic Factors
in the Production of Riots
 6 I The Geography and Demography of Riots

INTRODUCTION

C ollective violence such as riots rarely-indeed almost never-engulfs

an entire town or city of any substantial size. Nor does it ever include
all elements in the city's social organi:r.ation, whether defined by class,
caste, religion, or other cultural community. This obvious and elementary
character of collective violence is usually ignored in studies that identify the
sites of violence as the cities or towns in which they occur. The literature com-
monly refers to racial or religious riots as having occurred in Detroit or Chi-
cago or Newark or, in our case, in Bombay or Hyderabad or Aligarh. Paired
comparison and ecological studies almost invariably also take the social, eco-
nomic, and demographic characteristics of the entire city or town as the bases
for comparison. Rare is the study, Michael Keith again providing a stunning
exception, that identifies precisely the site within the urban area in which vio-
lence occurs and within which nobody living within its boundaries could
escape noticing, if not participating in, the action. Keith's study goes even
further and takes us down to the street and the community centers in which
the action occurs. He describes not just that action occurring when the vio-
lence breaks out, but the action that is continuous and that lays the basis for
the explosive violence that follows. l Only in that way can the dynamics of riot
production be uncovered.
In this chapter, I will first identify the broad areas of the city in which riotous
violence is concentrated. I will then identify specific areas which have repeat-
edly been sites of violent outbreaks, and specify the types of demographic,
social, and economic environments in which they occur. In the second chap-
ter in this part, I will consider the economic bases for communal conflict and
violence in Aligarh.

149
 150 / The Geography and Demography of Riots

GEOGRAPHY AND DEMOGRAPHY

There are several geographical and demographic features of Aligarh City that
are of great importance in understanding political change and communal
relations there since Independence. The first is the fundamental division cre-
ated during colonial rule between the Civil Lines area and the old city. The
Civil Lines is a chara(.ieristic British creation that existed in most of the towns
and cities where there was a significant presence of British administrators.
These areas were cantonments for British armed forces and posh residential
areas for its administrative officers, usually created out of rural space on the
outskirts of existing towns. In Aligarh, in addition-and ultimately of greater
importance-the Civil Lines provided the site for the establishment of the
Aligarh Muslim University (AMU) and, with its expansion, for the residential
bungalows of its faculty and staff.
Largely because of the presence of the AMU and its predominantly
Muslim faculty in the Civil Lines, the population of the Civil Lines ward was
more than 50 percent Muslim in 1951. Insofar as the university community
itself is concerned, the vast majority of whom live in the Civil Lines, 86 to 88
percent are Muslim today.2
The division between the Civil Lines and the old city is sharply marked,
physically, emotionally, demographically, culturally, and in lifestyle. Physically,
the Civil Lines is separated from the old city by tlle railroad line and the
Ramghat Road, which meet to form a triangle comprising the northeastern
sections of the city, and by the flyover (viaduct) at Katllphula, over which
one travels to reach the old city (see Map 1),3 Most of the land lying to the
northeast of the railway line, "except the Railway Colony and the Indian res-
idential mohallas on the eastern side"4 is included in the Civil Lines. Lelyveld,
in his reconstruction of the life and times of Aligarh at the founding of the
AMU, remarked on the contrast existing even then between the old city and
Civil Lines with its newly founded college as follows: "Nothing could be greater
than the contrast between the chaos of KoH's [Aligarh's] winding streets and
the deliberate order of the Aligarh College."5 Hirt described the Civil Lines
area in 1955 as follows.

It is characterized by large, widely spaced upper class residences and adminis-

trative and judicial buildings arranged along wide, predominantly straight
streets. The Muslim University occupies a large tract ofland in the northwestern
part of the area. 6
 The Geography and Demography of Riots / 151

~foIonY
Med~ollege
•

Gular R o " ' - - - 4 l

New and newly incorporated mohallas

MAP 1. Aligarh City

 152 / The Geography and Demography of Riots
In more recent times, especially in the last decade or so, new business money
and old aristocratic money have been used in the Civil Lines to build fantas-
tic houses-huge in their dimensions, with spacious lawns, sometimes
grotesque in architecture and color-for members of former landlord fam-
ilies and businessmen from the old city who have amassed fortunes from their
commercial and industrial activities. (See Figure 6.1.) Many of these houses
are located on the relatively wide Marris Road.
The division is also an emotional one. Most people living in the Civil Lines
never travel across to the other side, while most of those living in the old city
would have no reason to do so. Some people do cross from the Civil Lines
to the old city to shop and some people from the city side cross over to the
Civil Lines to engage in meiniallabor. Recently, also, as just noted, those who
amass sufficient wealth may move out of the filthy slums on the city side to
spacious new bungalows on the Civil Lines side.
A further emotional division arises out of the presence of the AMU in
the Civil Lines and the attitudes towards it that exist on the other side. The
importance of the AMU as a symbolic presence in the town-a living lieu de
memoire,7 conveying a multiplicity of strong emotional meanings in the
present-has already been noted and will be discussed further below. For the
moment, it is sufficient to note that, despite its emotional significance, there
is little actual contact between the AMU, its faculty and staff, and the popu-
lation in the city. In fact, it was the original intention of the founder of the
university to wall in the students, in the manner of an English cloistered uni-
versity, to provide them with a peaceful, sequestered atmosphere to pursue
their studies and to foster a communal life inside the campus walls separate
from that of the outside world. 8
However, as noted in Chapters 3 and 4, there have been dramatic occasions
when crossovers have taken place in politics and in the enactment of com-
munal violence. Faculty from the university have at times, most notably in
1962, entered the Legislative Assembly contest for Aligarh City. Hindu stu-
dents in the university have sometimes gone to the city to recruit fellow Hindu
youths to redress insults and grievances they have claimed to suffer in AMU
student politics at the hands of Muslim students. Cadres from militant Hindu
organizations, student crowds from the degree colleges in the city, and gangs
of rowdies and criminals have for other reasons also from time to time crossed
the boundary to create violent disturbances around the university. At times,
the university hostels have became havens for criminals from the city. Partly
for such reasons, the walls originally established to provide a communal, col-
legiate atmosphere have recently been raised to block off the university still
 The Geography and Demography of Riots /153

FIG. 6.1. Posh new house in the Civil Lines area, March 1999

further, but now in order to wall off the university grounds from unwanted
intruders.
The second demographic feature of great importance in understanding
politics and communal violence in Aligarh is the fact that the bulk of the pop-
ulation of the city and most of the Muslim population lives south-southwest
of the railway line in enormously congested mohallas. Although the area occu-
pied by the Civil Lines is larger than that of the rest of the city south of the
railway line, the population distribution is highly skewed in the opposite direc-
tion. In 1951, only 12 percent of the population of Aligarh lived in the spa-
cious Civil Lines area.
The major population concentration in Aligarh is in a section of the city
known as Upar Kot (Upper Fort) (Map 2), although there is no ward or other
census division designated by that name. At the highest site in Upar Kot stands
the Jama Masjid (Friday Mosque; Maps 1 and 2 and Figure 6.2), the principal
mosque for the Muslim population of the old city. In this respect, too, there
is a distinc.:t division between the old city and the Civil Lines, in this case among
Muslims, for the AMU has it~ own Jama Masjid on the university campus (Figure
6.3), which makes it unnecessary for faculty, staff, and students ever to visit the
other. In the old city itself, however, the more significant fact is that the mosque
and the Muslim-majority mohaUas around it are "located in the centre ... and
are surrounded by Hindu majority and mixed [Hindu-Muslimllocalities."9
1541 The Geography and Demography of Riots

MAP 2. Upar Kot

 The Geography and Demography of Riots / 155

FIG. 6.2. Mosque, Upar Kot

A third demographic feature of great political importance, especially in

recent years, has been a substantial expansion of the population of the city
into outlying areas on both sides of the railway line, including the Civil Lines
areas. This expansion has included the incorporation of former villages and
their transformation into urban mohallas, the creation of new "colonies" or
abadis (as new urban settlements are often called in India), and the migra-
tion of persons from rural Aligarh in to these areas. Many of these new mohal-
las are not included in the 1951 census nor are there any data on them in later
censuses, which use artificial agglomerations rather than living neighborhoods
as their units. In many cases, however, I have been able to count or estimate
the proportion of Muslims and Hindus and others from recent voters' lists
(Table 6.3) and to place them on the maps.
Many of the new mohallas (see Map 1) are populated entirely or predom-
inantly by Muslims. They include areas to the north and nortlleast such as
Jamalpur, Jamalpur Mati (not on map), Jamalpur Nagla (village), Badamnagarl
Hamdardnagar, and Jiwangarh (Map 1). On the southern outskirts of the city
lies predominantly Muslim Bhojpur. To the west is Shah Jamal (no. 69, Map
1), for which no census information is available, but which is overwhelmingly
Muslim, adjacent to a locally famous dargah (no. 70, Map 1) and to the Idgah
(Muslim festival celebration site). To the northwest, between the railway line
 156/ The Geography and Demography of Riots

FIG. 6.3. Mosque, Aligarh Muslim University campus, 1962

and the university, almost at the outer boundary of the city near the old Aligarh
Fort, is another cluster of new mohallas, of which the largest is Firdous Nagar.
Another mohalla in this duster, attesting to its newness and its Muslim char-
acter, is Maulana Azad Nagar, named after the famous Congress nationalist
Muslim leader; two others are Shahinshabad and New Abadi (New Settlement).
There are also several newly incorporated areas where the Hindu popula-
tion has expanded, some of which are entirely populated by Hindus (includ-
ing Scheduled Castes and others). These lie mostly to the east and southeast.
One such area comprises Pala Sahibabad, Nagla Pala Sahibabad, and the older
locality of Mahendranagar (number 49, Map 1). This grouping is populated
entirely by Hindus and others. A second newly populated area, Chhavani,
lies to the north of the previous grouping; its population consists entirely of
Hindus, Scheduled Castes, and others, with no Muslims.
Several of these areas will feature prominently in the discussion to follow
later in this chapter concerning the extension of riot sites in the great riots
of 1990-91. They will also feature in the electoral analysis to follow in Part IV.
To anticipate briefly, the political significance of these newly settled and incor-
porated areas lies in the fact that the predominantly Muslim areas have been
included in the boundaries of the Aligarh City Legislative Assembly con-
stituency, while several of both the older and newer overwhelmingly Hindu
 The Geography and Demography of Riots / 157

areas have been excluded, leading to a substantial increase in the percentage

of Muslim voters in the constituency as a whole that has, in recent elections,
been transformed as a result into a Muslim-majority constituency. However,
in the municipality of Aligarh, and consequently in the corporation elections,
the predominantly Hindu localities are included, with the result that Hindus
(including Scheduled Castes and others) retain a majority therein.

WARDS AND MOHALLAS

The Wards of Aligarh

The 1951 census divided the city into eight wards and 241 mohallas. Since that
census, however, the number of census wards has been increased several times
while the mohallas have been replaced by "enumerator's blocks" as the pri-
mary census units. As far as I have been able to determine, there is no corre-
spondence between these blocks, listed by serial number only, with the living
neighborhoods of the city. Nor does there appear to be any correspondence
between the political division of the city into wards and the census division.
We are, therefore, compelled to use the 1951 census as the basic source for the
information of most concern in this book, namely, the caste/communal com-
position of the population. The data from that census have, however, been
checked against voters' lists in my possession for later periods, to detect any
significant discrepancies in the proportions of persons in the Wards and mohal-
las by religion, and to provide relevant information on the communal com-
position of new neighborhoods.
Table 6.1 below provides the distribution of the city population wardwise
by religion and caste from the 1951 census. Three wards then had a Muslim
majority: Civil Lines, Kanwariganj, and Turkman Gate. However, the bulk
of the Muslim population (76 percent) in the city as a whole was concen-
trated in the more heavily and densely populated wards of Kanwariganj and
Turkman Gate in the old city. The Hindu population, the majority in five of
the eight wards, was more evenly distributed throughout the city.
Politically, as indicated in Chapter 2, the Aligarh municipality is now
divided into 60 wards that encompass the original 241 mohallas as well as sev-
eral newly incorporated areas. The correspondence between the old and new
wards as well as the distribution of municipal corporators by new ward num-
ber are shown in Table 6.2. The table shows that the BIP in 1995 dominated
completely all the wards in Achal Talab, which, in 1951, had a Hindu popu-
lation above 80 percent. The BJP is also the predominant party in all the other
Hindu-majority areas, namely, Raghubirpuri, Mamubhanja, Shahpara, and
 TABLE 6.1. Population of Aligarh City wards, 1951

Muslims Scheduled Castes Others Total Population

No. Name No. Percent No. Percent No. Percent No. Percent U

Civil Lines 8,461 50.35 1,027 6.11 7,315 43.53 16,803 100.00
2 Achal Talab 540 3.52 2,401 15.67 12,381 80.81 15,322 100.00
3 Raghubirpuri 6,601 34.44 2,497 13.03 10,071 52.54 19,169 100.00
4 Mamubhanja 4,273 25.17 1,803 10.62 10,898 64.20 16,974 100.00
5 Shahpara 3,401 17.53 1,877 9.68 14,118 72.79 19,396 100.00
6 Jaiganj 4,462 29.54 2,340 15049 8,303 54.97 15,105 100.00
7 Kanwariganj 14,030 60.22 644 2.76 8,622 37.01 23,296 100.00
8 Turkman Gate 7,128 52.05 1,562 llo4l 5,004 36.54 13,694 100.00
TOTAL 48,896 34.99 14,151 10.12 76,712 54.89 139,759 100.00

'All percentages rOllnded to 100.00.

 TABU 6.2. Correspondence between 1951 and 1995 wards
and the distribution of corporators by new ward number

Old Ward New Ward

Name a Numbers EIP ESP SP Ind.

Civil Lines 9,31,39,40,43,44, 9,39,40,44,51 31,43,47, 52

47,49,51,52,57 49,57
Achal Talab 1, 11, 15, 17,22, 1, 11, 15, 17,22,
32,33,34,42,60 32,33,34,42,60
Raghubirpuri 12,14,16,21,27, 12, 14, 16, 50 27,37
37,39,46,50 21,39,46
Mamubhanja 1, 10,27,39,58,59 1,39,58,59 10 27
Shahpara 5,6,12,13,35,45,55 5,6,12,13 55 35,45
Jaiganj 4,5,6,19,23, 4,5,6,25,60 19,23,24,55
24,25,55,60
Kanwariganj 20, 39, 54, 55, 20,39,58,59 55,56 54
56,58,59
Turkman Gate 2,3,8,9, 12, 14, 9, 12, 14, 16 2,3,19,38,56 54 8
16,19,38,54,56

"There is some overlap in the case of several wards that lie on the boundaries between the old wards.
 160 I The Geography and Demography of Riots

Jaiganj. In contrast, in areas where the Muslim population is above 50 per-

cent, namely in the old Civil Lines, Turkman Gate, and Kanwariganj wards,
the SP, BSP, and independent corporators are well represented, though, even
in Kanwariganj, the BJP has stronger representation than other political forces.

The mohallas of Aligarh

Many of the mohallas of Aligarh have names that suggest their past histori-
cal function, their present function, or the dominance or prominence of par-
ticular castes or baradaris in their past or present. At the extremities of the
inner part of the old city are a series of "gates" or, in Hindi, darwazas, which
were probably at one time the entry ways into what must have been a walled
city.lO These gates are also the names of the mohallas around them. Most of
them are situated on or near main roads that take one out of the city. Forty-
four of the mohallas of Aligarh bear the name Sarai, reflecting the fact that
they originated as caravan series for travellers in earlier times. l l Nine of the
Aligarh mohallas carry the Urdu word bazar, which, of course, implies exactly
what it says, namely, a market for the sale of goods. A few carry the Hindi
word mandi in their name, which means the same thing, though it more com-
monly refers to a wholesale market. For example, the main vegetable market
in Aligarh, as in every city in north India, is called Sabzi Mandi.
Sensitive, Riot-Hit, and Crime-Prone Mohallas. It was shown in Chapter 3
that there has been an increase in the intensity of riots in Aligarh since
Independence by the measure of the number of deaths in those five-year peri-
ods in which there were riots with deaths. (Refer to Table 3.1 and Figure 3.1.)
A further feature of Aligarh riots has been an increase in their locational spread.
Figure 6.4 selects from Table 3.1 the seven riots in which there have been five
deaths or more since 1961. It is immediately apparent from the figure that there
is an extremely close relationship between tlle intensity of riots measured by
deaths and their spatial spread. Indeed, tlle correlation coefficient between
number of deaths and number of localities affected is .97, giving an R2 of 94
percent. Further, although there have been fluctuations in the intensity and
spread of riots, there has been a long-term trend towards spatial spread, notable
first in 1978 and then dramatically so in 1990-91. Although the number of deaths
in 1961 and in 1971 reached 15 and 17, respectively, by official count, most of
the killing and other riotous activity occurred in four locations. In 1961, rioting
began on the A..\1U campus and spread to the Shamshad Market (Map 1) on
the main arterial, the Anupshahr Road, that adjoins it, after whicll rioting tllen
occurred for utterly different reasons in two localities in the old city, namely,
 The Geography and Demography of Riots / 161
100

80

60

40

!; 20 Deaths
.c
S
Z'" 0 Sites
1961 1971 1978

Riot

FIG. 6.4. Number of deaths and number of sites of riotous activity

in Aligarh riots in which there were five or more deaths

Manik Chauk and Mamubhanja (Map 2). In March 1971, although the riot-
ing was associated in some accounts with the agitation at AMU concerning
the demand for restoration of its minority status, all rioting occurred in the
old city, particularly in the four localities of Achal Talab, Babri Mandi, Phul
Chauraha, and Sabzi Mandi. (See Appendix Table A.2 and Map 2.)
It is necessary to pause here for a moment to note particular features of
these localities in the old city. Manik Chauk and Mamubhanja are mohallas
dominated by the Varshney caste, centers of militant Hindu and anti-Muslim
hostilities. Achal Talab (Achal Sarowar on Map 2) contains the most impor-
tant Hindu temple in Aligarh (Figure 2.1) as well as three degree colleges on
whose faculty are several leading RSS organizers and whose students are mobi-
lized in all large-scale riots for processions and attacks against Muslims and
their property. Indeed, although Achal Talab is not mentioned as a site of
riotous activity, in 1961 and in most later riots, excluding 1995, it is prima-
rily because Hindu rioters move out from this locality to march, protest, and
attack elsewhere in the old city. Babri Mandi (Map 2), in contrast, is a local-
ity dominated by the Muslim baradari of Qureshi, the most numerous of
the Muslim baradaris of Aligarh. Phul Chauraha (Map 2) is at the epicenter
of many riots because it is a major crossing and market, situated between
Hindu and Muslim mohallas. Finally, Sabzi Mandi (Map 2) is another mar-
ket situated at "the intersection of the major bazaar streets."12 In short, the
latter two are focal points where streets, crossings, and bazaars converge,
toward which provocative processions generally proceed at times of Hindu-
 162 / The Geography and Demography of Riots

Muslim confrontation, and where the heavy concentration of shops of all

kinds provides targets.
The first great extension of riotous activity in Aligarh occurred in the riot
months of 1978, when 28 deaths were officially recorded and some 27 mohal-
las were riot-hit. The intensity, duration, and spatial spread of this rioting in
Aligarh was, as we have seen above (Chapter 3), widely commented upon in
the press and by the politicians and led to official and nonofficial post-riot
inquiries. By then, also, the local authorities had identified for their own pur-
poses, presumably for establishing curfew and for determining deployment
of forces, a list of riot-prone mohallas. At the same time, the Minorities
Commission also published in its report a list of "disturbed areas" that it vis-
ited in the course of its inquiries. Finally, an unpublished report by the Centre
for Research in Industrial Development, "Communal Violence," compiled
a list of some 29 mohallas divided into five categories: (communally) sensi-
tive; riot-hit, but not crime-prone; crime-prone, but not riot-hit; riot-hit and
crime-prone; sensitive and crime-prone, but not riot-hit.13
Most of the riotous and other criminal activity of the city takes place in
these 29 mohallas,14 comprising 12 percent of the total number listed in the
1951 census. They are listed in Appendix Table A.l with the available caste!
communal data on their composition. Spatially, all 29 mohallas are located
in the central area of the old city where the Muslim population is most heav-
ily concentrated, especially in the broad area known as Upar Kot (see Maps
2 and 3). The term "Upar Kot" means, simply, "upper fort," referring to the
high ground of the city once occupied by a fort, but now the site of the grand
mosque of the city, the Jama Masjid (Friday mosque). Although the term "Upar
Kot" is often used to refer to the site on which the mosque is located and the
entire area surrounding it, it also refers to a specific mohalla; I will, therefore,
use the term "Upar Kot" to refer to the broad area and the term "Upar Kot
proper" to refer to the specific mohalla where the mosque is actually located.
Upar Kot, as Mann has noted, "is almost entirely populated by Muslims, while
encircling it are mostly Hindu mohallas, with some of mixed Hindu/Muslim
populations. "15 This heavily congested area lies across both the railway line
and the Gurudwara Road, which run parallel to each other and divide the new
from the old city. It also lies to the west of the Grand Trunk Road that joins
the Gurudwara Road and the railway line at Kathphula and the footbridge.
One reaches the old city (see Map 1) from the Civil Lines area in which the
university is located by ascending, at Kathphula, a steep hill that crosses over
the railway line. The Railway Road that brings one up from the new city and
down into the old city ascends again into the heart of the old city, up to Upar
 The Geography and Demography of Riots / 163

Kot proper and the Jama Masjid, which occupies the highest ground. Side by
side, to the east of the mosque, is the kotwali or police station (Map 2), which
occupies a site approximately equal in size to that of the mosque. The two
buildings are separated by a broad street. Both are adjacent to the Railway
Road. The latter road is busy with all kinds of foot and vehicular traffic, but
is fairly broad up to the mosque.
On either side of the Railway Road, lanes extend in every direction into
the qasbah that is old Aligarh, a maze of alleys filled with human, animal,
and vehicular traffic of every description through which one picks one's way,
whether by foot, bicycle rickshaw, or car, with great difficulty and with con-
stant risk of jostling others or being hit by bicycles or other wheeled vehi-
cles. Most of these lanes are unsanitary in the extreme, with open sewage
flowing in channels dOWll the streets, and excreta, animal and human, dot-
ting the lanes and alleys to such an extent that one must keep one's eyes firmly
focused upon the ground and upon each step one takes, while simultane-
ously, through peripheral vision, attempting to avoid being struck by pass-
ing vehicles. It is in these filthy, choked lanes, which also contain subterranean
passages, that most of the criminal and riotous activity and violence are per-
petrated. The authors of the Centre for Research report on riots in Aligarh
remarked that, in these mohallas, "the continuous tension permeating daily
life is nearly tangible, as people go about their business, alert to detect the
slightest disturbance."16
Let us move now to consider two clusters of mohallas that have been
identified repeatedly as centers of riot production and victimization. Moving
south on the Railway Road to where it forks, and following the fork through
the locality of Mamubhanja (Map 2), we reach the famous four-way cross-
ing knOWll as Chauraha Abdul Karim. To the southeast of that crossing is
another famous crossing knoWll as Phul Chauraha (Flower Crossing; Fig-
ure 6.5). Police pickets are posted constantly at both these crossings. To the
north of Phul Chauraha are Sarrafa Bazaar (Jeweller's Market) and Sabzi
Mandi, and further to the north is the infamous mohalla of Manik Chauk.
Manik Chauk is communally mixed, being 76 percent Hindu, 24 percent
Muslim. It is the mohalla of Aligarh most identified with the Varshney com-
munity, which has always provided strong support for the RSS, the Jan Sangh,
and the BJP. There is also a large concentration of Agarwals, the second promi-
nent Hindu trading caste. 17 Manik Chauk is also the home and business site
of Krishna Kumar Navman, the Varshney businessman and BJP politician
considered by the authorities to be one of the principal instigators of com-
munal riots.
 164 / The Geography and Demography of Riots

Hindus
and others
L _ _ _ _ _ ..........................................................._
II
Majority

....................................._ ..•. _ _ _._

Muslim NA
lI
----------------------------~
I

MAP 3. Location and population composition of mohallas officially designated

as communally sensitive, riot-hit, and/or crime-prone
 The Geography and Demography ofRiots / 165

FIG. 6.5. Phul Chauraha

Manik Chauk figured prominently in the 1961 riots, as indicated in the

account below.

A major incident took place [during the 1961 riots] in the Manak Chowk
[sic) locality which lies on the Agra-Aligarh road. There was some dispute regard-
ing property between Lachhman Das, a well-to-do [Hindu] businessman, and
Sharafuddin [a Muslim]. Litigation, so far, had been in favour of Sharafuddin.
Taking advantage of the situation, [Lachhman] Das and his men accompanied
by a mob of students raided the houses of Sharafuddin and other Muslim res-
idents living in the area. One IAS Joint Magistrate who was posted in that dis-
trict on training later told the case-writer that the police [were1present at this
spot but did not make any effort to stop Lachhman Das and others from burn-
ing and looting the houses.... This alone accounted for seven killed and 21
injured, all Muslims. ls

Even in 1974, when the sparking incident that precipitated riots occurred far
from Manik Chauk and the old city, in the university area, there was violence
in this mohalla. 19
Both Manik Chauk and Phui Chauraha were at the center of the riotous
activity of 1978, which began in Phul Chauraha when, after the death of the
Hindu wrestler Bhura at the hands of an assassin, "a crowd of 30 or 40
 166/ The Geography and Demography of Riots

Hindus ... removed the body from the hospital" and "carried it provocatively
through Muslim localities," shouting the Hindi slogan khoon ka badla khoon
se lenge (blood for blood). Rioting predictably broke out "in certain areas of
the town." In Phul Chauraha, "police opened fire in which one Muslim was
killed." The report of the Minorities Commission continues as follows.

The Commission found bullet marks on the walls of houses some of them sit-
uated deep inside the Muslim localities and also on the walls inside the mosque
and on the wooden ceiling of the verandah of the mosque near Phool Chauraha,
indicating that firing was done either from the entrance of the mosque or after
entering the mosque. The Commission also found blood stains inside the houses
of some Muslims which indicated that the Muslims were shot while they were
inside their houses. On the same evening there was large scale arson and loot-
ing of Muslim houses at Manek Chowk [sicJ.20

The report of the Centre for Research also pinpoints Manik Chauk as a major
flash point for the riots both in October 1978 and May-June 1981.

With the phenomenal rise in the price of land in the Aligarh town, the ten-
dency to force the Muslim artisans to vacate their houses and shops located in
the business centres or more developed residential areas dominated by Hindu
Banias has been steadily gaining strength. This tendency manifested itself promi-
nently in the riots of October, 1978, and May-June, 1981. During these riots
killings in Manik Chowk [sic 1and Dahi WaH Gali are known to have been orga-
nized. by the local roughs in league with the land hungry rich of these localities. 21

These two account., of the sources of riotous activity in Manik Chauk provide
different perspectives. In that of the Minorities Commission, the emphasis is
on the deliberate provocation of anti-Muslim feelings by Hindu crowds, who
insisted upon carrying the body of Bhura in procession through Muslim local-
ities, and on the anti-Muslim activities of the PAC, whose forces fired indis-
criminately upon Muslim crowds, as well as inside their houses and mosques.
The emphasis is clearly upon Hindu-Muslim animosities, upon Hindus and
Muslims as collectivities. The Centre report, on the other hand, reduces the
sources of conflict in Manik Chauk and similar areas to the greed of rich busi-
nessmen and to the activities of a specific segment of the Hindu community,
namely, the Varshneys, who use hooligans to harass, intimidate, and kill Mus-
lims to get rid of them in order to gain control of their land and property.
Violette Graff's account of the events at Phul Chauraha and the attack upon
 The Geography and Demography of Riots / 167

Muslims in Manik Chauk in 1978 also emphasizes the role ofVarshney busi-
ness interests, the use of goondas, and the complicity of the PAC. However,
she adds a further dimension, namely, the participation of known militant
Hindu political figures.

The worst carnage ... took place in ... Manak Chowk [sic1, a mohalla where
a small pocket of 10-15 Muslim houses was surrounded by Barahseni (Varshney)
houses.... Two young men were burnt alive. And it was not done by goondas.
Most conspicuous at the head of the assailants was K. K. Navman, the Janata
Chief in Aligarh, and a host of persons "known to be members of the RSS."
Among them, although he was later to demonstrate that there was no evidence
against him, was Dr. B. D. Gupta, a lecturer from AMU who was in the fore-
front in the fight against the Minority Bill. 22

Graff's observations on the riots of October 1978 extend the ring of complicity
so that it now includes Hindu traders, specifically of the Varshney caste, local
criminals in their employ, and local militant Hindu politicians, all protected
by the police and all out to attack Muslim property and religious places and
to kill Muslims.
Graff's account is supported in several respects by my own interviews con-
ducted in 1983 and 1991. Specifically, for example, with regard to the roles
played by known individuals, most especially Navrnan, the following remarks
from one of my interviews are pertinent.

And Navman, since very beginning, has been creating problem like last time,
'78, I know about that incident, what happened in Manik Chauk. He was there,
standing himself and trying to, and directing people to destroy the houses and
to burn the houses of the Muslims. And, he has a group and he has a RSS cadre?3

Graff's report concerning the links between Hindu businessmen and the police
is also supported by Zoya Hasan.24
Further observations on the origins of the 1978 riots and the centrality of
Manik Chauk in them come from Elizabeth Mann, whose account is con-
sistent with the others, but adds yet a further dimension, namely, the suspi-
cion of a direct link between the militant Hindu organization, the RSS, and
the local police. She describes a series of incidents that followed upon the ini-
tial tussle at the wrestling match, after which a Hindu was stabbed here, a
Muslim there, with each side blaming the other. In the midst of these inci-
dents, she notes as follows.
 168/ The Geography and Demography of Riots

There was a function held in mohalla Manak Chowk [sic] by the ... RSS.... The
RSS are strong patrons of many AJigarh akharas (though not all akharas are patro-
nised by the RSS), where drills and organised fights (sakhas) are held. At this RSS
function was a Senior Superintendent of Police (SSP), P. S. V. Prasad, who took
part in the religious ritual. The SSP subsequently said he had attended at the invi-
tation of personal contacts, a statement which gave Muslims little confidence in
the unbiased attitude of the AJigarh police. Moreover, the mohalla of Manak
Chowk [sic] is a sensitive one, bordering the area separating Hindu from Muslim
localities. It had formerly been a largely Muslim mohalla, but concerted efforts
of the Varshney and Aggarwal [sic] (Bania) castes had been made to purchase
property there and transform it into a Hindu mohalla. Muslims claimed, and
Hindus confirmed, that pressure, including intimidation and threats, had been
put on Muslim householders to sell at. concessional rates. A further aggravation
to Muslims was that the Bania. castes are strong supporters of militant Hindu
groups such as the RSS and the Vishwa Hindu Parishad (VHP).25

With Mann's account, we enter a still broader ring of complicity, one sug-
gesting the active participation of specific organizations in the events sur-
rounding Hindu-Muslim riots that are so often attributed to Hindu-Muslim
passions and animosities. The account also suggests that it is no accident that
the police played a partial role in supporting Hindu violence and attacking
Muslims, that the sympathies of the chief police officer of the district at the
time were with the RSS and militant Hindus.
When, more than two weeks later, on October 3, the Hindu wrestler Bhura
was stabbed and died in hospital on the following day, the nexus of goondas,
politicians, criminals, and police facilitated the provocative actions of the Hindu
crowd, itself composed primarily of goondas from a group of thugs known as
the Golden Gang as well as "volunteers of the RSS," in carrying Bhura's body
in procession to the areas of the old city where revenge and retaliation could
best be enacted, namely, at the two crossroads of Abdul Karim and Phu!.

The violence grew once the procession reached the Muslim-populated area of
Chauraha Abdul Karim, and the arson and looting reached a climax. Simul-
taneously, trouble flared up in Manak Chowk [sic] . ... tension continued
throughout October into the succeeding month. On 5 November, violence flared
again in Manak Chowk, the main targets of attack being the poorer Muslims
of that mohalla, most of whom were fruit -sellers, members of brass band teams,
or mazdoori labourers. The trouble was sparked off by an unrelated incident
in Phul Chauralla, an adjacent mohalla. 26
 The Geography and Demography of Riots /169
The collusion between local Hindus in Manik Chauk and the police, suggested
in several of the accounts given above, was also confirmed by "an outside team
sent in by the Peoples Union for Civil Liberties to report on events" during
the 1978 riots.

The Provincial Armed Constabulary (PAC) resorted to firing ... it fired delib-
eratelyat the Muslim houses.... According to Iqbal Ansari of AMU, a Hindu
gentleman of the area was asked on October 5 by the PAC men to identify the
Muslim houses so that they could shoot them from a vantage point. In Manak
Chowk [sic], it was estimated that of 40 Muslim houses with about 200 resi-
dents, ... 16 were burnt down and 20 residents remained after this incident,
the rest having fled t.o Muslim mohallas. 27

The complexity of the accounts given above of the incidents that took place
in this one mohalla of Manik Chauk and nearby areas suggests two aspects
of riot activity in Aligarh that deserve stress. One is the multiplicity of moti-
vations attributed to the actors. Second is the evidence of considerable
organizing activity by known individuals and organizations. Third is the exis-
tence of a network of relationships that comes into play before and during
riots, involving specific individuals, organizations, economic interests, crim-
inals, politicians, and the police, a network that is partially active at all times
and in which many of the participants have distinctive roles to play. These
roles include, among others, the organization of processions, the arousing
of sentiments, the gathering of crowds, the recruitment of specialists in vio-
lence, the transmission of information between organizers and the authori-
ties, particularly the police, to whom are assigned the uItimateworkofkilling
Muslims, the principal victims in most riots.
Manik Chauk figured prominently yet again in the series of incidents
between 1979 and 1981, especially those centering on the construction by a
Hindu businessman of a new cinema hall exiting into a Muslim mohalla. Hindu
and Muslim accounts in my interviews agree that violence was centered in
Manik Chauk. 28
Although a large number of people were killed in the 1979-80 violence,
nearly all Muslims, the police were acknowledged even by Muslim respon-
dents to have played a different role on this occasion, acting "properly,"
that is, impartially, to end the rioting rather than providing cover to Hin-
dus to attack Muslims and killing Muslims themselves. It is also notewor-
thy that Manik Chauk is perceived in accounts of these riots as a storm center
from which originate riots that will spread to other parts of the city unless
 170/ The Geography and Demography of Riots

such proper police action is undertaken. I was told, for example, that Varsh-
ney businessmen eager to gain control of Muslim property were respon-
sible for violent attacks against Muslims in this mohalla in 1980 in which
eleven persons were killed, but "the police acted properly, at once, so it was
not spread all over [the 1city.29 There is, therefore, a double reduction in
such accounts, from the elevated heights of general Hindu-Muslim ani-
mosities in Aligarh and beyond, to a specific mohalla where commercial
greed appears as the primary motive rather than communalism, and where
proper action taken by the police can confine and control violence at its
source.
This tendency to localize the significance of riots in Aligarh also appears
in many police accounts. For example, in 1983 the SSP, wbo appeared to me
to be impartial on Hindu-Muslim matters, isolated Manik Chauk along with
Upar Kot as the principal areas of riot activity from time to time and reduced
the conflicts further to conflicts between segments of the two communities:

Upar Kot and Manik Chauk are the two most riot -prone areas. Both have equal
number of people from both the communities and generally the trouble starts
with quarrels between Varshneys among the Hindus and Qasai among the
Muslims. 30

Leaving Manik Chauk now and moving south, following the broad road
known as tile Agra Road on the east, we soon reach another cluster of mohal-
las (Map 2), with Sarai Sultani just to the west of the Agra Road, surrounded
by the localities of Brahmanpuri, Madar Gate, Sarai Rai, and Barai (Map 2).
Sarai Sultani itself is overwhelmingly Muslim (89 percent according to the 1951
census, the remaining 11 percent being Hindu and others with no Scheduled
Castes listed). However, the surrounding mohallas are either overwhelmingly
Hindu-majority or mixed Hindu-majority mohallas. Strangely enough, none
of these mohaUas are classified as sensitive, riot-bit, or crime-prone, with the
exception of Madar Gate, which is classified as communally sensitive. Yet,
Sarai Sultani has been at the center of perhaps the most vicious killings in
Aligarh since Independence and has been caught up in nearly all the post-
Independence riots.
There is ample secondary documentation of the riotous activity in Sarai
Sultani during both the 1978 and 1990 riots. Insofar as the 1978 riots are
concerned, the report of the Minorities Commission on that riot noted that
"on 8th November the situation worsened and the police again opened fire
at Turkman Gate, Sarai Kaba and Sarai Sultani, during which five Muslims
 The Geography and Demography of Riots / 171

were killed in Sarai Kaba."31 Hasan notes that when the curfew was, "rather
unadvisedly, lifted" on November 9, "a number of Muslims were killed in
police firing and violence spread like wildfire," the "worst affected areas"
being "Manak Chowk [sic], Phul Chauraha, Turkman Gate, Sarai Kaba and
Sarai Sultani."32 Mann also lists Sarai Sultani as among the worst affected
mohallas in the 1978 riot and notes as well the partiality of the police in these
mohallas in searching Muslim houses, patronizing "known anti-Muslim
organisations," firing indiscriminately at Muslim houses, and also firing with
discrimination directly at Muslim houses identified by Hindus in the area. 33
Sarai Sultani was also at the center of the great riot of 1990. In fact, the
riots of 1990 began in Sarai Sultani and Upar Kot and only after the initial
incidents there on December 7 did the rioting move to other areas of the city.
According to hearsay and eyewitness accounts, discussed in Chapter 4, the
initial precipitating incident occurred when a bomb was thrown at a mosque
in Sarai Sultani, following which, in one account, "people from the mosque
came rushing towards the kotwali to complain." Their movement towards
the police station, interpreted by the PAC as an impending attack upon them,
was greeted with police bullets. The crowd responded by "snatching" the rifle
away from the "particular jawan [who] fired the first shot." Further police
firing followed in which two persons were killed. According to some reports,
"a Hindu mob joined the PAC in attacking the Muslims." The next morn-
ing, the PAC took revengeful action and "fanned out into the sensitive Muslim
areas" near Sarai Sultani, where an additional nine persons, all Muslims, were
killed. 34
The Peoples Union for Civil Liberties CPUCL) team heard a similar
account. Moreover, in their visit to the mohalla of Sarai Sultani, further infor-
mation was provided suggesting that the Muslims were attacked by both
Hindus and the PAC. Below are excerpts from the PUCL report on the 1990
riots in and around Sarai Sultani.

The area of Sarai SultaniiSarai Rai is largely composed of Muslims but is

surrounded by Hindu localities. Muslims who gathered in the Mosque for Friday
prayer on 7.12.90 [December 7, 19901 were alarmed when they saw hundreds
of people on roof tops of Hindu hOllses throwing bombs at Muslims and their
houses and at the Mosque. We saw impacts of bombs [they must be crude
bombs1on the houses in the locality and many houses destroyed in PAC firing
and rioting. We were also told that the ... PAC provided a cover to the riot-
ers by preventing the Muslims from approaching the Mosque to rescue their
friends and relatives; that some Hindus with country-made pistols were mov-
 1721 The Geography and Demography ofRiots

ing along with PAC jawans; that Muslims were confined to their houses and
inside the Mosque. If any Muslim dared to come out the PAC is alleged to have
fired at them.

Further evidence of Hindu participation in the actions against Muslims in

this area was provided to the PUCL team in their visit to the surrounding
Hindu areas.

In a Hindu area Jogipara, adjacent to Sarai Sulum, we met some Hindus, ...
but they were just not willing to speak to us. However, one elderly Hindu lady
told us that a huge crowd had gathered on the 7th morning from the Hindu
mohallas. Muslim neighbours were scared. Many of them came to her area;
she and some of her neighbours helped 20/25 escape through the backdoor
unhurt. She also told us that she too was scared because the mob was unruly
and they were unkind to even those people like her who were helping
Muslims.

In Sarai Sultani and other nearby areas visited by the PUCL team, its mem-
bers heard repeatedly "more or less the same picture, namely PAC resorting
to firing to kill indiscriminately, Hindus being helped by PAC cover, Mosques
attacked, houses destroyed, people subdued but angry."35
Several of my interview informants who had some contact with people in
Sarai Sultani had similar accounts to relate. According to one of them, a
Muslim professor at AMU who is involved in running a small school for
Muslims in Sarai Sultani, this mahalia-and several others in the city where
there were small pockets of Muslims, or mohallas such as Sarai Sultani that
are completely surrounded by predominantly Muslim mahallas-was delib-
erately selected for destruction. This informant referred specifically to Sarai
Sultani as an example of the manner in which deliberate, concerted attacks
were launched against Muslims in areas where they were in small numbers
surrounded by Hindu populations.

The same pattern is in Sarai Sultani. Every night they are throwing bombs
and this is an area surrounded on all sides, it is a small pocket. I go there, I used
to go about twice in a month there .... Nobody can come out because ... all
around they are being surrounded .... Now, PAC also, sitting on the top,
rooftops of the residents and then they attack. Pattern is they throw a bomb,
the people will come out, they will think that some attack is being made, and
then the PAC will start shooting them. Sometimes, a real attack, sometimes
 The Geography and Demography of Riots / 173

they have gas [?]. They throw the gas [?] in such a way that the people think
the mohalla is being attacked, they come out and the PAC start. This happened
in Sarai Sultani.

This informant connected this general pattern of attack to the incident at the
mosque that precipitated the first killings. However, he added an element to
it that I heard nowhere else, namely, that there was some kind of dispute over
this mosque, which had been repossessed by Muslims in Sarai Sultani after
having been abandoned for some time. It was at this mosque that the bomb
was thrown during Friday prayers, allegedly by people (presumed to be
Hindus) who were trying "to get hold of this mosque."36
I heard another account at the residence of the mufti of Aligarh in Upar
Kot proper at the very end of the 1990-91 riots when Muslims were walk-
ing up the hill to tell their tales of victimization in the days of rioting. The
stories that were brought to the mufti concerning Sarai Sultani were that
the area was attacked by Hindus, supported by the police, creating panic
among the Muslims there. On December 7, the frightened residents of Sarai
Sultani had been forewarned that there would be an attack on the mosque
during the jama namaz (Friday prayers), so they went to the mosque to pray
an hour before the proper time in order to avert the attack. Nevertheless, as
they were returning home from the mosque, they were attacked. They found
PAC men surrounding the roads. The crowd was fired upon and bombs were
thrown.37

THE SPATIAL ENLARGEMENT AND DISTRIBUTION OF

RIOT SITES IN THE GREAT ALIGARH RIOTS OF 1990-91

The survey of riot-hit mohallas presented above covers only the localities in
the old city that traditionally have been the centers of riotous activity. We
must now refer again to Appendix Table A.2 and Figure 6.4 to consider once
again the great riots of 1990-91. It is evident from the table and figure that
there was a vast increase in the number of localities hit by these riots com-
pared to all that preceded it: 55 compared to 27 in the 1978 riots, previously
the most widespread in Aligarh's history.
But it is not just the increase in the number of sites that is of interest here,
but their spatial location (see Map 4). A considerable number of riot-hit sites
were never hit by riots before, including a large number for which we do not
have census data comparable to that for the other mohallas, because they are
areas newly populated or newly incorporated into the municipal limits and
 114/ The Geography allei Demography

RasalganJ Road

Turkman Gate

R.iot Sites

.M.AP 4. Riot sites, 1990--91

 The Geography and Demography of Riots / 175

TABLE 6.3. Major A1igarh mohallas not included in the 1951 census

P.S.Nos. Voters Muslim voters

Mohalia Ward No. (1989-91ya (1984/1995Jb (%)'

PREDOMINANTLY MUSLIM

Bhamola Mati 46-49 3,632 73.13

Bhojpur 23 184-87 1,186 63.41
Firdous Nagar, 31 Notinduded 4,543 97.80
Kila Road,
Shahinshahabad,
Maulana Azad
Nagar
Hamdardnagar 36 Notinduded 3,193 100.00
Hamdardnagarl 41 Not included 3,306 97.22
Badamnagar
Jamalpur, 30 40-45 8,190 87.34
Jamalpur Mati,
Jamalpur NagIa
JauharBagh 43 17-20 d 7,132 100.00
Jiwangarh 43 22-27 3,230 93.22
Jiwangarh 47 22-27 7,036 100.00
NagIa Mallah 43 21 1,766 64.04

PREDOMINANTI.Y HINDUS AND OTHERS (INCLUDING SCs)

Chhavani 28 Notinduded 6,865 0

NagIa Pala Sahibabad, 29 Notinduded 3,441 0
Pala Sahibabad,
and Nai Abadi

'Polling station numbers, according to the 1989 and 1991 de.limitation of the Aligarh City
Legislative Assembly constituency.
bAll figures in this wlumn are from the 1995 voters' list except Bhojpur.
'Figures in this column are, in some cases, exact counts, in others estimates based on
a 1 percent random number count.
dlndudes Kela Nagar and Krishi Farm.
 176/ The Geography and Demography of Riots
lie on the outskirts of the city, not in the center. They include areas whose
newness is indicated by their names, that is, areas named after prominent polit-
icalleaders of India since Independence, such as Indira Nagar (after Indira
Gandhi) and Jagjivan Rampur (after the famous Scheduled Caste politician,
Jagjivan Ram), areas that have received an influx of Muslims from the sur-
rounding countryside, such as Hamdardnagar, and outlying villages as well,
such as Bhojpur and Jamalpur (see Table 6.3 and Map 4). The AMU author-
ities also released a qrclostyled sheet that listed some of the areas just men-
tioned as well as others near the university and on the outskirts of the city
that suffered disturbances, violence, and killings, based partly on reports that
reached them from university employees living in those areas. The localities
so listed were the Anupshahr Road, Badamnagar IHamdardnagar, Bhamola,
Dhaurara, Ektanagar, Jamalpur, Jiwangarh, Nagla Mallah, the Medical Col-
lege Colony near the Medical College, and Sir Sayyidnagar (see Map 4). No
census or other information is available with regard to Sir Sayyid Nagar or
Ektanagar. Dhaurara was listed only in the 1951 census as a village in Koil
tahsil, but is not included in either the Aligarh municipality or the Legislative
Assembly constituency. Census and electoral data are, however, available for
all other areas included in the AMU list.
Let us now examine the available evidence on these riot-hit areas and what
was said to me concerning the happenings therein. Insofar as Nagla Mallah
is concerned, there is no census information whatsoever on this village up to
the present. It is located on the northeastern outskirts of the city (Maps 1 and
4) along a road by the same name to the east of the AMU campus. According
to the voters' lists for 1995, the mohalla was communally mixed; out of 1,766
voters, 1,131 or 64 percent were Muslim (see Table 6.3), the rest Hindus and
others. One of my informants referred to it as a small village inhabited, as
the name would suggest, by "a community called Mallahs." Mallahs are a
somewhat notorious caste in northern India, fishermen and boatmen by tra-
ditional occupation, many of whom live on or near the banks of rivers, who
are said to engage in numerous forms of criminal activity such as looting and
smuggling, in addition to or instead of their traditional occupation. The same
informant previously cited alleged that, from tl1is village, Mallahs came out
to attack Muslims in various places, such as Shah Jamal and Sarai Sultani,
both areas some distance from Nagla Mallah. He alleged further that this was
a deliberate, planned operation that was part of a general strategy to attack
"the Muslim community from every corner."38
Shah Jamal (no. 69 in Maps I and 4), whence rioting began on November
1, was mentioned very frequently in my interviews. No census information
 The Geography and Demography of Riots 1177

is available on this site either. However, it can be precisely located and

described. It lies to the south of the Khair Road leading westward out of the
city from the Delhi Gate. It is close to one of the most important religious
sites for Muslims in Aligarh City, the Ustad Sahib kaDargah (no. 70 in Map 1),
a shrine that "attracts a wide following from Aligarh District, among Hindus
and Muslims."39 In 1990-91, however, despite its intercommunal appeal-
not at all uncommon for such shrines throughout northern India-the area
in which it was located was, according to numerous accounts, a site for delib-
erate attack by Hindu rioters and the PAC. The informant just cited included
this area among several others never attacked before to argue that there was
"a plan, ... a pre-plan because ... in every corner of the city, even this area
[AMU campus], behind Medical College, Jamalpur area, and then ... around
the whole city, Shah Jamal area, ... all these small pockets, they were just
burned and even Qazipara, ... in the last forty-three years of Independence,
they never attacked." Insofar as the Shah Jamal area is concerned, this inform-
ant alleged that a concerted attack was launched upon it by persons living in
the new colony ofIndira Nagar (no. 68 in Maps 1 and 4) just across the Khair
Road from Shah Jamal, along with PAC personnel.
The mufti shahar also referred to the violence in Shah Jamal and noted
that communal violence erupted in that area on the very night on which cur-
few was imposed in Aligarh.40 Still another Muslim informant reported to
me that "some element of [the] Hindu community attacked the Shah Jamal
community and, in return, there was pelting [of] stones from both the com-
munities. And when the police action was started, the police action was one-
sided." The police even attacked and damaged some portions of a mosque
in Shah Jamal. 41 Six persons, this informant claimed, including two women,
were killed in police firing in this area.
Also of great interest is the fact that two new colonies in the northeastern
outskirts of the city and just to the northeast of the AMU campus as well,
Hamdardnagar and Jamalpur (Maps 1 and 4), came under heavy attack.4Z
Hamdardnagar comprises all of ward no. 36 and had 3,193 voters in 1995, all
of whom were Muslims; the adjacent ward, no. 41, comprises Badamnagar
as well as Hamdardnagar, the two settlements being lumped together in the
voters' list. The corporator for ward 36 was a Muslim, elected as an inde-
pendent. The BIP had no candidate in this ward. The corporator for ward 41,
also a Muslim, was elected on the SP ticket. The BJP polled a mere 59 votes
out of 2,350 cast in this ward.
Two villages were listed in the 1951 census under the name Jamalpur:
Jamalpur Mua11 (Ma11) and Jamalpur Sia. The latter has remained in the rural
 178/ The Geography and Demography of Riots

tahsil, while Jamalpur Mafi has been incorporated in the municipality along
with another village not listed in any census, called Jamalpur ka Nagla (Map 1).
The population ofJamalpur Mati in 1951 was recorded as 1,027, of whom 259
(25.22 percent) were Scheduled Castes. The two Jamalpur villages comprised
all of ward number 30 in 1995. The total voting population in this ward in
1995 was 8,190 (Table 6.3). Seven candidates contested the 1995 corporation
election, of whom five were Muslim. The winning candidate, a Muslim inde-
pendent, gained 1,446 out of the total 4,737 valid votes cast (30.53 percent).
The BJP candidate, a Hindu, polled second from last with a mere 148 votes.
The second Hindu candidate polled better than the BJP, but still quite mis-
erably, with only 323 votes. So, the vote for the two Hindu candidates com-
bined accounted for less than 10 percent of the total valid votes polled (9.94
percent). The locality is overwhelmingly Muslim. It is also said to have a large
population of Qureshis. 43
In Jamalpur, I was told that 7 or 8 people were killed in police firings. 44
The PUCL report, cited earlier, also took note of the spreading of the riots
in 1990-91 to areas in the Civil Lines near the university, including the Zakaria
Market, Hamdardnagar, and Jamalpur. Hamdardnagar, the report noted, "is
a purely Muslim area" that was attacked by a "Hindu mob from nearby vil-
lages." Here, also, "a number of Muslims were killed by PAC firing." Jamalpur
was described in the same report as "a Muslim area with a small Hindu
pocket-about 95% Muslims and 5% Hindus." In the rioting here, "one Hindu
was stabbed to death and another injured" and, confirming my own inter-
view just cited, "a number of Muslims were killed in PAC/Police firing."45
The Times of India also remarked upon the incidents in these two localities
as follows.

Violence spread to the outskirts of the city t.oday when members of two
communities clashed at Hamdard Nagar and Jabalpur [JamalpurJ, hardly
half-a-km from Aligarh Muslim University (AMU). The Provincial Armed
Constabulary (PAC) personnel opened fire t.o disperse the rioters, killing at
least five persons. Officially, however, only three deaths were reported from
these localities....
Contrary to official reports, bloody clashes took place on the city outskirts
this morning. A mob of 2,000 people set afire many houses in Hamdard Nagar
locality. Residents of the area narrated gory tales of PAC brutality. When this
reporter visited the house of two brothers, who were shot by the PAC per-
sonnel in their residence, the victims' fresh blood stains were still visible. The
widows said their husbands, both rickshaw-pullers, were at horne when six PAC
 The Geography and Demography of Riots / 179
jawans barged in and despite repeated pleas shot them from point blank range.
"They were innocent. Please remove the PAC or they will kill all of us," one of
them said, wailing uncontrollably. Their neighbours expressed the same appre-
hension. Scores of residents began to migrate to safer places after the firing,
with their scanty belongings.
The scene at the AMU's Jawaharlal Nehru Medical College was heart-
rending. The bodies of the five killed in Hamdard Nagar were still lying in the
hospital with those of victims from other localities. 46

These incidents near the AMU occurred a day after the massacres of Muslims
on the Gomti Express and a day before the circulation of the false newspaper
stories concerning the killing of Hindus at the A..\1U Medical College Hospital.
These new areas are sites of an influx of Muslim migrants from the rural
areas, whose presence has been particularly noted by Krishna Kumar
Navman. Navman estimates Muslim migration to these new areas altogether,
including other sites such as Jiwangarh, which lies just to the east of Nagla
Mallah, and Bhojpur on the southern side of the city, has amounted to twenty
to twenty-five thousand persons. Most important from Navman's point of
view is his claim that, although some of these new areas lie outside the bound-
ary of the Aligarh City Legislative Assembly constituency, the Muslims have
succeeded in registering to vote illegally in the constituency and have voted
against him. 47 It would stray too far from what is known to say that this infor-
mation suggests preplanning on the part of Navman and the BJP to teach
these Muslims a lesson or to encourage their out-migration. It can only be
said for certain that these new areas of Muslim migration were hit hard in
1990-91, that they were hit for the first time in the history of Aligarh at that
time, that their presence in the city limits is resented by militant Hindus,
and that, therefore, they are sites of political contestation for control of the
constituency that, as we will see in succeeding chapters, has been closely fOUght.
Jiwangarh in 1951 was an insignificant place with a small population.
However, forty years later, it had become a huge mohalla with an electorate of
10,266, of whom 98 percent were Muslim (Table 6.3). It comprised six polling
stations in the Legislative Assembly constituency and the bulk of two wards
in the municipality. Navman polled a mere 84 votes out of 1,919 cast in 1989,
and zero votes in the six polling stations in the 1991 elections. Of course, I can-
not determine whether or not any of the votes cast in these polling stations
were illegal, but they certainly posed a serious threat at that time to N avman's
and the BJP's continued predominance in the constituency.
Let us consider now the locality of Bhojpur, also hit for the first time in
 180/ The Geography and Demography of Riots

the riots of 1990-91. It is located on the southern outskirts of the city below
and to the southwest of the Sasni Gate (Maps 1, 2, and 4). In the 1951 census,
it was listed as a village in Koil tahsil with a total population of 451, of whom
99 (21.9 percent) were Scheduled Castes, but there was no breakdown by reli-
gion. The voters' list for 1984 listed 1,186 voters, of whom I have been able to
identify positively 752 Muslim names, giving a Muslim population percent-
age of 63.41 percent (Table 6.3).
Before 1967, Bhojpur was included within the Koil Rural Legislative
Assembly constituency. Since 1967, it has been included within the Aligarh
municipality and in the Aligarh City Legislative Assembly constituency.
Because of its relatively small population, it was lumped together with other
mohallas in the polling stations of which it was a part, with seven other mohal-
las in the 1996 and 1968 delimitations and with two other mohallas in the 1980
and 1985 delimitations. However, by 1989, its population and number of vot-
ers had grown sufficiently for Bhojpur alone to comprise four polling sta-
tions. I have the polling station results for two elections conducted after and
based upon this delimitation, the 1989 and 1991 Legislative Assembly elections,
which sandwiched the great riots of 1990-91. They are of great interest from
several points of view. In 1989 and 1991, 1,443 and 1,604 votes, respectively, were
cast in these polling stations. Of these votes, 75.26 and 78.05 percent, respec-
tively, went to the Janata Dal candidates, who were, in both elections, Mus-
lims, while 15.66 and 17.02 percent went to the BIP candidate, who was, in
both elections, Krishna Kumar Navrnan.
However, there was a marked difference in the results among the four
polling stations. In polling stations 184 and 195, the BJP polled 29.90 per-
cent in 1989 and 35-19 percent in 1991, while the Janata Dal polled 58.67 per-
cent and 59.24 percent, respectively. However, in polling stations 186 and
187, the BJP polled a mere 1.91 percent and 1.62 percent in the two elections,
while the Janata Dal polled 91.28 percent and 94.01 percent. In these elec-
tions, the Janata Dal was seen by Muslims as their main protection against
the militant Hindu onslaught against the Babri Masjid and as the party most
likely to defeat the BJP and its candidate, considered by Muslims to be the
leading riot-monger in the Hindu community; as will be shown in detail in
later chapters, Muslims voted overwhelmingly for this party in nearly all
mohallas of Aligarh City. It can be safely assumed that virtually all Muslims
in these polling stations voted for the Janata Dal and that most Hindus voted
for the BJP. In short, in the elections that preceded and followed the riots
in Aligarh in which Bhojpur was badly hit, the electorate in this area was
politically polarized.
 The Geography and Demography of Riots / 181

In 1995, Bhojpur comprised approximately half of ward 23 in the munici-

pal elections, which were won hy the Samajwadi Party of Mulayam Singh
Yadav with a Muslim candidate. In an election in which there were ten can-
didates, all but one of whom were Muslim, the winning candidate garnered
1,286 votes out of 4,066 votes cast (31.63 percent). The lone Hindu candidate,
contesting on the ticket of the BJP, polled 518 votes (12.74 percent).
It is clear, therefore, from all the demographic data and the election results
available that pinpoint this locality, that it is preponderantly Muslim, that its
population has increased greatly since 1951, presumably by in-migration from
the rural areas, that the BJP has little strength here and probably none from
Muslims, and that the Hindu and Muslim populations in the locality are
sharply divided politically.
There were two reports in the Times of India concerning rioting in Bhoj-
pur. One referred to <Ca[n] ... outbreak of violence in the Bhojpura [sic]
locality" on the night of December 28. The news reports linked the violence
here to two factors. The first was a visit to the city on December 25 by the
chief minister, Mulayam Singh Yadav. ''A senior police official" was reported
to have said that, as a reaction to his visit, "there appeared to be a 'deliber-
ate attempt' to engineer violence in certain parts of the city." The second
"major factor" was said to be "the relationship between the Muslim com-
munity and the Provincial Armed Constabulary (PAC)" that had "now
definitely reached 'boiling point.''' The disenchantment of Muslims with the
PAC had led, it was said, to "militancy in a section of the Muslims, who are
seeking 'desperate solutions.'" So, it went on, when violence broke out in
Bhojpur on the night of December 28, "the PAC opened fire and, according
to official figures, two persons were killed. But local residents claim that at
least another three persons are missing and accuse the PAC of disposing of
their bodies."48
A second report from the same newspaper stated that there was violence
again in this locality on the night of January 1, when "police resorted to firing
at Bhojpura [sic] locality when some miscreants set five houses on fire. Two
of them received bullet injuries. They hurled bombs at policemen in retalia-
tion. Two persons were stabbed to death in Bhojpura."49
I have comments in my interviews on the happenings in Bhojpur during
these riots from the mufti shahar and from riot victims who were beaten and
who lost relations. The mufti, who had stressed that most of the violence in
these riots had been directed against Muslims on the outskirts of the city,
noted also that such violence had occurred on the southern and western out-
skirts in areas such as Laria, Jangal Garhi, and Bhojpur (Map 2). I also heard
 182 / The Geography and Demography of Riots
victim accounts of the happenings in Bhojpur from several residents, includ-
ing two bearded Muslims, who said that the police entered their house, pulled
their beards, and "beat them mercilessly" on the night of December 31. I asked
if there was any reason for the police behavior and received the following reply.

The police opened the door and asked us, "Where are your sons? They were
firing from here." But none of us was firing and we don't have anything. So,
just pretending about the firing, they said, "Beat them up."
PRS: But there was no firing from the house at all, nor firing from the roof?
INTERLOCUTOR: No.
PRS: Nothing in the locality?
INTERLOCUTOR: There was firing in the locality, but it was police firing and
one person was killed by the police, a boy of nine years.
FRS: And it's not possible the police were making some mistake, only they were

doing this just to harass, not making some mistake?

INTERLOCUTOR: No, no, they are harassing. They're doing it deliberately.
PRS: And this is police or PAC?
INTERLOCUTOR: He's saying-he's a very old man-the's saying] I can't
differentiate.... He's saying they broke open the door. He's crying with pain.
FRS: From being beaten?
INTERLOCUTOR: Yah, yah. With the rifle butts. And these are the [people].
[Laughing sardonically.] According to them, according to police and the
PAC, these are the rioters. These people and that woman, these are the riot-
ers. [Laughs ironically again.] They can't walk, they can't stand up straight.
You see. [Again laughing ironically.] They can fire and they can make the
riots. What a fun. 50

In reflecting on these events in Bhojpur, it should be noted that no connections

have been explicitly made to economic factors, caste-communal rivalries in
the mohalla, or any other factors by those who commented upon them in the
press and in my interviews. All we know for sure is that, like other localities
on the outskirts of the city that were hit for the first time during tl1ese riots,
Bhojpur has been a site of greatly increased in-migration of Muslims and a
site also of great political concern to the BIP for that very reason. If, how-
ever, it was the intention of anti-Muslim rioters, whoever they may have been,
to provide negative incentives to the Muslim residents of Bhojpur and other
such areas to move out of these areas and back to their villages, it failed. Indeed,
on the contrary, one of my respondents claimed that there had been a con-
siderable further influx of Muslims from the rural areas into these localities
 The Geography and Demography of Riots /183

that turned the balance against the BJP in successive elections, leading to its
ultimate defeat,;l which will be analyzed in subsequent chapters. According
to my estimates from the 1984 and 1995 voters' lists and the corporation elec-
tion count of actual voters, the number of voters, most assuredly mostly Mus-
lims, tripled in one decade in this locality, a matter of the greatest political
concern to the BJP.
Two other areas close to the university were mentioned in the AMU list
of affected sites, Anupshahr Road and Bhamola. Anupshahr Road was classed
as a mohalla in the 1951 census. It had then a population of 1,105, of whom
930 (84.16 percent) were Muslim. In 1995, Anupshahr Road was included in
ward 57, University, which comprised, as suggested by its name, other local-
ities in the neighborhood of the AMU. The number of electors in this ward
was 3,686 at tllat time, of whom 3,500 (94.95 percent were Muslim). The area
also comprised six polling stations in the 1989 and 1991 elections. The Janata
Dal was the favored party in both elections, winning 85.99 percent of the vote
with Khwaja Halim as the candidate in 1989 and 80.85 percent with Moham-
mad Sufiyan as the candidate in 1991. The BIP was of little consequence in
these polling stations, where Navrnan polled less than 4 percent in 1989 and
less than 5 percent in 1991.
Bhamola, unlike Anupshahr Road, fits better in the category of the new
and newly incorporated areas that have experienced an influx of Muslim res-
idents and voters. In 1951, it was a village in Koil tahsil Witll a total popula-
tion of 880, of whom 229 (26 percent) were Scheduled Castes. No information
was provided on the Muslim population. In 1995, Bhamola comprised the
largest portion of ward 53, along with another village called Nagla Munda.
Bhamola was then predominantly Muslim. It comprised all of four polling
stations and half of anotl1er shared witl1 Nagla Munda. In the former four
polling stations, the percentage of Muslim voters ranged between 55.57 and
88.89 percent, whereas Nagla Munda, on which incomplete information is
available, appears to have been either completely or overwhelmingly Hindu
and others. The ward, therefore, was divided in its communal composition,
which was in turn reflected in the election results. The winning candidate, a
Muslim independent, was elected with less than 40 percent of the vote.
In the 1989 and 1991 elections, Bhamola alone comprised all of four
polling stations, numbers 46 to 49, in which, as in Anupshahr Road, the Ianata
Dal candidates polled higher botl1 times, but with less strong majorities of
60.57 and 66.61 percent, respectively. Navman had some strength here, hav-
ing polled 28.09 and 21.76 percent, respectively, in the two elections.
The addition of Anupshahr Road and Bhamola to the long list of areas
 184/ The Geography and Demography of Riots

that experienced unprecedented communal violence in the northern half of

the city brings out another feature of the 1990-91 riots, namely, that the entire
university area was encircled by riotous violence. It cannot be said that such
violence was instigated entirely by Hindus or by RSS/B]P/Bajrang Dal activists.
In fact, there was even a suggestion in the above-mentioned AMU list of dis-
turbed sites that the safety of "non-Muslim teachers" was threatened in the
Medical College colony. It was noted therein that these non-Muslim teach-
ers "were extremely apprehensive of [their lives) and property particularly
because of the infiltration of large number of refugees where it was felt that
anti-social elements have also got mixed up." The AMU authorities, there-
fore, requested the DM to post a CRPF force therein, which was done on the
night of December 9.
A further factor mentioned to me in interviews concerning these riots con-
cerned the alleged participation of persons from Scheduled Castes, on the
instigation of non-Scheduled Caste Hindus, in attacks on Muslims, as well
as alleged Muslim attacks on Scheduled Castes. In particular, the locality of
Qazipara (Map 2) was mentioned, which lies in the heart of the old city but,
as noted above by one respondent, had never been hit by riots since Inde-
pendence, until these riots. The same respondent alleged that Scheduled
Castes, who had in previous riots sided with Muslims, had been instigated,
incited, bribed, and provided with liquor by "RSS people" to attack Muslims
in this and other localities. It was even alleged that some masked persons,
presumably Hindus masquerading as Muslims, deliberately killed Scheduled
Caste persons in order to incite them to attack Muslims in retaliation. It was
said also that the attack on Qazipara was launched in this way from a new
colony, on which I have no information, known as Jagjivan Rampur after the
famous Scheduled Caste leader.52
While there is no way to prove or disprove these allegations, two localities
adjacent to Qazipara-Sarai Qazi and Samnapara (Map z)-are overwhelm-
ingly populated by Scheduled Castes, their proportion in these localities in
1951 having been recorded as 93.16 percent and 88.50 percent, respectively.
Moreover, similar accounts were provided in the PUCl report, which noted
that Qazipara, like several other mohallas in the old city that were attacked
during these riots, was surrounded by predominantly Hindu areas and that
"Kazipara [sic) was attacked by a mob from the nearby Harijan colony." The
report also referred to instigation of Scheduled Castes by persons from the
Hindu "trading community to attack Muslims" in another locality called Sarai
Hakim Takia. It was also noted that the Harijan colony in this area had been
attacked by Muslims and Harijan houses had been burnt and looted. 53
 The Geography and Demography of Riots /185

The Times of India also reported that "a Harijan youth" had been killed
on November 24 in Ghuria Bagh locality in the old city, another mohalla
with a large population (above 20 percent) of Scheduled Castes, and that
the incident "had led to renewed tension in some parts of the old city" on
November 25. 54
Finally, it also deserves to be noted that the participation of Scheduled
Castes in attacks on Muslims-and on Sikhs in the 1984 riots in Delhi-was
reported from other areas of the country as well during this period. In one
case, in Kanpur City, of which I have personal and direct knowledge and infor-
mation, these attacks were led by a Scheduled Caste leader who was at the
time a member of the BJP and who was made into a hero by the RSS and Jan
Sangh leaders in the city for his activities during the post-Ayodhya riots there
in December 1992.55
All the evidence available, therefore, suggests that the enlargement of the
spatial spread of disturbances in 1990-91 compared to all previous riots in the
history of the city may have been related to two factors: first, the in-migration
of Muslims to outlying areas of the city, which has been resented by militant
Hindus and has posed obstacles to the maintenance of the recently ascen-
dant BJP in the political life of the city; second, the recruitment of Scheduled
Castes into the process of riot production, probably not for the first time,
but in a much more significant way at this time. That evidence also suggests
the possibility openly stated to me by respondents and reported by the PUCL
and the press that these enlarged attacks on new areas of the city were delib-
erate and preplanned.56 I cannot say that proof has been provided on these
matters, for proof has been made impossible by the failure of the authorities
to appoint an inquiry commission that might have explored these matters in
Aligarh, as in most other parts of the country}7
Moreover, the circumstantial evidence that indicates selective targeting of
mohallas is somewhat mixed. All the mohallas on the ou tskirts of the city that
were attacked in 1990-91, except Hamdardnagar, were part of the Legislative
Assembly constituency in the two elections of 1989 and 1991. But there are
several other overwhelmingly Muslim mohallas located in ward no. 31,
Firdous Nagar (Map 1), located to the northwest of the AMU and close to
the very hard-hit area of Jamalpur. In addition to the mohalla by the same
name, there are three other localities in this ward: Kila Road (part), Shahin-
shahabad, and MaulanaAzad Nagar (Map 1). The total voting population of
the ward in 1995 was 4,443, of whom the estimated proportion of Muslim
voters was 98 percent.58 The corporator elected in 1995 was a Muslim on the
ticket of the SP, the nemesis of the BJP. However, this area did not suffer in
 186/ The Geography and Demography of Riots

the 1990-91 riots. While there may be other reasons for its having escaped
attack, it deserves note that the area was not a part of the Legislative Assembly
constituency. On the other hand, Hamdardnagar, which was part of the con-
stituency, was attacked.
On the whole, however, it is clear enough that, though some areas of
Muslim in-migration and population expansion did escape attack, all the out-
lying areas that suffered were areas of concentrated Muslim population and
in-migration. Such evidence of deliberate targeting also further demonstrates
the importance of going beyond ecological analysis of demographic factors
in the search for the causes of rioting. It is not that an ecological analysis
would ignore the differences in, say, the communal/caste/ethnic composi-
tion of different sites or the proportion of migrants-very common sources
of explanation, indeed, for riots everywhere in the world-as explanations
for their occurrence here and not there. Rather, it is that such analyses are
useful only to describe the sites of disturbances, but cannot be used to explain
why they have occurred here and not there. Social science research that stops
at the statistical demonstration of an association, even when it asserts that
the finding of an association does not demonstrate causality, provides raw
material for uninformed speculation. Only through on-site observation at
the micro level can we begin to grope towards the truth of things, towards
the actualities of human agency, purposive activity, and active production
of violence.

A Visit to Sarai Sultani

In April 1999, I visited Sarai Sultani and interviewed persons in the locality
concerning the history of communal violence there. My primary informant
was a Qureshi Muslim bicycle lock manufacturer, nephew of the municipal
corporator from that area. His lock business has a turnover of three million
rupees annually and employs forty persons who produce twelve hundred locl<s
per day. He sends all his production to the Delhi markets. He sells his locks
for Rs. 14 each; they are sold on the retail market at around Rs. 32 each. A
Muslim manufacturer, his locks are sold under the brand name "Krishna,"
the name of a Hindu deity.
I had actually not planned to interview this informant, but his uncle, the
municipal corporator. It was only by chance, while waiting for his unde,
who never arrived, that in the course of conversation we discussed the com-
munal situation in the area. I summarize below the gist of that very lengthy
conversation.59
 The Geography and Demography of Riots / 187

FIG. 6.6. Lock manufacturing, Sarai Sultani, Aligarh, 1999

He said that only Muslims live in this mohalia of Sarai Sultani. There are no
Hindus living in the mohalia. However, there are Hindus living on all sides of
Sarai Sultani and Hindu organizations as well, as the tongue is between all the
teeth. Pointing in different directions, he said that is the side of the Bajrang Dai,
that is the side of Vishwa Hindu Parishad, that is the side of the Hindu
Mahasabha office, that is the side of the RSS, that is the main circuit of the riots.
Several Muslim baradaris live in Sarai Sultani, including Qureshis, Dhobis,
Ansaris, and Kurnhars, among whom the Qureshis constitute about 40 per-
cent of the total population.

Conversation concerning riots in the area began in response to a general ques-

tion that I put to him concerning whether or not there had been riots in this
mohalla.

He began quite emphatically by referring to his personal experiences in 197960

when he himself was beaten by the police, as a consequence of which parts of
his body had swelled and never returned to normal, including a knob on his
head that he showed me as well as a knee injury. Further, his father, his broth-
ers, his uncle, the whole family were beaten and jailed by the police. For two
days, on the 19th and 20th in the month of June [invariably a month of hell-
 188/ The Geography and Demography of Riots

ish heat and humidity], they were held without being given water to drink. When
they asked for water, they were told to piss and drink. 61

I questioned him repeatedly concerning whether or not they had done any-
thing, provoked the police in any way, given them any ground whatsoever
for beating and jailing them.

That night they were just sleeping and the police came and surrounded the whole
area and they called him up, called up his neighbors also and they all, all the
police force, took them away to the police station. There were five police sta-
tions in the area and they took them to one only 50 meters away from his house.

I asked again, "But they were sleeping, doing nothing?"

Yes, nothing. They were doing nothing but sleeping that night. I asked then
why the police came and called them out. He said that, because the riot was
going on around the area and there was no other Muslim mohalla anywhere
in this area, so they came to this mohalla.

One might question how the entire family could have been sleeping bliss-
fully, doing nothing, in the midst of the turmoil of a riot in the area. How-
ever, the family was in no position to do anything else, unless they were
deliberately seeking trouble, since curfew had been imposed two days earlier
on the 17th of June.

So, I asked again, what did he think, what was the reason the police came here?
The main factor was communal feeling. They saw that here there are Muslim'>,
so why not drive them away from here? They feel that there are two Muslim
parts in this [broad) area. One is Upar Kot and the other is this place. So why
not drive them away and kill those Upar Kot people also?

He repeated again that all his family members were beaten bitterly and their
bones were broken.
I asked yet again, trying to find some justification for such police behavior:
the Muslims in here did not come out and fight?

They were just sleeping and when they got up, they found that the whole area
is surrounded by the police, and particularly by PAC, and with them, there were
people of Hindu organizations like Bajrang Dal, Hindu Mahasabha, and RSS.
 The Geography and Demography of Riots /189

I then said, "But, you know, if I talk to the police, they will say the Muslims
come out, they come out in the street and we have to shoot to protect our-
selves and that the Muslims are aggressors and so forth. [There was1no attack
on the [police], they were not angry because Muslims had come out and
attacked them anywhere?"

No, no, no. Our total population is only 2,000, and that includes women and
children, so how can Muslims, being such a small population, go and attack
anywhere or attack anybody? It's difficult to defend ourselves, how can we be
offenders? So, we defend, we always try to defend ourselves; never are we the
aggressors.
PRS: SO the police came here and dragged you out, [and beat you] with mthis.

RBSPOND.ENT: Yes, yes. The police called all the young men from their houses

and they collected them all. There were around fifty-five. The police gath-
ered them near the main entrance [to the mohalla).
PRB: And beat them there?

RESPONDENT: While they were gathered there, there were two groups of police

on different sides. One group was calling them [residents] to this side and
the other was calling them to that side. While they [the residents) were going
to one side, the other group [of police J was saying you are going to that side,
you are trying to run away, and so they started picking on these people bitterly.
INTERLOCUTOR: So, all the parts of their bodies were broken and their bodies

started bleeding, including his body and his uncle's. He had seven stitches.

At this point, he again showed me his knee that had been cut with a bayonet
and a big bump on his head from a lathi blow.
The 1990 Riots. This respondent had much to say also about the 1990 riots
in Sarai Sultani. Recall that several secondary sources and my informants in
1991 had referred to the incidents in Sarai Sultani, precipitated by a bomb
thrown at a mosque. The mosque is actually not in Sarai Sultani, but in the
adjacent, mostly Hindu mohalla of Barai. Here is what he had to say about it.

Nineteen-ninety was the biggest riot. Here the nearby masjid [mosque) was
burned. The imam [prayer leader) was in the mosque at the time and was burned
alive. One M. M. Jain had a furnace 62 nearby. Twenty-two Muslims were taken
there and all of them were thrown in the furnace and burned alive. This was
done in the presence of the Station Officer, one Sharma, from Sasni Gate police
station. He is a very infamous person because of this incident.
One person [whom I soon met] survived this incident, but all his family
 190 I The Geography and Demography of Riots
members were killed. Since that time, he has become mad and a drunkard. One
of the survivors of this incident also had a sister, who had just been married
when the riots broke out and she was raped by the police for seven days, after
which she was killed.

Once again, I asked: "And that time also there was no provocation on the part
of the Muslims here? There was curfew. Muslims stayed in their houses or
the Muslims had come out and the police came?"

No, there was no provocation on the part of the Muslims. They didn't pro-
voke anybody. On this side, in Barai, there are only five houses of Muslims and
there are 5,000 Hindus, so how can anybody imagine that people from these
five houses go and provoke and challenge those 5,000? He also had his own
personal story to tell about the 1990 riots.
At that time, his factory was being constructed. There were five Hindu labor-
ers working in this factory, so, when the riots began, he called the police and
handed over all the five workers to the police.
But at that time, his own brother [cousin-brother, that is, mother's sister's
son], a ten-year-old boy at the time, was shot in his hands and a bomb was
thrown at him and exploded on his stomach, as a consequence of which that
part of his body was burned. [I later met the young man and saw the wounds;
see Figure 6.7.J At the same time, two Muslim men, one of them a prominent
district Congress officer, were also shot and injured. He himself was just sit-
ting here at his factory when his brother was injured. The injuries to his brother
occurred at the entrance to the mohalla, not far from the factory, quite near
the police chowki. He went to the site and had a confrontation with the police
and the district authorities who were there, including the District Magistrate
and the SSP. He asked the police there how could it happen that the Hindus
shot [sic] his brother while you were here, but they had no answer. He then
had a heated exchanged with the SSP and the DM, but they also had no answer.
He was just asking how come the whole administration is here and the Hindus
are killing my brother? What are you doing here?

Yet again, trying to find some reasonable explanation for these incidents,
I said to him, but there was curfew, what were your brother and you doing
out during curfew? He replied that the curfew was off for two hours when
his brother went out and was shot. In fact, he said he asked the administra-
tion, the DM and the SSP, "How come there was still a half-hour remaining
 The Geography and Demography of Riots / 191
and you imposed curfew a half-hour early?" It appears that, while the cur-
few was off for those two hours, there was a big fight and the curfew had been
reimposed.
There was yet a further atrocious incident in Sarai Sultani in 1990 that this
respondent recounted, which occurred while he was at the AMU Medical
College, where he had taken his injured brother for treatment.

According to his account, during the 1990 riots, there was curfew for 21 con-
tinuous days. And there was no water, no electricity, no milk, and the children
were weeping because they did not even have any milk to drink. [This seems
to contradict his earlier statement that curfew was lifted for two hours each
day.) And so, what happened was that two women entered the Hindu area
and took their children, two children, aged four and five years, also with them,
to the entrance point ofthe mohalla. At the entrance point, the PAC men asked
them why they had come there. Then the women said that we have come for
milk, we have to get milk. So these PAC men took the two children with them
and the milk pot and went away while the two women were standing there wait-
ing for the milk, but what happened actually was that these PAC men burned
those two children alive.
PRS: Burned them alive!
RESPONDENT: Yes.
PRS: How did they know that they were Muslims?
RESPONDENT: They were sure that, because they were coming from the side
of Sarai Sultani, these women had masked themselves, but were actually
Muslims. Besides, there were Bajrangi men sitting there and they knew that
they must be Muslim women. This incident of the burning alive of the two
children has been [officially) recorded.

Other Riots in and around Sarai Sultani. Having listened to this respondent's
accounts of riot activity in the two most severe riots in post-Independence
Aligarh, I inquired whether or not he had experienced any other difficulties
in the years between 1979 and 1990. He replied that he had many difficulties
during that long period. He said that there had been several disturbances in
the early 1980s. However, there was a period of relative calm in the mid-1980s,
which he attributed to the judicious administration of District Magistrate K. M.
Punia, a Scheduled Caste man from the Jatav caste, whom I knew personally
as well. 63 Punia, being a Jatav, was favorable to the Muslims, supported the
Muslims, and appointed several Muslims, including the respondent, to peace
 192 / The Geography and Demography of Riots
committees of police and citizens. This respondent referred to Punia in English
as "my best friend."

After Punia left, a new DM was appointed, one Gupta, during whose tenure
a riot occurred in the nearby flower market. This respondent described this
riot as being directly connected to economic competition between Hindu and
Muslim flower sellers. The problem here was that there were some Hindus as
well as Muslims in that market, in the flower business, and they sold flowers
at Rs. 500 per [quantity not given] whereas Muslims sold the same quantity
of flowers at only Rs. 200. So, that led to the tension and the entire Muslim
market was burned by Ram Bhakts [devotees of Ram], Bajrang Dal people.
When the Muslims went to see the district administration to ask what these
Ram Bhakts were doing there and why were they allowed to burn the whole
market, the DM said: "What can we do? We can't do anything." And, in that
trouble, about Rs. 50 lakh [5 million] worth of property was destroyed.

I asked then whether there had been any problems in Sarai Sultani since
the great 1990 riots. He said there had been disturbances in 1992 during the
Babri Masjid movement, but that there had been no problems in Sarai Sultani
thanks to the actions of the local police commanding officer, one Subhash
Baghel, who, like Punia, was also a Scheduled Caste man. The respondent had
a personal acquaintance with him, a good relationship with him.

He was sent to this area since this was a known trouble spot. Baghel sealed the
entire area around Sarai Sultani and there was no incident of communal vio-
lence. Further, he told the population of the area that, if there is any distur-
bance, there will be great trouble for the whole population. And he told the
policemen that, if even a single incident occurs, [they] will be responsible for
that. So, no incident occurred at that time.
He went on to remark that, normally, whenever curfew is imposed in the
city, it is only the Muslims who have to undergo all kind of troubles and
suffering. But, if you go into any Hindu area at such times, you won't feel
there is any kind of curfew. But, when this Subhash Baghel was in this area,
then the Hindus for the first time realized what is curfew because he sealed
the whole area.
At that time, there was only one small incident at Adda Hathras, where there
is a small mosque and these Ram Bhakts attempted to burn the mosque. But,
this Sub hash Bagllel put a stop to it and tlle masjid was not burned. And, in that
incident, 25 people, Hindus, were arrested and they were beaten.
 The Geography and Demography of Riots / 193

At the end of my interview with this respondent, he expressed his hope-

lessness and fear for the future if he continued to live in Sarai Sultani.

He's not very hopeful. He fears his future at this place. He says it's actually hope-
less. He says that he will ultimately have to leave this place. He's not safe here,
he does not feel safe here. Always, these days, these Bajrang Dal people and
VHP people are increasing their activities, different kinds of activities, different
kinds of programs. They sometimes start processions in all these areas, so these
people feel very threatened. He gave one example that, during kite flying sea-
son, these people write very abusive things against the Muslims on their kites,
saying Muslims go away from here, and they write abuses against mothers and
sisters. There have also been incidents of some pamphlets issued by the Bajrang
Dal and VHP people saying that you, the Muslims, leave this place, you have
no right to live over here. When the Muslims here report these matters to the
police, the police say what can we do? We cannot do anything. But the Muslims
are hesitant to leave this place. They think that if they leave these places, their
mosques will be deserted. Already, there are areas from which the Muslims are
leaving: for example, they have vacated five or ten houses in Barai and in other
places. And, in those areas, there are around 25 mosques and they are deserted,
nobody goes there to offer prayer. In this area also, there are six mosques. They
think that, if they leave the place and go somewhere else, these mosques will
be desecrated. Nevertheless, they are thinking of leaving this place and going
to the university area because they do not feel safe here.
What the Hindus want is to provoke the Muslims and kill them. They want
some kind of pretext to kill them, so they try to provoke Muslims and kill
Muslims. They try to make the Muslims react and then they kill. Suppose, for
example, there is some small incident, somebody is beaten or somebody is killed,
and the news spreads like wildfire all over an area. Suppose this incident occurs
anywhere in the city, it will spread everywhere and the whole area will be dis-
turbed and then the killing starts. Many people are killed. It can happen any-
where in the city. Any Muslim can be anywhere around here in this area or in
the Hindu area mainly. While they are [walking 1, since this kind of news spreads
so fast, they may be killed in any spot here or there. So, they do not feel safe in
this area. There is no hope.

After this very long interview, I was taken through the mohalla to visit the
sites where the several incidents occurred to which the respondent referred.
I visited the mosque and met the survivor from the group of 22 people who
had been thrown into the furnace in 1990, all of whose family members had
 194 / The Geography and Demography of Riots

been part of the number killed. I was shown newly constructed houses that
had replaced the ones that had been burned.
I also visited the mosque in the adjacent mohalla of Jogipara where the
imam had been burned to death in 1990. At this spot, I interviewed a Muslim
man who had been taking care of the mosque and who formerly owned five
houses adjacent to it. I was told that he had sold four of the houses to Hindus.
The reason given was that the Hindus were trying to push him out by threat-
ening him, so he had to sell these houses at low prices. He still maintained a
small cloth shop there. I was told that this was the very person whose family
members had been burned to death and whose daughter and sister had been
raped and who had gone mad since then. I was told also that he had given
many interviews of this type and had shown these places to other people, other
intellectuals also.
I was shown also a small garden park that this Muslim man owned, which
the Hindus are also pressuring him to sell. However, since it is just next to
the mosque, he feels that it will be a threat to the mosque if he sells it to the
Hindus, for it will come under their control and they will disrespect it. And,
if the Hindus did so, nobody would come to the support of the Muslims,
neither the administration nor the police nor anybody. I was taken to a shop
owned by a Muslim man and shown a formerly Muslim house that had been
purchased by Hindus for a very small amount. I was shown a small community
hall that had been used by persons from the Muslim community for mar-
riage ceremonies or other parties. I was shown a small platform and walls
that had been damaged by Hindus. The place was occupied by buffaloes
belonging to Hindus, who are trying to seize the property. I also met the
younger cousin-brother of the respondent, who raised his shirt to show me
the wounds he had suffered from the bomb blast when he was a youth. (See
Figure 6.7.)
During the latter half of this tour, a young Hindu man appeared in an
alleyway on a motorcycle. He idled his cycle and stopped to watch us as we
moved about. In a brief exchange of words with him, he urged me to meet
with the Hindu municipal corporator from the area. I made an appointment
to do so and met the corporator the next morning, but he had virtually noth-
ing to say to me.
Summary. How can we summarize the situation in Sarai Sultani? First, it
is a context in which Muslims live in a locality in which they feel and are in
fact surrounded by Hindu-majority localities. Second, it is not just an amor-
phous Hindu population that surrounds them, but a range of Hindu mili-
tant groups and organizations located at points all around their mohalla. Third,
 The Geography and Demography ofRiots / 195

FIG. 6.7. Victim of bomb blast in 1990 riots

these groups are said to engage and I believe do engage in provocative actions
directed against the Muslim population. Their very presence is a provoca-
tion since they are anti-Muslim organizations, promote a cult of violence,
and have an organization, the Bajrang Dal, specifically devoted to the prac-
tice of violence against Muslims. 64 Fourth, it is not only the militant Hindu
groups who are perceived as hostile to the Muslims, but other elements in
the population. My interviews contain references to specific acts in "normal"
times such as the alleged writing of nasty anti-Muslim curses on kites.
Further, the interviews in Sarai Sultani and elsewhere point to the existence
of other elements in the population who seek to take advantage of Muslim
discomfiture to acquire their property at cheap prices and to compel them
to move elsewhere. Fifth, the authorities, notably the police, are generally per-
ceived to be anti-Muslim in normal times, extracting more than the usual
share of graft from Muslim businessmen compared to that taken from
Hindu businessmen.65 In a word, the entire atmosphere in the area in usual
times is permeated with provocation, threat, insult, and the fear of injury or
sudden death in an outburst of violence. Finally, most important, it is sim-
ply not the case that riots occur only occasionally in such places as Sarai Sultani.
It was not just in the years 1978-80 and in 1990-91, in the riots reported in
 196/ The Geography and Demography of Riots

FIG. 6.8. Riot damage, Sarai Sultani, 1999

the newspapers, that violence occurred in this and adjacent mohallas. On the
contrary, riotous violence has occurred on many other occasions before 1978,
between 1978 and 1990, and after 1990, not all of which have been reported
in the press.
What about the how part of the question? How does violence break out
and how does it come to be directed against Muslims who are doing noth-
ing but sleeping in their beds during curfew? How do such ghastly atrocities
akin to Nazi violence occur in democratic India? First, the police are pre-
dominantly Hindu, many of them imbued with the same anti-Muslim feel-
ings as the general population. Second, there is a special police force, the PAC,
that is notoriously anti-Muslim. Third, repeated accounts in Sarai Sultani and
elsewhere make it dear that the police and tlle PAC work hand in hand with
members of the militant Hindu organizations in beating and killing Muslims.
But what about the atrocious character of some of these acts of violence?
How can these be explained? It is idle to seek explanations of that type. They
 The Geography and Demography of Riots / 197
can no more be provided in a convincing manner than can explanations for
equally atrocious acts committed in Nazi Germany and around the globe in
today's world in Cambodia, Rwanda-Burundi, Sierra Leone, and Kosovo. The
most one can say is that such acts in some cases take place within a context,
a discourse, that isolates and demonizes a particular group of people-in this
case Muslims, their religion, and their social practices-and characterizes
them as a physical danger to Hindus and to the unity of the country. In such
contexts, not only will "ordinary" injuries and killings by stabbing, bomb-
ing, and gunshot take place, but so also will major atrocities.
But there is something further to be said concerning how such atrocities
and riotous behavior in general can be allowed to continue in places such as
Sarai Sultani. What about the higher authorities in the district, leaving aside
the state authorities for the moment? We have in this respondent's account
a rather clear indication that social origins and the attitudes of the authori-
ties matter a great deal. It is not the case that, if caste Hindus occupy the posi-
tions of DM and SSP or CO, violence against Muslims will be allowed, and
that if Muslims are in those positions, violence against Muslims will not be
allowed. Indeed, there have been occasions in which Muslims in authority
have appeared as helpless to prevent violence against Muslims as their Hindu
counterparts. It is, however, a matter of great interest that violence against
Muslims was said to have been prevented on two occasions in the history of
Sarai Sultani when Scheduled Caste men were in positions of administrative
and police authority. What these men were said to have done, moreover, was
what many caste Hindu officers can and will do when their own sense of pro-
priety or orders from the state government lead them to it. They made it clear
that they would not tolerate acts of violence, that they would use necessary
force against Hindus as well as Muslims to prevent them, and that every police
officer would be held responsible for preventing acts of violence or contain-
ing any that occurred under their eyes.
There is one other important question to be answered, namely, as to the
timing of the larger, uncontrolled riots. The two largest events of this type,
as we have seen, were the long series of incidents between 1978 and 1980, on
the one hand, and 1990-91, on the other hand. The first took place during
the period of turmoil between the elections of 1977 and 1980, when politics
at the Center, in the state, and in the districts were undergoing major trans-
formations that included the fall and return to power of Indira Gandhi and
the Congress at the Center, Janata rule in the central and state governments
in which former Jan Sangh members and RSS cadres were included, and cor-
responding changes in Aligarh City, in short, a period of intense interparty
 198/ The Geography and Demography of Riots

competition. The 1990-91 riots took place at another dramatic turning point
in national, state, and local politics, namely, during the mass mobilizations
around the issue of demolition of the Babri Masjid that the BIP used to fur-
ther the downfall of the Congress and the defeat of its other set of rivals in
the Janata Dal. These were times, also, when the local administrations in
Aligarh and elsewhere in the state were not given clear instructions to pre-
vent riots, or where the local authorities themselves were sympathetic to Hindu
mobilization that had clear anti-Muslim aspects. So, two elements seem to
be required to explain large-scale rioting: intense political competition in which
Hindu-Muslim relations are involved and in which voting by community is
expected or encouraged, on the one hand, and an administration that does
not have instructions or its own will to act decisively to prevent or control
rioting.
Before turning to a detailed analysis of the broader political context in
which large-scale riots take place, however, I want to consider further the role
of economic factors in the riots in Aligarh, which my informants and the doc-
umentary sources have repeatedly mentioned as among the important causes
of rioting. I will show in the next. chapter that economic factors must be
included in any comprehensive account of the underpinnings of Hindu-
Muslim violence in Aligarh, but that they, too, comprise only a part of the
whole.
 7 I The Economics of Riots
Economic Competition and Victimization

THE ECONOMY OF ALIGARH

A ligarh, like several other towns in western U.P. that are also riot-
prone, has developed in the twentieth century from what is known
. in the Indian literature as a qasbah town dominated by a rentier class
of rural landholders-from both the Hindu and Muslim communities-with
an economy "based on small shopkeeping, service occupations, and cultiva-
tion of crops," to a semi-industrialized economy in which "manufacturing,
processing, servicing and repair operations" now provide jobs for approxi-
matelya third of the workforce. l Modern industrial activity began in Aligarh
during British rule when cotton cultivation expanded for the export trade in
the nineteenth century in the rural areas of the district and ginning and press-
ing factories were established in the town by both British and local Hindu
and Muslim entrepreneurs. 2 However, the cotton industries declined and
mostly disappeared long ago. Aligarh has provided a site for a few other large
industries, including "dal [lentil] and vegetable oil mills, an instrument fac-
tory, and the Government of India Press."3 A Glaxo factory was also estab-
lished on the outskirts of Aligarh in 1961.4 The large-scale industrial sector is
not, however, a major force in the economic life of the city either in its contri-
bution to production or to the workforce. Nor are there any large-scale trade
unions in the city.s
The predominant industry in the town now is lock manufacturing, for
which Aligarh has long been famous. 6 This industry, which also began in the
town in the late nineteenth century, has traditionally been dominated by
Muslims, with its manufacturing units heavily clustered in tlle predominantly

199
 200/ The Economics of Riots

Muslim center of the old city known as Upar Kot. Lock manufacturing has
also been the principal industry in which Muslims have been involved: half
of all Muslim-owned industries in Aligarh in the early 1980s were engaged in
lock manufacturing? Nowadays, however, Hindus participate in the indus-
try as both workers and owners of manufacturing establishments. 8 Although
small- and medium-scale factory production oflocks has probably increased
in recent years, the work traditionally has been done and is still largely done
by "artisans at home. "9
Insofar as "the marketing and supply of raw material" is concerned, how-
ever, this aspect of the business "is controlled mainly by Hindu Bania mid-
dlemen."lo The metal scrap dealers who provide a good portion of the raw
material for the lock industry have in the past all been Banias, from two castes
in particular: Agarwals and Varshneys.ll As noted above, these two castes have
often been considered to be antagonistic to each other in social life, to have
for some time followed different political paths, and also to have been in com-
petition with each other in trade. Persons from the Varshney caste have, as
we have seen above, provided the principal leadership for militant Hindu pol-
itics and anti-Muslim agitations. Recently, Mann noted that "two Muslim
families" had entered the scrap trade, "hoping to break the Bania monop-
oly"lZ in that business.
Before the scrap metal reaches the lock manufacturers, it must pass
through foundries, "which are operated by Kolis," a Hindu Scheduled Caste,
and then through "rolling mills, owned by both Muslims and Hindus. "13 There
is, therefore, in this most important industry in Aligarh, what Mann describes
as "considerable interdependence ... between Hindus and Muslims."I4
However, intercommunal relations among Hindus and Muslims involved in
the lock trade in the town are virtually confined to commercial activities. Only
when persons from these two communities have attained a level of prosper-
ity and sophistication to move to the Civil Lines area of the town do their
relationships extend beyond business to social ties.15
Also of relevance both to the lock industry and to other commercial and
industrial enterprises in the city is the fact that "almost all" moneylenders are
"Punjabi Hindus. "16 Punjabi Hindus, many of them refugees or descendants
of refugees from Pakistan at the time of partition in 1947, are, along with
Varshneys, the principal supporters of militant Hindu ideas and organizations.
A further-and often-cited-example of Hindu-Muslim economic rivalry
and competition in Aligarh and one that has also been related to communal
riots in some of my interviews concerns the "trade in meat and hides." The
slaughter of animals-particularly, of course, of cows-is anathema to many
 The Economics of Riots 1201

Hindus in India. Even where it is not religiously proscribed, as in the case of

the slaughter of goats, whose meat is relished by many Hindu castes, the occu-
pation of slaughtering is held in the lowest esteem. It is no surprise, there-
fore, that this occupation has been traditionally monopolized by Muslim
butchers from the caste known as Qasais and by low castes, formerly called
Untouchables. In Aligarh, hostility between these two communities arising
out of competition to "control the lucrative trade in meat and hides"17 has
been often reported. Is Further, there are some areas of the town that are riot-
prone in which Qasais and low-caste Hindus live side by side. It appears that,
at times of riotous violence, these two groups do come into conflict with each
other.19 In some riot-prone areas, Qasais also live in mohallas adjacent to large
Varshney populations. Some interview respondents gave greater weight to
Varshney-Qasai conflicts as riot-precipitants than to conflicts between Qasais
and low castes.
Competition between Hindus and Muslims in several other trades and
industries in Aligarh has been reported, but as often as not there is as much
or more competition between sections of Muslims as there is between Hindus
and Muslims. Nevertheless, in her interviews in Aligarh, Mann found that
the rise of various Muslim baradaris and their entry into various parts of the
growing urban economy of Aligarh have been considered to be among the
main causes for Hindu-Muslim violence in the city.:w
Beyond the lock industry, the two most important industries in Aligarh in
which Muslims are involved are building fittings and brass casting and hard-
ware. 21 Most of the Muslim-owned factories are located in the Upar Kot area
(see Map 2),22 notably in the mohallas ofTurkman Gate, Tantanpara, Usman-
para, Delhi Gate, Bani Israilan, Sarai Mian, and Chauk Bundu Khan, all of which
save the last two are classed as communally sensitive, riot-hit, and/or crime-
prone (see Map 3) but there are many other mohallas in the old city as well as
in the Civil Lines and the new Industrial Estate in which Muslim-owned factories
are located. Muslims also have been increasing their share in small-scale indus-
tries and have a small share in the "handloom and weaving sector" as well. 23
Like every other town of comparable size in India, Aligarh also has its share
of bazaars of many types, some carrying a range of goods, others specializ-
ing in particular types of goods. In the former category is the extensive bazaar
known as Rasalganj,24 which begins on the southern side of the railway and
extends up the Rasalganj Road into the Upar Kot area. In the more special-
ized category are included bazaars such as Bara Bazaar, specializing in cloth
and shoes; Dube Ka Parao, the vegetable and grain market; Mahavirganj, with
its grain and grocery shops; Sarrafa Bazaar, with its jewelry shops; Phul
 202/ The Economics ofRiots

FIG. 7.1. Muslim medical clinic in Sarai Hakim

Chauraha (Flower Crossing), the flower bazaar; Sabzi Mandi, as its name indi-
cates, the wholesale vegetable market; Sarai Hakim, containing many Muslim
medical clinics (see Figure 7.1 and Map 2); and Centre Point, a new market
area where many different types of goods are sold and containing some rela-
tively modem shops and restaurants. Some markets, such as Barahseni Bazaar,
are named after particular castes. 25 Some of the bazaars are dominated by
Hindu, others by Muslim shopkeepers. Most of the shops in Phul Chauraha
and Sarai Hakim are owned by Muslims,26 those in Dube Ka Parao, Sarrafa
Bazaar and Mahavirganj by Hindus. 27
Communal riots often begin in these extraordinarily crowded bazaars,
which are also commonly the principal targets for arson and looting once
riots begin. Among the worst-affected areas in the riot of 1978, for example,
were Sarrafa Bazaar and Phul Chauraha. 28 Indeed, the first clear signal that
a riot has begun or is about to begin is the sudden slamming down by shop-
keepers of the iron shutters that enclose their shops at the end of the day.

ECONOMIC FACTORS IN RIOT PRODUCTION IN ALIGARH

There have been many suggestions in the literature on riots in India in gen-
eral and in Aligarh in particular, some of them noted in the previous chap-
 The Economics of Riots / 203

ter, that they have their main foundation in economic competition and rivalry
between Hindu and Muslim businessmen, on the one hand, and in attempts
by Hindu commercial interests and real estate developers to grab valuable
urban land owned or occupied by Muslims. Thus, the Minorities Commission,
in its report on the 1978 riots, noted that it had "been suggested that the eco-
nomic rivalries between groups of businessmen of the two communities who
thrive on the lock-making industry at Aligarh may be the reason behind the
tension" that led to the riots.29 The press also reported on the 1978 riots the
view of "senior officials" in the district that "business rivalries between
Muslims and upper caste Hindus have played their role in engendering feel-
ings of hatred" that in part explained the spread of riots in previously
unaffected areas of the city.3 0
Similarly, the Centre for Research study noted that "the most sensitive
and intensely riot-hit areas are located in the CBD [Central Business District J,
where there is a preponderance of Muslim houses and where most of the
trading and commercial activities are in the hands of the Hindus."31 The
Centre study also remarked that, owing to "the phenomenal rise in the price
of land in the Aligarh town," Hindu Banias had been seeking to force "Mus-
lim artisans to vacate their houses and shops located in the business centres
or more developed residential areas" such as Manik Chauk and Dahi WaH
Gali, where, as noted above, "killings are known to have been organised by
the local roughs in league with the land hungry rich of these localities. "32 The
Centre's information was based on its study of "the riots of October, 1978,
and May-June, 1981."33
Many Muslims in Aligarh concur in the economic explanation for the
recurrence of riots in Aligarh, seeing them as a form of minority persecution
for the sake of Hindu profit. It is said, for example, that Banias from the busi-
ness classes, many of whom are Jan Sanghis,34 attack Muslims in order to
expand their business in the city. They precipitate riots in order to try to force
the Muslims to leave and sell their houses at low cost-or even at high cost-
but mainly to get their property somehow. So, they create communal tension
and then, during riots, attack Muslim houses.
It is said that Hindu big businessmen also have other economic motiva-
tions concerning Muslim laborers in their own establishments. They turn labor
troubles into communal troubles so they can avoid giving proper pay and
bonuses to their own workers.35
Perhaps the most common explanation of an economic type that I have
heard repeatedly over the years concerns the AMU and the way it is perceived
in the town, especially by teachers and students in the local colleges that are
 204 / The Economics of Riots
primarily Hindu-run and have mostly Hindu students. It has been said repeat-
edly that there is deep resentment in these local colleges among the teachers
there over the advantages Muslim professors at the university have, particularly
salaries that are much higher tllan tll0se prevailing in the colleges.36 Since most
of the riots noted above involve teachers and students from the local colleges
in one way or another, this explanation is quite satisfying to many observers
of them.
The most bizarre explanation of this type concerned the vicious rumors
circulated during the 1990-91 riots that Hindu patients were being slaugh-
tered in the AMU Medical College Hospital. The rumors, as we have seen,
were totally false, but the question remained as to why they were circulated
in the first place. The economic explanation was that private medical clinics
instigated these rumors about the AMU to make money by diverting busi-
ness to them!37
While not all these explanations can be discarded entirely, none of them
constitute anything close to a reasonable explanation for riots of the scale
that have been produced in Aligarh from time to time since 1961. At best, they
suggest the operation of economic motivations that impel some participants
to participate in such riots, to take advantage of them when they occur. While
economic interests have been pursued in such a way in Aligarh as to precip-
itate riots, notably the incident of the cinema hall that constituted the start-
ing event in the incidents that occurred in May-June 1979, such interests
cannot account for most of the events that tllen follow.

THE ECONOMICS OF RIOTING

IN MANIK CHAUK MOHALLA

The names of one community in Aligarh, the Barahsenis (or Varshneys, as

they are commonly known), and of one man from that community, Krishna
Kumar Navman, recur repeatedly both in my interviews and in published
accounts of the origin and course of riots in Aligarh since Independence.
Varshneys are heavily concentrated in the locality of Manik Chauk, which
we have seen is the locality most often mentioned as a center of riotous activ-
ity. Many Varshneys also live in the locality of Brahman Tola. It was noted
above that they have their own market as well, called Barahseni Bazaar. The
Varshneys and Navman are also associated in the minds of many with the
RSS and with the Jan Sangh/BJP. Navman is himself a businessman, a pub-
lisher of commercial books and calendars,38 whose large warehouse is located
in Manik Chauk.
 The Economics of Riots / 205

Varshneys and other business castes in Aligarh have often been economic
and political rivals, competing with each other in similar business lines and
supporting different political parties. Agarwals, in particular, have been busi-
ness rivals of the Barahsenis. Many Agarwals, in the early years after
Independence, supported the Congress. In the famous 1962 election for the
Aligarh Legislative Assembly constituency, a Varshney candidate, Madan Lal
Hiteshi, contested as an independent and placed third behind the Republican
Party and the Congress, while Tota Ram Vidyarthi, a Marwari Bania, con-
tested on the Jan Sangh ticket, coming in fourth place. From 1989 through
1996, Krishna Kumar Navman was the BJP candidate for the assembly con-
stituency, while all his principal rival candidates were Muslims. By then, most
Hindus from the business classes had become BJP supporters.39 Navman was
elected in three successive elections in 1989, 1991, and 1993 before he was finally
defeated in 1996. (See below, Chapters 9-12.)
Muslim respondents single out the Varshneys among Hindu castes,
including other business castes, as exploiters of poor Muslims, and always
mention Navman when they discuss the activities of the members of this
caste, as in the following quotation from one of my interviews with an AMU
professor.

There is a community called Barahseni, Banias. These people ... all belongs
to RSS ... and that is also a special factor in this [communal/riot situation].
Agarwal Banias are not so bad, but in Aligarh these people [Barahsenis] are-
this Navman comes from the same community-and they are actually very
moneyed people. They have money, they use that money [implying their use
in riots J. 40

The names of the Varshney community and of Navman are particularly

prominent in accounts of the 1978 riots. Although its origins had nothing to
do with the community, Navman, or the locality of Manik Chauk-having
arisen out of the rivalry between two wrestling akharas-it has been reported
that "the extremely communal minded section of the Hindus," that is, the
Varshneys, "appeared to have planned attacks on Muslims" after the death
of the Hindu wrestler, Bhure Lal. This same report records the allegation that
"rich Varshney Hindus" provided the "money for this plan" and that Navman,
then "president of the Aligarh City Janata Party and his dose associates, by
all accounts, seem to have played a very prominent role in these plans and
their later execution."41
This report of the Minorities Commission does not, however, explain why
 206/ The Economics of Riots

the rich Varshney businessmen decided-made a "plan"-to attack Muslims

in Manik Chauk. Other reports, however, do provide such an explanation,
a quite specific one that has nothing to do with extreme communal-
mindedness. It is said, rather, that as a consequence of a "phenomenal rise
in the price of land" in Aligarh in the years preceding the 1978 riots, there
developed a "tendency to force the Muslim artisans to vacate their houses
and shops located in the business centres or more developed residential areas
dominated by Hindu Banias."42
One of my own interview respondents gave an account of the origins of
these incidents that fully supports the above documentary reports. These inci-
dents that occurred between 1978 and 1981 were not actually riots, he said.

There was a-actually-not a riot. A section of the-Manik Chauk, predom-

inantly of the-Banias, yah, pred.orninantly of the Jan Sangh community, busi-
ness class community-in Manik Chauk-attacked ... the Muslims.

When I asked why, this respondent continued as follows.

Why? Just I will tell you. Because there is a very few Muslims in this mohalla,
and they are dominating-Hindus [are J. And they want the Muslims [to leave 1
this place. Because-they-they, ... for expansion of their business in city. City
is very costly: lands, and houses. Particularly in this-in that area, in Manik
Chauk. And they could not expand their business, their houses. [So J there is
the costly lands. Very costly houses. They want to harass the Muslim so
Muslim[s] will [leave] this area and sell their houses on-very low cost-yah,
any cost, low or high cost; the-there is no problem of the money; they want
to purchase their houses, they want tocome-mm-set[tle] ... -on this area,
particularly. [SoJ this is the major. Then-when the communal tension [is]
created, they attack ... the Muslim houses. 43

However, this attribution of specifically economic motives to the busi-

nessmen of Manik Chauk and Dahi Wali Gali localities is not a sufficient expla-
nation, for it must be asked why only Muslim artisans were targeted, not Hindu
artisans. Either there were none of the latter, which is unlikely, or we are back
again to a communal explanation, at least in part. A wholly satisfying expla-
nation for the 1978 attacks on Muslims is that the communal atmosphere and
the opportunity to mobilize people for retaliation against Muslims because
of the death of Bhure Lal provided a cover for and the opportunity to take
possession of Muslim property. To do the same against small Hindu propri-
 The Economics ofRiots / 207
etors would expose "the land hungry rich" as merely that, and would arouse
hostility against those "known" to have organized the attacks. In effect, there-
fore, the communal and the economic explanations for riotous activity here
merge. The explanation is satisfying in the sense that it is complete and leaves
no unexplained motivations; it is not, however, necessarily the truth, and cer-
tainly is not the whole truth for rioting that extended beyond Manik Chauk.

RIOTING FOR PROFIT

There is no doubt that, in Aligarh as well as in many other cities and towns
in India, including notably Bombay, urban land-even the smallest portion
in the nastiest slum-commands a high premium. 44 Mann has noted that, in
general, "the urban development of Aligarh has accelerated building activ-
ity, while growth of the urban population and subsequent pressure on exist-
ing housing and related facilities have created a high demand for residential
and business premises."45 This demand opens up many kinds of potential
conflicts that are exacerbated whenever persons from one community are
affected by the demand for such residential or business premises by persons
from the other community. For example, the issue of control of scarce land
in Aligarh also featured in the incidents that followed upon the opening of
the Chandra Talkies cinema in June 1978, which was protested by local Mus-
lims and by Muslim politicians, including Khwaja Halim "and a volatile young
advocate, Sujatullah Khan, local President of the Muslim League. "46 In this
case, however, the issue was not one of Hindu businessmen gaining control
of Muslim property, but of a Hindu businessman opening a cinema hall in
an alley that opened up into a Muslim locality, as a result of which, it was
claimed, the privacy of the residents, particularly of the women, would be
threatened by the exiting of large crowds into the alleys late at night.'*?
Furthermore, in this case, it appears that the prime movers in enlarging the
scope of the issue were Muslim politicians. This set of incidents, therefore,
does not provide much support for the view that economic motivations are
primary in Hindu-Muslim rioting.
The economic explanation for riots that focuses upon the land grabbing
of the Varshneys and other rich Hindu businessmen falls into the category
of the rich exploiting the poor and vulnerable, "the business classes against
the poor Muslims,» as one of my respondents put it. A variation on this theme
was provided by the same respondent and by others, who alleged that "big
business houses" took advantage of communal riots to solve their labor prob-
lems; they, in effect, turned their labor problems into communal problems,
 208/ The Economics of Riots

thereby making a profit by not giving their laborers "proper money, proper
finance, proper bonus, et cetera," during times of "communal troubles."411
Quite similar statements were made to me by a Hindu station officer at Sasni
Gate, a major area of communal conflict. This officer also provided a list of
other alleged advantages gained by businessmen during riots. He was respond-
ing not to a question concerning the causes of communal riots, but to the
question of who benefits from persistent communal riots in Aligarh. Though
the question was, therefore, loaded, insofar as it assumed that some people
must benefit, I left the respondent free to choose the beneficiaries. He chose,
first of all, businessmen.

Sometimes during the riots certain people have good business. Laborers on daily
wage have no job during curfew. They go to the industry bosses and ask for
financial help, which is offered, but the bosses exploit the situation afterward.
They may give a low wage, and the laborer has an obligation because he was
helped during bad days. so it becomes almost like bonded labor. Also, the labor-
ers prefer to work at factories that can still run during riot or curfew, so again
they are forced to work for lower wages. So business owners benefit.
Sometimes, during the riots, there are certain people who have good busi-
ness; they may be Hindu and Muslims. What they do is that they have labor-
ers on daily wages. And during curfew, these people go out of-er-they don't
have any job or any work to do. So-and they go to their ... [employees1...
and promise financial help to them.
And exploit the situation afterwards, because they may give them ... not
the proper wage, and aftelwards-it also puts them under obligation because ...
they always tell them that, "We helped you during the bad days." So later on
they also employ them on lower wages. So somehow they become bonded labor-
ers. And the other thing is that there are certain people who manage to run
their installations, and during that time because of the daily wages, he depended
upon his wages, he prefers to work there. And again, because of the fact that
he does-he doesn't work there, so then he might not get that job, because
there are other people who are willing to work. So then again he is supposed
to work on lower wages. So some-that way, the business owners-they benefit
a lot in this kind of situation. 49

RIOTS AND ECONOMIC COMPETITION

Other economic explanations refer to more general competition between

Hindus and Muslims in particular economic niches. There is, for example,
 The Economics of Riots 1209

the claim that some of the rioting in Aligarh is attributable to "increasing com-
petition between [Muslim ... butchers] and Harijans in the beef and meat
business and, with their mohallas opening into each other, incidents are more
and more frequent."5 0
As in the case of Hindu businessmen of different castes who are in compe-
tition with each other, so there is competition among Muslims from different
segments of the Muslim community-lineages and baradaris-for leader-
ship within the Muslim community as a whole. Success in "business and indus-
try" for such Muslims provides them with an opportunity for recognition
within their communityY However, Muslims of all baradaris in Aligarh in
general have been far behind Hindu business castes in the economy of the
city. In recent years-that is, since the 1970S and 1980s-Muslims have entered
into direct competition with Hindu castes in particular economic niclles dom-
inated by HindusY
Even where there is "interdependence" rather than competition in busi-
ness relations between Hindus and Muslims in Aligarh, it does not extend
beyond "commercial expediency," at least in the old city. 53 Moreover, even
in such situations of interdependence, let alone competition, the organiza-
tion of business and the structuring of business relations tend to follow kin-
ship, caste, and baradari lines.54 The caste and baradari-not just the religious
affiliation-of every business owner in Aligarh will certainly be known by
everyone who deals with him. Further, once a member of a caste or baradari
becomes established in a particular branch of business or industry, others of
his community will soon follow,55 among whom relations of cooperation are
more likely to develop than with persons from other status groups. This does
not mean, however, that only persons of the same caste, lineage, or baradari
as the owner(s) will be employed. On the contrary, as will be shown momen-
tarily, employees often come from members of the other community and from
different subgroups within each community.
Whether or not economic competition between Hindus and Muslims is
a contributory cause in the production of communal violence in Aligarh,
communal violence and the tensions that have intensified between the two
communities as a consequence have adversely affected "economic relations,"
even leading "in some cases to severance of long-established business links. "56
As early as the 1950S, Pars Ram, reporting on the results of interviews with
fifty Hindus concerning their relations with Muslims in his famous UNESCO
Study of communal tensions in Aligarh, done in collaboration with Gardner
Murphy, remarked that 32 of the 50 met with Muslims "as business partners
and as acquaintances, and some of them at meals, but they entertain an inward
 210 / The Economics of Riots
image of the Muslim as dirty, a cow killer and a cruel person."57 Since then,
as Varshney has noted,

Religion has continually interfered with business. For some time, trade and busi-
ness associations of the city tried to weather communal tensions and integrate
businessmen from both communities, but the effort did not succeed. Trade
associations were broken along political lines, rather than business lines. 58

Furthermore, the fact that a proprietor employs or depends for his supplies
or finished products upon artisans or traders from the other community does
not necessarily mean that his prejudices against the other community are
diminished. Here again the Varshney community enters the picture, at least
in the mind of one of my Muslim respondents.

Artisans are Muslims. They are making locks and they are supplying them and
these Banias are getting richer and richer on the poor Muslim artisans. They
[Barahsenis] do nothing.... They [Muslims] make locks for them, they
[Barahsenis] take the locks, they collect the lock [s], ... and then they sell [them J.
But, they're all [the locks] done by Muslims.
Now probably, now they are more afraid because now the educated Muslim
boys are coming up and now they are trying to hold that business in their own
hands and that is also agitating them. Now, the educated boys from Aligarh
University, they have started businesses. Very few, but they are entering into
this and they want to do business themselves, they don't want to go to the Banias
and say, "Please give us five hundred rupees advance and we'll do," and then
they become their, under their control for all the life to come. So, probably this
is also maybe a slight cause in their mind, is now the business is going out of
their hand.
There was one businessman called ______ Kumar, he was a top
businessman of locks and entire force, manufacturing force, was Muslim, and
he was an RSS man. And the reason [for his RSS affiliation] I could not under-
stand. He used to be in Marris Road, a very big man. Still he has a very big bun-
galow there. And he used to say to us, oh, I'm so secular, my entire business is
because of them and whatever I am is because of Muslims and whatever money
I have is because of them, but he was a man of RSS.

This respondent went on to contrast the situation in Moradabad, where

Muslim competition with Hindus in traditional industries-in this case, brass
utensils-was more intense as a consequence of the entry of educated Mus-
 The Economics of Riots 1211

Hms into the sale of these utensils previously monopolized by Banias. He

saw a similar process developing in Aligarh. Moreover, he also revealed how
economic advancement of Muslims and their competition with entrenched
Hindu business groups may spill over into politics. He mentioned the name
of a former student of his in the AMU, who was now doing very well in the
lock industry. Having become established in this industry, he and others
thought the former student should now make his mark in Aligarh politics as
well.

We asked him [to] please contest for the mayorship or chairmanship of the
municipal board and he gave a very tough fight because the money mattered,
votes are bought. So, he was able to fight with a Bania, O. P. Agarwal, he is RSS
man, and he [the Muslim] was only defeated by ... one vote. 59

In fact, there has been a dramatic increase of Muslim participation as

owners in several industries during the past three decades. In the lock indus-
try, traditionally dominated by Muslims, the number of registered units (that
is, factories rather than small-scale shops) owned by Muslims increased from
4 in 1965-75 to 33 in 1976-80 and to 87 in 1981-85, a very substantial increase,
indeed. The second most important industry into which Muslims have
advanced as owners is building fittings (hardware). Muslim unit ownership
increased in these three periods from 2 to 9 to 24, but nearly 90 percent of
such units were Hindu-owned in 1982.60 Such increases in Muslim owner-
ship do not necessarily involve direct competition with Hindus, for the com-
munal division of labor in the industry is more complementary than
competitive. Muslims have traditionally produced the locks, either as arti-
sans or, increasingly nowadays, as laborers in medium-size factories. Hindus
have provided the raw materials through their dominance of the scrap trade
and have purchased the finished locks for sale ou tside of Aligarh.61 Even today,
the trade in locks is mostly controlled by Muslims, while the manufacturing-
artisanal or factory-is carried on by MusHms. 62
In the 1996 Legislative Assembly elections, Navman was defeated by Abdul
Khaliq, a second-generation Muslim lock manufacturer, graduate of the AMU,
and owner of a factory employing about one hundred persons and produc-
ing a thousand locks per day. Abdul Khaliq gave an ambiguous response to
questions concerning whether or not communal tensions in the city had
affected business relations between Hindus and Muslims. At first, he claimed
that Hindus and Muslims in the lock business had good relations, being
"involved with each other," and denied that there was any business conflict
 2121 The Economics ofRiots
between Hindus and Muslims as such in this regard. When I pressed him fur-
ther, however, he said that not only in Aligarh, but in India as a whole, when-
ever Muslims "dominated" economically, the "BJP, or say RSS, want to disturb
things."63 In other words, in this man's mind, the cause of communal ten-
sion did not arise out of intercommunal business relations, but out of polit-
ical action on the part of militant Hindu organizations. 64
Muslim employment in the small-scale sector, outside the lock industry,
is quite low, probably less than 3 percent. 65 Nevertheless, the labor force in
some small and medium enterprises is communally mixed. Hindu owners
employ Muslim artisans and workmen; so do Muslim owners employ Hindus.
For example, a Muslim proprietor of a small building-fittings enterprise in
Vpar Kot employed 15 people. Of these, seven were Hindus, including by his
own account his "main labor[er] who is guarding my factory and is [on]
twenty-four-hour duty." The other six were employed in casting and mold-
ing his brass items. He was emphatic that he experienced no problems with
either his Hindu employees or with Hindu competitors. His business rela-
tions in Aligarh were only with laborers and traders. He claimed that the ques-
tion of competition with Hindu rivals in the city did not arise at all since he
was supplying his goods only to exporters in Bombay. He acknowledged that
some of the Hindus of V.P. are "communal-minded," but said that "outside
of V.P., they are not [sol communal-minded. "66
Our lock-manufacturing informant in Sarai Sultani who gave the exten-
sive account cited in the previous chapter emphasized communal hostilities
on the part of Hindus more than economic factors as an explanation for the
recurrent rioting in Aligarh and in his mohalla in particular. At one point,
however, I said, "All right, [you say1the Hindus feel why not throw the Muslims
out of here? Maybe this was the reason. But was there any rivalry, jealousy,
some Hindus here making locks like [you dol, who want to take over [your]
business, or some such reason?" He acknowledged that there certainly was
competition between Hindu and Muslim lock makers in Aligarh and gave me
a long explanation of its several aspects. However, there did not appear to be
a direct connection in his explanation or in his mind between this economic
competition and communal violence. At the same time, he gave examples of
how the police victimized Muslim businesses, including his own, and helped
his competitors.
He explained that there was significant price competition between Muslim
and Hindu lock sellers, the former selling much more cheaply than the lat-
ter. He added that they were able to do so because their household expenses
were less than those of the Hindus and they could, therefore, subsist on a
 The Economics of Riots / 213

smaller profit margin, and because the Muslim manufacturers did their own
manufacturing while the Hindu sellers contract the manufacturing out to
laborers. For the latter reason, the Hindu traders faced heavier capitalization
costs for raw materials that had to be purchased, then contracted out, while
the Muslim manufacturers used the same raw materials in a much shorter
time period-two or three days compared to 15 or 16 days for the Hindus.
All the raw materials come from Hindu suppliers, but the respondent did
not complain of any discrimination from this quarter. Finally, the Muslim
manufacturers sell to nearby Dellii, a market that they monopolize, while the
Hindu sellers have huge transport costs to their more far-flung markets in
south India.
Given this description of economic competition between Hindu and
Muslim lock manufacturers and sellers, I asked if this respondent thought
that the lock industry's Hindu owners were encouraging the police to beat
them, or the police were beating them just because they don't like Muslims.
He replied as follows.

These Hindus bribe the police to oppress the Muslims. For example, a [Muslim]
worker in his own factory had to pay Rs. 2,000 extortion money to the police
only two or three months previously. He was just a worker, an ordinary worker,
a laborer, earning Rs. 2,500 per month.

I asked how this could be and was given the following explanation.

Suppose a Muslim worker is working in a Hindu factory. He works there for a

month and gets paid forthat. For some reason or other, he quits and then comes
to his factory and works here. Then, the Hindu factory owner will come to know
where his former worker is working, will come here, ask for that worker, and
claim falsely that the worker owes him Rs. 5,000. Then, that Hindu manufac-
turer will go to the police station and he will return with the police and, in their
presence, will demand the Rs. 5,000 from the worker. Then, to settle matters,
the Muslim proprietor will have to pay Rs. 5,000 to the Hindu factory owner
and the worker will have to pay the Muslim proprietor back monthly, off his
wages. This was the case with the worker who had to pay Rs. 2,000, in fact, but
such situations have occurred several times. The purpose of this harassment is
to discourage the Muslim manufacturers, to get them out of the business.

Such extortion directed against Muslim manufacturers was said to be beyond

the ordinary police extortion that both Hindu and Muslim businessmen have
 214 / The Economics of Riots

to pay. The police, he said, come sometimes two or three times a month for
50, 100, or even 500 rupees. During the year, he will have to payout around
Rs. 5,000 to 8,000 in this manner.
Here we have a businessman, a secular person, not involved in Muslim
religious or fundamentalist organizations or causes, a person in fact who has
been a Congress man all his life and remained so as of the date of the inter-
view, long after the Congress had ceased to be a significant force in the pol-
itics of the city. As a Muslim businessman living in a predominantly Hindu
country, he has no compunction-like so many Jewish businessmen in
European history-in making compromises and adjustments to his envi-
ronment in order to sell his products, which carry the name of the Hindu
god Krishna. He, like other Muslim lock manufacturers, must buy his raw
materials from Hindus. He does not mention any discrimination on the part
of his Hindu raw materials suppliers, with whom, as Mann has pointed out,
Muslim lock manufacturers exist in an interdependent economic relation-
ship. There is also competition between him and Hindu lock sellers, though
they both manage to survive by selling their locks to different markets in
India.
Economic competition among lock sellers does not, however, appear to
be a primary cause of the riotous activity and anti-Muslim killings that have
occurred in Sarai Sultani and adjacent mohallas. A more relevant factor is
the pressure on the part of local Hindus to gain control of Muslim-owned
property, to put pressure on the Muslims, intimidated by the militant
Hindu groups that surround them, to sell their property at cheap prices.
However, I do not believe that this economic factor provides the motiva-
tion for the intimidation of Muslims that has existed in this area for decades.
This is a matter of judgment on my part, but it is my perception that Hindu
businessmen and real estate operators are taking advantage of a situation
created for other reasons, rather than creating the situation in order to take
advantage of it.

CONCLUSION

What can we conclude, then, concerning the role of business and business
relations-including dependency relations as well as interdependence, coop-
eration, and competition-between Hindus and Muslims in Aligarh as a fac-
tor in riot production? Modern business and industry were for long dominated
by Hindus. Muslims for the most part occupied positions as artisan manu-
 The Economics of Riots / 215

facturers and wage earners, producing and working for Hindu businessmen.
The increasing entry of Muslims into trades, occupations, and industries pre-
viously dominated by Hindus has certainly been noted by Hindus and
resented and feared by some of them. There is little evidence, however, that
such competition has been a factor in the production of riots. Rather, there
is more evidence that, in particular localities, competition between Hindus
and Muslims in the same traditional trades-such as butchering-has been
a factor providing recruits from particular castes and baradaris for riotous
activity against each other. There is also substantial documentary evidence
that rich Hindu businessmen have taken advantage of poor Muslim artisans
at times of communal riots and, as we have seen, have allegedly deliberately
planned their attacks upon the latter for economic advantage. There is the
further allegation that both Hindu and Muslim businessmen take advantage
of their laborers during riots to settle labor disputes, create dependencies upon
the owners, and reduce wages and other benefits in exchange for secure employ-
ment. Finally, there is resentment as well as competition arising from the fact
that some educated Muslims from the hated AMU have entered into direct
competition with Hindus in the lock industry and in transportation, among
other possible business lines.
However, none of these economic factors implicated in riot production
stand by themselves as sufficient explanations. They occur in a communal
discursive context and are almost always associated with political activities.
The communal context predates Independence, feeds upon the presence in
Aligarh of the AMU as a symbol of Muslim treachery to the projects of Indian
and Hindu nationalism, and derives further sustenance from traditional
prejudices instilled in Hindus and Muslims as part of their upbringing. This
discursive context in turn sustains political movements that thrive upon the
promotion of communal tensions even when they do not set out deliber-
ately to produce riots. The series of riots that occurred between 1978 and 1980,
the ones in which business interests have been most directly implicated,
occurred at a time of dramatic political change in India, of turmoil and trans-
formation in interparty relations and conflict in the state and in the coun-
try. That turmoil was reflected in Aligarh, where it took on a distinctive
character nourished by the historical breeding of Hindu-Muslim commu-
nal tensions, in which both Hindu and Muslim politicians played roles.
Navman and his Varshney supporters came into the limelight during the
October 1978 and May-June 1979 riots, while Khwaja Halim and other local
Muslim activists stood forth as the protectors of Muslim in terests against the
 216/ The Economics of Riots

Hindu owner of the Chandra Talkies cinema. These riots did not arise out
of business conflicts per se. Rather, Hindu businessmen gained economic
advantage and a Hindu politician-businessman gained political advantage
under the cover of communal action for the alleged benefit and protection
of the Hindu community, while Muslim politicians gained advantage by
standing forth as the protectors of poor Muslims against avaricious Hindu
businessmen.
 PART IV

Riots and the Political Process

 8 I Riots and Elections

INTRODUCTION

W iIkinson has argued that the principal cause of Hindu-Muslim

riots in India, indeed of ethnic riots in general, is a "close elec-
toral race in which one party believes it can win by appealing to
ethnic majority-group voters."l Although I have argued above that the search
for causes of such large-scale and multiplex events as riots is problematic, I
will nevertheless treat this argument seriously here, to discover what can be
said precisely about the relationship between riots and party/electoral poli-
tics in Aligarh in general and the intensity of interparty competition in that
city in particular. The opposite possibility will also be considered, namely,
that riots produce intense interparty competition. I have already established
the case for a close association between riots and politics in Aligarh in the
preceding chapters and elsewhere in my previous work. The association will
be further demonstrated here.
The analysis in succeeding chapters will focus on the Legislative Assembly
constituency, comprising most of Aligarh City, and the city segment of the
parliamentary constituency. A chapter will also be devoted to demonstrat-
ing the relationship between riots and the communalization and polariza-
tion of the electorate in particular mohallas. However, on the basis of the
discussion in Chapter 6, it is not to be expected that there will necessarily be
a direct relationship between riots and the intensity of electoral competition
at the mohalla level, for it has already been demonstrated that riots may be
produced from mohallas in which militant Hindu sentiment is predominant
and where that predominance is reflected in voting for the BJP. It has also
been shown that the mohaUas targeted for riot production may be over-
whelmingly Muslim or mixed, that there may be deliberate efforts to attack

219
 220 / Riots and Elections

and intimidate Muslims in places where the demographic balance provides

insurmountable obstacles to the BJP through the normal electoral mecha-
nisms and where other-than-political calculations are also involved. To put
the matter in a nutshell, riots are produced in specific mohallas, but the polit-
ical context that matters most is citywide, where the stakes are greatest. Thus,
riot production, whether concentrated in a few or spread in many mohallas,
influences to lesser or greater extent, respectively, the political balance in the
city as a whole, not just in the mohallas that experience violence.
The gist of my argument on the relationship between party politics and
riots was stated in one of my earlier works as follows: "there is a continuum
from political rivalry leading to communal riots to political rivalry feeding
on communal riots."z The continuum may, however, start at either end, that
is, from political rivalry to riots as well as from communal riots to intensified
political rivalry. However, the sequence in Aligarh has been primarily in the
latter direction, that is, communal riots have preceded and have led to inten-
sification of interparty competition. The mechanisms that lead to this intensi-
fication arise from the tendencies that follow from riots to foster increased
communal solidarity and polarization, which in turn are promoted by polit-
ical parties and/or individual candidates who stand to benefit from such sol-
idarity and polarization. The resultant communalization and polarization in
turn reduce the electoral prospects of parties and candidates who stand for
secular political practices, intercommunal cooperation, and class or castel
baradari mobilization rather than communal mobilization.
The operation of these mechanisms is, of course, affected profoundly by
the social composition of the electorate. Communalization and polarization
obviously tend to offer greater political and electoral advantages where the
numbers of any two potentially cohesive groups, in our case Hindus and
Muslims, are of a sufficient size. We expect neither riots, nor electoral advan-
tage to be gained from them, where there is an insufficiently large popula-
tion of one of the communities. There is no precise ratio between Hindus
and Muslims that can be stated to incline towards the production and polit-
ical use of riots. It depends as well upon other factors, particularly the divi-
sions within each community on the basis of caste or baradari and how
deep-seated these are. But there is no doubt that numbers matter. Indeed,
it will be demonstrated later that, in the given conditions of Aligarh-which
have tended for decades to produce from time to time riotous violence, com-
munalization and polarization of the electorate, and intense interparty
competition-important alterations in the boundaries and in the commu-
nal composition of the electorate have recently changed dramatically the elec-
 Riots and Elections /221

toral contests and the party configuration, reducing the political value of
riotous violence.

THREE PERIODS IN THE PARTY/ELECTORAL

HISTORY OF ALIGARH CITY

Table 8.1 gives the election results for 28 elections in Aligarh City held
between 1952 and 1998 for which I have information, including Legislative
Assembly, parliamentary, and mayoral contests. The mayoral and Legislative
Assembly constituency boundaries are somewhat different, the former com-
prising the entire municipality of Aligarh, whereas the assembly constituency
does not include some parts of the municipality, though it is entirely urban.
The Aligarh parliamentary constituency comprises five segments, each of
which is a separate Legislative Assembly constituen<..-y. Wherever possible in
the table, I have reported the results of the election for Parliament for the
Aligarh segment of the parliamentary constituency only, since the other four
of the five segments that constitute the parliamentary constituency are rural.
The Aligarh City segment of the parliamentary constituency and the Aligarh
City Legislative Assembly constituency correspond exactly.
The Legislative Assembly elections have constituted the dynamic center
of Aligarh politics for the most part since Independence, the mayoral elec-
tions having been held only sporadically. There were, however, citywide may-
oral elections in 1991 and in 1995. Until recently, therefore, the Legislative
Assembly constituency was the only arena that included most of the city. At
times, also, the parliamentary contest in the Aligarh City segment has been
as or more hotly contested than the Legislative Assembly contest, though the
results for the parliamentary constituency are inevitably decided by the votes
from the four predominantly rural segments. The results in the Aligarh seg-
ment are always quite different from those in the rural segments. They some-
times bring out even more clearly than in the assembly contest the communal
and political identities of Hindus, Muslims, and Scheduled Castes.
The table provides at a glance a summary view of the development of
party politics in the city over the past half century, which can be divided
roughly into three periods. The first period is the first decade of electoral
politics after Independence, marked by Congress dominance in the city.
Although that dominance was destroyed by the Republican Party in 1962,
the latter party faded away thereafter. (However, in the 1980s the Bahujan
Samaj Party, also a party representing primarily the lower castes, emerged
to capture the same support base of its predecessor party.) The second period,
 222 / Riots and Elections

TAB LE 8.1. Winning party or independent candidate, Aligarh City Legislative

Assembly, 10k Sabha, and mayoral elections, 1951-98a

Winning Party
Year Election Type or Independent
PHASE 1: Congress dominance

1952 Assembly Congress

1952 LokSabha Congress
1955 Assembly (bye-election) Socialist
1957 Assembly Congress
1957 LokSabha Congress

PHASE 2: Congress dominance contested by two rival parties

1962 Assembly Republican Party

1962 LokSabha Republican Party
1967 LokSabha Independent
1967 Assembly Jan Sangh
1969 Assembly Congress
1971 LokSabha BKD
1974 Assembly Jan Sangh
1977 LokSabha BLD
1977 Assembly Janata Party
1980 LokSabha Congress
1980 Assembly Janata Party (SC)
1984 LokSabha Congress b
1985 Assembly Congress

which followed immediately after the Congress defeat in 1962, lasted through
the 1985 election. It was marked primarily by struggle between the Congress
and two other rival parties: the militant Hindu Jan Sangh and the party led
by the peasant leader, Charan Singh, which took different names-.8KD,
Janata Party (SC), Lok Dal, and .8LD. In the table, Charan Singh's party is
shown as the winner in the 1971 Lok Sabha, 1977 Lok Sabha, and 1980
Legislative Assembly constituencies. The strength of militant Hinduism was
greater in this period than the table indicates because the Janata Party can-
didate in the 1977 assembly election was supported by the former Jan Sangh.
The Congress underwent a temporary resuscitation at the end of this period
 Riots and Elections / 223

TABLE 8.1. (continued)

Winning Party
Year Election Type or Independent
PHASE 3: Reconstitution of interparty struggle between the BJP and the JD/SP

1989 Assembly BJP

1989 LokSabha Janata Dal
1991 Assembly BJP
1991 LokSabha BJP
1991 Mayoral BJP
1993 Assembly BJP
1995 Mayoral BJP
1996 Assembly Samajwadi Party
1996 LokSabha BJP
1998 LokSabha Samajwadi Partyc

aResults in all 10k Sabha elections except those for 1952, 1957, 1971 and 1977 were for the
Aligarh City segment of the parliamentary constituency only. Segment-wise results were not
available in the other years. Since the parliamentary constituency consists of five nested seg-
ments corresponding to individual Legislative Assembly constituencies, the wirmer in the
parliamentary constituency as a whole mayor may not have won a plurality of votes in a
particular segment, in this case the Aligarh segment.
bThe constituency as a whole was won by the Janata Party (S).
<The constituency as a whole was won by the BIP.

after the assassination of Indira Gandhi and the rise to power of her son,
Rajiv Gandhi, as prime minister of the country. The third period, from 1989
to 1998, is marked primarily by the disappearance of the Congress as a seri-
ous contender in Aligarh politics and the reconstitution of interparty struggle
as one primarily between the BIP and the SP, the party of militant Hinduism
and the party representing the interests of backward castes and Muslims,
respectively.
Figure 8.1 shows the number of deaths in riots during the three political
periods in Aligarh. Since the 1988 riots occurred in the time between politi-
cal phase 2 and political phase 3, but closer to phase 3, they have been included
in the latter period. Because the periods are of different length, no statistical
assertions can be made from this figure concerning the relationship between
the configuration of party competition and the number of riots. The figure
merely illustrates the facts, namely, that (1) the lowest number of deaths in
riots occurred during the nine-year period of Congress dominance from 1952
 224 / Riots and Elections

1952-61 1962-88 1989-98

Political Period

FIG. 8.1. Number of deaths in riots by political period

to 1961, when there were three riots, two with no fatalities and one, in 1961,
with 15; (2) during the period of struggle between the Congress and rival par-
ties in the 16 years between 1962 and 1985, the number of riots with deaths
and the number of deaths both increased, to five and 62, respectively; (3) in
the ten-year period between 1988 and 1998 marked by the disappearance of
the Congress and the reconstitution of interparty struggle between the BJP
and the SP, there were four riots with deaths, and the number of officially
recorded deaths reached the maximum of 102.

ELECTORAL TURNOUT,
INTERPARTY COMPETITIVENESS, AND RIOTS

From the history of the electoral contests, we may learn much about the
prominent political aspirants, their political and ideological affiliations, their
caste and communal identifications, the intensity of popular political involve-
ment and of party competition, and the relationship of all these, particularly
the latter, to the history of communal riots in the city. Let us consider first
the overall pattern of party competition and displacement in the 14 Legislative
Assembly contests during the 45 years for which we have the results. The two
principal contending forces over most of this long period have been the
Congress and the militant Hindu parties. In the 14 elections, the Congress
 Riots and Elections / 225

TABLE 8.2. Winning party in 14 Legislative Assembly contests,

Aligarh constituency, 1952-96

Year Congress Militant Hindu Other

1952 X
1955 X
1957 X
1962 RPI
1967 BJS
1969 X
1974 BJS
1977 JP
1980 BKD
1985 X
1989 BJP
1991 BJP
1993 BIP
1996 JD

has won five times, the militant Hindu party six times, and other parties three
times (see Table 8.2).
Figure 8.2 shows the history of the constituency for the two leading polit-
ical forces from 1952 to 1996,3 namely, the Congress and the militant Hindu
parties. The latter term includes the Bharatiya Jan Sangh (BJS) from 1952 to
1974, the Janata Party in 1977, and the BJP in succeeding years. It is evident
from the chart that the electoral contest between these two parties in the con-
stituencies divides neatly for the Legislative Assembly contests into the same
three periods identified above: Congress dominance in 1952 and 1957, intense
competition between the two party forces from 1962 to 1985, and finally dis-
placement of the Congress by the BJP as the stronger force after 1985. This
long-term dual contest for predominance between the Congress and the mil-
itant Hindu parties has, however, been interrupted from time to time by par-
ties representing the lower and backward castes. In 1962, the seat was won by
the low-caste party, the Republican Party, and in 1980 and 1996 by parties
identified especially with the interests of the backward castes, the BKD in 1980
and the ]anata Dal in 1996.
In order to consider more precisely the relationship between interparty
competition and riots, it is necessary to look more closely at the temporal
 226/ Riots and Elections

70~----------------------------------------~

60
50
40
30
20 /
INC
5' 10 .--"', /
~ 0 '-, BJSlJP/BJP
1952 1962 1969 1977 1985 1991 1996
1957 1967 1974 1980 1989 1993

Election Year

FIG. 8.2. Vote shares for Congress and militant Hindu parties,
Legislative Assembly elections, 1952-96

proximity of riots and elections, and at their juxtaposition-that is, whether

riots precede elections with intense interparty competition or follow them.
In other words, if a causal connection is to be established, we need to know
whether interparty competition is a factor in riot production or whether riot
production intensifies interparty competition. Two standard indicators of the
intensity of electoral competition are turnout and the interval between the
winning candidate and the runner-up. Each will be examined in turn below.
In addition to these, two others suited to the specific situation in Aligarh will
be considered, namely, the degree of communalization and polari7..ation of
the electorate-that is to say, the extent to which voters vote as communal
blocs, and the extent to which, in particular elections, they vote as antago-
nistic, polarized communal blocs.

Turnout

Electoral turnout has varied considerably between 1952 and 1996, as revealed
in the table and charts showing turnout rates in Aligarh compared with those
in the state as a whole. Looking first at the turnout figures for Aligarh City
alone in Table 8.3, it is evident that there has been a considerable range in
turnout rates, from a low of 45.08 percent in the 1985 election to a high fig-
ure of 71.15 percent in the 1974 midterm election. There have also been not-
able fluctuations in turnout, with two rising slopes and two declining slopes
(Figure 8.3).
 Riots and Elections I 227
TABLE 8.3. Comparison of turnout rates in Aligarh and Uttar Pradesh
Legislative Assembly elections, 1952-96 a

yearh Aligarh turnout Stnte turnout

1952 46.54 37.88

1957 57.89 44.92
1962 64.78 48.58
1967 62.39 50.96
1969 70.35 52.22
1974 7l.l5 55.17
1977 56.II 44.87
1980 45.99 44.00
1985 45.08 45.22
1989 55.14 48.49
1991 54.10 47.20
1993 62.08 55.83
1996 52.13 54.91

a Figures are valid votes turnout.

bThe atypical 1955 by-election has been left out of this table.

Between 1952 and 1974, the overall trend in Aligarh turnout was upward,
with only a slight downward deviation in 1967. The trend in Aligarh in those
years corresponds for the most part with that for the state as a whole, with
the significant difference that the turnout figures are very much higher for
Aligarh City than for the whole state. The first concern, therefore, must be
to explain why the increasing interest in electoral politics between 1952 and
1974 was so much greater in Aligarh than in the state as a whole 4 and what,
if anything, the greater turnout in Aligarh had to do with riots.
The first declining slope in turnout rates occurs in the period between 1974
and 1985. The slope is downward in this period for both Aligarh City and the
whole state. Moreover, the two lines converge in 1985. The convergence, how-
ever, is momentary; both lines move upward again between 1985 and 1993
while the Aligarh slope regains some distance between it and the line for the
whole state. However, the distance between the two slopes is not so great as
in the earlier period. Our second problem, therefore, is to explain Aligarh's
divergence again-not from the general upward trend, but from the general
average.
 228 / Riots and Elections

80 ~----------------------------------,

70

eo
,..
\
50 -- -- ,/

\
\
VaUd votes turnout
(Aligam)
c \
~ '--- VaHd votes turnout

& ~1952+-~--~~--
1962
__-'--~-T--~~--~-'~
1969 1977 1985 1991 1996
(U.P.)

1957 1967 1974 1960 1989 1993

Election year, legislative assembly

FIG. 8.3. Valid votes turnout in Aligarh City Legislative Assembly elections, 1952-96

The second declining slope occurs between the 1993 and 1996 elections,
when the decline is much greater for Aligarh than the whole state. Further,
turnout in Aligarh in the 1996 election crosses that for the whole state for the
first time since 1962, but on this occasion moving below the state line. The
third puzzle, then, is to explain the considerable decline in turnout between
the two most recent elections, in this case a decline that is not really paral-
leled by the much more modest downward movement in the state slope.
The three divergences in turnout rates between Aligarh and the whole state
cannot be explained solely with reference to the incidence of communal riots
and a corresponding state of communal tension in the town. The assump-
tion behind such an explanation would be that communal tension is mir-
rored in interparty conflicts as the different parties blame each other for it
and the violence associated with it, thus increasing interest in the elections
that occur after riots. That explanation is supported insofar as the first ris-
ing slope is concerned by the fact that there was a sequence of riots in the
town during this period, beginning with the October riots of 1961 before the
1962 General Elections and including the 1971, 1972, and 1974 riots (Table p).
It is also supported by the rising slope between 1985 and 1993, for it was between
these two dates that the buildup of communal tension associated with the
Ayodhya movement occurred in the state as a whole, but probably to a greater
 Riots and Elections / 229

extent in Aligarh than in most other cities and towns in the state. The great
riot of 1990-91 in Aligarh provides evidence sufficient in itself for the latter
statement. On the other hand, this explanation is not consistent with the
plunge in turnout rates in Aligarh between 1974 and 1985, a period that wit-
nessed the long series of riots and violent communal incidents over the three
years between 1978 and 1980.

Interparty Competitiveness

We must, therefore, look further into the election results to consider what
kind of relationship exists, if any, between communal riots and party/
electoral politics. One possibility, already stated above, is that there may be
a relationship between the intensity of interparty competition, on the one
hand, and communal tension and riots, on the other hand-that is, that com-
munal riots may contribute to an increase in interparty competitiveness or,
vice versa, that increased electoral competitiveness may contribute to com-
munal tension and riots. There does in fact appear to be a closer correspon-
dence between interparty competitiveness-measured by the size of the
interval between the two main parties, as shown in Figure 8.45-and the inci-
dence of riots. It appears especially close in the long period of intense inter-
party competition between 1974 and 1993, when six of the major riots in which
there were deaths occurred.
Leaving aside the exceptional circumstances of the 1955 bye-election,6 the
period in Aligarh between 1952 and 1962 was, as we have seen, associated with
the dominance of the Indian National Congress and with a huge interval
between it and the second-place party. Interparty competitiveness then
increases significantly in the first election held after the 1961 riots, remains at
roughly that level during the next two elections, increases greatly once again
during the 1974 elections (a year in which riotous violence occurred, but was
contained), and remains high throughout the next decade. During the decade
of the 1980s-free of large-scale rioting-competitiveness increased in 1985,
but declined in 1989, then increased again in both the 1991 and 1993 elections
during the period of rioting and extreme tension associated with the 1990
riots and the Ayodhya movement. Finally, in the absence of major rioting
since 1991 and the decline of communal tensions generally in the years since
the destruction of the mosque in 1992, interparty competitiveness declined
in 1996 to its lowest level since 1957.
In effect, therefore, interparty competitiveness is more strongly associ-
ated on the whole with rioting and communal tensions than is turnout.
 230 / Riots and Elections

OO~--------------------------------------------,

40

30

20

-'
~ 10
0::
~
~ O+-__~__~__~__~__~~~~__~--__--__--~--~--~
1952 1957 1957 1974 1980 1989 1993
1955 1952 1959 1977 1985 1991 1998

Election year, legislative assembly

FIG. 8.4. Percent interval between winning and runner-up candidates,

Aligarh City Legislative Assembly elections, 1952-96

Moreover-and curiously-interparty competitiveness acts independently

of turnout, with which the correlation is negative( -.12). That is, the common
positive association in electoral systems generally between competitiveness
and high turnout does not hold in Aligarh. Evidently. therefore, considering
both these measures in relation to the incidence of riots, riots in general are
not necessarily followed by elections in which there is a high turnout.
However, riots or near-riotous mobilization have preceded elections in
which interparty competitiveness was intense, and interparty competitive-
ness has declined in the absence of riots.
Intense electoral competition is associated with riots in Aligarh and usu-
ally, though not always, precedes them. Moreover, riots have often occurred
in close temporal proximity to elections. The association is illustrated more
clearly in Table 8.4, which gives the dates of riots and elections in Aligarh from
1946 to 1998. Thus, the riots of March, 29, 1946, occurred just after the elec-
tions of February-March 1946. The February 1962 General Elections were pre-
ceded by the October 1961 riots, four months earlier. The series of riots in the
years 1978 to 1980 sandwiched the elections in January and May 1980. The
contained riots of November 10, 1989, occurred in the midst of the 1989 elec-
tion campaign. The great riots of December 1990-January 1991 preceded by
four to five months the elections of May 1991.
 Riots and Elections / 231
If, however, one tries-as Wilkinson has done 7-to make a precise cal-
culation regarding the proximity in time and the precedence in sequence of
riots and elections, the relationship is partly obscured. First, there is the prob-
lem of counting the number of riots, especially in the two long periods of
rioting in 1978 and 1980, which extended sporadically over montlls in both
cases. These riots could be counted as many riots or as one. I have listed these
riots on the charts as two separate, ongoing disturbances. If we do so, the
number of riots on the list is 24, of which 22 are dated. Using Wilkinson's
criterion of six months on either side of an election as a proximity measure,
12 of these riots occurred beyond the six-month period before or after an
election (coded NP for non-proximate), whereas 10 occurred within the
proximity period, including 3 that occurred in the month of the election
(coded S for simultaneous), thereby eliminating any obvious correlation
between riots and elections. Further, of the 10 riots within Wilkinson's prox-
imity period (including the 3 "simultaneous"), 5 occurred before and 5 after
an election.
Even if we confine the analysis to the 12 most intense riots, those which
produced deaths, there is a similarly ambiguous division. Six occurred
beyond the proximity period and 6 within it, including 2 in the month of the
election. Of the remaining 4 within the proximity period, 3 occurred before
the elections, 1 afterward.
Although the statistical association, therefore, is indeterminate, there is
no doubt that riots in Aligarh have been implicated deeply in the political
process, including both electoral politics and the politics of mass mobiliza-
tion, as in the Ayodhya movement that intensified and led to violence there
just before the December 1990-January 1991 riots. Thus, in the case of the
worst riots in Aligarh's history, they were preceded by a statewide mass mobi-
lization of militant Hindus out to destroy the mosque in Ayodhya, creating
havoc in Aligarh as well, and were followed by the elections the following May.
This type of association supports more broadly the Clausewitzian argument
I made 25 years ago in connection with the August 1967 Ranchi riots in Bihar,
namely, that in India, riots are a continuation of politics by other means, 8 of
routine by violent politics, but the political ends are the same. Those ends
are what politics are so much about in so many places, plainly and simply
the capture of power through scapegoating and intimidation of others,
rather than persuasion of voters on issues of public policy. The means, whether
routine or violent, are to consolidate the votes of one ethnic/religious group
by portraying another as a dangerous threat and, as we will see in later chap-
ters, by blaming the violence upon the victimized group.
 232 I Riots and Elections
TABLE 8.4. Riots and elections

Riots Elections Proximity

February 13-March 25, 1946 b
March 29,1946" Sf
March 4, 1950· NP
January 3-25, 1952 (LA/LS)'
June 6, 1954 NP
Till July 31, 1955 d
September 14, 1956 6B
February 24,1957 (L5)
March 12, 1957 (LA)
October 1-3, 1961 a 4B
February 19,21, and 25, 1962
(15/LA)
1966 NA
February 15-28, 1967(15)
1969 February 9,1969" NA
March 1, 1971 (LA/LS)
March 2, 1971· S
June 1971 A3
June 1972" NP
August 1972 NP
February 24 or 26, 1974 (LA)
October 3-5, 1974 NP
March 20,1977 (L5)
June 10, 12, or 14, 1977(LA)
October-December 1978" NP
May 1979' NP
June 17, 1979 NP

Communalization and Electoral Polarization

A third variable to consider in the history of electoral contests in Aligarh City

in relation to communal tension and riots may be called "communalization."
To what extent has the electoral contest in the city itself become a contest
between spokesmen for the Muslim and Hindu communities, respectively,
in which other factors such as economic and other noncommunal issues, or
even intercaste relations, have been ofless significance? A fourth variable may
be called "polarization." To what extent has the electoral contest eliminated
 Riots and Elections / 233

Riots Elections Proximity

January 6, 1980 (LS)

May 28 or 31, 1980 (LA)
August-November 1980' A3
March 1983 NP
December 28,1984 (15)
March 6,1985 (LA)
June 1987 NP
October 1988 a NP
November 10,1989 November 22,1989 (LA/LS) S
November 1990-January 1991" 4B
May 20, 1991 (LA)
December 1992" NP
November 21,1993 (LA)
March 1994 A4
December 1995a 5B
May 7, 1996 (LS)
October 9, 1996 (LA)
March, 1998 (LS)

"Riots with deaths.

bElections for different types of constituencies were held on different days throughout this
period; P. D. Reeves et a!', A Handbook to Elections in Uttar Pradesh, 1920-195-1 (Delhi: Manohar
Book Service, 1975), p. 245.
"LA :::: Legislative Assembly; LS :::: Lok Sabha.
dr have not been able to find the date on which this midterm election was held, only that
it was held before July 31, 1955.
eUttar Pradesh, Office of the Chief Electoral Officer, Results of Mid- Term General Elections
(1969) to the Legislative Assembly of Uttar Pradesh (Allahabad: Superintendent, Printing and
Stationery, 1969), p. 75.
f S :::: riots occurring simultaneously with election; A :::: riots occurring after election; B ::::
riots occurring before election. Numbers refer to months preceding or following the closest
election: NP = non-proximate; NA :::: not available.

other factors than communal identity as predictors of voting behavior and

concentrated the votes of the two communities in more or less total oppo-
sition to each other to such an extent that no other factors and no parties or
candidates representing counter-tendencies of any sort could influence the
outcome? Most important for our purposes, to what extent have the degrees
of communalization and polari7..ation in particular elections been influenced
by previous or continuous intercommunal violence before or during the elec-
tion campaign period itself?
In order to answer these questions, I have created an extensive dataset for
 234 / Riots and Elections

12 elections, including both the Aligarh City Legislative Assembly con-

stituency and the parliamentary segments, by polling booth. This dataset
makes it possible to identify the specific polling booths and mohallas in which
each candidate polled his or her best and worst votes. Reaggregation of the
polling booths to conform to mohalla boundaries also made it possible to
run correlations and regressions to determine the precise association between
the religious and Scheduled Caste populations, on the one hand, and the
vote shares of every candidate by mohalla or groups of mohallas, on the other
hand. The data are too extensive to be presented easily within the compass
of this volume. Consequently, while the overall correlations will be reported
by community/caste and party for all the 12 elections for which I have the
data, the detailed results will be provided only for tl10se elections that illus-
trate most clearly the relationship between riots and party electoral politics. I
have selected for the latter purpose the elections of 1962, held in the aftermath
of the October 1961 riots, and the sequence of elections in 1989, 1991, and 1993
which occurred after the great Aligarh riots of 1990-91 and the contained riot-
ing that occurred after the destruction of the mosque in Ayodhya. The over-
all results for the 12 elections will be presented in the remainder of this chapter,
the detailed results on specific elections in the following chapters.

THE SUPPORT BASES OF THE BJP

AND THE CONGRESS FROM 1957 TO 1985

The major strength of the HIP, of its predecessor party, the Jan Sangh, and
of candidates seen as spokesmen for militant Hindu nationalism is the con-
sistency of the support base that underlies it. Table 8.5 suggests that, at least
from 1957 until 1985, these candidates have been able to rely upon a stratum
of militant Hindu sentiment that has persisted through time. The size of the
support base has certainly fluctuated, but there can be no doubt about the
existence of a steadfast seam within the Hindu population of the city that
has been mined successfully to some degree in every election throughout
this period. Every election for which station-wise polling data are available,
whether for the Legislative Assembly or for Parliament, has produced high
correlation coefficients at significance levels of p=.ooo, with the range in the
correlations from a low of .47 in the 1980 Lok Sabha election to an extra-
ordinary high of .96 in the 1962 Lok Sabha election. Even in 1980, the some-
what lower correlation between the Janata Party vote and Hindus and others
is misleading, since there were two militant Hindu candidates in this elec-
tion. Krishna Kumar Navrnan ran as an independent at this time, dividing
 Riots and Elections / 235
the militant Hindu vote. Even so, the correlation coefficient between his vote
and percent Hindus and others was .38 (P=.005).
In contrast to the relative consistency of the support base for militant
Hinduism, that for the Congress in relation to Hindu and Muslim popula-
tion concentrations has fluctuated drastically from election to election, as indi-
cated in Table 8.6. It is evident also that the fluctuations depend heavily upon
the religion of the Congress candidate. The strongest negative correlation
coefficient with percent Hindus and others, as well as the strongest correla-
tion with percent Muslims, occurred in 1969 when the Congress candidate
was a Muslim, Ahmad Loot Khan. The second strongest set of negative/posi-
tive correlations occurred in the 1967 Legislative Assembly elections when,
likewise, the Congress candidate was a Muslim. On three occasions when the
Congress fielded Muslim candidates, namely, in the 1962 Lok Sabha and the
1980 and 1989 Legislative Assembly elections, the correlations with percent
Hindus and others were positive, but the significance levels were low. Con-
versely, the four highest positive correlations with percent Hindus and others
occurred when the Congress fielded Hindu candidates, in the 1962, 1985, and
1991 Legislative Assembly and the 1984 10k Sabha elections. On all those occa-
sions, the significance level of the Congress correlation with Hindus and others
was above .002.
The contrast between the consistency of the militant Hindu support base
and the fluctuating support base of the Congress is brought out in Figures
8.5,8.6, and 8.7. Figure 8.5 graphs the results of the struggle between the
Congress and militant Hindu candidates for the Hindu vote. Figure 8.6 dem-
onstrates the relative stability of the militant Hindu vote shares among all
three population categories, while Figure 8.7 reveals the instability of the
Congress vote share among the same three categories. More specifically, while
the militant Hindu candidates' correlation coefficients with percent Hindus
and others remained above -47 throughout the entire period from 1957 to
1985 (Table 8.5 and Figure 8.5), those for the Congress took a very steep drop
from the 1962 election onwards until 1984 and 1985. During the Rajiv Gandhi
landslide in 1984, the Congress candidate was Usha Rani. In the Legislative
Assembly election that followed, the Congress candidate was identified with
Rajiv Gandhi. In these two successive elections, the correlations for the
Congress returned to their highest points since the 1962 elections, nearly
equalling those for the BJP. Then, of course, everything changed again with
the rise of the Ayodhya movement in the four elections that followed, for
only two of which we have the detailed polling station data. The Congress
correlation with percent Hindus plummeted to .06 with a Muslim candi-
 TABLE 8.5. Correlation between votes for militant Hindu candidates and local
population composition across Aligarh mohallas in selected elections, 1957-91

Election Year and Type (Legislative Assembly [LA] or Lok Sabha [LS]),

1957 (LA) 1962 (LA) 1962 (LS) 1967 (LA) 1967 (LS)
(N=39) (N=44) (N=44) (N=41) (N=41)

Community! Shiv Kumar Shiv Kumar

Caste Ian Sangh lanSangh Shastri lanSangh Shastri

Hindus
and others .59** .58"" .96"" .76"" .77""

Muslim -.50** -.44** -.86"" -.65"" -.68**

Scheduled
Castes -.15 -.30* -.11 -.22 -.18

.. = .05 or less .
.... p == .01 or less.

TABLE 8.6. Correlations between votes for Congress and local population
composition across Aligarh mohallas in selected elections, 1957-91

Election Year and Type (Legislative Assembly [LA]

1957 (LA) 1962 (LA) 1962 (LS) 1967 (LA) 1967 (LS)
(N::::39) (N=44) (N=44) (N::::41) (N=41)

Ravind
Community! AnantRam AnantRam larmr YusufRashid NetraPal
Caste Verma Verma Haider Khwaja Singh

Hindus
and others -.00 .73** .22 -.74"* -.66**

Muslims -.04 -.74** -.31" -.74** .66**

Scheduled
Castes .05 .12 .29* -.06 -.05

"p = .05 or less.

"*p = .01 or less.
 and Party or Name of Principal Militant Hindu Candidate

1969 (LA) 1980 (LS) 1980 (LA) 1984 (LS) 1985 (LA) 1989 (LA) 1991 (LA)
(N=41) (N=53) (N=53) (N=65) (N=66) (N=70) (N=70)

lanata
lanSangh Party RIP RIP RIP RIP RIP

.57** .47** .6S** .59** .63** .76** .73**

-.52** -.53** -.62** -.4S** -.51 ** -.75** -.73**

-.12 .22 -.OS -.26* -.24 .06 .07

or Lok Sabha [LS]), and Candidate

1969 (LA) 1980 (LS) 1980 (LA) 1984 (LS) 1985 (LA) 1989 (LA) 1991 (LA)
(N::;:41) (N=53) (N=53) (N=65) (N=66) (N=70) (N=70)

Ahmad Ghanshyam Ahmad Baldev Mohammad Bhanu

Loot Khan Singh Loot Khan UshaRani Singh Furkan Pratap

-.90** -.42** .23 .51** .59** .06 .36**

.91** .45** -.32* -.49** -.62** -.09 -.40**

-.01 -.13 .23 -.00 .13 .09 .14

 238/ Riots and Elections

1.0

.5
I
I
t
\
, //
----, \
I \ / \ /
I \ I \ /
0.0 \ I "
\ I
\ I
\ I
\ I
-.5 \ I Correlation
§

1
'0
\
\ -- ......, /
,/
/
",itllmilllanlHIndu

·1.0 withCongreS$

~ ~'" ?", ~... ~... , , , ?~ " , ?~..

~~~~q~~q~~~~~~
Election yearltype

FIG. 8.5. Correlations for Congress and militant Hindu candidate vote shares
with percentage of Hindus and others, 1957-91

1.5r-------------------------------------~

1.0

.5 Correlation

0.0 ,,- with % Hindus/other

---_/ ..-
with % Muslim

with % SCs

Election year/type

FIG. 8.6. Correlations for militant Hindu party vote shares

with percentage of religious/caste groups, 1957-91

date in 1989, then rose again to .36 with a Hindu candidate in 1991. In those
elections, its overall vote share also plunged, while the electoral contest
became polarized between the BJP candidate, Krishna Kumar Navman, and
his invariably Muslim opponent from either the Janata Dal, the BSP, or the
SP-which latter finally defeated him.
The evidence provided so far, therefore, is overwhelming and persuasive
 Riots and Elections 1239
1~r--------------------------------------'

I .•

. ~ .... ~ . -.. '''....

.• ,/ ~',.".,
,... ..,..:.:: '" ,.'
' CorrelatiOn
0.0 '--- '....... /...;' ............ _----~L..-

I '.' \\~-------.~;""
with % Hioduslothels

with % Muslim
E \. _... --., ~
o ""' ... ~~ with %SC.
o ·1~ ~'--'"",:":"~-:::-~:-:-:::C-:"--"""-:::-:-c=:'~-::::::-' .~~__~__-I
1957 (LA) 1962 (LS) 1967 (LS) 1960 (lS) 1984 (lS) 1989 (LA)
1962 (LA) 1967 (LA) 1989 (LA) 1980 (LA) 1985 (LA) 1991 (LA)

Election yearltype

FIG. 8.7. Correlations for Congress vote shares with percentage

of religious/caste groups, 1957-91

on two points. The first is that major riots before an election are followed
by-and most probably produce-intense interparty competition. Second, such
collective violence produces not only intensified party competition, but a com-
munalized and polarized electoral result. It will be shown in the next chapter
how the politicians in Aligarh contribute to the intensification and commu-
nalization of party competition through their involvement in several ways
in the production, exploitation, interpretation, and use of collective violence
for political advantage.
 9 / The Practice of Communal Politics

A lthough I have argued that, insofar as Aligarh City is concerned, riots

that precede elections intensify interparty competition and provide
a basis for it through communalization and polarization of the elec-
torate, Aligarh is also part of a broader framework of interparty competition
at the state and national levels that must also be kept in mind in considering
the overall relationship. In the case of the mass mobilization that accompa-
nied the three successive communalized elections of 1989, 1991, and 1993 in
Aligarh, interparty competition intensified only after the great riots of 1991.
However, it was the intense struggle for power at the Center and in the state
of U.P., accompanied by the MandaI agitation and the Ram Janmabhoomi
movement, that initiated the process. As one of my respondents put it just
after the 1990-91 riots, they were a consequence of the "particular game ...
going on in Delhi, how to win votes and how to become the prime minister.
I think that was one of the important reasons of [the communal riots 1in the
whole U.P. state and Aligarh."l
Moreover, as this same respondent put it, "the riots did not occur in a
day or so," but involved three years of continuous preparation and rehearsal
on the part of the SIP, the VHP, and other parts of the Sangh Parivar, punc-
tuated by the shilanyas of 1990 and the vast rioting that was associated with
it in Aligarh, followed by the destruction of the mosque on December 1992
and the vast rioting associated with it throughout most of the country. Nor
can there be any doubt that there was deliberate provocation of Hindus to
attack and kill Muslims, available in the form of tape cassettes of the fiery
speeches of the BJP female "sannyasi," Sadhvi Rithambara. Although we
do not have any recorded speeches of RSS-BJP-VHP leaders in Aligarh, it
cannot be doubted from what they have said to me in polite encounters with
a foreign scholar that they spoke even more directly to their followers and
 The Practice of Communal Politics 1241

to their voters in the communalized Hindu mohallas of Aligarh than they

did with me.
Furthermore, the militant Hindus have their own object of hatred in Aligarh,
their own Babri Masjid, as it were, in the presence of the AMU, which has
been used for the past half century "as a symbol that this is a Pakistani cen-
ter, antinational center,"2 and was used so effectively in the great rumor about
the AMU Medical College Hospital killings in 1991. So, the AMU, like the Babri
Masjid, was "very much a part of this whole game" of competition for votes.
Throughout this broader vote competition, Mulayam Singh Yadav and the
Ianata Dal appealed directly and successfully to the Muslim voters, standing
forth as their protector and as the protector of the mosque. Although his police
saved the mosque from destruction in 1990, they did not save the Muslims
of Aligarh from the riots that descended upon them in its aftermath. Further,
it was the BJP that emerged victorious, riding the electoral wave produced
by the shilanyas and its riotous aftermath, in both Aligarh and in the state as
a whole. Indeed, the May 1991 elections themselves were accompanied by addi-
tional waves of rioting in many constituencies in the state. Moreover, those
elections brought the BJP to power in the state for the first time} That gov-
ernment in turn created the conditions that made possible the destruction
of the mosque in December 1992.
The respondent previously quoted remarked to me at the tail end of the
1990-91 riots in Aligarh, "This is a political riot, ... a communal-cum-
political riot."4 Indeed, the whole period from the first stirrings of the VHP
mobilization over the mosque in Ayodhya in 1984 until its denouement in
December 1992 was a deliberately provocative buildup of communal tension
that intensified political competition. It also produced-predictably-waves
of communal rioting that in turn built heightened communal solidarities in
constituencies with large Muslim populations deliberately targeted by the BJP
for special efforts; the rioting in turn fed back into the competition for votes,
in a circle that ended only in the aftermath of the destruction of the mosque,
when the returns from communalization of the electorate and the produc-
tion of communal violence rapidly diminished.

THE POLITICIZATION OF COMMUNALISM AND RIOTS

AS A CONTINUATION OF COMMUNAL POLITICS
BY OTHER MEANS

My interviews are replete with explanations that consider the political process
central to the communalization of Hindu-Muslim relations in India. It also
 2421 The Practice of Communal Politics

happens to be my own preferred contextualization, the one I have used and

presented over many years in several of my writings. It is not, however, my
own invention. One can find it articulated by persons at many levels in Indian
society, particularly from those in the secular parties and factions associated
with left and agrarian politics in north India, and especially the moderate and
radical left parties that used to carry the words "socialist" or "people" some-
where in their title (such as Samyukta Socialist Party [United Socialist Party1,
or 10k Dal [People's Party1), and nowadays generally carry the word "janata"
in their title (as in Janata Dal [People's PartyJ).5
My interviews also provide evidence to support the general point of view
that riots are an integral part of the process of political/electoral competi-
tion. Politicians and political parties benefit directly from the politicization
of communal affiliations, in which the fomenting of communal riots plays
an important part. The responsibilities of particular groups and parties in
promoting riots and the extent to which they are preplan ned and organized
in advance is often difficult to demonstrate, but it is equally often clear enough
that there is a direct connection between the politicization of communal
differences and the killing that takes place in the riots that result from them.
Moreover, the politicians all know this very well.
It is in this sense, among others, that communal riots in India are a con-
tinuation and extension of communal politics by other means. 6 That does
not mean that the murderers, arsonists, looters, and rapists are themselves
political people. Sometimes politicians or political activists are found among
the perpetrators, but most are not. Nevertheless, communal riots play upon
preexisting communal antagonisms, many of which have in turn been cre-
ated by previous political movements. One cannot say, therefore, that there
is an independent bedrock of communal prejudices having nothing to do with
politics that lies slumbering, waiting to explode. These antagonisms are partly
created by political mobilizations, nurtured in apparently peaceful periods,
and drawn upon by political leaders and parties when an emotional issue,
such as the status of the Urdu language or the alleged existence of a mosque
built upon a Hindu temple destroyed by Muslims in centuries past, becomes
salient or is made salient and is presented in order to divide and polarize
Hindus and Muslims politically, and to gain political/electoral advantage
thereby.
The riots that often follow from such politicization of communal difference
and mobilization of community support are then used by political parties to
discredit one another, to gain the support of one community, and to con-
 The Practice of Communal Politics / 243

solidate a community's votes. The electoral strategies of political parties and

groups may change dramatically after major riots in a district such as Aligarh
or in the state as a whole. Such was the case in Aligarh District as early as
1962, when the Congress's virtual monopoly of power was broken by the
alliance between newly mobilized Scheduled Caste voters, on the one hand,
and Muslim leaders from the AMU angered by the riots of 1961, on the other
hand. The latter, as will be described in detail in the following chapter, joined
forces with the Republican Party, representing the Scheduled Castes, to defeat
the Congress in the contest for the Aligarh parliamentary seat and in several
of its component Legislative Assembly constituencies. Such was the case in
the elections of 1991 in the entire state, when the combined issues of reser-
vations for backward castes and Ayodhya, played upon by the SIP, created a
consolidation of upper-caste Hindu voting, particularly in constituencies with
large Muslim minorities, which the BJP specifically targeted for communal
mobilization and Hindu consolidation.

THE INSTIGATION AND JUSTIFICATION

OF COMMUNAL VIOLENCE

At the center of the Aligarh organization of communal violence is the local

network of the RSS family of organizations, and allegedly in its forefront are
Krishna Kumar Navrnan and his son. Navman's name has appeared many
times in the accounts above. It is time to pay closer attention to his role in
these events, which became ever more important with the rise of the mili-
tant Hindu movement in the 1980s.
Mr. Navman was born and raised in Aligarh City. He was born in 1936
and joined the RSS as a young man, sometime in his teenage years. When
asked why he joined the RSS, he said that it "build[sl up your character" and
is "a nationalist organization." He was educated up to the Intermediate stan-
dard and graduated after taking the Hindi Visharat examination from the
Hiralal Barahseni Inter College.
Navrnan, as noted earlier, is a big businessman engaged in the publica-
tion of commercial books. His factory, warehouse, and home are in the riot-
prone mohalla of Manik Chauk. Two-thirds of the population of the mohalla
are Varshneys by caste, to which Navrnan himself also belongs. He believes
that "Aligarh politics are basically dependent on the relations between Hindus
and Muslims" and that the Muslims are responsible for what he character-
ized in 1983 as a "virus" of communal violence that was "spreading all over,"
 244 / The Practice of Communal Politics

FIG. 9.1. Krishna Kumar Navman. November 1997

not only in Aligarh. From 1970 to 1974. Navman was secretary of the Jan Sangh.
In 1974, he contested his first election in a rural constituency of Aligarh Dis-
trict, Barauli, as the Jan Sangh candidate, and came in third with H.I percent
of the vote, not enough to retain his security deposit. When I first met Navman
in 1983, he said he was a member neither of the Jan Sangh nor of the B]P,
though he proudly acknowledged his association with the RSS. In fact, how-
ever, he had defied the discipline of the RSS in the 1980 parliamentary elec-
tions. The RSS supported the Janata Party in that election, but Navman chose
to contest as well as an independent. He came in fourth in a field of 26 candi-
dates, but polled only 7,039 votes, amounting to a vote percent of 2.12. Nav-
man's vote share in the Aligarh segment of the constituency, though somewhat
higher, was still not impressive; he polled 3,735 votes, which gave him a vote
share of 5.74 percent. His candidacy had no effect on the election outcome
and he again lost his security deposit. In 1983, Navman was "still angry with
the RSS, and the RSS [was] still angry with him."7
Even the local police inspectors in riot-prone areas named Navman as "the
main fellow among Hindus" instigating people to riot against Muslims. 8
Navman and his defenders, however, insist that he has never instigated any
riots and that, on the contrary, he has been repeatedly victimized by a biased
administration and falsely arrested and implicated in cases arising out of riots.
For example, during the 1978 riots, at a time when the Janata government
 The Practice of Communal Politics I 245

was in power in the state, and Kalyan Singh of the BJP, elected from another
constituency in Aligarh District, was a minister in that government, Navman
was "implicated" (allegedly falsely) in "eight to ten murder cases" as well as
some arson cases.9
When I first met Mr. Navman in 1983 in a prearranged interview appoint-
ment, I found him waiting for me upstairs in an office in his business prem-
ises, a quite large business and warehouse complex. He was sitting with a
folio-size copy of the Koran in front of him in a trilingual edition (Hindi,
Urdu, and English), from which he quoted me several passages that to him
clearly indicated the evil character of the Muslims and their religion, which
he considered to be the root cause of the communal problem and of riots
in Aligarh. Among the passages he found particularly relevant was one that
he read to me in English: "Mohammedans should not make any non-
Mohammedans their friends.» The establishment of friendships between
Muslims and non-Muslims is a danger to Islam because Muslims then may
become kafirs. Therefore, he interpreted the Koran as saying that "such people
must be found and killed.» I tried repeatedly to interrupt his readings and
the conclusions he drew from them, in order to get him to focus specifically
on the causes of and circumstances surrounding riots in Aligarh, but he
insisted that these passages and the Koran itself were "the source of all riots"
in Aligarh and elsewhere in India. He was particularly offended by what he
perceived as the Muslim condemnation of all kafirs and by the term itself.lO
Navman felt personally aggrieved, also, that Muslims of Aligarh were out
to get him, as indicated in my 1983 interview with him.

PRB: Why are these cases against you?

KKN: They want to denigrate the RSS through me. They say, Navman belongs to RSS,

Navman started the riots, so the RSS started, or is the main cause of the riots.
I'RB: Who started t.he cases?
KKN: Muslims at the instigation of AMU.

I'RB: Who in AMU?

KKN: No one in particular.

I'RB: What are the charges against you, specifically?

KKN: Murder, dacoity [gang robbery], robbery. They don't say I personally did it, but

that at my direction others did these things.ll

When I met Navman a third time in 1997, having interviewed him also in
1993, he greeted me as an old friend. I asked him how old he was. He said he
was 61, to which I replied that we were exactly the same age. He then revealed
 246 / The Practice of Communal Politics

to me a sense of weariness at the difficulties he has had to face in his life. He

began by saying he had "faced lots of problems in [hisllife."12 By then, he
had been to jail about 25 times by his rough reckoning, of which the last times
were during the Ramjanmabhoomi movement in December 1990 and again
in January 1991.
In December 1990, during the great riots, Mulayam Singh Yadav was the
chief minister and, according to Navman, Mulayam Singh himself sent
instructions to the district administration for Navrnan's arrest. He was actu-
ally arrested initially on October 12, presumably as a preventive measure at
a time of growing tension in the city. A first "installment" of rioting took
place between October 18 and November 6, during which time Navman was
in jail. He was released a few days after the end of this first phase of rioting,
so Navrnan was present in the city during the main phases of rioting in
November and December.13 When rioting again broke out and curfew was
imposed, orders again went out for Navrnan's arrest. Although the city was
under curfew, Navrnan was "roaming around this area" (that is, Manik Chauk)
until the police came to arrest him at 4 A.M. one morning, but they could
not do so because, he claims, "more than 20,000 persons" came out of their
houses in the midst of the curfew, with the result that the police "could not
get the courage to arrest him." He was not finally arrested until January 20,
1991, when he was on his way to give a speech at the Ram Lila fairgrounds.
From there, he was taken straight to the Fatehgarh Central Jail and locked
up for two months under the National Security Act, after which he was released
without any proceedings being instituted against him.14
Navman's son also was said by one informant to have been quite active
during these riots. His son is considered to be even more militant than he,
said to be "a Bajrang Dal chief." Navman himself virtually acknowledged his
son's activities, although at first, when I asked where his son was during these
riots, he replied that "he was in his home." When I pressed further and
remarked that people had said his son had also played a big role in these riots,
going around mobilizing the people (that is, the Hindus) to attack Muslims
and the like, and that he had become famous, he responded, "People believe
that if they have any difficulty or if they are in a dire hour of crisis, then he
would come and help them. So, maybe he was approached by a few people
and went out to help them." Navrnan's son was arrested as a consequence of
his helping activities in the first week ofJanuary 1991.15
N avrnan was present in Ayodhya at the time of the destruction of the mosque
on December 6, 1992, and gave me his own eyewitness account of it. When he
arrived in Ayodhya just after noon on December 6th, he said, "there was total
 The Practice of Communal Politics / 247
confusion and a lot of hulla [sic] going on there" and people had already climbed
up to the domes. He claims that L. K. Advani personally admonished the crowd
and attempted to stop them from their work of destruction.

NA VMAN: But the people were in no mood to listen to him.... They scolded him
back and they told him, ki, you go and give your speeches and let us do our work.
Actually, the people were very tired, being called there again and again and being
dispersed and called again and being told to go back and called again and told to
put some sand somewhere or perform some fJuja somewhere and then go back.
They were not willing to listen to anybody.
I'RB: And how did y.ou feel, at that time?
NA VMAN: I was happy t.o see it.
I'RB: SO y.ou were on the side of the crowd, that wanted t.o take d.own that mosque
right then and there and nut follow Advani's advice?
NA VMAN: I felt every bit like the crowd. It is a headache and it is a national humilia-
tion and an issue like this ought t.o be ended once and for all. 16

In my second interview with Navman in 1993, I asked him to explain to

me why there had been such a terrible riot in Aligarh in December 199.0, but
no riot at all in the city after the destruction of the Babri Masjid. He gave the
following response.

Whenever there have been riots in this city, it has been because the Muslims
have taken the initiative. In 1971, there was a fact-finding commission appointed
and the decision which the commission gave was that the MuslinIs had started
the ri.ot which the commission was investigating. So, always in Aligarh City, it
is the Muslims who begin the riot. N.ow this time, the Muslims were a little
scared because they kept getting the feeling that previ.ously when Mulayam
Singh's g.overnment was in p.ower and it was a g.overnment which was sympa-
thetically inclined towards them, even then they had t.o put up with so many
hardships and this time now that the BJP had been in p.ower, they were sure
that maybe they w.ould be put to grief if they indulged in riots.

It is evident from these interviews with Navman over a period of fourteen

years that he harbors strong resentment against Muslims, that he was happy
to see the destructiDn of the mosque at Ayodhya, that he believes Muslims
always start riots when they can, and that he believes one of the benefits .of
BJP rule is to make Muslims sufficiently aware and fearful that they will not
even think of starting any riots. Navman, as we have seen, was elected to the
 248 / The Practice of Communal Politics

Legislative Assembly for the first time in 1989 on the rising wave of the Ayodhya
movement. Behind Navrnan, himself "a very big businessman," as one of my
respondents described him, is the organization of traders who support his
election campaigns and communal activities.17 But it is not only Navrnan and
his coterie who provide leadership and organization to militant Hindu activ-
ities in Aligarh. Another RSS man, one in fact not friendly to Navrnan, was
secretary of the Vishwa Hindu Parishad for the state as a whole at the time
of the Ayodhya movemene3
It is, however, Navman whom the BJP chose to contest four consecutive
elections in a constituency with a large-and lately a majority-Muslim popu-
lation. How did it happen that the BJP chose to allow Mr. Navrnan to be its
candidate in successive elections? Before the rise of the Ram Ianmabhoomi
movement, Navrnan had been an unsuccessful local politician. In his first elec-
tion contest, in the 1974 election, as just noted, he ran as a Jan Sangh candi-
date and lost his deposit; in the 1980 Lok Sabha elections, he ran as an
independent against the candidate supported by the BIP, polled miserably,
even in the Aligarh segment, and lost his security deposit once again. It is
clear that his relations with the local RSS were strained at that time. But
Navrnan had assets that were valuable to the BIP, when the time came for
increased militancy, that other RSS men did not have. He had a long history
of activity in mobilizing Hindus during riots. He was a member of the VHP
as well as the BIP and his son became the local Bajrang Dal chief. He has been
a prosperous businessman of the Varshney caste, the solid core of BIP sup-
port in the city. Although there are many other Varshney RSS men, some of
whom have aspired to the ticket that has gone to Navrnan in four successive
elections, he has had an additional asset, namely, the support of the traders,
who have been "united behind him. "1<} Thus, when the time carne for the mobi-
lization of Hindu militancy, including the deliberate "playing of the Hindu
card,"zo the instigation of Hindu-Muslim confrontation, and the provoca-
tion of the Muslim population, Navman was the man in Aligarh City who
had the right combination of personal and financial resources: a reputation
for devotion to the militant Hindu cause, strong support from the leading
Hindu caste in the city, considerable personal wealth, and the financial back-
ing of the traders.

THE POLITICAL CONTEXT OF COMMUNALISM

Although Mr. Navman and his supporters have been of central importance
in the communalization of politics that has contributed to the production
 The Practice of Communal Politics / 249

of riots in the city, he has been far from alone. Nor should one conclude that
the politicians whose actions promote riots from which they then benefit polit-
ically are extraordinary. Once the political calculations are made and the game
of politicization of communal identities begins, all politicians get caught up
in it, either willfully or simply because it is the only available context in which
they can appeal for votes. Muslim politicians are no different from Hindus
in this respect. All that differs is the rhetoric used to justify their entry into
the electoral process, the means they use to garner votes, and the ways in which
they seek to benefit from riots. Consider the following nine statements made
to me in various interviews by Hindu and Muslim politicians over the years.
The first is from an interview in 1983 with a person associated with a conser-
vative and allegedly fundamentalist group in university politics. This group
is in turn associated with the agitation to maintain the minority character of
the AMU, and hostile to the Congress because of its policies, which are con-
sidered to have undermined that status. He later contested two Legislative
Assembly elections and is cited again below.

I supported BJP in Aligarh City in the last election [1980 I and before that also
with the intention to patch up the differences between Muslim community
and Hindu community, to remove communalism from the minds of these two
communities.... So we support Jan Sangh, BJP, so that they may come dose
to us, and we may understand what they want, with that intention to have a
communal harmony in the city, we supported them. Frequently there are riots
here, and political parties exploiting the situation. Not one party; we cannot
blame Jan Sangh that Jan Sangh is the only party which is exploiting situation
in Aligarh often now. But sometimes Congress (I) also exploits. When Congress
(I) comes in power, Jan Sangh exploits, other parties exploit. If, say, Jan Sangh
or BJP comes in power [orl Janata comes in power, then Congress (I) exploits
the situation. 21

These remarks, made in 1983, followed upon a discussion of university pol-

itics. I had not brought up the question of riots in the city. The respondent
introduced the matter as well as his support for the BIP on his own. Though
he was at that time among the discontented, anti-Congress Muslims who had
either joined or supported the BIP, it is obvious that he did not do so out of
sympathy with that party's ideology. He did so, he avers, for the sake of pro-
moting communal harmony in a city with a history of riots.
The second statement is from an interview in 1983 with an advocate, and
ex-president, BIP, Aligarh District.
 250 / The Practice of Communal Politics

PRB:Now you said you joined [politics and Jan Sangh] in 1966-especially in view
of the communal conditions in ... Aligarh.
RESPONDENT: Ah,.-communal conditions, circumstances and conditions then preva-

lent at Aligarh-
PRB: What were those conditions, exactly?
RESPONDENT: Is it very pertinent to ask me?
PRB: Yes!
RESPONDENT: The very fact has been that the policies and principles of the govern-
ment, whosoever had been ruling, unfortunately the Congress (I) ... has been
divided in two. Unfortunately, as well, Aligarh has been not exactly predominant
by Muslims, but had a substantial percentage of Muslims, induding-mane- . ..
coming to about 37 percent in 1966 and 1967.
PRB: In the city?
RESPONDENT: In the city. Unfortunately, ... Aligarh Muslim University-I always
thought to be a snake pois-a snake-pit of communalists. All communal riots,
all communal hatreds, all communal bias, et cetera-and even all communal
riots here perpetrated, plans organized, any start of a problem, the proceedings of
the Aligarh Muslim University with the help-of innocent Muslims ... -er-
persons of the town.
PRB: I see.
RESPONDENT: That-that has been my firm conviction. And therefore I would in the
same perspective and relevance-request you to kindly go through the find-
ing[s] ... of Aligarh Inquiry Commission, which was headed by his Lordship
Justice ... Mathur. You probably-you might not ... be remembering that there
was a major communal riot, at Aligarh, on the 2nd of March 1971, wherein Mr.
Mohammed Sharif [sic j actually Mohammad Yunus Saleem], who was at the [time1
Deputy Minister in the Railways, at Center, had contested here .... It has been a
repeated history of Aligarh to have communal riots after five or ten years-every-
every time. That has been unfortunate. And then, ... the property of Hindus they
are lost. I was also accused of instigating communal riots in Aligarh. It was alleged
against me that I organized a big ... procession, I collected a rally of the students,
took the procession through the market, burnt the city-mane-houses and build-
ings of Muslims; I was also responsible for the murder of certain-Muslims, for
looting of their property, burning of their shops, etc. For-in that, the Government
of U.P.... was pleased in appointing a Commission, Aligarh Riot Commission,
which was headed by his Lordship, Justice ... Mathur ... And he has-openly
R5S and Jan 5angh, myself and Shri Inder Pal Singh and one Mr. Ranjan Pal
Singh were accused. He has himself written that the riot was sponsored and
arranged by Aligarh Muslim University, ill order to get Mohammad [Saleem]
 The Practice of Communal Politics / 251

arrested [sic; elected J. And the charges leveled against-against RSS or Jan Sangh
are absolutely ... fantastic.
PRD: Did you-did you organize any procession or anything of the sort?
RESPONDENT: Absolutely fantastic!
PRD: Nothing?
RESPONDENT: Absolutely fantastic!
PRD: Completely all [false]?
RESPONDENT: All-all [falseJ. All concocted version. And that is a judicial finding.
Which has been published.
PRB: Hm. Hm! (Pause) So-this was the ... 1971 riots?

RESPONDENT: Yes.... And I was also arrested in connection with everything.

PRB: Achcha?
RESPONDENT: HaN. On the spot. Because I had myself seen-Muslims firing from-
from within the mosque upon innocent Hindus, burning their shops, etc. And the
district administration was watching like-like a tomfool. Or a dumb fool organ-
ization. When I became very hot they had no other-And the Hindus also got agi-
tated, on seeing that the Jewelry Market ... was being burnt and looted. They had
no other option excepting arresting me. So if that is-if that is communalism, I
am opposed....
PD:Hm.
RESPONDENT: If that is communalism-I am proud to be communalist. If that
is Hinduism, I am proud to be a Hindu. That's my positive, concrete, pattern,
principle.... 22

This second, long set of remarks made by a former BJP district president
is of great interest from several points of view. First, to the general question
concerning why he entered politics in 1966 and specifically why he joined the
Jan Sangh, he gave the single response that it was because of "the communal
conditions" in the city. Second, there can be little doubt that his attitude
towards Muslims is that of a rabid communalist. He cites precisely the pop-
ulation percentage of the Muslims in the city as a matter of concern. He speaks
of the AMU as a snakepit of Muslim communalism, the source of "all com-
munal riots" in Aligarh. Like Mr. Navman, he distinguishes between edu-
cated and "innocent Muslinls," the latter being led astray by the educated elite
among them. Third, he points to a direct connection between riots and elec-
tions, specifically the riot of March 2, 1971, that occurred on the day after
polling for the Legislative Assembly and Lok Sabha elections of 1971, which
he implies was related to the candidacy of Congress candidate Mohammad
Saleem, supported by people at the AMU.
 252/ The Practice of Communal Politics

Fourth, he lists several charges against him of instigating a communal riot

and participating in criminal activities during the riot, for which he was arrested
"on the spot," but denies vehemently their accuracy. Here I want to say that
I believe this respondent's denials on the specific points that he described as
"absolutely fantastic." But then, he also reveals that his actions during this
riot were far from innocent. Let us note that, by his own admission, he was
out in the street in the midst of the turmoil in the riot-prone mohalla of Sarrafa
Bazar [which he calls the Jewelry Market], far from his home or office. He
admits that he "became very hot" and "the Hindus also got agitated," and
finally that the police "had no other option excepting arresting me." Let us
also keep in mind that this respondent was not then an ordinary citizen of
Aligarh. He was president of the BJP at the time. He was a big lawyer in the
town, whose clients included both Hindus and Muslims accused in riot cases.
According to the administrative and police authorities of the district, he
defended Muslim criminals involved in riot cases. While that fact would seem
to suggest his lack of communal bias, my interviews with police suggest oth-
erwise, namely, that these Muslim criminals were useful in starting riots desired
by self-proclaimed, proud Hindu communalists such as this respondents.
All these facts and implications point to one conclusion. The respondent,
an intelligent and educated lawyer, did not commit criminal acts. His role
was different. He is a prime example of what I call a "conversion specialist,"
the person who plays the role of mobilizing people, including local crimi-
nals, leading them into potentially riotous confrontations, bringing them to
the brink of criminal activity, and tllen, depending upon the specific cir-
cumstances, most notably a competitive political context, giving the signal
for the riot to begin. That signal may be getting "hot" oneself, yelling out that
the Muslims are burning and looting, or getting oneself arrested.
Shortly after my interview with this respondent, I interviewed two local
police station officers, one of whom identified succinctly the political con-
text for riot production and also named the three most active Hindu mili-
tants in producing tllem. Insofar as the political context is concerned, the
inspector described the matter simply as follows.

During and before the election, for sympathy and votes, some politicians may
start a rumor or a small riot to gain support of one side, then that community
goes largely on that basis for the politician.23

The inspector described the roles of the three militant Hindu communalists
as follows.
 The Practice of Communal Politics I 253
Krishna Kumar Navman is the main fellow among the Hindus. Shiv Hari Singhal
is the lawyer for most Muslim criminals. Manga Ram first looks at the situa-
tion, then goes to the colleges, instigates the students, and they go there. These
three minds play the most important parts during riots.

We know enough about Navman so that further commentary is unnec-

essary at this point. Manga Ram, however, is another case, a professor at one
of the local degree colleges, who openly declared to me the most extreme hos-
tility to Muslims. Among his important roles during the production of riots
is the specific task of instigation, collection, and mobilization of students to
potential riot sites. Shiv Hari Singhal, a BJP leader and a "conversion special-
ist," allegedly has another task after the riots, of defending the criminals who
engage in riotous activity. Whether or not the charge is true in his case, there
is no doubt that this is a specific role in riot production that makes it possi-
ble for criminals to loot, kill, and burn during riots in the knowledge that
they will escape prosecution for their criminal a<.is. The same role is performed
by all the politicians of the city, who go to the police stations after the arrests
of their supporters during and immediately after riots to demand their release.
The latter role is described in my second interview with a station officer
in 1983. When I asked him about the part played by politicians in riots in a
tape-recorded interview, I offered to shut off my tape recorder. He said it was
not necessary because "everyone knows this."

These politicians ... do matter, because whatever they say carries a lot of weight.
But mostly the politicians are not around during these times [of riots J. And
even if they are around, ... they don't give ... good advice. And ... in case
they had given good advice, the situation might not turn to-big riots.
[For example, a constable] met some Jatavs, who were all drunk. And they ...
had a small altercation among themselves. And the constable came back to the
thana and reported the matter. So the S.O. went there himself, without the force.
And as he entered the mohalIa, ... the rumor spread around that ... the
police has come to arrest the people. So they started. throwing stones at him,
and he got injured. So then there was a-small jhagara [fight or quarrel] between
the people and the police. And then ... a lot of politics got ... in it and-
[Interruption from outside].

PRB: Yes, yes, go on. What kind of politics?

SSY: And the people-then said that police has ... been against the Jatavs, and they
approached the officers and they approached the politicians.
 254 / The Practice of Communal Politics

PRB: Which politicians?

SSY: Local politicians of the mohalla. These leaders interfere in little, little things all
the time. [ ... J The local politicians always interfere. And if you don't listen to
them, then they get annoyed; in case you listen to them, then the other party gets
annoyed. 50 the working of the policeman is a very tough job. And ... the politi-
cians are always eager to exploit the whole situation.24

The third statement is from an interview in 1983 with Krishna Kumar Nav-
man, then considered by the civilian and police administration as the princi-
pal riot-monger among militant Hindus, but not yet the successful politician
he was to become.

Aligarh politics are basicany dependent on the relations between Hindus and
Muslims. 25

Enough has been said about Mr. Navman so that his comments at the time
need no interpretation. All that need be noted is that we have here an exam-
ple of the self-fulfilling prophecy; anyone who might have then disputed
Navman's contention has been proven mistaken in no small measure in con-
sequence of his own efforts.
The fourth statement is from an interview in 1983 with a successful Muslim
candidate on the Janata Party ticket (precursor of the BJP in Aligarh) for the
Legislative Assembly in 1977.

Now-my aim, when I had joined the Jan 5angb, was to see what happens when
we make a new political experiment. My aim primarily was somehow or the
other to put an end to the riots in Aligarh, see if this new political experiment ...
open[s] up some avenues where we can go further, hand in hand, the two sis-
ter communities. But this program could not be continued, because the
Emergency came in [1975].26

The fifth statement is from an interview in 1983 with a professor of psy-

chology, AMU, then vice-president of the BIP, Aligarh, and an RSS man.

Dr. Manga Ram. Raj Puri Navrnan. These are the Hindu leaders who are
responsible-for-er- ... for preparing the people for retaliation. That also I
can tell you. I am very clear, and very clean on this matter, that they get ... politi-
cal capital out of it. They get recognition by Hindus. Because otherwise Jan Sangh
is a dead party. Madhok's Jan 5angh. They get recognition by Hindus only, only
 The Practice of Communal Politics I 255

for this reason, that they take the code [cause?] of Hindus at their-during com-
munal riots-in their heads. And they prepare the people to attack Muslims, in
case-er-more than six or seven Hindus-Hindus are killed by them and the
patience is lost of Hindus. Hindus also go to them, for complaint, ki, "No, this
our man is killed. See, our man is killed. While we are not able to do anything."
This is the situation of the politicians. They want that communal ... tension
should prevail. If one-hundred or two-hundred Hindus or Muslims are
killed, they don't bother for that. Because they know that this situation is going
to give them-Muslim votes. Because vote banks is with ... Muslim-
Muslims only. Hindus are divided, among different parties, because Hindus
are intellectuals basically. And they-they know where to join according to
their philosophy of life. Muslims have got no philosophy of life, frankly
speaking, because they are backward, illiterate, and uncultured people, most-
most of them. So they follow their leader. That suppose the leader says, "Oh,
all right, YOll are-under-tension. You are-mm-mm-insecure. You
should go for Congress (1)."27

This interview is of particular interest for several reasons. First, the respon-
dent himself is what I call a "fire tender," someone who moves about the city
uncovering incidents-such as the elopement of a Hindu girl with a Muslim
boy, to be discussed below-ostensibly to prevent such incidents from turn-
ing into communal riots. However, the actual effects of his actions and those
of others who play this role is to keep the embers of communal hostilities
from dying out.
Yet, at this time, when the militant Hindus of Aligarh were divided orga-
nizationally, between the most extreme communalists who remained with
the rump Jan Sangh, on the one hand, and the vast majority who joined the
SIP, on the other hand, this AMU professor was with the more "moderate"
SIP. Although he acknowledged that the two forces joined together when
there were riots, he considered the rump Jan Sanghis he named-Manga Ram
and Raj Puri Navman, Mr. K. K. Navman's son-as themselves guilty of riot
instigation. These men, he said emphatically, prepare "the people for retali-
ation," prepare them "to attack Muslims." He himself, he stresses, is free from
blame on the matter, he is "very dean," but they "get political capital out of
it" in the form of "recognition by Hindus." So are the Hindus in general free
from blame. They are so patient that, even after "six or seven Hindus" are
killed by Muslims, they do not retaliate unless prodded to do so by the likes
of Manga Ram and Raj Puri Navman. It is to the latter persons also that the
Hindus go to report when a Hindu has been killed in a riot.
 256/ The Practice of Communal Politics

This professor, diligent fire tender though he is, extends blame further to
all the politicians. He says outright that they want communal tension to pre-
vail, whereas he sees himself as calming communal tensions. But the persons
whom he calls politicians do not care how many persons from their own com-
munity are killed as a result of the riots that follow from their efforts. Hindu
and Muslim politicians are both guilty, but the Congress politicians are most
guilty, because the Muslims constitute an ignorant vote bank, easily deluded
by the very politicians who are maintaining the communal tensions into vot-
ing for the Congress (I) as their protector.
The sixth statement is from an interview in 1997 with the Muslim candi-
date who defeated Mr. Navman in the 1996 elections.

PRB: How did you come into politics?

RESPONDENT: I was simply a businessman when I felt that the communal forces are
disturbing the Aligarh peaceful city. Then I [was] feeling that I should join the
politics and then I came to try MLA-ship with the BSP party.28

The seventh statement is from an interview in 1997 with an unsuccessful

Muslim candidate for the Legislative Assembly on the BSP ticket in 1996.

PRB: And what is the communal situation now in Aligarh in the old city?
RESPONDENT: At present, it is all election gimmick, that Hindus are trying to exploit
the Hindu sentiments and the Muslims are exploiting that, some political parties,
Muslim political parties, and also the person in power is also just trying to con-
solidate Muslims.29

The eighth statement is from an interview in 1999 with a retired professor

of education, Barahseni Degree College, Aligarh, a lifelong member of the
RSS, former secretary of the VHP.

P RB: And what is the reason [for persisting riots in Dahi Wali Gali and various other
mohallasJ?
RESPONDENT: Efforts are being made even now, efforts are being made to create a
riot in the town. This is also politicaL Because Kalyan Singh is supposed to be a
RSS man and who is JeaderofBajpa [BJP], is chief minister ofU.P.30

The ninth statement is from an interview in 1999 with the unsuccessful Muslim
Janata Dal candidate for the Legislative Assembly in 1993.
 The Practice of Communal Politics 1257
PRB: How did you come into politics?
RESPONDENT: In '90, in Aligarh, there was a communal riot. In that communal riot,
in the whole city, there were around two hundred casualties, and this was how. I
mean, I was very much affected, I was quite sensitized, and I got inspiration, I saw
the atrocities committed by the police, their support to the other community, and
I stood against that, my voice, and the people came behind me, people gave me
support, supported my voice. That's how I entered the politics. People knew me
and I thought to contest the election and enter into politics.
So that was the entry point for me and then I joined party, I formally entered
into politics, I started some constructive program on the [part-time] basis, and I
found that the Janata Dal, which was on rise at that time, is the better alternative
of Congress. This was the party which was mainly concerned with the problems
of the Muslims. It had genuine concern for the people. So, I thought it had also
good programs for the people, so I thought to join this party and work from Aligarh,
and this is how I entered in politics,31

It should be noted that in several of the interviews cited above, I had not
yet broached the communal question at all. As in nearly all my interviews with
active politicians, I began with general questions concerning the respondent's
background, including always the question, "How did you come into politics?"
When a respondent first mentioned that he had joined a particular political
party at some point, I generally asked why he chose that party. In all such cases
cited above, respondents began with the communal answer to a neutral ques-
tion. The successful Muslim Janata Party candidate in 1983 declared that it was
his intention, in the strange position of a Muslim joining with the former Jan
Sangh, to do some good, to bring an end to communal riots in Aligarh. The
winning Muslim candidate in 1996, himself considered by Hindus as a firebrand
likely to appear with a large group of people at a disputed site, averred that he
came into politics because "communal forces" were disrupting the peace of
the city. The unsuccessful Muslim candidate in 1993, who does not have a rep-
utation as a firebrand, and entered politics only after the great riots of 1990-91,
also said that he did so because of his reactions to "the atrocities committed
by the police" and the latter's partiality in favor of the Hindus in those riots.
Further, he avers that he chose the Janata Dal because, of aU the alternatives
to the Congress, this was the party most sympathetic to Muslim concerns.
In the two cases, when I was first to broach the communal question or the
reasons for rioting, the responses placed them promptly in a political con-
text. The Muslim BSP candidate in 1996 characterized the communal ques-
 258/ The Practice of Communal Politics
tion as nothing but an "election gimmick," exploited by persons from both
communities. The RSS respondent in 1999 attributed the persistence of riots
to political factors and maintained that efforts were being made even as we
spoke "to create a riot in the town" to embarrass the then BJP government
led by Kalyan Singh, himself a man of the RSS.

ALIGARH'S INSTITUTIONALIZED RIOT SYSTEM

From several of the interviews cited and commented upon above, we have
detected aspects of what I call an "institutionalized riot system," by which
I mean a perpetually operative network of roles whose functions are to main-
tain communal hostilities, recruit persons to protest against or otherwise
make public or bring to the notice of the authorities incidents presumed
dangerous to the peace of the city, mobilize crowds to threaten or intimi-
date persons from the other community, recruit criminals for violent
actions when it is desired to "retaliate" against persons from the other com-
munity, and, if the political context is right, to let loose widespread violent
action. That network has been in existence for a very long time. In my own
experience, it has been in existence from the time of my first visit to Aligarh
just after the October riots in 1961. However, it extended vastly over the next
thirty years, displaying its expanded powers in the great riots of 1990-91,
which were marked by two features that distinguished them sharply from
all earlier riots. The first was the ability of militant Hindus to produce vio-
lent anti-Muslim crowds at a multiplicity of sites in different parts of the
city at roughly the same time. The second was the reported extensive par-
ticipation of large crowds of ordinary people in the cheering, looting, and
killing that went on.
Riot systems exist in both the Hindu and Muslim communities of Aligarh,
but the militant Hindu riot system is a widespread, well organized, contin-
uously functioning one that bears no comparison with the mostly local Muslim
networks. As we have seen from the many quotations provided above from
interviews, the entire system that produces Hindu-Muslim riots is sustained
by the political framework in which nearly all successful politicians have oper-
ated since the destruction of the Congress organization in Aligarh City in the
1962 election. It is sustained, in other words, by a political discourse that legit-
imizes it, even while condemning it, as nothing but routine politics. The com-
munal politics that provide the context are said to be mere "election
gimmicks." The politicians who are playing the game may not even be "com-
muna]" persons, they may even be "decent" men, as one respondent put it,32
 The Practice of Communal Politics I 259
but the game they are playing is the game of communal politics, which has
often fatal consequences for many of its victims.
The inner workings of the organization of riot systems is not something
that can be described in detail. To penetrate it thoroughly, one would have to
be a Bill Buford, make friends with organization members and join in the action
oneself, something quite beyond what a foreigner could do even if so inclined.
What I have seen and tried to present here are but glimpses of a system that
operates at many levels, from the high level of the communal discourse to the
lowest level of the criminals deliberately recruited and paid to kill.
I want to conclude this discussion with two further such glimpses taken
from interviews previously cited, with two respondents, the first being an
example of the "conversion specialist," the second the "fire tender." The for-
mer described to me an incident that had occurred in 1961 in which he was
personally involved: "There was one incident in 1961 when the son of a Punjabi
Khatri was arrested for committing a theft in AMU. However, apparently,
he was badly beaten at AMU before being arrested and we had to rush to the
AMU to rescue him."33
A Hindu thief is arrested in the AMU. It is said he was badly beaten. This
is clearly a matter for the police, but somehow the news reached this respon-
dent and his associates, who then "had to rush to the AMU to rescue him."
Revealed here is something that occurs repeatedly in Aligarh: the existence
of a news and action network concerning all incidents that involve Hindus
and Muslims. The news is transmitted instantly from the scene to persons in
the RSS and in the Jan SanghlJanata Party/BJP-whichever political forma-
tion represents the RSS ideology-who then decide whether or not it is wor-
thy of their attention and action. If it is deemed so-that is, if the political
context is such that "political capital" can be gained from it-then a group
win be gathered to rush to the scene. They will not necessarily rush with the
intent to begin a fracas, but they go to observe, to confront, and, if necessary,
to decide whether or not further forces should be mobilized.
At times, it is possible to witness the operation of the communal discourse
in practice as well as in speech, as illustrated by my observation of and inter-
view with tlle AMU professor. I observed him first by chance in a visit to the
office of the senior superintendent of police (SSP), where, when called upon
by the SSP to state the reason for his presence, he said he had come to report
the alleged kidnapping of a Hindu girl by a Muslim. The SSP immediately
challenged the use of the term kidnapping and said, "You mean eloped."
With this professor, who was neatly dressed in Western clothes, was a scruffy-
looking bunch of dirty people whom he told me later were all citizens of
 260 / The Practice of Communal Politics
Aligarh, including another professor and some people from the local branches
of the State Bank of India and the Bank of Baroda.34
When I asked him later in a formal interview why he had gone to the police
station with these people to report the elopement, he said that he believes
that, when such things happen, they should immediately come forward and
seek police help in order to prevent a communal riot before rumors start to
spread. When I asked what interest the people with him had in the matter,
the reply was they were all Hindus, so they came for the Hindu girl. When I
asked him how he himself had come to know about it, he said the girl's brother
came to him and he immediately arranged for the people to go in the form
of a delegation to the SSP.
This gentleman admitted on my pressing him that the girl was not kid-
napped, she was "lured"; he used the word kidnapping to get the attention
of the SSP. She came from a very poor family, he said, which did not have
two meals a day, so she might have been lured for money. He revealed to my
assistants before I interviewed him and to me during the interview that, if
the girl was recovered, she was likely to be killed by the members of her fam-
ily. He believed that was the proper thing to do in the circumstances because
it would save hundreds of lives in Aligarh, preventing a riot that might oth-
erwise occur; therefore, it was better that one person be sacrificed to save the
lives of many. The girl was seventeen and a half years of age at the time. He
stressed, however, that "the Hindus will do nothing with tlle Muslim man
because that will lead to a communal riot. It is up to the Muslims what to do
with him. There are some good-thinking Muslims in Aligarh and they should
punish that boy."
My respondent also made the argument that, if the girl married the Muslim
man, nobody would mind, but the man-or boy-was already married.
However, he then went on to say that if she went to the family of a Muslim,
she would produce only Muslim children and they (Hindus of Aligarh and
India) don't want the Muslim population to increase because, once they are
in majority, they will behave with their minorities as they do in Iraq and Iran.
Thus does a trivial incident involving two youths become projected spatially
and temporally to encompass the worlds of Hindu India and of an imagined
Islamic universe, in a dire future in which the 85 percent majority of the sec-
ond largest country in the world becomes a minority dominated by prolific
Muslims.
While this man's speech is likely to seem to us the ravings of a lunatic, it
would be a mistake to believe that he is unrepresentative of the views and
sentiments of a large and increasingly outspoken segment of the Hindu pop-
 The Practice of Communal Politics 1261

ulation of north India. Equally important is the specific role he is playing in

this context, one which I have seen repeatedly in different forms in other places
in north India. He proclaims that he seeks to prevent a riot, whereas it would
seem his actions are designed to promote one. In fact, I do not believe he
sought either to prevent or to provoke a riot. He was, rather, performing a
critical role in the perpetuation of the communal discourse and communal
tensions in Aligarh, a role that is being performed in most cities and towns
in northern and western India these days-namely, keeping the fires of com-
munalism tended so that, when the opportunity arises, the fire may be stoked
into a flame, into one of those apparently spontaneous conflagrations that
we all hear about in connection with riots everywhere.
 10 / Communalization and Polarization
Selected Constituency-Wise Results for Aligarh Elections

T he effects of riots on the processes of communalization and polar-

ization in Aligarh elections are dramatically illustrated by the elec-
tions that followed the communal riots in October 1961 and by the
series of elections associated with the Ram Janmabhoomi movement and the
riots that occurred along with it. The effects of a decline in communal vio-
lence on the electoral process in Aligarh are illustrated by the elections that
followed after the ebbing of the great wave of violence in India and in Aligarh
after the destruction of the mosque in December 1992; they will be discussed
in Chapter 12.

THE 1962 GENERAL ELECTIONS

Legislative Assembly
The entire pattern of electoral politics in Aligarh City changed dramatically
in 1962, a major watershed in the political history of the city. As indicated in
Figures 8.3 and 8.4, turnout and interparty competitiveness (leaving aside the
1955 bye-election) reached new highs. The elections of that year, fought in
the aftermath of the October 1961 Hindu-Muslim riots, saw the merger of
two discontented segments of the electorate-Muslims and low caste
Chamars/Jatavs-in an alliance under the banner of the Republican Party
of India (RPI). The leader of the low castes in Aligarh was B. P. Maurya, him-
self a Jatav and the RPI candidate for Parliament. His running mate in the
Aligarh City Legislative Assembly constituency was Dr. Abdul Bashir Khan,
professor of law in the Aligarh Muslim University and the leader of a group
in university politics commonly dubbed as the "communalist" group, as
opposed to the so-called Communist group. It would be less pejorative to
say that Dr. Abdul Bashir was a conservative, attached to his identity as a

262
 Communalization and Polarization /263

TAB LE 10.1. Election results for Aligarh City Legislative

Assembly constituency, 1962

Percental
Turnout Castel Votes Total Valid
No./% Candidates Community Party Polled Votes

51,308
68.25% Abdul Bashir Khan Muslim RPI 21,909 42.70
Anant Ram Verma Kayastha INC 16,164 31.50
Madan Lal Hiteshi Barahseni IND 5,584 10.88
Tota Ram Vidyarthi Bania BJS 3,270 6.37
(Marwari)
Bhoj Raj NA IND 1,949 3.80
Kanhiya Lal NA PSP 881 1.72
Deo Datt Kalanki NA IND 844 1.64
Devld Nandan Sharma NA SWA 316 0.62
K. K. Trivedi Brahman IND 157 0.31
TotaRam NA IND 118 0.23
Haflsur Rehman Dausi Muslim IND 116 0.23

SOURce: Result sheets in the A1igarh District election office.

Muslim, who stood for the defense of Muslim interests that he saw endan-
gered by the October riots, during which he was proctor of the university.
While he was proctor, the Students Union elections had been communalized
by the mobilization of Hindu and Muslim candidates behind candidates of
their respective communities, leading to a clash that in turn was followed by
anti-Muslim rioting of Hindu mobs from the city-both in the old city and
in areas adjacent to the university-in which fourteen or fifteen Muslims were
killed (see Chapter 3). Among Dr. Bashir's complaints at the time was that
the police failed to provide adequate protection to Muslims and to univer-
sity employees when these crowds attacked in the vicinity of the university.
In the election of 1962 in the city, most Muslim voters mobilized behind
the candidacy of Dr. Bashir, who also was supported by Chamars. With this
dual support base, Dr. Bashir won a decisive victory against the Congress
candidate, Anant Ram Verma, who polled less than a third of the votes (Table
10.1). The Congress was also weakened by the continuing factional feuds within
the organization that led some Congress men to sabotage the campaign of
their ostensible candidate by remaining aloof or working against hinI. There
 264 / Communalization and Polarization

was, therefore, a communal and caste mobilization behind the candidacy of

Dr. Bashir as well as the beginnings of a polarization between Hindus and
Muslims. The polarization was not, however, complete. The Congress can-
didate did not speak for Hindu interests. The Jan Sangh candidate, who did
so, placed fourth and lost his deposit with only 6.37 percent of the vote. In
fact, the Jan Sangh dropped to fourth place in this election, polling less than
half the vote it secured in 1957. Third position in this election was occupied
by another Hindu candidate, Madan Lal Hiteshi, a prominent Hindi jour-
nalist, Barahseni (Varshney) by caste, who drew considerable Hindu support
away from the Jan Sangh.
A professor of political science at the Barahseni College in Aligarh remarked
on the impact the latter two candidates had upon the election contest in 1962
in the following manner, in the notes from my interview with him.

There are two business communities in Aligarh-Barahsenis and Agarwals-

and there is a tug-of-war between these two communities; they don't see eye
to eye in the public life of the city. The Barahsenis have their own market-
Barahseni Bazaar. However, when the news went around that Dr. Bashir was
going to win, all the Hindus united around Anant Ram Verma. The Jan 5angh
candidate, Tota Ram Vidyarthi, is also a Bania (Marwari); he is a lawyer, man-
ager of a school (Maheshwari Intercollege), and connected with a number of
sectarian institutions. All those who would have voted for these two candidates
voted for Anant Ram Verma. l

While the distribution of the votes for these four candidates by polling station,
as well as the correlation coefficients between party vote shares and percent
Hindu and Muslim, do support the view that the election was communalized,
they do not support this respondent's assessment of the election as completely
polarized between Muslims voting for Dr. Bashir and Hindus uniting behind
Anant Ram Verma. Rather, they indicate that, in 1962, Hindus were not fully
consolidated and remained at least partly divided along lines of caste antag-
onisms and party affiliation within the Hindu community.
The communalization of the electorate in 1962 is evident from both a com-
parison of the leading party and independent candidate votes in 1962, shown
in Table 10.2, and from a comparison of the distribution of votes among the
parties between 1957 and 1962. Table 10.2 shows starkly the concentration of
RPI votes in the high Muslim population mohallas and the virtual irrelevance
of this party in the high Hindu population mohallas. The difference between
Dr. Abdul Bashir Khan's vote in his top five polling stations and his bottom
 Communalization and Polarization /265
TABLE 10.2. Vote shares for party candidates in their top and bottom five polling
stations, 1962 Legislative Assembly elections, and demographic data for the mo-
hallas included. in them, according to the 1951 census (all data in percentages)

Best/Worst Scheduled
Polling Stations Vote Share Muslims Castes Hindus/Others

RPI top 5 91.39 88.99 353 7.48

RPlbottom5 1.84 1.07 1.44 97.49
INC top 5 67.18 4.94 2.08 92.99
INC bottom 5 2.57 81.98 0.91 17.11
IND top 5 44.35 11.98 3.93 80.07
INDbottom5 0.56 84.55 6.32 6.06
BJStop 5 24.74 11.68 0.42 87.90
BJS bottom 5 0.17 72.37 10.05 17.59

five is nearly 90 percent. The difference in the Muslim and Hindu popula-
tions in those polling stations is nearly 88 percent for the Muslim population
and 90 percent for the Hindu. It is also noteworthy that five of the 14 mohal-
las contained within Dr. Bashir's five best polling stations were among those
later classified as communally sensitive (see Chapter 6), of which three are
considered both riot-hit and crime-prone. The latter included Turkman Gate,
where there is a three-way division, demographically, among Muslims (38.70
percent), Scheduled Castes (29.08 percent), and Hindus and others (32.21 per-
cent), the most explosive population mixture imaginable at times of rioting
in Aligarh. It also included Sabzi Mandi, where there is a mixed Hindu-Muslim
population ratio of 71-29 in favor of Hindus, and Bani Israilan, nearly 100
percent Muslim. 2
The spread in the Congress vote share between its candidate's top and
bottom polling stations and the Muslim and Hindu populations is also huge:
there is a difference of nearly 65 percent in its vote share and a spread of above
77 percent in the size of the Muslim population and above 75 percent in the
Hindu population. The case is sinIilar with the Hindu independent candi-
date and with the BJP candidate as well; the vote share spread between the
top and bottom polling stations is much smaller than in the case of the two
leading candidates, but only because their total vote was also much smaller.
However, the spread in the percentage of Hindus and MuslinIs between the
top and bottom polling stations for these two candidates is also huge.
 266/ Communalization and Polarization

The damage done to the Congress candidate-the same person who con-
tested in 1957-by the communalization of the electorate is also evident from
a comparison of the distribution of his vote shares in his best and worst polling
stations in the two elections. In 1957, the interval between Anant Ram Verma's
vote shares (a spread of 43.67 percent vote share between his best and worst
five) was very much less than in 1962 (spread of 64.61 percent between his
top and bottom five polling stations), as was the spread in communal com-
position of those polling stations with respect to the proportions of Hindus
and Muslims in them} Further, the polarization of the vote between the two
leading candidates in 1962 was very high, as indicated by the correlation
coefficient for the votes of the RPI and the Congress in all 96 polling stations
(-.78, P = .000). There was, therefore, a polarization of votes but not of can-
didates representing the two communities. Muslims were clearly united behind
Dr. Bashir, but Hindus were divided among three candidates: Congress, inde-
pendent, and Jan Sangh.
Indeed, the outcome in this assembly election might well have been
different had Hindus united behind a single candidate. The degree of com-
munalization of the electorate as a whole, the division of the Hindu vote
among three candidates, and the concentration of Muslim and Scheduled
Caste votes are illustrated by examination of the results in particular polling
stations. For this purpose, I have selected the polling stations in which the
intensity of interparty competition was greatest on the interval measure,
namely, polling stations 55 (men) and 56 (women). The interval between the
first- and second-place candidates (Anant Ram Verma and Dr. Abdul Bashir,
respectively) was 2.64 percent. The polling stations comprised six mohallas
in which Hindus and Scheduled Castes were the two largest categories, with
a population distribution as follows: Hindus and others 59.85 percent;
Scheduled Castes 30.63 percent; Muslims 9.53 percent. The Congress candi-
date polled 38.65 percent of the vote, the RPI candidate 36.01 percent.
However, the distribution of the vote for all the leading candidates mirrored
very closely the population percentages, such that the vote for the three Hindu
candidates was 59.46 percent (virtually identical to the population of Hindus
and others), whereas the RPI vote was 40.16 percent (four percentage points
above the combined population of Muslims and Scheduled Castes). Had the
Hindu vote consolidated behind one of the Hindu candidates, the vote in
these polling stations would have been nearly 23 percent above that for the
RPI. In the constituency as a whole (Table 10.1), had the Hindu vote of the
three leading Hindu candidates been combined around a single Hindu can-
didate, the margin of victory would have been around 6 percent.
 Communalization and Polarization 1267

LokSabha

The 1962 elections for Parliament were held simultaneously with those for
the Legislative Assembly. They were even more intensely fought in the
Aligarh segment than those for the Legislative Assembly seat. In the dataset
of 21 elections, this contest ranks fourth in interparty competitiveness in the
history of assembly and parliamentary elections in Aligarh. The interval
between the winning and runner-up candidates was only 2.44 percent.4
There are similarities, but also some differences between the results at the
two levels because of the somewhat different caste and communal identifi-
cations of the candidates. As already noted, the successful RPI candidate for
Parliament was Mr. B. P. Maurya, the most important Republican Party leader
in the state ofU.P. He led also in Aligarh City, polling only slightly more votes
than his party's candidate for the assembly (see Table 10.3). Maurya was the
true architect of the 1962 electoral transformation in the Aligarh parliamen-
tary and Legislative Assembly constituencies. He had worked for the Jatavs
since he began his political career in the 19405. He had then joined the
Scheduled Caste Federation of Dr. B. It Ambedkar, the famous "Untouchable"
leader and founder of the Republican Party, and had since led several agita-
tions in Aligarh District and in other parts of the state for the satisfaction of
Jatav demands and for the conversion of Jatavs to Buddhism. 5 He appealed
specifically to caste antagonisms oflower-caste groups against the upper castes,
particularly Brahmans.
The runner-up candidate in the constituency as a whole as well as in the
parliamentary segment was Maurya's opposite in every way. A Hindu Rajput,
he was the candidate of the Arya Samaj, the preeminent Hindu religious reform
movement of northern India since the late nineteenth century. The Arya Samaj
has always stood for unity of the Hindu community and unity of the coun-
try as well. At this time, it even favored the elimination of the federal system
and the establishment of a unitary state, the adoption of Hindi as the sale
official language of the country, and the combatting of both "parochial"
attachments to any "particular part of India" and "extra-territorial" loyal-
ties. The latter obviously referred primarily to Muslims allegedly attached
by feelings ofloyalty to Pakistan rather than India. Shiv Kumar Shastri's mil-
itant Hindu position won him the support of the Jan Sangh as well as broad
support among Hindus in Aligarh City, enabling him to win 40.31 percent
of the city's vote share, displacing the Congress, whose candidate came in
second in the assembly contest, to third place in the Aligarh parliamentary
segment (Table 10.3).
 268 / Communalization and Polarization
TAD LE 10.3. Election results for Aligarh City segment
of Aligarh Lok Sabha constituency, 1962

Percent
Turnout Castel Votes of Total
No./% Candidates Community Party Polled Valid Votes

51,907 Buddha Priya Maurya Jatav RPI 22,190 42.75

NA Shiv Kumar Shastri Rajput IND 20,296 40.31
Jarrar Haider Muslim INC 7,260 13.99
Nahar Singh Jat IND 731 1.41
Vasant Rao Oak NA SWA 568 1.09
Jaganath Prasad NA RRP 232 0.45

SOURCE: For the data, as for table 10.1. For party affiliations, Government of Uttar Pradesh,
Office of the Chief Electoral Officer, Uttar Pradesh, Results of Third General Elections (1962) to
the House of the People from Uttar Pradesh (Allahabad: Superintendent, Printing and Stationery,
1962), p. 20. Dates of poll: February 19,21,25,1962, are the dates given for phased polling for the
whole 10k Sabha constituency.

The third-place Congress candidate, a Muslim, was an old Congressman

who, however, had not been active in district politics in recent years. He was
a lawyer in Aligarh town and, like both Dr. Abdul Bashir and B. P. Maurya,
had close connections with the AMU, in his case as its legal advisor. He was
a non communal, secular Muslim who had been opposed to the Pakistan
movement and the partition of India. In AMU politics, he was identified with
the group variously called secular, progressive, or Communist, and opposed
to the policies of the group variously termed conservative, traditional, or com-
munalist (that is, the group associated with Dr. Bashir). He claimed that
Muslims who were angry about the October 1961 riots were opposed to his
accepting the Congress nomination. 6 In short, Jarrar Haider was just the sort
of candidate who fit Congress policies at the time, a man with a national, non-
parochial, noncommunal outlook, whom the national leadership of the Con-
gress would have welcomed in Parliament. He was Muslim in a predominantly
non-Muslim constituency-that is, in the parliamentary constituency as a
whole-a noncommunal Muslim in an environment where Hindu-Muslim
relations were antagonistic, and a man unattached to local factions. However
commendable his selection by the Congress under the circumstances, hi" qual-
ities were a certain recipe for a severe defeat. He polled a mere 13.99 percent
 Communalization and Polarization /269
of the vote in the Aligarh City segment, compared to 31.50 percent for his
"supporting" assembly candidate.7
The correlation coefficients for all important candidates in the two elec-
tions, assembly and parliamentary, for all 96 polling stations, are of extraor-
dinary interest. First is the nearly perfect correspondence between the vote
for the two RPI candidates, Dr. Bashir and B. P. Maurya, the correlation
coefficient between their respective vote percentages being .9963. yielding
an R Square of 99.26 percent. This correlation is most striking at a time in
Indian political history when party loyalties were generally weak, the par-
ties themselves were highly factionalized, and there was a relatively low cor-
respondence between party vote shares in assembly and parliamentary
contests held simultaneously, because of caste and communal cross-voting
by ethnic rather than party identification. Also of great interest is the fact
that there were no positive correlations between the vote share for the RPI
for either Legislative Assembly or Parliament with any other party. In fact,
all correlations with all other parties were not only negative (ranging from
-.36 to -.94), but all were significant (at p=.ooo). These correlations suggest
the following conclusions about caste/communal voting behavior in Aligarh
City in 1962. There was extremely high solidarity between the Muslim and
Scheduled Caste voters in this election and virtually total vote transfer across
caste!communallines, that is, virtually all Scheduled Castes who voted for
B. P. Maurya for Parliament also voted for Dr. Bashir in the assembly con-
test. Further, virtually all Muslims who voted for Dr. Bashir in the assembly
contest voted for B. P. Maurya for Parliament.
A further point of great interest for which the correlation coefficients pro-
vide evidence is the very high degree of polarization of voting that occurred
in the Lok Sabha contest, much higher than that in the assembly contest where
the Hindu votes were divided among three candidates. In the parliamentary
contest, by contrast, the vast majority of the Hindu votes dearly were con-
centrated behind Shiv Kumar Shastri. The correlation coefficient between his
vote share and that for B. P. Maurya is -.944; the correlation between his vote
and that for Dr. Bashir is virtually identical at -.941. In contrast to the virtu-
ally total solidarity of Muslim and Scheduled Caste voting, however, the divi-
sions in the Hindu vote at the assembly level were reflected in the correlation
coefficient between Shiv Kumar Shastri's vote for Parliament and the com-
bined vote for the Jan Sangh candidate and Madan Lal Hiteshi in the assem-
bly contest, which stood at .70. At the same time, despite the divisions in the
Hindu vote in the assembly contest, this correlation suggests the existence of
 270/ Communalization and Polarization

a very considerable support base for politicized militant Hinduism in both

electoral contests, one that displayed itself in cross-party voting. That is, most
Hindus who voted for the Hindu Congress candidate, Anant Ram Verma, in
the assembly contest did not vote for the Muslim Congress candidate for
Parliament, Jarrar Haider.
Another set of data that throws further light on the communalization and
polarization that took place in the 1962 elections and adds additional preci-
sion to it comprises correlations and regressions between party vote shares
and the percent population Muslim, Hindus and others, and Scheduled Castes.
A set of data for 44 electoral/census units in 1962 was generated by grouping
polling stations and mohallas to correspond exactly, each one comprising one
or two polling stations and several mohallas. Correlation coefficients were
then run for all the leading parties and independents for both the assembly
and Lok Sabha contests; they are presented below, arranged for each com-
munity in such a way that the highest positive correlations are shown on the
left, the lowest on the right of Table 10.4.
Looking first at the party/independent vote share correlations with per-
cent Muslims, there are several notable results. The first are the very high cor-
relations between percent Muslim and the vote for the RPI. Second, confirming
the findings above from our other data, the correlations for the RPI candi-
dates for both Legislative Assembly and Parliament are very close. Third, the
correlations for both Congress candidates are negative, bringing out clearly
the loss of Muslim support for that party. Fourth, however, it should also be
noted that the negative relationship for the Congress parliamentary candi-
date, Jarrar Haider, is much lower than that for the assembly candidate, Anant
Ram Verma. This difference suggests the possibility that at least some Muslims,
but certainly a very small minority among them, preferred to vote for a Muslim
candidate, not the Jatav candidate, B. P. Maurya, on the RPI ticket. Fifth, the
correlations with militant Hindu candidates are also of interest from several
points of view. They demonstrate that the most disliked candidate in Muslim
areas was the independent contestant for the Lok Sabha seat, Shiv Kumar
Shastri. After Shiv Kumar Shastri, the least popular candidate in the Muslim
mohallas was the Congress nominee, followed tlten by the independent, Madan
Lal Hiteshi, and the Jan Sangh candidate, Tota Ram Vidyarthi.
Looking next at the correlations with Hindus and others, they are virtually
a mirror image of the correlations with percent Muslim. The only important
differences are the strikingly high correlations at both the positive and nega-
tive ends, with Shiv Kumar Shastri on the positive side and the RPI candidates
on the negative. These correlations are, once again, extraordinary in social sci-
 Communalization and Polarization 1271

TABLE 10 .4. Correlation coefficients of party vote shares with percent

population Muslim, Hindus and others, and Scheduled Castes,
1962 Legislative Assembly (LA) and Lok Sabha (15) elections (N :::: 44)

MUSLIMS

Congress Jan Sangh Congress

RPI(IA) RPI (IS) (IS) (LA) IND(LA) (LA) IND(IS)

.90** .88** -.31* -.44** -.47** -.74** -.86**

HINDUS AND OTHERS

Congress JanSangh Congress

IND(LS) (LA) (LA) IND(LA) (LS) RPI (IS) RPI (LA)

.96** .73** .58** .53** .22 -.95** -.95**

SCHEDULED CASTES

Congress Congress /imSangh

(IS) (LA) RPI (IS) RPI (LA) IND(LA) IND(LS) (LA)

.29* .12 .04 .01 -.10 -.12 -.30*

*P=.05 or less.
**p=.Ol or less.

ence research. They explain approximately 92 percent of the variance in the

vote for these candidates. It can be said with assurance that Shiv Kumar Shastri
was the overwhelming favorite candidate for Hindus of Aligarh in the 1962
elections and that he received virtually all his votes from Hindus, especially
militant Hindus whose dislike for the RPI candidate was intense.
Another striking result comes from stepwise regression equations run with
the RPI assembly candidate's vote share as the dependent variable and per-
cent Muslim and percent Scheduled Castes as the independent variables. The
first equation gives a correlation with percent Muslim of .896, which trans-
lates to an R Square of 80 percent, leaving only 20 percent of the variance
unexplained. In the second step, the addition of percent Scheduled Caste to
the equation yields a Multiple R of .951 and an R Square of 90 percent. Thus,
half the unexplained variance is accounted for by the addition of Scheduled
Castes to the equation, leaving a mere 10 percent of the variance unexplained.
 2721 Communalization and Polarization

All the evidence available, therefore, indicates clearly and decisively the
extraordinary high degree of communalization of the electorate in 1962, as
well as a polarization between Hindus on the one side and Muslims and
Scheduled Castes on the other side that was less than total only because there
were two militant Hindu candidates in the assembly contest. The results in
general also point without doubt to the connection between communal antag-
onisms and the communal riots of October 1961 and the results of the elec-
tion. Muslims turned away from the Congress and towards the RPI out of
resentment against the Congress because of the riots that occurred under
Congress rule. The alliance with Scheduled Castes turned this resentment into
a massive defeat for the Congress, from which it never fully recovered. For
their part, Hindus turned in larger numbers than ever towards either the Jan
Sangh or independent candidates who stood for militant Hindu feelings.

THE LEGISLATIVE ASSEMBLY ELECTIONS

OF 1989, 1991, AND 1993

A second illustration of the relationship between riots and elections comes

from the three successive Legislative Assembly elections in 1989, 1991, and 1993,
which were saturated with communal antagonisms. In all three Legislative
Assembly elections, the HJP candidate was Krishna Kumar Navrnan. N avrnan's
rise to prominence in Aligarh politics and the thoroughgoing communali-
zation of the electoral contest occurred simultaneously with the intensification
of the Ayodhya movement that took place in the years between 1989 and 1993.
Furthermore, the intensity of the Legislative Assembly contest increased with
each successive election in this triad, as reflected in the interval measure, which
moved from 8.17 percent in 1989 to 4.87 percent in 1991 to 1.42 percent in 1993,
the most closely contested Legislative Assembly election in the city's history.
The Ayodhya movement had its most serious consequences for Aligarh
in the vicious communal riots of 1990-91 (see Chapter 4 ). In the midst of
this highly charged and violent atmostphere, Navrnan won three consecu-
tive elections, an unprecedented feat in the electoral history of the city. He
won two of these elections, in 1989 and 1993, because of a division of the
Muslim votes, but won the 1991 election even without such a division.
Rather, in Aligarh as elsewhere in U.P. in 1991, the HJP benefited from the
decline of the Congress as a serious competitor and its own capture of the
bulk of the Congress's upper-caste Hindu vote base. That is, a thoroughgo-
ing consolidation of the Hindu vote behind the HJP swept it to victory in
1991 despite a countervailing Muslim consolidation. Notwithstanding the
 Communalization and Polarization /273

division of the Muslim votes in two of these elections, all three elections were
marked by a high degree not only of communalization but of polarization,
with Muslims failing, however, to consolidate as effectively as did the Hindus
behind the BIP.

The Legislative Assembly Elections of 1989

In the 1989 election, Navman stood alone as the only strong Hindu candi-
date against three Muslims: Khwaja Halim, on the Janata Dal ticket, Moham-
mad Furkan, the Congress candidate, and Habibur Rahman, on the ticket of
the BSP (Table 10.5), which entered the electoral contest for the first time in
this election. Although the two other Muslim candidates besides Khwaja
Halim lost their security deposits, they polled more than enough votes to
deprive Khwaja Halim of victory. It is unlikely that the BSP candidate received
any significant share of the Muslim vote, but it is probable that the Congress
candidate did so.
Mohammad Furkan, who made his first and last appearance in a Legislative
Assembly contest in Aligarh City in this election, was born in the town of
Jalali in a former zamindar family of Aligarh District from the Qureshi
baradari. He attended the Aligarh Muslim University, where his political career
began when he took a prominent part in student union politics at the time
of the mass student agitation directed against Professor Irfan Habib, over a
newspaper article in which the latter was quoted as criticizing severely the
decline in standards and student discipline at the university.8 During that
period, Furkan served as vice-president, then president, of the Aligarh
Muslim University Students Union (AMUSU). He was, however, a moder-
ate among the student militants in that agitation and ultimately worked to
bring it to an end by supporting the vice-chancellor's efforts to restore the
educational environment at the university. After his graduation from the
AMU, Furkan joined the Congress and was appointed one of the joint sec-
retaries of the U.P. Congress (I) in 1988. He was selected the very next year
to contest the Aligarh City Legislative Assembly seat.
Furkan is a noncommunal person of the type commonly associated with
the Congress in the past. In my interview with him, he emphasized that when
he was a student at AMU, "he worked equally for the Muslims and the Hindus"
and had no communal feelings. He claimed that it was because of his non-
communal outlook that he garnered as many as 12,000 votes in the 1989 elec-
tion. However, his candidacy suffered from the communal atmosphere that
pervaded the election of 1989, which occurred in the midst of the shilanyas,
 274/ Communalization and Polarization

T ABLB 10.5. Election results for Aligarh City Legislative

Assembly constituency, 1989

Percent of
Turnout Castel Votes Total Valid
NO./%4 Candidates Community Party Polled Votes

112,061 Krishna Kumar Navman Varshney BJP 51,982 46.39

55.14% Khwaja Halim Muslim JD 42,834 38.22
Mohammad Furkan Muslim INC 12,146 10.84
(Qureshi)
Habibur Rahman Muslim BSP 2,182 1.95
Others (19) 2,917 2.60

'Valid votes turnout. Total votes turnout was 56.70 percent (115,220 total votes in an
electorate of 203,no).
SOURC.E: U.P. Election Office, Lucknow.

the mass movement to bring consecrated bricks to Ayodhya for the building
of a new temple to Ram on the site of the Babri mosque. The decline of the
Congress in U.P. as a whole was already in progress and was further intensified
by this movement, in which it was sidelined. During that time, also, the Shahi
Imam of the Jama Masjid in Delhi issued a Jatwa advising Muslims to vote
against the Congress for its failure to defend the Muslim right to possession
of the site. Finally, as noted above (Chapter 3), the election outcome was
affected by the November 10 riots in the city and the turmoil over the poster
campaign on and around the AMU campus, which contributed to the
Muslim consolidation of its votes behind the Janata Dal candidates for both
the Lok Sabha and the Legislative Assembly.
Table 10.6 shows the distribution of the vote for the three leading candi-
dates in their five best and worst polling stations. The spread in the vote share
for the BIP and the Janata Dal is among the highest ever witnessed in any post-
Independence ele(,tion in Aligarh City. As usual, the BIP's best and worst polling
stations were in Hindu and Muslim mohallas, respectively, whereas the reverse
was true for the Janata Dal candidate. It remains of interest to note once again
also that the best and worst Congress polling stations were much more mixed
than those for either the BJP or the Janata Dal. However low its vote share,
the Congress has remained in Aligarh a party that has been able in most elec-
tions to remain in the political and demographic center on the communal issue.
Of equal interest in revealing the extent of communalization and polariza-
 Communalization and Polarization / 275
TABLE 10.6. Vote shares for party candidates in their top and bottom five polling
stations, 1989 Legislative Assembly elections, and demographic data for the mo-
hal/as included in them, according to the 1951 census (all data in percentages)

Best/Worst Scheduled Hindus

Polling Stations Vote Share Muslims Castes and Others

BJP top S 93.68 9.36 4.62 86.02

BJP bottom S" O.IS 7S.62 11.40 12.98
JD top Sb 9S.86 92.48 1.48 6.04
JD bottom S 1.38 21.00 S.56 73.44
INC top S 64.3S 40.29 9.21 SO.49
INC bottom S' 0.1S 67.71 0.94 31.36

"The demographic data are for the mohallas of Tila, SUnet, and Sarai Rahman only.
Census data for the mohallas of Rorawar, Shah Jamal, and Qayyumnagar in polling station
139, Tauhar Bagh in polling station 18, and Jiwangarh in polling station 22 were not available.
However, these are all predominantly Muslim areas. Rorawar was formerly a village included
in rural Koil tahsil (pargana Koil); there were 88 registered voters here in 1984, of whom 55
(62.5 percent) were Muslim. Shah Jamal is the site of the oldest Mu&lim dargah in Aligarh
District. The new Idgah (prayer ground for the celebration of the Muslim holy day of Id) also
lies in this area. Tauhar Bagh and Jiwangarh, with a combined number of registered voters of
10,362 in 1995, were 100 percent Mu&lirn. Assuming for purposes of a rough estimate only that
the proportion of Mu&lim voters approximates the proportion of Mu&lims in the population
and vice versa, then it is probable that the actual percentage of the Mu&lim population and
voters in these polling stations was 91.69 percent, or closer to that figure than to the figure in
the table of 75.62 percent.
bThe demographic data are for the mohallas of Tila, Sunet, Chowk BUlldoo Khan, and
Sheikhan only. Census data for the mohaUas of Rorawar, Shah Jamal, and Qayyumnagar in
polling station 138 were not available. However, as indicated in the previous footnote, these
are predominantly Muslim cultural areas.
'Census data for mohallas in polling stations 138, 139, and 187 not available, that is, for
mohallas Rorawar, Shah Jamal, Qayyumnagar, Mullapara, and Bhojpur.

tion of the Hindu and Muslim electorate in this election is the presence or
absence of communally sensitive, riot - and/or crime-prone mohallas in the best
and worst polling stations for the three main candidates. The top five BIP polling
stations contained fourteen mohaUas, all but one with Hindu majorities, rang-
ing from 54 to 100 percent, of which two were classed as communally sensitive
(Madar Darwaza [Map 3] and Rafatganj [not on map]), and two riot-hit and
crime-prone (Gular Road and Katra [Map 3]). These four mohallas are also
communally mixed: Katra is the least so, with a Muslim population of 9 per-
cent but the others contain large Muslim populations: Gular Road with 37 per-
cent, Madar Darwaza with 32 percent, and Rafgatanj with 21 percent. Only one
 276 / Communalization and Polarization

small mahalla, Sarai Kalan, had a Muslim majority (above 98 percent). It also
deserves note that four of these mahallas (Madar Darwaza, Tamolipara, Sarai
Pakki, and Brahmanpuri) are in the group that forms a circle round Sarai Sultani,
which, we have noted, has suffered greatly in most of the severe riots in Aligarh.
Although Khwaja Halim was said to have benefited also from communal
sentiments in the aftermath of riots, the distribution of the sensitive and
riot/crime-prone mohallas suggests that the BJP candidate was the prime
beneficiary. Among the seven mohallas in Khwaja Halim's top five, one,
Sheikhan, was classed as communally sensitive. Also in this candidate's top
five, but in the BJP's bottom five, was the mahalla of Shah Jamal (no. 69 in
Maps 1 and 4) that was to be attacked in the great riots of 1990-91. Among the
ten mohallas in the Congress candidate's top five, one was classed as com-
munally sensitive (Tantanpara, Maps 2 and 3), a second as riot- and crime-
prone (Atishbazan, Maps 2 and 3). In this candidate's bottom five were
included two of the mohallas that were targeted in the 1990-91 riots, Shah Jamal
and Bhojpur (Maps 1, 2, and 4). It is, therefore, a matter of considerable inter-
est to note the relative importance of these outlying mohallas in the top and
bottom five polling stations for all the leading candidates. We cannot say that
these Muslim-majority mahallas, whose very presence in the city limits has
troubled Navrnan and the BJP, were targeted for revenge in the 1990-91 riots,
but their importance in this election provides further evidence linking riotous
activity with electoral competition in mohallas in which the results are per-
ceived as critical to the electoral outcome. However, in this case, it is the elec-
toral competition that precedes the targeting of a mohalla for later attack, rather
than a riot in the locality that precedes intensified electoral competition.
The relatively small vote share for the Congress had little effect on the high
degree of polarization between the BJP and Janata Dal candidates, displayed
graphically in Figure 10.1. The huge white spaces on that chart, notably between
polling stations 13 and 49, again between polling stations 97 and 133, and at
several other places, indicate that, in large parts of the city, voters voted over-
whelmingly for one or the other candidate. The contiguity of the polling sta-
tions in which such polarization occurred also reveals that the polarization
is regional within the city.
The extent of two-party polarization is further demonstrated by the num-
ber and percent of polling stations in which one or the other of the two main
parties polled above 75 percent of the vote. That was the case in 116 of 223
polling stations or 52 percent. In 96 polling stations or 43 percent of the total,
one or the other party polled 80 percent or more. In 60 polling stations or 27
percent, the larger party polled 85 percent or more, and, in 28 or 12.56 per-
 Communalization and Polarization /277

120.----------------------------------------,
100

60

40
BJP89PCT

JD89PCT

Polling station number

F lG. 10.1. Vote shares for two leading parties, 1989 Legislative Assembly election

cent, the larger party polled above 90 percent. The interparty correlation of
-.91 between the BJP and Janata Dal vote shares adds still further evidence
of the polarization between the two parties.
Finally, the correlation coefficients indicate that the polarization was
extraordinarily high between Hindus and Muslims, with Scheduled Castes
on the sidelines in this election (Table 10.7). The positive correlation of .82
between the Janata Dal and the Muslim vote percentage is the third highest
in any election since Independence, barring the watershed election of 1962
and the election of 1969. Similarly, the positive BIP correlation with percent
Hindus and others equals its previous high correlation in the 1967 election.
The sidelining of the Scheduled Castes is indicated by the unusually high cor-
relation between the vote for all other parties combined and the Scheduled
Caste population (.43). By far the largest share of the other-party vote was
for the BSP candidate, Habibur Rahman, who polled 2,182 votes, 43 percent
of the total for other parties. His highest vote in a single polling station was
122 votes, garnered in a polling station containing the mohalla of Sarai
Lawaria (Map 2), with a Scheduled Caste population of 25 percent. His sec-
ond highest vote in a single polling station was 120 votes, from a polling sta-
tion containing a newly incorporated village known as Rasulpur, in which
the 1951 census showed a Scheduled Caste population of 64 percent.
Although, therefore, the two other Muslim candidates most likely drew
away the votes that would have defeated Navrnan and given the victory to
Khwaja Halim, the presence of the Congress candidate only partly softened
the high degree of polarization between the two leading candidates. The
 278/ Communalization and Polarization

TABLE 10.7. Correlation coefficients of party vote shares with percent

population Muslim, Hindus and others, and Scheduled Castes,
1989 Legislative Assembly elections (N :::: 70)
MUSLIMS

Ianata Dal BIP Congress (I) Others

.83** -.74** -.09 -.20

HINDUS AND OTHERS

BIP Congress Others Ianata Dal

.76** .06 .03 -.80**

SCHEDULED CASTES

Congress BIP Ianata Dal Others

.09 .06 -.16 .43**

**p=.01 or less.

Congress vote share correlated only at a very low level with Hindus, Muslims,
and Scheduled Castes, indicating the likelihood that its candidate drew some
votes from all three groups. The other Muslim candidate, contesting on the
ticket of the BSP, most likely gained nearly all his votes from Scheduled Castes.

The Legislative Assembly Elections of 1991

The General Election of 1991 is one of the most famous and important in post-
Independence Indian history, another great watershed election for the coun-
try, the state ofU.P., and Aligarh as well. It is the election known in popular
and journalistic parlance as the "Mandal-Mandir" election because of its dual
focus on the issues of backward castes reservations in public sector employ-
ment recommended in the Mandai Commission report and approved by the
Janata Dal government before the ele(.:tion, on the one hand, and the mili-
tant Hindu demand to replace the mosque in Ayodhya with a Hindu temple
(mandir). The conflicts engendered by the Mandai decision and by the kar
sevak movement of volunteers journeying to Ayodhya to remove the mosque
and build a temple to Ram on the same spot, and the fallout from the latter,
 Communalization and Polarization 1279

namely, the actions of Mulayam Singh Yadav and the communal riots that
occurred before, during, and after the kar sevak movement, provided the issues
and determined the new alignment of political forces which dominated the
parliamentary and Legislative Assembly elections in U.P. in May-June 1991.
So-called Hindu-Muslim communal riots were of extraordinary impor-
tance both before and during the election campaign. They provided both a
general context in the state as a whole and a specific context within partic-
ular districts and cities that framed the interparty struggle. In Aligarh, riots
occurred before the election. In this case, as in some others in the state, mas-
sive riots provided a background for the election campaign, helping to frame
it in a way which worked to the advantage of the BIP. The Aligarh riots
occurred during the kar sevak movement. They fed into the BJP propaganda
theme of Muslim instigation, state police protection and pampering of Mus-
lims, and Hindu martyrdom at the hands of "Muslim miscreants" despite
the fact that here, as elsewhere, many more Muslims than Hindus were killed.
Although there were many districts in U.P. that remained relatively
unaffected by communal rioting and somewhat less affected by the general
atmosphere of communal hostility in the state, the communal atmosphere
in U.P. as a whole and the widespread occurrence of Hindu-Muslim riots
and confrontations affected profoundly the results of this election. This was
not an ordinary election. The 1991 election in U.P. stands apart from all others
held since Independence, constituting a serious deterioration in the quality
of the democratic process.
The riots that occurred before and during the campaign had three conse-
quences in the state as a whole, all of which worked to the advantage of the
BIP: concentration of Hindu voting for the BJP, a high turnout among Hindus
as well, and a communal polarization far greater than anything that has
occurred since Independence, for which one has to look back to 1946 for a
precedent. In Aligarh, the results paralleled those in the state as a whole inso-
far as communal and polarized voting are concerned. Where the Janata Dal
and the Samajwadi Janata Party (SIP, predecessor of the SP) did not contest
against each other, Muslims voted overwhelmingly for the Janata Dal. Where
the two parties were divided, so was the Muslim vote. In the latter case, the
high degree of polarization between caste Hindus and Muslims was partly
diluted, though communalization was at a peak. Also mirroring results else-
where in the state, caste Hindu mobilization overwhelmed Muslim solidar-
ity and the combination of many backward caste and Scheduled Caste votes
with the Muslims against the BIP.
In the 1991 election in Aligarh, the first after the great riots of December
 280 / Communalization and Polarization

TABLE 10.8. Election results for Aligarh Legislative

Assembly constituency, 1991

Percent of
Turnout Castel Votes Total Valid
No./%" Candidates Community Party Polled Votes

111,353 Krishna Kumar Navman Varshney BJP 52,670 47.30

54.10% Mohammad Sufiyan Muslim JD 47,249 42.43
(Qureshi)
Bhanu Pratap NA INC 6,903 6.20
Hafiz Usman Muslim SIP 2,445 2.20
Shamim Ahmad Muslim SSP 728 0.65
Others (19) 1,358 1.22

aValid votes turnout. Total votes turnout (113,816) was 55.29 percent. Total electorate was
205,838.

1990-January 1991, Navman again was the only Hindu candidate with enough
support not to lose his security deposit. He improved slightly his total votes
and his vote share to 47.30 percent (Table 10.8). The runner-up candidate on
the Janata Dal ticket, Mohammad Sufiyan, a Qureshi Muslim, polled 42.43
percent of the vote. Although the Congress candidate was this time a Hindu,
the Congress decline that began all over U.P. in 1989 deepened and its can-
didate won only 6.2 percent ofthe vote. The fourth-place candidate from the
SJP was also a Muslim, Hafiz U sman. Although he polled well in a few polling
stations, his overall vote total and vote share were too small to influence the
outcome in the constituency as a whole.
N avman gave the following analysis of the voting pattern in this election.
He claimed that the Muslim candidate on the SJP ticket, the party of M ulayam
Singh Yadav, got only the Ahir (Yadav) votes, that is, the votes of the com-
munity (defined by the MandaI Commission as a backward caste) that solidly
supports that party throughout the state. He averred that his main opponent,
Sufiyan, the Janata Dal candidate, got the votes of Scheduled Castes; how-
ever, because Sufiyan "himself was a Qureshi," a baradari of low status
amongst Muslims, Navman claimed that some "upper-caste" Muslims as well
as Shia Muslims voted for him (Navman),9 This analysis, however, does not
stand up against the large number of votes received by Mohammad Sufiyan.
Although it is of interest that Navman takes a position consistent here with
that of the HJP generally, that some Muslims support the party, it does not
 Communalization and Polarization /281

lWr-----------------------------------------,

BJP91PCT

JD9 I PCT

13 37 61 85 109 133 157 181 205

Polling station number

FIG. 10.2. Vote shares for two leading parties, 1991 Legislative Assembly election

appear likely that there was a significant division of the votes among the
Muslim candidates. The election was, rather, an almost straight fight in a thor-
oughly communalized and polarized electoral contest-held in the aftermath
of the first assault on the Babri mosque in November 1990-in which Nav-
man emerged victorious once again, with a somewhat larger vote share, but
a smaner margin between him and the runner-up.
However, it must be stressed that a straight fight in the constituem:y as a
whole does not mean a straight fight in the mohallas. On the contrary, in Aligarh
it means a regional, that is to say, locality-wise polarization of the votes.
Polarization between the two leading candidates was even more extreme than
in the previous election. The chart showing the dispersal of the BJP and Janata
Dal vote shares of 1991 (Figure lO.2) again contains great white spaces, reflect-
ing the huge spread in the vote between the two candidates, particularly in
predominantly Hindu and predominantly Muslim polling stations. The num-
ber and percent of polling stations in which one or the other of the two lead-
ing candidates polled above 75 percent of tlle vote was also somewhat higher
this time, 122 out of 223 polling stations, comprising 55 percent of the total.
As in 1989, one or the other of the two candidates polled above 85 percent in
60 polling stations in 1991 as well. The correlation coefficient between the vote
shares for the two parties was an extraordinarily high -.97 (N=223).
Polarization of the communal vote was also evident in the distribution of
the votes for the leading candidates in their best and worst polling stations
(Table 10.9). Especially notable in this election was the number of bottom
polling stations in which the candidates polled zero votes or close to zero.
The BJP candidate got no votes at all in nine polling stations, tlle Janata Dal
 282 / Communalization and Polarization

TABLE 10.9. Vote shares for party candidates in their top and bottom polling
stations, 1991 elections, and demographic data for the mohallas included
in them, according to the 1951 census or 1995 voters' lists' (all data in percentages)

Best/Worst Scheduled .Hindus

Polling Stations Vote Share Muslims Castes and Others

BJP top 5 95.06 17.62 2.53 79.85

BJP bottom 4h 0.00 78.67 9.40 H.93
JD top 5 97.58 79.24 9.03 11.73
JD bottom 5 0.33 11.09 5.78 83.13
INC top 5 30.25 12.23 6.11 81.66
INC bottom 6 c 0.00 77.11 2.94 77.11
JPtop 5d 26.11 98.79 NA NA
JP bottom 17' 0.00 29.27 7.56 63.17

•All data are from the 1951 census except where otherwise noted.
bThe BJP polled zero votes in nine polling stations, for which census mohalla data were
available for only four.
'The Congress polled zero votes in seven polling stations, for which census mohalla data
were available for only six.
dMuslinI percent in this row is an estimate from a random number sampling of the 1995
voters' list.
eThe Janata Party polled zero votes in 18 polling stations, for which census mohalla data
were available for only 17.

candidate polled only 0.33 percent of the vote in his bottom five, the Congress
candidate polled zero in his bottom seven, and the Janata Party candidate
polled nought in eighteen. At the top end, the vote shares for the two lead-
ing candidates were extremely high: the HJP polled 95.06 percent of the vote
in its top five polling stations dominated by Hindus while the Janata Dal can-
didate polled an even higher 97.58 percent in his top five polling stations dom-
inated by Muslims. As for the third- and fourth- place candidates, the
Congress Hindu candidate did relatively well only in Hindu-dominated polling
stations. It is a matter of great interest to note that all the mohallas in the top
five polling stations for the Muslim Janata party candidate are located in the
northeastern corner of the city, center of in-migration of Muslims and the
bane of the HJP and Krishna Kumar Navman for that reason. 10 The estimated
percentage of Muslims in these areas is, moreover, nearly 90 percent. In con-
trast, this candidate's bottom seventeen polling stations, in which he got no
votes at all, were in predominantly Hindu mohallas.
 Communalization and Polarization /283

As usual, the BJP's best polling stations included a high proportion of com-
munally sensitive and crime-prone mohallas: four out of eight, including
Rafatganj, Manik Chauk (Navman's home mohalla), Gular Road, and Madar
Darwaza. Also notable is the mohalla-wise distribution of the nine polling
stations in which the BJP polled no votes. In addition to the four mohallas
already mentioned for which we have the 1951 census data, there were four
other mohallas located in the same northeastern corner of the city where
the Muslim Janata Party candidate polled his best, namely, Jauhar Bagh,
Jiwangarh, KelaNagar, and Krishi Farm (Map 1). These mohallas are all located
in two wards, numbers 43 and 47, which are SP strongholds and where the
winning SP candidates in the 1995 corporation elections both were Muslims.
In other words, the BJP was wiped out in precisely the same localities in which
one of the Muslim candidates polled his best, in the area of Muslim in-
migration. In contrast to the BJP, the Janata Dal candidate's top five polling
stations contained only one riot-prone mohalla out of eight, while the
Congress candidate's top five contained none.

The Legislative Assembly Election of 1993

The 1993 election, held after the destruction of the Babri Masjid on December
6, 1992, was in some ways a repeat of the election of 1991, but in other ways
foreshadowed the dramatic change that was to occur in 1996. As in 1991, there
was a high degree of communalization of the election, but polarization was
not as extreme because of a division of the Muslim vote between two Muslim
major party candidates. This division, however, presaged the transformation
of the constituency from a Hindu-dominated to a Muslim-dominated one,
for the combined votes of the two leading Muslim candidates surpassed that
of the victor, Krishna Kumar Navman. The vote shares for the four leading
candidates are given in Table 10.10.
Although the Muslim vote was divided, Krishna Kumar Navman barely
defeated his principal opponent, Abdul Khaliq, a Saifi businessman of AHgarh,
owner of a lock factory, who contested on the ticket of the BSP. Mohammad
Sufiyan, contesting again on the ticket of the Janata Dal, was displaced by
Abdul Khaliq, but gained enough votes to ensure the latter's defeat and the
victory of Navman. Here is how Mohammad Sufiyan himself described the
voting pattern in this election.l l He attributed his poor showing compared
to 1991 to a division of the Muslim vote in 1993, which was not the case in the
previous election. That division in turn was a consequence of what is gener-
ally described as strategic voting on the part of the Muslim community, in
 284 / Communalization and Polarization
TABLE 10.10. Election results for Aligarh City Legislative
Assembly constituency, 1993

Percent Total
Turnout Castel of Votes Valid
No./%U Candidates Community Party Polled Votes

142,0871 Krishna Kumar Varshney BJP 58,027 40.84

62.08 Navman
Abdul Khaliq Muslim BSP 56,005 39.42
(Saifi)
Mohammad Muslim JD 15,338 10.79
Sufiyan (Qureshi)
Vivek Bansal Agarwal INC 10,121 7.12
Others (54) 2,596 1.83

'Valid votes turnout. Total votes turnout was 1450364 (63.51 percent).
SOURC!!: Uttar Pradesh, Mukhya Nirvachan Adhikari, Vidhan Sabha Samanya Nirvachan,
Uttar Pradesh. 1993 (Allahabad: Government Press, 1996), pp. 507-08. Date of poll: November
21,1993.

which Muslims vote as a bloc for the candidate perceived to be the most likely
to defeat the BIP. Sufiyan did not use the term, but his account of what hap-
pened is consistent with that strategy. He remarked that in 1991 the Janata
Dal "was emerging as the single largest party in India," whereas in 1993 it "was
on decline and he was contesting from JD. So, in 1993, the Muslims thought
that, if they all come together under the banner of BSP and vote against these
BJP candidates, then theywiU get their candidate elected. That's why my share
of the vote in 1993 declined." In other words, since the Janata Dal was on the
decline, "there was a common perception that it's only the BSP that can give
figl1t to the BIP. That's why Muslims all came together and they voted against
the BJP candidate."
I asked Sufiyan then if he got any Muslim votes at all. He said that he got
"some percentage of the Muslim vote," mostly "personal votes" that are in
his "pocket," "based on his personal contacts." When I asked where he got
his best votes, he listed the following mohallas: Delhi Gate, Sarai Mian, Khai-
dora, and Sarai Bibi. Three of these mohallas-Delhi Gate, Sarai Mian, and
Khaidora-are predominantly Qureshi mohallas, two are classed as crime-
prone mohallas-Delhi Gate and Khaidora-and all are predominantly
Muslim mohallas. It is clear, however> that Sufiyan got his "pocket" votes pri-
 Communalization and Polarization /285

marily from the Muslim voters of his own baradari. Otherwise, he acknowl-
edges that Muslims voted solidly for Abdul Khaliq.
Congress placed fourth in this election. Its candidate, Vivek Bansal, polled
only slightly more votes than the Congress candidate in 1991, and, like his
predecessor, lost his security deposit. Bansal, an Arya Samaji Hindu and an
Agarwal Vaishya by caste, is a manufacturer of brass tacks and a manufac-
turer and exporter of brass art ware.12 He is a young man, born and brought
up in Aligarh, who has lived all his life in a mansion on the poshest road in
Aligarh, the Marris Road. He is a completely noncommunal person, a grad-
uate of AMU, who has personal relations with Muslims from both the
former landed aristocracy of Aligarh and from the faculty of AMU. He employs
50-55 persons in his factory and claimed, in response to my question con-
cerning the numbers of Hindus and Muslims in his workforce, never to have
considered tlle matter.
Bansal gave the following explanation for his poor performance on the
Congress ticket. First, he said that he was himself new to politics, a "green-
horn," as he put it. But second, the "[Congress] party's position was awful;
this happens to be predominantly a Muslim constituency, ... and minorities
were very, very, very angry with the party on account of ... the demolition
of Babri Masjid in 1992."
Though he fared poorly, Bansal was hopeful of becoming a successful can-
didate on the Congress ticket for the Legislative Assembly or Parliament in
future. In contrast to other candidates, who admitted the restricted castel
communal composition of their support bases, Bansal could not give any list
of polling stations in which he fared well, saying that he did not think he had
any particular strongholds, but got "votes from every booth irrespective of ...
how many I got." Further, he stressed that he believed in and had worked for
Hindu-Muslim "communal amity" in the past, notably during the riots of
March 1996 (sic; presumably March 1994) in the city, when he claims to have
helped persons in the Sarai Sultani mohalla. When I questioned the ability
of a Hindu to win in this communally polarized, predominantly Muslim con-
stituency any longer, Bansal argued that Navman's tactics of mobilizing mil-
itant Hindus was an "easy solution" to winning the election, but that "people
tend to vote for the person who is able to deliver the goods, in the long run."
In short, Bansal was talking in the manner of the traditional, secular Congress
politician of times past and elections past, but the contours of electoral pol-
itics remained communalized and, though the Muslim vote was divided, polar-
ized between the leading Hindu and Muslim candidates.
 11 I Communal Solidarity
and Division at the Local Level

SELECTED MOHALLA-WISE RESULTS

FOR ALIGARH ELECTIONS, 1957 TO 1995
ManikChauk

L et us now have a look at the available election results for the infamous
mohalla of Manik Chauk, center of so many of the riots in Aligarh
since Independence and home and business premises of Krishna
Kumar N avman. Recall first the demographic characteristics of this mohalla.
Even in 1951, it was a large mohalla, with a total population of 3,848 persons.
It was also mixed in its communal composition, although the Hindu popu-
lation was predominant with 75.55 percent of the total, Muslims comprising
24.01 percent and the Scheduled Castes having negligible representation.
According to the 1984 voters' list, there were 2,485 voters in the mohalIa, of
whom, however, only 13.17 percent were Muslims. According to the 1995 vot-
ers list, the number of voters was somewhat less than in 1984; 2,375 voters
were listed, of whom 345 or 14.53 percent were Muslim. It is not clear whether
we can extrapolate from these figures to the total population of Muslims in
the mohalla and assume that Hindu efforts to intimidate Muslims during riots
here have succeeded in reducing their percentage in the population. It is, how-
ever, likely that something of the sort has happened. It is also quite possible
that the number of Muslim voters in the mohalla has been deliberately under-
counted. Let us recall also that Manik Chauk falls into the most severe cate-
gory of riot-proneness, being both riot-prone and crime-hit. This is also a
mohalla in which the predominant community is the Varshney caste. The
mohalla, along with Sarai Barahseni and Patthar Bazaar, falls entirely in ward
no. 58, where the successful BJP candidate in the 1995 municipal corporation
elections was Krishna Gopal Varshney.

286
 Communal Solidarity and Division at the Local Levell 287

In 1957, the first election for which I have polling station data, Manik Chauk
was included in polling station numbers 42 (men only) and 43 (women only),
along with two other mohallas, known as Sarai Bairagi (not on map) and Sarai
Nawab, both located some distance from Manik Chauk (see Map 2).1 The
total population of these two mohallas was 1,293 persons, thus comprising a
quarter (25.15 percent) of the total population of the polling station. Sarai
Bairagi was overwhelmingly comprised of Hindus and others (90.13 percent,
while Sarai Nawab contained a mixed population, 60 percent Muslim, 28.91
percent Scheduled Castes, and 11.09 percent Hindus and others. In this elec-
tion, no doubt because of the mixed character of the polling station, num-
ber 42 (men only) ranked first in the intensity of interparty competition by
the interval measure, a mere two votes (0.28 percentage points) separating
the first-place candidate from the second. In first place in polling station 42
was the Jan Sangh candidate, who polled 38.29 percent of the vote here in
comparison with his showing of 13.53 percent in the constituency as a whole,
whereas the Congress candidate polled 38.02 percent compared to his show-
ing of 50.09 percent in the constituem:y as a whole. We know enough about
these mohallas, especially Manik Chauk, by now to say with assurance that
the Hindus in polling station 42 voted overwhelmingly for the BIP even at
this early date in its history in the constituency. More interesting, however,
from the perspective of hindsight in relation to the post-riot elections that
followed this one, is that, despite the sharp division in Hindu and Muslim
voting, Muslims clearly chose for the most part not to vote for dIe only Muslim
candidate in this election, who polled only 31 votes in polling station 42.
Instead, it is virtually certain that they cast their votes for the Congress can-
didate. So, in this first election in this polling station, we see the seeds of later
Hindu consolidation in its militant Hindu epicenter of Manik Chauk, along
with a communalized electorate in which Hindus and Muslims here-but
not in the constituency as a whole-must have voted for different candidates.
Further, the Muslims adhered to the Congress, then considered in principle
and practice to be a secular political party.
From 1962 d1fough 1991, the available polling station data are for the
mohalla of Manik Chauk by itself. Figure 11.1 compares the vote share for the
militant Hindu candidates in the Aligarh constituency as a whole with that
for the Manik Chauk mohalla for the six Legislative Assembly elections in
that period. There are three significant things to note from the chart. First,
insofar as the line for the mohalla is concerned, there are two peaks, the first
in 1962 following the October 1961 riot, the second in 1991 following the great
riots of 1990-91. Second, the slopes of the two lines are, with the exception
 288/ Communal Solidarity and Division at the Local Level

100r-----------------------------------~

---
60

60

40
Militanl Hindu vote

AJigam constituency

Manit< Chauk mohalla

1967 1960 1965 1969 1991

Election year

FIG. 11.1. Militant Hindu vote share in Aligarh Legislative Assembly

constituency and Manik Chauk mahalia, 1957-91

of 1962, virtually identical. Third, however, the line for the militant Hindu
vote share is consistently higher in Manik Chauk than that for the constituency
as a whole. Thus, in effect, insofar as the militant Hindu vote is concerned,
the results in Manik Chauk mohalla were exceptional in 1962 only. Thereafter,
the rest of the constituency followed the statistical and political lead ema-
nating from this mohalla, but at a lower level of Hindu consolidation. Fourth,
the militant Hindu vote in this mohalla reached extraordinary heights in the
1989 and 1991 elections, the years in which the great militant Hindu mobi-
lization took place on the issue of the Babri Masjid. The vote for Krishna
Kumar Navman, the BJP candidate in these two elections, reached 88.74 and
93.36 percent, respectively. These results, of course, are exactly what we should
expect from everything that we have learned about this mohalla.
Figure 11.2 compares the Congress vote share in the constituency as a whole
and in Manik Chauk mohalla. The results are partly a mirror image of the
results of the comparison of the militant Hindu vote in the constituency and
the mohalla. That is, the slopes are parallel, but that for the mohalla is for the
most part below that for the constituency as a whole, with one significant
exception, namely, the 1985 election. In the latter election, the Hindu vote
was divided between the BJP and the Congress. Although the BJP candidate
polled 45-19 percent of the vote in this mohalla compared to 42.92 percent
for the Congress candidate, the latter did better in the mohalla in 1985 than
in the constituency as a whole, where he polled only 39.49 percent. There are
two likely reasons for this result. First, the RSS at this time was divided and
many members were supporting the Congress, which, under Rajiv Gandhi's
leadership, was adopting stands similar to that of the BJP on both the issues
 Communal Solidarity and Division at the Local Level / 289

~r-----------------------------------'

50
,
40 1\
\
"/
30 " I
" , /
W '--___
--~-_.J
/
/
,
\
COngress vote
\
\ Atigarh constituency
1:: 10

! 0
19~57~--~I%~2~--~1%~7~--~19~W~--~19~8S~--~19=~~--~1991
\_-
--- Manik Chauk mohalla

Election year

FIG. 11.2. Congress vote share in Aligarh Legislative Assembly

constituency and Manik Chauk mohalla, 1957-91

of national unity and the right of Hindus to worship at the Babri Masjid.
Second, it is quite likely that the Hindus in Manik Chauk as well as in other
parts of the city engaged in a form of strategic voting here, opting for the
Congress candidate as the one most likely to defeat Khwaja Halim, the Muslim
candidate of the Lok Dal.
Finally, a look at Figure 11.3, which gives the results for the militant Hindu
candidates and the Congress in all elections for which I have polling station
data, including the Lok Sabha elections, rounds out the picture of voting in
this mohalla over thirty-five years. The history of voting in the mohalla for
the time series for which I have data divides clearly into three periods. In the
1962 Legislative Assembly contest, the voters of Manik Chauk preferred a mil-
itant Hindu candidate to either the Congress or the Jan Sangh, namely, the
Hindi journalist Madan Lal Hiteshi, who polled 44.25 percent of the vote com-
pared to only 15.86 percent for the Jan Sangh candidate and 21.62 percent for
the Congress. If, therefore, we combine the votes for the two militant Hindu
candidates, as has been done in Figure 11.3, the interval between their com-
bined votes and that for the Congress is 38.49 percent. The gap in the 1962
Lok Sabha election is even greater, the voters of Manik Chauk having given
75.67 percent of their votes to the independent militant Hindu candidate, Shiv
Kumar Shastri, and only 8.60 percent to the Congress candidate.
This phase in the history of voting in Manik Chauk mohalla lasts from
1962 through the 1980 Legislative Assembly election. This is the period
framed by the riots of October 1961 and the beginning of the 1978-80 series
of riots. It is marked by a huge upsurge in voting for militant Hindu candi-
dates and a vast gap between their vote share and that for the Congress. Krishna
 290 I Communal Solidarity and Division at the Local Level

100 . . . - - - - - - - - - - - - - - - - - - - - - - - ,

80

60

40
"" \

------""'"
................ \_-
\
\ Militant Hh>du

O~-~-~----~ -~~~~-~~~~~-~-~
1962 (LA) 1967 (LA)
__
1900 (LA) 1984 (LS» 1989 (LA)
Congress

1962(LS) 1967 (LS) 19l1O (LS) 1985 (LA) 1991 (LA)

Election y••rltype

FIG. 11.3. Vote shares for all militant Hindu candidates and for the Congress,
Manik Chauk mohalia, 1957-91

Kumar Navman entered electoral politics for the first time as an indepen-
dent candidate in the 1980 Lok Sabha election. Although he polled poorly in
the city as a whole, where Hindu votes were divided between the Congress
and the Janata Party candidates, he won a plurality of the vote in Manik Chauk
(36.21 percent). If his votes and that of the Janata Party militant Hindu can-
didate are combined (as in Figure 11.3), then the gap between their combined
votes and those for the Congress is above 33 percent.
In the second phase, which includes only the two elections of 1984 and
1985, the Congress vote share rises above that for the militant Hindu candi-
dates for reasons indicated above. In the third period, the distance between
the vote for the militant Hindu candidates and the Congress is enormous.
We have here in Manik Chauk a distilled and refined picture of the com-
munalization of a segment of the Hindu electorate that has been the advance
guard of militant Hindu nationalism and anti-Muslim sentiment in Aligarh
since Independence. Not since 1962 has there been any significant deviation
in the expression of that sentiment. The anomaly of the 1984 and 1985 elec-
tions itself confirms its persistence, since the militant Hindu voters of Manik
Chauk divided only in their belief that the Congress had come closer to the
RSS view of the Indian nation as a Hindu nation.
But these results raise a methodological issue. Till now, I have argued that
there has been a relationship between riots and interparty competition (on
the interval measure), such that pre-election riots produce an intensification
of party competition. While that remains true for the city as a whole, it does
not apply to each and every mohalla where the other variables previously intro-
duced, namely, communalization and polarization, reveal their importance
 Communal Solidarity and Division at the Local Level / 291
in the overall process. That is to say, communal riots have a dual effect in
towns such as Aligarh where there are sizable populations of both Hindus
and Muslims. In particular localities, where there is often a preponderance
of members of one community, riots intensify communal solidarity (com-
munalization), which translates in the context of the city as a whole into
intensified interparty competition that, at its most intense, also involves a
polari7.ation in which religious and party identities merge. So, there is here
a clear causal sequence as follows:

communal riot > communal solidarity > intensified

interparty competition > communal polarization

In particular localities, the entire sequence will reveal itself only if there
is a communal balance within them, in which case an interval measure of
interparty competition will correspond at the local level to that at the level
above it. From a methodological point of view, the principal point here is
that any kind of electoral analysis that concentrates on a single level to explain
a phenomenon such as the relationship between riots (a societal occurrence)2
and interparty competition (a political event) is flawed. I have demonstrated
elsewhere that in a society as heterogeneous as India's, with multiple layer-
ings of potential identities to kin, caste, clan, language, and religion, the level
of ethnic/communal identification and competition will, in the normal
course, depend upon the size of the political arena. 3
The Legislative Assembly contest in Aligarh, or, for that matter, virtually
all such constituencies in the state ofU.P. and India as a whole, can be framed
in a multiplicity of ways. The most common frame is a combination of inter-
caste and communal competition. There is in fact a predominant tendency
towards a degree of fragmentation that corresponds in each constituency to
the number of castes and communities of a size that provides a sufficient voter
base to affect the result, usually in U.P. somewhere between three and five
candidates from different castes and religions. To build a broader vote base,
parties must either make intercaste or intercommunal deals or they must find
an appeal that transcends the identities of particular castes and baradaris. In
contemporary U.P. politics, there are three such transcendent appeals at work:
to all backward castes, to all lower castes, or to religious community. It is
because of the heterogeneity of castes and baradaris, each with its own par-
ticular interests and sense of separateness, that such broader appeals are so
difficult to construct, so difficult in fact that only very dramatic events such
as riots can provide an effective basis for them.
 292 / Communal Solidarity and Division at the Local Level

Sarai Sultani

Let us now compare voting behavior in the Muslim-dominated mohalla of

Sarai Sultani, which, as we have seen above, has experienced virtually the whole
gamut of rioting in Aligarh since Independence. According to the 1951 cen-
sus, the total population of this mohalla was 1,303, among whom there were
1,154 Muslims (88.56 percent) and 149 Hindus (11.44 percent); it contained
no Scheduled Castes. According to the 1995 voters' list, there were 1,813 vot-
ers, of whom 281 were non-Muslim (15.51 percent). Once again, therefore,
the figures suggest that the 1951 census data remain reliable indicators of
caste/communal population proportions for most inner-city mohallas. They
also suggest that, in the face of the alleged pressures from militant Hindu
groups in the surrounding mohallas to intimidate the Muslim population to
leave, the mohalla has remained predominantly Muslim. The predominant
Muslim baradari in Sarai Sultani is Qureshi. Thus, in all tlle above respects,
Sarai Sultani is the mirror image of Manik Chauk, that is, in its caste/com-
munal population ratios and in the presence in each mohalla of the community
alleged to be the most active during communal riots, Varshneys in Manik
Chauk and Qureshis in Sarai Sultani.
Unfortunately, we cannot trace the electoral history of this mohalla as far
back as for Manik Chauk since, until 1980, the polling stations that encom-
passed Sarai Sultani included several other mohallas, some predominantly
Hindu, others predominantly Muslim; further, the mohallas joined with Sarai
Sultani changed at each election. The results for the 1995 corporation elec-
tions also cannot be broken out for Sarai Sultani alone since it shared ward
no. 35 with another mohalla. It is a matter of interest, however, that the cor-
porator for this ward elected in 1995 was Abdul Rab, uncle of our principal
informant concerning the riotous history of the mohalla. Abdul Rab won this
election as an independent.
It is possible, however, to compare the election results for Sarai Sultani
with Manik Chauk and Aligarh constituency as a whole for six elections
between 1980 and 1991, of which four were Legislative Assembly elections and
two (in 1980 and 1984) were for tlle Lok Sabha. Figure 11.4, showing the mil-
itant Hindu vote shares for these elections, gives as clean and clear a picture
of these contrasts as one might expect from all that we have learned so far.
The slopes of the three lines are virtually identical, but that for Manik Chauk
is high above that for the constituency as a whole, while the one for Sarai
Sultani runs considerably below.
There is much less clarity, however, when we compare the Congress vote
 Communal Solidarity and Division at the Local Level / 293

IOG,------------------------------------,

A1igarh constituency

20 ManikChauk

...I o~----____----~~--~~~~~~~~~
1980 (LA) 1980 (LS) 1.984 (LS) 1985 (LA) 1989 (LA) 1991 (LA)
SwmSw~i

Election yearltype

FIG. 11.4. Militant Hindu party vote shares in Aligarh constituency and Manik
Chauk and Sarai Sultani mohallas, 1980-91

100

80

60
-

40
/~/\ .. \\ A1igarh constituency

'5
20
-- "-
'----'---
ManikChauk

~ 0 ~- - - - - Sarai Sultani
1980 (LA) 1980 (LS) J984(LS) 1985 (LA) 1989 (LA) 1991 (LA)

Election year/type

FIG. 11.5. Congress vote shares in Aligarh constituency and Manik Chauk
and Sarai Sultani mohallas, 1980-91

shares in the three units under examination (Figure 11.5). The slopes all pro-
ceed in tandem for the first two elections, but that for Sarai Sultani crosses
the other two steeply downward in 1985 (LA), steeply upward again in 1989
(LA), and close at near zero in 1991 (LA).
We have already explained the deviations from the mean for Manik
Chauk; those for Sarai Sultani occurred partly for the same, partly for other
reasons that are illustrated in the next figure. Figure 11.6 gives the party vote
shares for the three principal political tendencies in Aligarh for Sarai Sultani
alone. In the 1984 and 1985 elections, as we have seen, many militant Hindus
in Manik Chauk found the Congress a party more suitable to their proclivi-
ties than even the BJP. However, the voters of Sarai Sultani remained loyal to
 2941 Communal Solidarity and Division at the Local Level

.@r--------------------------------------,
Md. Su.yon

Elect;"" yearltype

FIG. 11.6. Pa.rty vote shares in Sarai Sultani mohalla, 1980-91

the Congress through the 1984 Lok Sabha election, but then most deserted
the Congress in favor of the Janata Party candidate, Khwaja Halim, in the
1985 Legislative Assembly election. In 1989, when the militant Hindu honey-
moon with the Congress abruptly terminated, the voters of Sarai Sultani
remained divided between Congress and the Janata Dal, but more this time
preferred the Congress. Finally, in 1991, as the Congress vote in this mohalla
plummeted to near zero, that for the Janata Dal candidate rose to 88.26 per-
cent of the total.
Further factors influencing the voting in Sarai Sultan i-aside from the
disaffection with the Congress and the move towards the Janata parties expe-
rienced by Muslims in most parts of the state as well as in Aligarh during the
Ram Janmabhoomi movement-were the community and baradari of the
candidates of the Congress and the Janata parties. In the 1980 Legislative
Assembly election, won by the Janata party (SC) candidate, Khwaja Halim,
Sarai Sultani voters preferred another Muslim candidate running on the
Congress ticket, Ahmad Loot Khan, who polled 55.80 percent of the vote in
the mohalla. In the two 10k Sabha elections of 1980 and 1984, mohalla vot-
ers remained loyal to the Congress despite the absence of Muslim candidates.
The Congress vote shares in Sarai Sultani were very high in these two elec-
tions: 77.54 percent for Ghanshyam Singh in 1980 and 73.63 percent in 1984
for U sha Rani. In 1985, however, mohaUa voters deserted the Congress Hindu
candidate in favor of Khwaja Halim, the Lok Dal candidate. In 1989, with the
option to choose from two Muslim candidates, Khwaja Halim again on the
Janata Dal ticket and Mohammad Furkan, the Congress nominee, Sarro Sultani
voters divided between the two, but preferred the Congress Muslim candi-
date. Finally, in 1991, although there were two other Muslim candidates, one
contesting on the SIP (SP), the other on the BSP ticket, mohalla voters con-
 Communal Solidarity and Division at the Local Level / 295

centrated their votes overwhelmingly behind Mohammad Sufiyan, a man of

their community and also a man of the predominant Qureshi baradan.
It should be noted, therefore, that, though there is a general tendency
throughout the period from 1980 to 1991 for the Muslim voters of Sarai Sultani
to prefer-other things being equal-a Muslim candidate, they are less com-
munalized than the Hindus of Manik Chauk. The latter have divided only
when two Hindu candidates have been seen as ideologically militant Hindu
or when the Congress itself came near the BJP ideology of nationhood. While
Sarai Sultani voters, too, have divided when offered a choice between two
strong Muslim candidates, their voting has also been influenced by party ties:
initially to the Congress, then to the Ianata parties. At the same time, it must
also be recognized that the party loyalties of the Muslim voters of Sarai Sultani
also reflect their communal interests to the extent-certainly considerable-
that they perceive particular parties (initially the Congress, then the Janata
parties) as the secular parties, seen at different times as the best protectors of
their rights, safety, and security, and the parties most able to defeat the BIP.
So, while we see the tendencies in both mohallas, as in the city as a whole,
towards communalization and, at times, polarization, there are evident
countervailing tendencies as well. Amongst militant Hindus, however, the
countervailing tendencies fall within a consistent ideological line of Hindu
nationalism, while amongst Muslims, the countervailing tendencies are of
party, community, and bamdan. It is the pressure of militant Hindu mobi-
lization and hostility to Muslims that feeds the tendency towards communal
polarization. When that pressure recedes, other kinds of divisions appear and
become politically important to Muslim voters and to Hindus as well, who
divide according to caste as well as party.
 12 / The Decline of Communal Violence
and the Transformation of Electoral Competition

F or a decade after the destruction of the.Babri Masjid in December 1992

and the negative reaction to it in the country and around the world,
there was a decline in communal violence in India as a whole and in
the state of U.P. and Aligarh as well. This decline was in no small measure
owing to the deliberate decision of the RSS and BJP leadership to move toward
an accommodating stance in its drive to gain, maintain, and consolidate its
power in particular states and in the country and the necessity for it to make
alliances with noncommunal parties to do so. That strategy has, however, pro-
duced mixed results. Although the BJP was in power in 2000 at the Center
and in several Indian states, it suffered a major defeat in the U.P. Lok Sabha
elections of 1999. In Aligarh City, the BIP has been displaced in both the Leg-
islative Assembly constituency and in the Aligarh segment of the parliamentary
constituency. In this chapter, I will demonstrate its decline in the Legislative
Assembly constituency and the reasons for it.

THE LEGISLATIVE ASSEMBLY ELECTION OF 1996

In the past decade, a major demographic shift has been occurring in Aligarh
City, the nightmare of all Hindu communalists, namely, a rise in the Muslim
population of the city. Demographic shifts in the population of Aligarh City
have combined with alterations in the delimitation of the boundaries of the
constituency to affect significantly the communal composition of the voting
population and the results of the last elections. These changes have involved
the exclusion from the constituency of Hindu-majority areas and the alleged
illegal voting of Muslims from another constituency in M uslim-majority areas.
Altogether, 16 mohallas from the old ward of Achal Talab, with a popula-
tion of 10,792 in 1951, which can be estimated to have reached 36,584 by 1991,1
 Decline of Violence and Transformation of Competition /297

were transferred out of Aligarh constituency after 1969 (after 1980 in the
case of Madar Darwaza). These numbers constitute 70.43 percent of the total
population of this predominantly Hindu ward, the ward with the highest
proportion of Hindus (80.81 percent) in 1951. The proportion of Hindus in
the 16 transferred mohallas was 88.02 percent (estimated total population
of 27,782).
In the meantime, as noted earlier, there has been a very considerable
increase in the Muslim population in outlying areas of the city, including many
former villages and new mohallas that have grown in part as a consequence
of migration from the rural hinterland. Table 6.3, discussed earlier, shows the
voting population of those areas for which I have data from the latest voters'
lists, categorized as Muslim or non-Muslim (Hindus, others, and Scheduled
Castes), and as to whether or not they are included in the 198911991 delimi-
nation of the Legislative Assembly constituency, the latest delimination in
my possession. Although some of the newly incorporated areas are pre-
dominantly Hindu, they are not included in the Legislative Assembly con-
stituency, whereas many of the large predominantly Muslim mohallas are
included. The proportion of Muslims in these included mohallas ranges from
63-41 to 100 percent. The total number of Muslim voters listed on the 1984
and 1995 voters' lists combined for those areas is 27,290, aU added to the new
Aligarh City Legislative Assembly boundaries.
It is, therefore, the case that the population balance in the Aligarh con-
stituency has shifted in favor of the Muslims in consequence of successive
deliminations of constituency boundaries and a substantial increase in the
Muslim population in the outlying areas. It is evident also from the polling
station-wise voting in the transferred mohallas discussed in the preceding
chapters that the BJP has suffered in consequence of these shifts. Further
evidence of the loss to the BIP comes from the results of the corporator elec-
tions. Seven of the eleven transferred Hindu-majority mohallas fall in wards
15 and 17 in the 1995 delimination. z The BIP won the 1995 elections in both
these wards, polling 47.29 percent of the vote in ward 15 and 39.83 percent
in ward 17. This and other differences between the boundaries of the assem-
bly constituency and the municipality also explain why the BIP remains dom-
inant in municipal politics, but has lost its dominance in the legislative
constituency.
Navrnan had noted these trends with some concern in an interview in
1993, when he remarked to me that "one difficulty he faced in the 1991 elec-
tion was that Muslims had been migrating to the city from the rural areas
from time to time and that, as a consequence, the number of Muslim vot-
 298/ Decline of Violence and Transformation of Competition

ers had increased." When I visited Aligarh in 1997, it was estimated that
Muslims then comprised a majority of the voters in the city constituency,
perhaps 60 percent of the total. As a consequence, by 1996, the Muslims could
afford even to divide their votes to some extent, as they probably did in that
election, and still provide the bulk of the support to the winning Muslim
candidate on the ticket of the SP, defeating Navrnan by a wide majority (see
Table 12.1). It was also reported that Navman had lost the consolidated sup-
port of the Hindu community as a consequence of his failure to take care
of his constituency.
The man who succeeded in unseating Navrnan was Abdul Khaliq. He had
joined the BSP in 1993 and, in his first attempt at the Aligarh assembly seat,
nearly won it in a close contest against Navman. Abdul Khaliq's decision to
fight on the ticket of the BSP had the effect of resurrecting the old alliance
between the Muslims and the Scheduled Castes that unseated the Congress
in the 1962 election, for Abdul Khaliq had the support of most Muslim vot-
ers, while the BSP was able to bring behind him the votes of most of the
Scheduled Castes. He himself articulated his decision to join the BSP in those
terms. When asked why he joined the BSP at that time, he said that he felt
that "Scheduled Castes and Muslim votes will win the seat."
By 1996, however, he thought better of it, "joined Mulayam Singh Yadav,"
contested the seat on the SP ticket, and defeated N avrnan by a margin of more
than 35,000 votes. Asked what the main issue in this campaign was, Abdul
Khaliq replied that the "main issue was that Aligarh must be withdrawn from
BJP." He described Navrnan as "a very communal person," which he noted
that everyone knew "very welL" He also attributed his victory in part to the
alleged fact that "Navman did not do any work in the city."
In the 1996 election, the BSP again put up a Muslim candidate, pursuing
the same strategy as in 1993. Their candidate this time was Kaisar Hayat, pro-
fessor of law at the AMU. However, the BSP vote dropped by nearly 36,000
votes from the previous election. In 1993, the SP left this seat to the Janata
Dal. Muslims preferred the BSP to the Janata Dal. Although the Janata Dal
candidate lost his deposit, he won enough Muslim votes to deprive the BSP
of its victory. In 1996, Kaisar Hayat won even more votes than the Janata Dal
candidate in 1993, more than 20,000 votes, enough to retain his security
deposit, but not enough to prevent the victory of Abdul Khaliq and the SP.
When I asked Abdul Khaliq if the Muslim vote had not been split in 1996,
he said that it had not, that Kaisar Hayat received Muslim votes only from
his own baradan, that is, the Qureshi community. Otherwise, Kaisar Hayat's
votes came primarily from the Jatavs, who voted not for him as a person, but
 Decline of Violence and Transformation of Competition /299
TABLB 12.1. Election results for Aligarh City Legislative
Assembly constituency, 1996

Percent of
Turnout Castel Votes Total Valid
No./%a Candidates Community Party Polled Votes

163,4071 Abdul Khaliq Muslim SP 86,570 53.10

52.65 (Saifi)
Krishna Kumar Varshney BJP 51,427 31.54
Navrnan
Kaisar Hayat Muslim BSP 20,294 12.45
(Qureshi)
Others (11) 4,756 2.91

'Valid votes turnout. Total vote turnout was 53.24 percent.

SOURCE: India, Election Commission, Statistical Report on General Elections, 1996, to the
Legislative Assembly of Uttar PTildesh (New Delhi: Election Commission of India, 1997), p. 561.

for the BSP as a party. He acknowledged that he had received a majority of

the votes of the Qureshis and the Jatavs, but claimed he had votes from all
castes and communities. As for the victory of Abdul Khaliq, Kaisar Hayat cited
two factors in particular. First, he pointed to the fact that the Muslim vote
had increased. Previously, the Muslim vote in the constituency had been
between 46 and 48 percent, but now Muslims were in a majority. Second, as
Abdul Khaliq himself had noted, "Hindus were not happy with the person
who was elected thrice from the constituency, Navman."3 Thus, in effect,
Kaisar Hayat acknowledged that Abdul Khaliq won because he had garnered
the votes of most Muslims in a Muslim-majority constituency while Navman
had lost his previously strong support among Hindus.
Mr. Navman gave several reasons for his defeat, virtually all related to the
faults of others. The most substantial reason he gave, however, was that the
voters' list had been manipulated in such a way that some 20-25,000 people,
Muslims from the Jamalpur and Hamdardnagar areas of the city, which are
actually part of another constituency, voted in Aligarh instead, thus greatly
increasing the Muslim vote against him. (Let us recall here again that these
two localities were targeted for violent attacks for the first time during the
1990-91 riots.) Finally, he said that, in a constituency where "the voting pat-
tern is based on religion," the Hindus had gotten too "confident about their
victory" and had not come out to vote in large enough numbers while the
 300 / Decline of Violence and Transformation of Competition
"Muslims got united on this basis that we have not been able to elect any-
body since a long time, so let us unite." He went so far as to claim that even
the "BSP candidate could not get Muslim votes," contrary to the statements
of both Abdul Khaliq and Kaisar Hayat. Rather, Navman claimed that Kaisar
Hayat got only the votes "from the BSP community," that is, the Scheduled
Castes. He did not even get the votes of his own Qureshi community. The
Muslims as a whole, he insisted, had decided not to divide their votes in order
to ensure the victory of a Muslim candidate.
When I remarked to Mr. Navman that I had heard that the upper-caste
Hindus did not vote for him this time, he acknowledged that this was true.
However, he did not attribute it to his own failings, but to their "laziness."4
They had become too confident of their victory, that is, of his victory as the
representative of the Hindu community, and did not turn out to vote.
"Otherwise,» he said, "if they would have gone [to vote], then they had no
option except to cast [their votes] for him."5 In other words, if the Hindus
wanted to defeat the Muslim candidates and retain the seat for the Hindu
community, they would have had to vote for him.
In fact, there was indeed dissatisfaction with Navman among Hindu vot-
ers, as well as oppositon to his candidacy from within the BJP and the RSS.
An older RSS worker had hoped to get the BJP ticket for this election, which
had been awarded to him by the nominating committee in Lucknow, but then
his name was dropped and that of Navman restored in the final selection by
the party leadership in New Delhi on the grounds that he had won the seat
in 1993 and should therefore be allowed to contest again in 1996.6 The disap-
pointed RSS leader attributed N avman 's defeat to three factors. First, the Hindu
community was dissatisfied with him, "with his character,» including the BJP
workers, who did not support him this time. So the Hindus did not turn out
to vote in sufficient numbers. At the same time, the second reason was that
the "Muslims were united." Third was the increased number of Muslims in
the constituency. In short, the election contest was polarized between Hindus
and Muslims, but allegedly because of Navman's bad character and his fail-
ures as a representative of the people, and because of the increased number
of Muslim voters, the BIP could not retain this seat. Despite the latter factor,
however, it is evident from many of the comments by the candidates them-
selves, including, especially, Navman's remarks about the "laziness" of the Hindu
voters in this election and other evidences of dissatisfaction from within the
ranks of the BJP and the RSS, that the intensity of interparty competition was
thereby affected. Such "laziness" and disunity within the militant Hindu ranks
does not occur when Hindus are being mobilized in the aftermath of riots.
 Decline of Violence and Transformation of Competition /301

The decline of communal violence since 1993 in Aligarh has been followed
as well by a decline in the intensity of interparty competition by the interval
measure used in Figure 8-4. The interval between the winning SP candidate's
vote share in the 1996 election and the runner-up, Navrnan, was 21.60 per-
cent, the fifth largest in the 20 elections for which I have interval data?
It is evident in Aligarh, as elsewhere in the state, that the BJP has failed to
attain the dominant position for which it has striven. It is also clear that, in
the absence of powerful mobilizing issues that unite the upper-caste Hindu
population and a major portion of the backward-caste population as well
against the Muslims, the BJP cannot achieve a dominant position either in
Aligarh or in the state as a whole, for that matter. Moreover, the demographic
shift in the population of the Aligarh legislative constituency makes it impos-
sible for the BJP to win an election there unless there is a significant division
between the SP and the BSP and deviation in the voting behavior of Muslims
and the Scheduled Castes. That being the case, there is no longer any incen-
tive or use to the BJP and its allied organizations in the RSS family to foment
riots in the city. That is even more the case since, in the municipality, where
Hindus are in a strong majority, the BJP is the dominant party. Thus, with
no need to make use of riots to retain its hold in municipal politics and no
use in its doing so any longer in the city, it is likely that communal riots will
not erupt on the scale and with the frequency as in the past. That does not
mean, however, that communal tension and violence will disappear from
Aligarh City, for it is endemic in certain parts of the town, as I have shown
in previous chapters, and because it remains in the interest of the BJP, in order
to retain its hold over the Hindu population, to maintain the institutional-
ized riot system that the RSS family of organizations has built.
We have here, therefore, discovered one of the principal precipitants of
large-scale riots, namely, the political manipulation of local conflicts and their
transformation into Hindu-Muslim confrontations whose function is to solid-
ify communal identities, communalize the electorate, and polarize the elec-
toral contest in order to achieve victory no matter the cost in human lives.
Absent a political advantage for such political manipulation, the risk of large-
scale riots is reduced. Other factors independent of party politics remain active,
however, and other interests and sentiments continue to keep alive the
potential for communal violence. Moreover, the history of Aligarh, of V.P.,
and of much of the rest of the country as well does not inspire confidence
that there will be no occasion in future for the resuscitation of movements
of mass mobilization, designed to mobilize Hindus, intimidate Muslims, and
move India further in the direction of a militant Hindu national state.
 PART V

The Process
of Blame Displacement

A scathing denunciation of the recent happenings in Bombay and

the conduct of those who had remained passive witnesses of the same
constituted the core of Gandhiji's address after the evening prayer
gathering at Rungta House today.
The news of these events, said Gandhiji, had filled him with shame
and humiliation and he hoped that they too must have felt likewise.
He hoped that none of those who attended the prayer had taken part
in these disgraceful happenings. But that alone would not entitle them
to congratulation. They had reached a stage when no one could afford
to sit on the fence or take refuge in the ambiguous middle.
One had to speak out and stand up. Inaction at a time of conflagra-
tion was inexcusable. That might appear to them to be a difficult ideal
to follow but that was the only course that would take them through
the present difficult times.
It had become the fashion these days, continued Gandhiji. to
ascribe all such ugly manifestations to the activities of hooligans.
He deprecated the habit of taking refuge in that kind of moral alibi.
Who were the hooligans after all? They were their own countrymen
and so long as any countryman of theirs indulged in such acts they
could not disown responsibility for it consistently with their claim of
being one people and one nation. It mattered little whether those who
were responsible for the happenings were denounced as goondas or
praised as patriots. Praise and blame equally belonged to them. The
only manly and becoming course for those who were aspiring to be
free was to accept either.-Free Press Journal, March 12, 1946, report-
ing on Gandhi's reaction to the Bombay riots that accompanied the
Royal Indian Navy employees strike.

303
 13 I Riot Interpretation, Blame Displacement,
and the Communal Discourse

THE INTERPRET ATION PHASE OF RIOT PRODUCTION

T he preceding parts of this book have focused on the first two stages
in riot production: preparation/rehearsal and enactment. It has
been demonstrated that preparation and rehearsal for the enactment
of large-scale riots are ongoing activities in which known persons and groups
are actively engaged and in which there is a specialized division of labor embed-
ded within an institutionalized riot system. All the principal explanations for
the outbreak of riots have been examined. It is not claimed here that there is
a single causal explanation that will encompass the enactment of all com-
munal riots. On the contrary, it has been shown that there are a multiplicity
of factors that may precipitate riots and that there are a variety of factors and
forces that come into play when the opportunity for producing a riot occurs.
However, it is claimed that large-scale Hindu-Muslim riots are primarily polit-
ical productions in which the precipitating incidents are pretexts and the enact-
ment is in large part organized. Further, it has been shown that the intensity
and scale of riot production in Aligarh has increased dramatically during the
past half century, exceeding anything that occurred in the years before and
during the partition of the subcontinent.
The fact that there remains a considerable diversity of motives and fac-
tors for participation in and production of communal riots leaves ample
room for contestation in the third phase of riot production, that of post hoc
explanation. In this part, therefore, we move once again from the "fact
finding" realm of hard data to the nebulous realm of interpretation "after
the fact." Once again, therefore, we review all the factors said to precipi-
tate and produce riots by those who produce, participate, suppress, control,
and seek to explain them. However, my approach to these "explanations"

305
 306/ Interpretation, Blame Displacement, and Communal Discourse

will be not quantitative nor merely analytical, but critical in two senses.
Having established a factual basis for understanding the dynamics of riot
production, including the roles played in them by specific persons and organi-
zations, we are in a position to distinguish fact from fancy, justification from
explanation, hypocrisy from truthfulness, and pathological fantasy from plau-
sible reality. Second, we are in a position, now, also to appreciate the signal
importance in a riot's production of the struggle for control of its meaning
and interpretation in the aftermath. Like any dramatic production, the grisly
play of communal rioting depends upon the ability of its perpetrators, abet-
tors, and observers to provide an interpretation that justifies repeat per-
formances. But there is a difference between the role of the critics upon whose
judgment the success or failure of theatrical productions depends and the
role of the interpreters of riots. In the former case, the play stands or falls
on a favorable balance of praise over blame. In the case of riot production,
the balance that ensures repeat performance is reversed: wide dispersal of
blame masks responsibility and diverts the public gaze from the mechanisms
that produce riots.

TYPES OF EXPLANATIONS OF RIOT PRODUCTION

In my research extending over thirty-eight years in Aligarh, I have come across

five or six distinct types of explanations for communal riots in response to
my own questions concerning how and why such events in general as well as
particular riots have occurred. Some of these explanations can be fit into
broader discursive formations in the sense that they reflect fundamental under-
standings of human nature or of political society or of the relations between
peoples as much as they do particular understandings of the events discussed.
There is, for example, a discourse of profit that operates in Indian society, as
elsewhere, that accounts for most human events in terms of the economic
or otherwise self-interested calculations of individual actors. It is present in
Aligarh as well, where several general and particular explanations of riots fit
into this kind of context.l have summarized and analy:.".ed those types of expla-
nations in Chapter 7.
One primary function of such economic explanations is to displace blame
from the authorities, politicians and political parties, the police, the general
public, and the poor and disadvantaged onto the privileged and dominant
classes. Sometimes, however, riots are seen to arise out of "understandable"
grievances of particular classes or groups against others. Thus, the most com-
 Interpretation, Blame Displacement, and Communal Discourse /307
mon explanation I heard in 1961-62 for the October 1961 riots was the griev-
ances of the faculty and students of the local degree colleges against the priv-
ileges and status of their counterparts at AMU. There is a third function served
by economic explanations as well, namely, to displace blame onto the basic
greed of human beings who are considered capable of any kind of evil action
for the sake of profit. Thus, the bizarre explanation of the rumors concern-
ing the killings at the AMU Medical College Hospital that fed the riots of
December 1990-January 1991, that they were spread by private medical clin-
ics to discredit the AMU Hospital and take business away from it that would
then come to them.
In this and the following chapter, I will first consider the opportunities
provided by riots for blame displacement, then discuss several other types of
explanations-of which the most important are those that fit within the dis-
course of Hindu-Muslim communalism-that recur in my interviews and
which have not been analyzed fully above. Let us consider, first, the oppor-
tunities that riots provide for blame displacement, brought out very clearly
in an interview cited earlier concerning the 1978 riots, in which our respon-
dent enumerated a multiplicity of interpretations that were given at that time
to explain the occurrence of the riots. Although he was actively involved in
monitoring the riot and in the relief activities afterwards, this respondent did
not make dear precisely which of the various interpretations he personally
considered most accurate. Moreover, although he is a Muslim, he did not
automatically accept the argument that it was the Jan Sangh and the RSS that
were mostly responsible for the riots. On the contrary, despite tlle fact that
he was an admirer of Indira Gandhi in several respects, he argued based on
the consequences, that the Congress (I) deserved the greater blame, reason-
ing as follows.

I'RB: But, you know, t.o go back to the '78 [riots], you said there were three interpre-
tations. One is that it just happened because of this, by accident, by-because of
this Shure Lal business.
JH: But, some person[s] was saying like that.
I'RB: And another interpretation is the-Navman and his people. And the third is
economic.
JH: Yes.
I'RB: What is your own view?
JH: I could not say, but after the riots, who were the beneficiary? The Congress (I).
My opinion is that.1
 308/ Interpretation, Blame Displacement, and Communal Discourse

But how did the Congress benefit from these riots? It is clear that both the
Congress at the local and national levels and Mrs. Gandhi personally did
benefit from them, judging by the ways in which they used both the 1978 riots
in Aligarh and several other riots and police-public confrontations to dis-
credit the Janata government that came to power after the 1977 elections. The
factions within the Janata Party at that time also indirectly helped Mrs.
Gandhi, by making an issue of the presence of RSS members from the for-
mer Jan Sangh in the Janata government, and by themselves attributing blame
for the riots to the local Jan Sangh/RSS cadre in Aligarh, particularly Mr.
Navman. The demand was made for the suspension of Navman from mem-
bership in the party, which was done a month after the riots. In the mean-
time, however, Mrs. Gandhi integrated the Aligarh riots into her portfolio
of charges against the government to the effect that Muslims and other dis-
advantaged groups in Indian society were being slaughtered under Janata
rule-thereby emerging as their "protector" and making possible the return
to the Congress in the 1980 elections of large numbers of Muslim voters who
had deserted that party and Mrs. Gandhi in 1977. The 1978 riots were, there-
fore, taken out of their local context and merged into a national context, while
a predominant section of the Janata government and the Congress organi-
zation as a whole accepted the explanation of the riots that most suited their
political purposes. 2
By so placing the riots of 1978 in a broad national context, blame is assigned,
a particular explanation is accepted, and a particular party and leadership
benefit from that resolution. One explanation is distilled from among all the
various factors that contribute to a riot, the one that is most useful politically
to the temporarily ascendant political party and political leader. All the other
contributing factors can then be ignored. But the partial explanation does
not satisfy, as my inconclusive interview on this point demonstrated when I
pressed this respondent further to elaborate his implication that the Congress
bore responsibility for the events in Aligarh in 1978.

P RB: But when you say, uh, Congress (I) benefited, implying that they must have had
something to do with it ...
JH: I'm not blaming, but I'm saying this, who was the beneficiary?
PRB: But you know, some people do say that, uh, there are Muslim Congress politi-
cians in Aligarh who are as mischievous as Navman. I'm not sure it's true, you
know, they mention nowadays Khwaja Halim.
JH: I don't, I cannot name any congressman of Hindus or Muslims, but I-who was
beneficiary?
 Interpretation, Blame Displacement, and Communal Discourse /309

PRB: Yah, well that's a-that's-sometimes you can find the cause from the effect.
From the consequences, yah.
JH: This much I can say.

But it is not clear that, as I suggested in the interview, one can find the cause
from the effect. What is clear is that what is left from the process of distilla-
tion of factors into a satisfying and useful explanation is that all the factors,
including the one identified as central, continue to operate. Those that are
neglected continue to operate because they have not been brought more clearly
into focus, whereas the political explanation that has come to the forefront
is treated as just that-a political explanation that requires a political response.
"No," say the RSS, the Jan Sangh, and later the BJP. "It is a lie that we are
responsible for these riots. It is the Congress that is responsible. Moreover,
whenever we have been in power, there have been no riots."
So the process of blame displacement continues without end, at the local
and national levels. This respondent illustrated how the process worked in
Aligarh in the immediate aftermath of the riots, at a meeting on the AMU
campus called by the district magistrate, at which, the respondent claimed,
university leaders and Hindu leaders from the town, including the president
and professors of Barahseni College, were present. Asked to speak, he focused
particularly on this very issue of blame displacement and the consequences
as he saw them.

Everybody's alleging everybody, another person [is at fault]. Either the alle-
gations are correct-if the allegations are correct, therefore, every sinner is in
this meeting-or the allegations are baseless. If it is so, then it is immoral for
us that, at the time of crisis, we are blaming each other. ... But really the dead
bodies are there, wounded persons are there, and we can apply our humanity
to solve this. My speech worked and really the atmosphere cooled down.

But then the solution adopted to cope with the dead bodies, this respondent
himself acknowledged, is itself inhuman, as he later revealed.

JH: There are some things which ... are the universal values and, uh, these are the
human values. Now they are [assessing] the misery, now they count that Hindus,
twenty, Muslims, thirty [have been killedl.lt is such [an) inhuman act, really, even
the dead bodies.
PRB: I know, this business of counting.
JH: Now, you see the misery of the conditions, whether Hindu widow or Muslim
 310 / Interpretation, Blame Displacement, and Communal Discourse

widow.... Really, I-this Aligarh communal riot has changed my whole ...
[ ••• J When I saw the misery of the people, the feeling of the people, the tragedy.
[ ••• J I cannot ignore that woman ... , I was ... when I was-because it is hor-
riJYing memory for me that we were in town hall ... the district officials were
there and the university authorities were there, we were providing 20,000 rupees,
check oho,ooo rupees, check to the widows and the persons who have ... died ....
I cannot say to you that-what was my emotional condition and everybody's
emotional condition, that one Hindu widow came with [her] ... father-in-law,
a young lady came, now the tears in the eyes of Hindu father-in-law and the
Hindu widow, that was a Hindu face. After that, we called a Muslim lady, she
came with ... her father. She was also young. I could not remember exactly, but
both are having one child ... with them, ... on their laps. Now what I suggest
is the same misery, same agony and pain in the eyes of-and we provided chec.ks
[to] both of them. The feeling was the same and no power on earth can translate
a cry of human being in a misery. You cannot translate it.

In place of the impossibility of translation, the failure oflanguage, the polit-

ical process provides financial compensation and contextualization. The first
itself requires no act of speech. Although I have never witnessed these finan-
cial transactions, I have two comments to make about them. One is that it is
obvious that any act of speech-beyond those required for bureaucratic
accounting purposes-to a bereaved person from a government official hand-
ing over a check, who does not know the recipients in anything but his official
capacity, would be superfluous and meaningless. Another is that the whole
process of counting the bodies, identifying them by religion, assigning right-
ful payments to the bereaved, and paying them, is clearly dehumanizing.
For many years, at the end of each major communal riot, local officials
determined a count of the dead by methods that are not clear to me, and the
accuracy of whose results are always contested. The requisite compensation
payments to the bereaved were then made. Such counts were then cumulated
nationwide and ultimately published in a report or reports by the govern-
ment of India, usually in tables of statistics as well, in which were noted that
in such and such a riot and in all cumulated riots for the period under review
so many Hindus were killed and so many Muslims. This kind of compensa-
tion and cumulation of data is dehumanizing not because it is in itself improper
but because of what is not and cannot be said and what is not done. The cry
of the bereaved cannot be recorded and published in a government docu-
ment, but surely the government could publish the name, the occupation,
and a brief biography of every single person killed in a riot.3 The government
 Interpretation, Blame Displacement, and Communal Discourse /311

could also report the circumstances of each killing and the measures taken
to discover and prosecute the killers. By individualization of the statistics,
people would at least come to know or be able to make some judgment con-
cerning whether those persons killed were themselves rioters, or innocent
people minding their own business, who suffered from the acts of violent mobs
directed against them for reasons known or unknown. But the government,
of course, cannot publish the circumstances of each killing and the measures
taken against the killers because, in the majority of cases, the killers are the
police, the agents of government, and in the overwhelming majority of cases
absolutely nothing is done to apprehend and prosecute them. It is most prob-
ably for these reasons that the government of India has not published such
statistics for the past eighteen years.
There is yet a further act of dehumanization that takes place in the after-
math of riots when the time for compensation to the bereaved arrives. This
is the charge that many of the bereaved are bogus claimants. Their sons and
husbands, officials say, have absconded so that the parents and wives can claim
the compensation for themselves. It is more than likely true that such things
happen in the aftermath of riots. But, once again, the blame is laid upon the
people, not on governments under whose rule such events occur and whose
leaders monetize the loss of life that occurs in them.
So, for the untranslatable cry of the bereaved, we have monetary com-
pensation, but also contextualization, the removal of the cries of the bereaved
from the reality of the tragedy of human lives lost, and of the sufferings of
the bereaved to the political realm of explanation, where not only language,
but rhetoric and symbols, fly and flourish. The factors that caused the riots
are enumerated and assessed, charges are made against individuals and
groups, some administrators are transferred, one or two policemen who mis-
behaved are temporarily sent to the lines, while all sides ready themselves for
the next events.

THE COMMUNAL DISCOURSE

By far the most common context into which explanations for riots are placed
by local and extra-local observers is the communal discourse, which in turn
has several variations. The most extreme form I encountered in my visits to
Aligarh over the years came from persons in the rump Jan Sangh as well as
the BIP, all of whom also have RSS backgrounds. Several of these respon-
dents characterized riots as a form of Muslim jihad. Such explanations
included generalizations about the nature of Islam and the so-called Semitic
 3121 Interpretation, Blame Displacement, and Communal Discourse

religions in general, in comparison with Hinduism, as well as specific state-

ments concerning the organization of riots by Muslims in the local context
of Aligarh and other places in India.
I will summarize here one example of the militant Hindu explanation,
which may be described as a kind of essentialism applied to Muslims and Islam.

Muslims are aggressive when they are dominant. Aggressiveness is built into
Semitic religions, in contrast to Vedic, which believe in coexistence. Communal
riots came to India only with the Muslims. Jews are small in number and
Christians have become civilized, but the Muslims remain backward and bar-
barous. Moreover, their aggressiveness is built into their beliefs, into the Koran
itself. Contrast this with the passivity of Hindus who, despite provocations such
as the construction of mosques at or near Hindu places of worship in Mathura,
Ayodhya, and Varanasi-still the Hindus do not cause communal riots even
though the RSS is equally dominant in these three places. Muslims are so aggres-
sive that they will even try to kill innocent Hindus who go into their mohallas
for innocent purposes, such as a person who went to read an electricity meter
or government house inspectors. Hindus need to learn from Muslims, espe-
cially concerning how they treat their minorities in Islamic countries, where
minorities must live according to the wishes of the Islamic state. 4

A second example of this type of explanation comes from a respondent

who applied it to the specific context of Aligarh. This man remarked that,
though Aligarh was not dominated by Muslims, it contained a substantial
percentage. which he put accurately at 37 percent. The real root of the com-
munal riots in Aligarh, he averred, was the Aligarh Muslim University. Riots,
he said, were organized there and innocent Muslims of the town were
brought into the fray afterwards. They are organized every five or ten years.
Hindus lose property in these riots.
I asked this respondent specifically about the 1978 riot. He responded that
it was planned and perpetrated by Muslims. Further, he remarked that the
district administration-despite the fact that the Janata party, with its for-
mer Jan Sangh elements, was in power in the state-had supported the
Muslims by lodging false reports against Hindus and falsely arresting them.
He referred specifically to Mr. Krishna Kumar Navman.
More generally, he blamed Mrs. Indira Gandhi-who had been prime min-
ister of the country throughout most of the years between 1966 and 1984-
for communal riots on the grounds that her policies of "appeasement"
towards the Muslims had emboldened them.
 Interpretation, Blame Displacement, and Communal Discourse /313

This respondent then went on to make stark essentialist comparisons

between Hindus and Muslims, summarized below.

Hindus, in contrast to Muslims, are meek and peace-loving, respect all reli-
gions, creeds, castes. Hindus retaliate only in extremity. However, Hindus never
agitate, commit murder, loot, or arson. Riots are invariably started by Muslims.
Nowadays, [riots1are started to preserve the minority character of AMU. Soon,
they [Muslims] will demand the division of India. Before Independence,
[riots1were started in order to get Pakistan. Already, there is a demand for reser-
vation of places in government service for Muslims despite the fact that
Muslims have all facilities. 5

A third example of the articulation of the communal discourse in explain-

ing Hindu-Muslim riots in Aligarh comes from the HIP's former three-term
MLA, responding to questions concerning the causation of Hindu-Muslim
riots in Aligarh. When I asked who killed whom in these riots, his reply did
not support the view that mostly Muslims are killed. He said, "The first few
killed are Hindus; after that, it's equal." When I asked if the wretched con-
ditions in the mohallas in which riots occurred had anything to do with Hindu-
Muslim riots, he replied that the "surroundings have nothing to do with it.
It is a question of mentality." As for the police, considered by many people
to be partial to Hindus and hostile to Muslims during riots, especially in west-
ern U.P., he responded to my question concerning tlleir role that they were
"mostly impartial," though he averred tllat "sometimes they overreact and
people are forced to make complaints." It is clear from the context of this
statement that this man meant that sometimes the police overreact and harm
innocent Hindus. I believe also that he had himself in mind as someone against
whom the police sometimes "overreacted."
This man's solution to the problem of Hindu-Muslim riots is also con-
sistent with his explanation of their origins: "Muslims must change their
mentality.» He said that he had done his best "to claim them as brothers,
but their interest [in such conciliation] is only superficial." He claims that
he "even fought alongside Muslims in 1973 when the administration wanted
to forbid the tanzim procession" during the annual Muharram holiday.
Further, he noted that he had been chairman of a Save Urdu Committee
in 1971.
Thus, it is clear in this man's mind that not only do Muslims instigate
riots, but no sincere effort on the part of a well-intentioned person such as
himself to cooperate with Muslims even in their struggle, and to work with
 314/ Interpretation, Blame Displacement, and Communal Discourse

them to maintain communal peace and provide relief to riot victims, is worth-
while. Muslims, at least those who are educated and literate, are virtually incor-
rigible because of their religious teachings and cannot be trusted or worked
with for any good purpose.
This man's description of the behavior of Muslims, from their religious
leaders down to ordinary believers, smacks of nothing so much as the blood
libel charges against Jews in European history. The following paraphrased
excerpt from my 1983 interview with him brings this out clearly.

I'RB: What do you think are the causes of [the] 1978 [riots]?
RESI'ONDENT: The Koran....
I'RB: But Muslims read the Koran every day, and there isn't a riot every day. Why are
there riots on certain days?
RESPONDENT: They can't kill every day. Situations are created where they can kill.
I'RB: What are the situations?

RESPONDENT: One case-on 13 September 1978, at a wrestling contest, an altercation

between Hindus and Muslims turned into a free-for-all. Then on 15 September a
Hindu was murdered. On 17 September was Gyanchand's attempted murder, and
then on 3 October Bhura Pahalwan was murdered. On 5 October, again two Hindus
were murdered. After this the Hindus had to retaliate. They have tried to impli-
cate me on communal charges, but my composing foreman, binding foreman and
sales manager are all Muslim.
I'RB: But they read the Koran. How do you know they won't try to kill you?
RESPONDENT: Ninety percent of Muslims don't know what's in the Koran.
I'RB: SO those doing the killing are actually the literate and educated, not the goondas?
RESI'ONDENT: These people give protection and financial help to goondas. On each
level, town, district, up to central level, they have muftis who have the right to declare
jihad on non-Muslims, and it is compulsory for Muslims to kill. If they don't they
are also considered kafir.
I'RB: SO the muftis are responsible, they tell Muslims to kill Hindus?
RESPONDENT: They instigate people and consolidate support to attack non-Muslims. 6

In this bizarre exchange, we confront the logic of the mad. For every ques-
tion designed to suggest the idiocy of the reasoning, the respondent has a clear,
quick, and logical answer. Moreover, the answers are irrefutable. They can-
not be falsified for there is no evidence that can be provided to shake such
fixed beliefs. We know that there is no such organization of Muslims, that
not all Muslims are primed to kill non-Muslims by their holy book and the
teachings and directions of their clerics, but how can we prove it? We doubt
 Interpretation, Blame Displacement, and Communal Discourse /315
that Hindus only retaliate, never instigate. We doubt it especially in the case
of the respondent himself.
But this BJP leader is not mad, nor are his views uncommon. They are views
contained within a discourse, a pathological discourse of nationalism, fear,
and resentment in which many Hindus in northern India are implicated and
to which Muslims must also react.
This BJP leader is a central figure in the maintenance and perpetuation of
Hindu-Muslim tensions and animosities in the town of Aligarh. His name-
and in recent years that of his son as well-inevitably appears in the news
during every riot in Aligarh. His name is kept on a list of communal trouble-
makers in the town that is passed on to every new district magistrate and SSP
posted to Aligarh. There are many others in the town, however, who make
their own distinct contributions. Virtually all are members of the RSS and/or
the BJP.
In Aligarh in 1983, there was a division among the militant Hindu nation-
alists, with some still maintaining allegiance to the original Jan Sangh. The
first interview cited above in this section was with a gentleman who belonged
to the Jan Sangh. I asked another man, then vice-president of the Aligarh City
BJP and an RSS man, what the differences were between the BJP and the rump
Jan Sangh, whose national leader was Balraj Madhok. He replied as follows:

They come together on the same platform if there is a communal riot, but oth-
erwise they are on a different platform. The BJP believes Muslims have got a
right to stay in the country, to progress in the country, but they [Jan Sangh]
believe that India should be a Hindu rashtriya [nation]. The B]P believes India
should be a democratic country in which all cultures should progress, but not
at the cost of other cultures. [According to the RSS ideology,] all Muslims in
India are born from the Hindu womb. If they come to us, we will embrace them.
They do not come from another country, they are Indian, t.hey are Hindus,
born from Hindu mothers and Hindu fathers. However, on a cultural level,
we should have the same culture, just as Catholics and Protestants in Britain
have the same culture, even Indians who migrate to Britain, so why shouldn't
Muslims believe in Indian culture? They should believe in Indian culture, not
Arab culture?

This statement is fully consistent with the RSS ideology of Hindutva. It is one
that, with only minor variations, can be cited by every RSS man. It carefully
distinguishes the Hindutva ideology-which is prodainled to be the only true
"secular" ideology in India-from its perversion in the form of a racist ide-
 316/ Interpretation, Blame Displacement, and Communal Discourse

ology that would exclude Muslims rather than include them in the modern
Indian nation-state as equal citizens-provided, of course, they accept the
RSS definition of the Indian nation, which includes religious and mythological
as well as historical figures. The fact that there are hardly any Muslims
respected in their own community who are willing to accept these conditions
for integration into the Hindu Indian nation matters not to the militant Hindu
nationalists. Nor can it provide much comfort to Muslims to know that, what-
ever the ideological differences between them, the more extreme or racist and
the more moderate or militant nationalist groups come together when there
are communal riots.
Moreover, this ideology that accepts Muslims as part of the Indian nation
is also consistent with an outlook that denigrates and disparages Islam and
discredits and seeks to dismantle Muslim institutions. This outlook is
reflected in statements by both this respondent and others concerning the
role of the Aligarh Muslim University (AMU) in India as a whole as well as
in Aligarh, where it is seen as involved in aiding and abetting communalism
and communal riots. Indeed, the AMU, the leading educational institution
of the Muslim in India, and its academic heads have been demonized not
only as being responsible for communalism and communal riots, but for
admitting thugs to the university to be used against Hindus in such riots. No
fable concerning the AMU is too far-fetched for militant Hindus in Aligarh
to believe, including one that the dead bodies of poor Hindu milkmen were
buried inside the university's student accommodations during the 1980-81
riots. S Over and over again, we hear the "big lie" repeated that, insofar as
Hindus are involved in such riots at all, it is only either as victims or while
engaged in retaliation against Muslims for their instigation of riots and their
attacks on Hindus.
This respondent also assigned blame for Hindu-Muslim tensions and vio-
lence to the political parties that appeal to voters as members of religious com-
munities, that is, as Hindus, Muslims, and Sikhs, and to the government of
India.

The government itself is interested in creating or keeping communal tension

in these cities; they know who are the people responsible for the mess here,
both Hindus and Muslims; the Intelligence people know, but they don't take
any action against them; they may occasionally arrest someone for a month or
two, but then they are released; only on the local level can a good administra-
tor encounter the leaders of the goondas in the university.
 Interpretation, Blame Displacement, and Communal Discourse /317

There are three features of this part of the interview that deserve note.
The first is that the government itself wants to keep communal tensions alive
for its own purposes, which are similar to those of the political parties,
particularly-in this respondent's view-the then-ruling Congress. While
this assignment of blame fits in with one of the general arguments of this
book, that riots are functionally useful for both political parties and the
Indian state, it partly goes beyond and partly says less than the functional
utility argument. To say that communal tensions leading to riots are useful
for governments is something different from saying that they create such
tensions and riots. In the latter case, when such tensions do lead to riots,
insofar as the government is implicated in them, then the term pogrom
becomes more appropriate than riots to describe the carnage that follows.
It is part of the argument of this book that, in fact, several of the riots of
the 19805 and 1990S have been closer to pogroms or massacres of minori-
ties than to riots, but it is not always the case.
This respondent's remarks also say somewhat less than the argument of
this book in confining blame to the then-ruling Congress, by implication leav-
ing the RSS and the BIP free of blame when, in fact, it is evident that most,
if not all, political parties in India make use of riots in political contestation
and that the fomenting of Hindu-Muslim riots in particular has been part of
the routine of party competition in northern India for decades. The third fea-
ture to note is the respondent's reliance on the local administration and on
encounter as the only way of eliminating communal tensions that lead to riot-
ing at the local level. The term encounter, in the Indian context, means a delib-
erately staged, extrajudicial confrontation between the police and alleged
criminals, terrorists, or riot-mongers in which the police have the advantage
and use it to execute the latter.
For this man, then, the solution to the problem of communal rioting in
cities and towns such as Aligarh was a purely administrative one, involving
extrajudicial execution. The local administration, he said, should be sup-
ported "blindly." In this case, he was referring especially to what he described
as "a good experience in this respect with the previous SSP, B. P. Singh,"
whose measures to prevent communal rioting in Aligarh were mentioned
earlier (Chapter 3). This SSP is alleged-and I have heard the story from
other independent commentators-to have resolved the communal prob-
lem in Aligarh for nearly a decade by rounding up more than 40 presumed
criminals and riot-mongers, taking them outside the city limits, and exe-
cuting them.
 3181 Interpretation, Blame Displacement, and Communal Discourse
This respondent generalized this "good experience" into a paradigm for
dealing with the danger of communalism and preventing communal riots.

Our party supported the SSP on one condition: if a Hindu kills a Muslim, you
kill the Hindu immediately, if a Muslim kills a Hindu, you kill the Muslim imme-
diately. Otherwise, if twenty-four hours pass, there will be a communal riot.
We do not want to go to court; there should be immediate encounter. And Mr.
B. P. Singh did that. In fact, however, there were no Hindu goondas, because
a Hindu never takes the lead, he only retaliates. Hindus never have started any
communal riot anywhere in India. It is always the Muslims and only the goonda
Muslims who are either starving for bread and butter or are creating trouble
under the protection of people like Rahman Ali Khan [then a member of the
law faculty of AMU]. So, in one year, the Muslim goondas were killed by the
administration and now we are living in a peaceful city. Before, this was a city
of animals, but now everybody is peaceful. We go to Muslim areas, people know
I am an RSS man, but no Muslim says anything to us because there are no goon-
das there now. All the goondas were taken by the government from Muslim
areas, and they were encountered. There were no Hindu goondas, and, if there
were one or two Hindu goondas, they were also encountered.
The communal situation would also be helped if the senior professors at
AMU, who believe India is not their country and who communicate this in the
classroom, who are goonda-mongers, who organize goondas for communal riots
in Aligarh or get goondas from other cities for killing Hindus, if these people
are stopped by the administration, the administration can encounter the goon-
das very well. These goonda-mongers should be exposed, sent to jail, treated
as spies by whatever method is appropriate for dealing with outsider spies, then
there would be no communal riot in Aligarh. Of course, any Hindu leader who
talks in a communal way also should be dealt with in the same way. These goon-
das are fanatics and they are paid heavily by these supercommunalleaders and
they are also told that they will get shabaah in the coming life if they kill these
Hindu kafirs. And the mullahs, pandits and mullahs, who talk in communal
terms-if you go to any mosque on Friday, you will hear communal talk only
by the mullahs-they should be taught not to talk in communal terms.

Once again, it is necessary to point out that, though this respondent's remarks
may sound extreme and far-fetched, they are neither unusual in relation to
police work in India nor inconsistent with RSS ideology. False encounters
have been a regular practice in Indian "police work" for three decades in
many parts of the country and were used systematically for years in sup-
 Interpretation, Blame Displacement, and Communal Discourse 1319

pressing the Punjab insurrection. So there is nothing distinctive in the fact

that an RSS man should favor such an approach in dealing with communal
riots. At the same time, this respondent's remarks are unrefined. They would
not be articulated to a foreign observer or in public at all by the senior lead-
ers of the RSS or the BJP or any other party. It cannot be said, therefore, that
this strategy of false encounters is part of the RSS ideology. What is consis-
tent with that ideology, however-and brought out starkly in this fascistic
account of how to deal with dangers to the peace of the state-is the thin
veneer of evenhandedness that it projects. In this unsophisticated account-
though it should be kept in mind that the respondent was a highly educated
person and a lecturer in a social science department of the AMU at the time-
the veneer of evenhandedness slips evidently as his phrases come tumbling
out on tape.
The killings of riot-mongers should be done irrespective of whether the
persons killed are Hindu or Muslim-but, of course, it is only the Muslims
who are riot-mongers. There were, of course, no Hindu goondas, but maybe
there were "one or two," and, if so, "they were also encountered." As for the
Muslims, the entire community is not condemned, only the "starving" ones
and those who are thugs under the protection of a Muslim professor of law
at the AMU. Now that these dangerous elements have been eliminated and
the Muslims cowed, an RSS man can walk with impunity through the Muslim
localities. But there remains still a dangerous source of communal tension
in one of the India's two national universities, the AMU. All the riot instiga-
tors have to be treated in an extrajudicial manner, if necessary, as foreign spies.
Once again, of course, the same treatment should be meted out to "any Hindu
leader" of a similar type, among whom in this case our respondent men-
tions the leaders of the rump Jan Sangh. In singling out these few extreme
Hindu communal leaders, this respondent affirms his sincerity and his
adherence to a policy of evenhandedness. But the veneer is too thin, as is the
line which separates the former Jan Sanghis from the reasonable men of the
RSS and the BJP.
One of the lines that has always separated the RSS and the BJP from the
most extreme adherents of the militant Hindu nationalist ideology of the
Hindu nation -state is their alleged openness to admitting Muslims into their
ranks. There have often, in fact, been one or two Muslim members of the
party at the national level and at the state and local levels, though I have yet
to meet in 38 years of field research in India a Muslim RSS man. In Aligarh
in 1983, there was one Muslim BJP member, also a member of the faculty of
a social science department at the AMU and a former BJP member of the
 320 I Interpretation, Blame Displacement, and Communal Discourse
U.P. Legislative Assembly. He described Aligarh as a sick society, in which
the "Muslim mentality," as he put it, was as much at fault as the Hindu. In
the Muslim mind, he remarked, "every Hindu is a Jan Sanghi," "the Jan Sanghi
is the enemy of the Muslim," "therefore, every Hindu is the enemy of the
Muslim." "In fact," he went on, "though RSS people, like anyone, may get
involved in the riots on sentimental grounds, actually it is false that the RSS
actually engineers the riots."
I pressed this respondent further on the charge that riots are instigated by
organizations such as the RSS and the Jan Sangh or by other political parties.
He responded with two rhetorical questions: "What purpose does it serve for
them?" and '~re the RSS people so morally corrupt they would go for polit-
ical gain at the cost of human lives?" He believed that riots served no useful
purposes for the RSS and that they were not so morally corrupt.9
Most politically knowledgeable people in India, however, would give dif-
ferent answers to this man's two questions. To the first, most would respond
by saying that Hindu-Muslim riots serve the purpose for the RSS and the BJP
of consolidating Hindu sentiment behind the RSS and the BJP and provid-
ing votes to the latter in elections, since these two organizations are seen as
the principal advocates of Hindu interests. A factual basis for this answer has
been provided in the preceding part of this book. To the second question,
concerning the moral character of the RSS, opinion would be divided, but
most Hindus would certainly agree with the respondent. In fact, on the con-
trary, not only the RSS, but many of the most famous leaders of the Indian
nationalist and Muslim separatist movements before Independence, as well
as countless local leaders before and after Independence, have knowingly and
willfully "sought political gain" by bringing tense situations involving Hindus
and Muslims to the brink of violent conflict, knowing full well what the con-
sequences were likely to be.

A "Sub-Discourse" of Conflict between Segments of Each Community

In my interviews concerning the causes of Hindu-Muslim riots in Aligarh

over the years, I heard frequently a specific explanation that displaced blame
from the Hindu and Muslim communities as entities onto specific segments
of each community. The groups named were the Hindu caste of Varshneys
(Barahseni) and the Muslim caste of Qasais (also called Qureshis), whose tra-
ditional occupation, as noted earlier, is butchery, though most Qasais nowa-
days are not butchers at all. Having heard such comments for many years
before engaging in a systematic study of communal riots in Aligarh, I
 Interpretation, Blame Displacement, and Communal Discourse /321
specifically asked all my informants interviewed for this study to respond to
this explanation even when they did not volunteer it themselves.
Let us begin first with the rather concise views of the SSP of Aligarh dur-
ing my visit in 1983. Riots in Aligarh, the SSP said, were precipitated by "abnor-
mal" Hindu white collar workers unreconciled to the idea and the fact of
Pakistan's creation. They begin in riot-prone areas, of which he named two:
Upar Kot and Manik Chauk. These two areas are riot-prone because the two
communities are equal in number in them and there is a tendency for Varsh-
neys and Qasais, who live side by side there, to quarreLlO
The SSP, a Hindu, displayed no communal prejudice in his laconic
account. Indeed, he placed blame more on the pathological ideology of a par-
ticular class of Hindus. Otherwise, however, he distributes blame equally upon
the two communities, but particularly the two segments of Varshneys and
Qasais discussed above. Similar views were provided to me by a Yadav sta-
tion officer at Sasni Gate. ll
In contrast, those who spoke in the discourse of communalism rejected
the idea that communal conflict in Aligarh was primarily a conflict between
segments of the two communities rather than the whole. They adhere to an
understanding of riots that starts from a general depiction of the essential char-
acter of Hindus and Muslims. Muslims are active, Hindus are passive. There
cannot even be a generalized conflict between segments of the two commu-
nities, for Hindus do not engage in power struggles on communal lines. Since
so many of the militant Hindu activists in Aligarh are from the Varshney com-
munity, my respondents were quick to defend the Varshneys, to refute the
charge that the Varshneys are responsible for anything, least of all commu-
nalism, and to eliminate any differentiation within the Muslim community,
which must be seen in the militant Hindu communalist mind as a unity.12
Mr. N avman also dismissed any alternative explanation of Hindu-Muslim
riots in Aligarh, beyond the single cause that Muslims start riots to kill Hindus
according to the injunctions of their holy book itself.13 He insists that Hindu-
Muslim riots are community-wide affairs instigated by Muslims and followed
by Hindu acts of retaliation.
Muslims also tended to downplay the significance of the alleged conflicts
between Varshneys and Qasais, as in one of my interviews with a politician
from the Qureshi community, a member of the law faculty of the AMU. He
denied that conflicts between Qureshis and Varshneys or the participation of
Qureshis in communal riots in general were of any significance. While acknowl-
edging that persons from among the Qureshi community were involved in riots
and were frequently arrested by the police, he gave an explanation that absolved
 322/ lnterpretation, Blame Displacement, and Communal Discourse

them from any responsibility for instigating or taking advantage of riots. He

noted, rather, that it was a matter of chance, of the particular spatial distribu-
tion of the Qureshis. It just happens that Qureshis are living on the borders of
mohallas with high Muslim population concentrations adjacent to Hindu local-
ities. So, when riots start, persons from the adjacent, Hindu localities "come
to the border, ... play their part, and then run away." The Qureshis, being on
the borders, therefore, get blamed for whatever happens at the borders of the
Muslim localities. Further, when the police arrive on the scene, make raids,
and "start arresting Muslims, Qureshis are arrested because they are living there,
whether they are innocent or not innocent."'"! Thus do we find yet another
kernel of truth in the mosaic of explanations that simultaneously implicate
and free from blame categories of participants in Hindu-Muslim riots.

THE ROLE OF THE ALIGARH MUSLIM UNIVERSITY

We have seen numerous references in the interviews cited above to the role
of the AMU in riots in Aligarh. The range of views includes at one extreme
those who argue that the presence of the AMU in Aligarh is a mere pretext
for the organization of communal riots by Hindu chauvinists unreconciled
to the creation of Pakistan, who also consider the Muslims of India a vast
army of fifth columnists. At the other extreme are those who consider that
the AMU has been very much involved in communal riots in Aligarh, not
merely symbolically but in the provision of arms and shelter to criminals
involved in riot activity.

Criminals at the AMU

Although at the time I considered the charge far-fetched, I pursued this ques-
tion of the direct involvement of the AMU in riots through support of crim-
inal activity in many interviews conducted in 1983. I was, therefore, quite
surprised to find that there was virtual unanimity among my respondents
that the charges were at least partly true. The police and the civilian admin-
istration were emphatic that the charges were accurate. Moreover, not every
policeman who accepted the truth of these charges appeared to be biased
against the university as such.
For example, the SSP of the district, whose interview was previously cited,
did not see the university as the source of the problems leading to riots, except,
he noted, that criminals are sometimes sheltered there. 15 Riots as such, he said,
began in the city, not in the university, and, insofar as they did affect the un i-
 Interpretation, Blame Displacement, and Communal Discourse 1323
versity, it was a consequence of a spillover effect from the city. Nevertheless,
he remarked, it was true that criminals from the city got shelter in the AMU
during riots as well as during normal times.
A second police officer in Aligarh, one whose own impartiality and hon-
esty I came to doubt, but whose testimony on this matter is nevertheless rel-
evant as reflecting a widespread police view in the district at the time, was
emphatic on this matter. He said that it was "a thousand percent correct"
that criminals are sheltered inside AMU itself. Those sheltered were not only
ordinary criminals, but people who indulged in communal incidents, arson,
and so forth, in order to get shelter there. 16
Two lower-level policemen from the city also supported the same view.
The Yadav station officer of Sasni Gate told me that, during and after riots,
wanted criminals get protection at the AMU.17
The police inspector of Bannadevi remarked that criminals who had stud-
ied previously with AMU professors got protection at AMU. 18 After riots, all
protection was provided by the AMU to these criminals. The inspector also
noted in his remarks that people in Aligarh did not benefit from the AMU,
reflecting thereby his identification with the antagonism of Hindus in the city
towards the Muslim University. Nevertheless, he said, the cause of riots was
not AMU as such. The cause, he said, was in the city.
The district magistrate, head of the civilian administration of the district,
also responded that the charge that the AMU protected criminals was true.
Even in normal times, he said, criminals take refuge there. He noted that the
charges had been confirmed during a raid on one student hostel where many
arms of all kinds were found. 19
Even on the university campus, the police charges were not entirely dis-
missed. One noncommunal Muslim academic, himself a political scientist,
gave the following response to the question concerning the AMU role in pro-
tecting criminals.

It is only twenty percent true that many criminals get refuge in the university
campus during riots. However, it is not only Muslim criminals, but Hindu crim-
inals also who get refuge there during riots. The attraction of AMU for crim-
inals obviously is because the police are not allowed to enter easily. However,
the issue has been blown up by the press, especially in the case of Muslim crim-
ina.ls. Nevertheless, it is true that the [university] administration definitely over-
looked the problem of criminals on campus during [vice-chancellor J Khusro's
times. However, since Syed Hamid's takeover [as vice-chancellor], most of the
criminals were apprehended and the administration received active support
 324 / Interpretation, Blame Displacement, and Communal Discourse

from the students union. Anyway, singling out AMU is not helpful because the
problem exists and is even worse at other universities.:w

Overall, therefore, there is a very considerable consensus in my interviews

that it is factually correct, on a scale somewhere between twenty percent and
"a thousand percent," that the AMU did provide shelter and refuge to crim-
inals, that arms were hidden on the campus as well, and that the problem
had not yet been solved by the time of my interviews in 1983. It is at the same
time noteworthy that the police and civilian administration treated this ques-
tion as a police question and did not view the AMU as such as the source of
the communal problem in the city. They argued instead that the source of the
problems was in the city. The political science professor also insisted that
the issue of criminal activity inside the AMU was not confined either to
Muslims or to the AMU, but was a general problem. Both Hindu and Muslim
criminals received protection at the AMU. Moreover, criminal activity in north
Indian universities, he noted, was a commonplace at this time, as it has been
in later years as well.21

The Aligarh Muslim University as a Precipitating

or Originating Factor in Communal Problems

A second rather broader issue concerning the role of the AMU in the com-
munal situation in Aligarh concerned whether or not the university was a
source in one way or another of communal problems in the city, including
riots, rather than the reverse. The Hindu communalist view is clear enough
on this matter. To persons of this persuasion, the AMU is the real root of the
communal riots in Aligarh. It is charged that "AMU people" not only
"inflame" the city population, they plan and organize riots in the city.n
Although several police officers interviewed agreed that shelter and some
arms were provided to criminals at AMU and that some planning was done
at the university, all insisted that riots start in the town and then may affect
the univerity. But the militant Hindu argument goes much further and turns
around completely the Muslim and secular intellectual charge that riots are
preplanned and organized by militant Hindu groups such as the RSS. The
militant Hindu argument rather is that everything is planned and organized
at the AMU in order to achieve specific Muslim objectives, such as maintaining
the minority character of the university, which, as we have seen above, had
been a controversial issue in north Indian politics generally during the 1960s
and 19705. That the trauma of Partition lies behind everything in the Hindu
 Interpretation, Blame Displacement, and Communal Discourse /325

communalist's mind is evident in the fantastic view that the Muslims are capa-
ble of demanding another "division of India." This can only be described as
a pathology of the historical consciousness, a phenomenon that is endemic
in India and in some other parts of the world today, a subject to which I will
return below.
It is clear that the Hindu communalist view of the role of the AMU in
riots differs substantially from the police views cited above. But the DSP,
Intelligence presented an intermediate view. He did not take the position that
people from the AMU planned and organized riots. Moreover, he said that
there was "never any violence at AMU, except once in 1979, which was not a
communal case."
Although the DSP was clear, therefore, that in his view the university was
"not a contributing factor to the communal tension in Aligarh," it did con-
stitute a symbolic presence that affel-'ted "the course of events." It had a "unique
place in the Muslim world" and had "become a symbol of Muslim aspira-
tions."23 Although he did not say so, its very representation of Muslim aspi-
rations distinct from those of the Hindus make it, in the words of the district
magistrate at the time, an "eyesore for Hindus."24
How much of an eyesore the AMU is for militant Hindus, if not neces-
sarily for all Hindus in Aligarh, was brought out forcefully to me once again
as late as November 1997 in an interview with a lifelong RSS worker,25 a man
of 66 years of age, a retired chain factory foreman, born and brought up in
Aligarh, and himself a graduate of AMU. He put the matter to me in a very
long discourse as follows.

RESPONDENT: As a Hindu worker and as an experienced person of this university, I

believe and I have written books on the subject that AMU was the root cause of
the partition of this country. [ ... 1We believe that this university is responsible
for the partition of this country.
I'RB: And you are old boy.
RESPONDENT: Old boy of university. And we Hindus are sorry that that mentality
of this university has not changed even after the partition of this country.... So
the basic question before the Hindus of Aligarh-and I think Hindus of the entire
nation-is that the Muslims are again acting-and perhaps this time they will
not demand partition, another partition of this country-and this is our hon-
est opinion and I think I'm expressing the feelings of all the nationalist Hindus
that this time they will not demand another partition of this country. Now, they
are out to make this country darou/-Islam. The danger is not before Aligarh town,
before the Hindus of this town, but the entire country is facing this danger: how
 326/ Interpretation, Blame Displacement, and Communal Discourse

to, how to save this country from these terrorists. The conditions of Kashmir you
know very well.
PRB: Of course.
RESPONDENT: The same conditions are being created here. Aligarh University, when-

ever gentlemen like you come to this country to know certain things about the sit-
uation of this country, they don't allow the real Hindu workers to meet those
gentlemen. They will prefer secular socialist Hindus.
PRB: This is why I have come to see you.
RESPONDENT: And these secular socialist Hindus are more dangerous than Muslims,
so far [as] we Hindus are concerned. They are more dangerous. We cannot believe
them. We cannot [accept] them as our leaders. They are our hidden enemies. They
will never tell you the truth. They will abuse Hindus. They will defame Hindus....
And this is the most unfortunate situation, that whenever outsiders or foreigners
come to this country and they try to meet citizens of Aligarh through this uni-
versity, the university does not give them a chance to meet the real persons or the
nationalists of the town. I'm surprised that you are sitting in the office of the RSS
and talking to those people who believe that the country is facing perhaps a greater
danger than they faced in 1946 and'47. They are openly calling their community
for jihad. And nobody is taking action. Now, this is the situation in brief, which I
would like to put before you.

And here is what I would like to put before the readers of this volume.
The remarks of this gentleman are not exceptional. That should be to some
extent evident from several other statements quoted above. The beliefs con-
tained in this statement are widely shared among Hindus in northern India.
They are presented in this respondent's remarks in a highly coherent and log-
ical form. They are presented sincerely, they appear to be believed strongly
by the speaker, and they elicit a certain amount of sympathy in the listener-
not for their truth but for the feelings expressed. They are, nevertheless, non-
sense. They display a package of sentiments that have been present in all kinds
of antidemocratic, paranoid, and fascist movements in the twentieth cen-
tury: resentments against an "other" combined with gross ignorance of the
other's beliefs and sensibilities; fear of and magnification of the danger rep-
resented by the other, even though the other is manifestly weaker; identifi-
cation of an internal enemy hated even more than the other, those who should
be of us, but have been corrupted by false ideologies; belief that the country,
the community, the nation, are all in danger and that no action is being taken
against the manifest presence and actions of hostile enemies operating openly
in our midst.
 Interpretation, Blame Displacement, and Communal Discourse /327

The resentment expressed is twofold. It is directed first against the Muslims

of India, and against one of the leading symbols of their presence in India,
in Aligarh itself, the AMU. But another kind of resentment also appears con-
nected with the first, namely, that the voice this respondent represents, of
aggrieved Hindus, is not heard by outsiders. His manifest pleasure at my pres-
ence and the opportunity that I represented in his mind, to be heard by the
West, most likely prevented him from expressing that other resentment I have
heard times without number in 20 or 25 visits to India over several decades:
why have we (the United States) supported Pakistan, therefore the Muslims,
therefore the enemy of the Hindus and of India?
Gross ignorance is displayed in the apparent belief, shared by so many mil-
lions upon millions of Hindus and Christians as well, that jihad presumes
only violent conquest. That ignorance is compounded by exaggerated, para-
noid fear that the Muslims, through jihad-whatever it means to the mili-
tant Hindu-and through "terrorist" actions, will be able to turn India into
an abode of Islam. How can such a thing be imagined in a country where
those people classed as Hindus comprise 85 percent of the population of their
country and a sixth of the population of the globe? It can be imagined only
by demonizing the other, by attributing to the other demonic powers that
increase his strength, and, simultaneously, by attributing to oneself, that is,
to the Hindus ofIndia, internal wealmess produced by disunity and the pres-
ence of enemies within one's own fold who are worse than the other. This,
in a nutshell, is the essence of the ideology of militant Hindu nationalism
shared by millions of RSS workers and many more sympathizers.
But we have come far from Aligarh in this analysis, without leaving it. We
are clearly back here in the midst of the communal discourse, but not just
there. We are in the larger world of not only Hindus and Muslims, but the
conflict of ideologies in the country as a whole, the even larger world in which
Hindus and India have not achieved their deserved place, are not heard by
foreigners and not respected, because they have been displaced from their
occupation of the rightful center of their own country by its partition. In this
world, the riots that comprise my theme and subject are but specks in a much
bigger frame.
 141 Police Views of Hindu-Muslim Violence

THE POLICE, THE PAC, AND THE STATE

M ost of the persons killed in most riots in India and in all riots in
Aligarh since Independence have been Muslims; moreover, most
have been killed by the police. That fact alone would normally
arouse the suspicion that what are called riots in India and in Aligarh are actu-
ally pogroms, that is, the deliberate killing of Muslims by the agents of the
state. However, matters are more complicated than that and raise doubts about
the validity of the distinction between riots and pogroms. In fact, what has
been emerging from so much of the material presented so far is rather that
riots are political productions, in which, indeed, the police playa role-but
that role varies depending upon a number of factors, of which the most impor-
tant for our purposes concerns administrative and political control of the
police. Neither the Indian state nor the province ofU.P. are anti-Muslim insti-
tutions whose police are engaged in either systematic or sporadic slaughter
of Muslims. Rather, the police act against Muslims or do not act against
Muslims, when riots occur, depending primarily upon the inclinations of their
administrative and political superiors, which in turn depend upon which polit-
ical party or coalition is in power. That does not mean that individual police-
men do not act according to their own inclinations as well, but their range
of freedom to do so during communal riots depends critically upon the given
state of administrative and political control over them.
It is important also in this connection to make a distinction that further
complicates matters, between the ordinary armed police and the PAC. Police
constables are recruited from within their own district and are transferred
only within the district. Superior officers generally come from other districts.
The senior superintendent of police (SSP) will usually be an educated per-
 Police Views of Hindu-Muslim Violence 1329

son from another district who has been recruited by examination into the
Indian Police Service. Transfers and postings of the senior officers are done
by the state government, that is, by the party in power. They are done fre-
quently and are influenced by political and patronage considerations. The state
government wants men whom it can trust as SSPs in the districts tllat are com-
munally sensitive. When the government changes hands from one party to
another, iliere are usually wholesale transfers of SSPs. Wiiliin the district, trans-
fers and postings are made by the SSP, who in turn may be influenced by ilie
desires of locally powerful politicians, particularly the local MLAs.
The PAC, on ilie oilier hand, is an entirely different unit iliat operates under
its own command hierarchy, but is ultimately beholden only to the state gov-
ernment. Very little is known about the organization of the PAC and about
the effectiveness of state government control over it. The PAC jawans and
their officers are not available to be interviewed by foreign scholars, nor have
iliere been any scholarly studies of ilieir functioning. It is commonly accepted,
however, that the PAC is an anti-Muslim force. 1 Since, as I have said, it would
be a simplification to consider ilie police as an anti-Muslim force, it is-given
the fact that both the police and the PAC are under ilie ultimate control of
the state government-an anomaly that the PAC appears to act consistently
against Muslims whereas the police do not. I cannot explain this anomaly,
but it is evident from it that it further confounds the distinction between riots
and pogroms.
I want to present in this chapter two sets of perceptions: those of the pub-
lic and the politicians about the police and the PAC, on the one hand, and
the perceptions of the police concerning the role of the people and ilie politi-
cians in riots.

PUBLIC PERCEPTIONS OF THE POLICE

Public perceptions of the police do not always distinguish between the

actions of the police and those of the PAC once the latter have been called
out. While Muslims do distinguish the PAC from the police as separate forces,
when they say there was a "police firing," it is not always clear whether it was
the ordinary police who fire or the PAC. Insofar as the major riots in Aligarh
are concerned, I have heard repeatedly from many respondents that the police
in general engage in various actions that are considered either partial to the
Hindus or outright and unprovoked assaults on Muslims with intent to kill.
These accusations were especially prominent in my interviews after the
1990-91 riots. Indeed, the mufti shahar of Aligarh in January 1991 said to me
 330/ Police Views of Hindu-Muslim Violence

that the violence that had occurred in the preceding months was not a riot,
but a police action. 2
I have already reported on the alleged behavior of the police in Sarai Sultani
during that riot and previous riots. But Sarai Sultani has not been exceptional
in this regard. Accounts I have heard again and again from Muslims contain
the following elements} If an attack breaks out between Hindus and Muslims,
such as stone throwing by mobs from the two communities, the police act
partially on the side of the Hindus, and shoot to kill Muslims. The police also
engage in unprovoked actions, including the destruction of local mosques.
The police kill and loot houses during riots, even in Muslim localities where
there is no tension. Police fire upon crowds of Muslims returning from Friday
prayers. Police enter Muslim houses without provocation or, alleging that
shots have been fired from a particular residence, molest women and arrest
them and their sons if they protest, or kill the sons. Police act during riots
upon requests by local Hindus or Scheduled Castes who have had quarrels
with Muslims, and shoot Muslims dead. Police deliberately kill even children,
reportedly between the ages of six and twelve. In these cases, for all of which
I have documentation from interviews, the respondents sometimes said that
it was the police who committed the attack, sometimes the PAC, sometimes
both combined, and sometimes that they could not differentiate the two. 4
Other interviews suggest that a Muslim wounded in a riot, if he is lucky
enough to make it to his home, cannot expect to receive any medical help. If
the police are called, the likely response will be that the family is sheltering a
rioter. Only if the family members have influential friends or if they are lucky
enough to find a "kind and polite" police office can they expect their wounded
relation to be taken to hospital.5

SOME POLICE VIEWS ON RIOTS

The documented misbehavior of the police and the PAC in countless riots
in India since Independence is said to be associated with a distinct police view
of riots. That view has been expressed most clearly in an unofficial, unpub-
lished, and undated report on "Communal Riots and Minorities," which pro-
vides an accounting of the numbers of Hindus, Muslims, "others," and police
killed between 1968 and 1980, and concludes with a set of statements sum-
marizing "the perception of the magistrates and senior police officers" in the
country concerning riots. These perception constitute, in effect, in the judg-
ment of the report's author(s), a police view of riots. I undertook in my own
research numerous interviews with police to ascertain their views of riots,
 Police Views of Hindu-Muslim Violence /331

which I will compare with the conclusions in this document. I give below,
first, a long quote, comprising the report's complete summarization of that
police view.

(a) Riots take place in such districts where Muslims are either in a major-
ity or they constitute a sizeable minority.
(b) Muslims are excitable and irrational people who are guided by their reli-
gious instincts. Hindus, on the other hand, are law abiding and cooperate with
the police in controlling communal violence.
(c) Riots are started by the Muslims and they invariably take the first oppor-
tunity to strike at the other community and at the police.
(d) In all other previous riots in the country before the current riot, Mus-
lims took the upperhand [sic] which resulted in huge loss to the Hindu com-
munity. Therefore, there is moral justification if in the current riot casualties
on the Muslim side are heavier.
(e) State Government attaches a great deal of importance in ensuring quick
control of rioting. Since Muslims are aggressive, therefore, in order to control
violence, it is necessary that Muslim mobs must be taught a lesson through
arrests, firing and third degree methods.
(f) Hindu casualties are as a result of Muslim mob action, whereas Muslim
casualties are due to isolated stray incidents. Because of this difference in the
nature of aggression by the two communities, more Muslims have to be arrested.
Very little evidence is possible to collect regarding Hindu aggressions and this
explains why the number of Hindus arrested for substantive offences is less. 6

In short, despite the known evidence that more Muslims than Hindus
are killed in most riots, the police view is said to be that Muslims start these
riots in places where they have large enough numbers to do so and Hindus
are for the most part victims of "Muslim mob action." The known facts that
more Muslims are arrested and killed in most riots are not denied, but are
explained away by placing the blame for them upon Muslims and freeing
Hindus from all blame. In short, this alleged "police view" of riots is per-
fectly consistent with the militant Hindu communalist ideology discussed
in the previous chapter.
The report cited above, published twenty years ago, also noted that the
attitudes of the PAC, an armed police force even then notorious among and
hated by all Muslims in the state ofU.P., were even more negative than those
of the local police-who actually come into contact with the people on a
day-to-day basis, in contrast to the PAC jawansl who remain in their
 332/ Police Views of Hindu-Muslim Violence

FIG. 14.1. PAC encampment, Aligarh, 1999

encampments (Figure 14.1) and are called out only for riot duty. Further, it
was noted that Hindu district administrators and officers who acquire the
confidence of the Muslims lose the confidence of the Hindus in the local-
ity, with the result that "pressure is mounted to get them transferred," mak-
ing it increasingly difficult "to have officers commanding the respect of both
the communities."s
The police are widely condemned in India for their often partisan behav-
ior during riots and particularly for the fact that mostly Muslims have been
killed in most post-Independence riots in Aligarh, as elsewhere in the coun-
try. It is a matter of some interest, therefore, to discover in interviews with
police officers at all levels that many of them present quite cogent interpre-
tations of the causes of riots that depart significantly from the summary view
given above and from the militant Hindu ideology. To the extent that there
is a common police view of riots, it can be summarized as one that combines
the idea that there is a pathology of ideology in Indian society and perhaps
of personality as well with demographic and communal factors. That police
view also includes the attribution of profit motivations to rioters. I will show
below, however, that there are important differences in police views of the
causes of riots between subordinate policemen such as station officers and
police inspectors, on the one hand, and more senior officers, on the other hand.
It is especially among the latter that Hindu-Muslim riots are attributed to
fundamental differences between the two communities.
 Police Views of Hindu-Muslim Violence /333

Views from Below

It was a matter of some surprise to me to find that local police inspectors and
station officers in the mohallas of AJigarh in the riot-prone areas expressed
their views on riots very coherently, fully, articulately, and sensitively, with
many nuances. 9 For example, I asked the inspector of police, police station
Bannadevi-very much a riot-prone area-what were the factors making the
area near hi" police station riot-prone. I give below in compressed form, partly
directly quoted and partly paraphrased, this inspector's response. lO

[The factors making this area riot-prone] are, first, high population density,
such that "if there is even a small altercation people come out in huge num-
bers, and without knowing the issue they fight." Second, "criminals ... exploit
the situation." They stab someone and, if it is a Hindu, Hindus say a Muslim
stabbed him; if is a Muslim, they say a Hindu stabbed him. Third, before and
during elections also, politicians may start a mmor or even a small riot to gain
the support of one side. They engage criminals to start a riot, then portray them-
selves as saviors of the people by distributing grain and doth to the people dur-
ing curfew. They collect money from different districts and distribute it among
the criminals; they have formed committees for this.

Thus, this police inspector pointed to demography, criminality, and extreme

political opportunism, t.)'llicism, and hypocrisy as a triad of factors produc-
ing riots and making particular areas of the city riot-prone. He elaborated
further, however, and added additional factors after giving an example of how
riots began in AJigarh in 1980 in the aftermath of the terrible anti-Muslim
riots in the nearby district of Moradabad in that year.

In 1980, an AMU professor and another used money to form a group of crim-
inals, held a meeting to organize, then spread through the area, first breaking
the lights so there would be darkness, then stabbing people, with each crimi-
nal taking charge of a particular area. This was a repercussion from the Mora-
dabad riot of 1980.

A fourth factor, therefore, now appears: the deliberate instigation of riotous

activity by professors from the AMU, who use criminals in the task of pay-
ing back Hindus for their actions in another district. The involvement of AMU
professors must, therefore, in the light of this inspector's account, be taken
seriously, since the charge is made without evident rancor or prejudice, in
 334/ Police Views of Hindu-Muslim Violence

contrast with the RSS respondents cited in the previous chapter. Yet another
factor was mentioned by the police inspector, what he calls the "mentality of
the people."

Even ordinary people become susceptible to violence over even trivial incidents.
For example, a child may fire a shotgun blast prematurely during Rarnzan to
confuse Muslims waiting for the blast to indicate their fast is over; this actu-
ally happened in 1982 and nearly caused a riot. People have also become inured
to riots, expect them, and even have extra children because they feel they will
lose some in riots.
It's the mentality of t.he people, how they live. Most families have 12-14
children. Most children are uneducated, with nothing to do. In a riot, they will
throw stones at the police jeeps. Rioting is the only form of entertainment avail-
able. Employment also is limited to the lock industry, with few jobs. The pop-
ulation must be reeducated. A new town must be constructed in which these
narrow alleys are done away with, and the hygiene is improved.
The nature of these areas also makes police work nearly hopeless. There are
underground hiding places and numerous by-lanes through which criminals
can reach the city outskirts. Moreover, in Muslim areas, the people support
the criminals. Muslim women come and stand in front of the houses and say
no one is inside and you cannot argue with them. Even if someone is caught,
chances are that Muslims of the area will come out in support of him and throw
stones at the police.

The number of factors continues to tumble out as this police inspector explains
the conditions of the people in the mohallas in which riots occur. Rioting has
become a routine aspect of the life of the people, arising out of how they live
and how they think about how they live, and what they have come to expect.
Consequently, riots can erupt spontaneously as a consequence of such triv-
ial incidents as a child's prank. In an area with limited employment oppor-
tunities, hordes of children, no hygiene, little money for entertainment, and
none available except for cinema halls that cost money, rioting becomes a
"form of entertainment." 11
This police inspector also believes, as would any Western urban planner
or any casual Western visitor to these indescribably unhygienic and crowded
areas, that these filthy slums must be torn down and "a new town must be
constructed." This idea, as far as I know, has not occurred to any of India's
planners, or, if it has, it has been kept a secret. Yet, within ten or fifteen
kilometers of this mohalla-which, in the seventeen years since this inter-
 Police Views of Hindu-Muslim Violence /335

view was done, has become even more unhygienic and so crowded that one
can barely move through it without coming into contact, often hard con-
tact, with people, animals, scooters, cars, and assorted other mechanically
and humanly powered vehicles-huge new buildings are being constructed
in open areas, many of which would take up as much space as that occu-
pied by hundreds of slum families, for the occupation of the newly rich (see
Figure 6.1). The police inspector did not make any such comparison
between his police station area and other, better-off areas of Aligarh. He
was merely noting the obvious: if people have to live like animals, they will
behave like them.
There is yet a further point in the police inspector's account that requires
special notice. It is about police work, especially in the Muslim areas. The
inspector was a Hindu, his area mixed Hindu and Muslim. He expressed no
obvious hostility to Muslims, which does not mean that he did not harbor
any that he might have expressed to persons other than a foreigner with a
tape recorder. But I detected no tinge of it in his remarks. Therefore, we must
take seriously his expressions of futility over the problems of police work
during a riot in a qasbah town with "underground hiding places and numer-
ous by-lanes" in which criminals are at work and where ordinary Muslims,
including Muslim women, come out to protect the criminals and "throw
stones at the police." In such circumstances, it is not difficult to imagine the
police losing control, opening fire in frustration and rage, and killing many
such "ordinary" people. Such actions, which have in fact occurred count-
less times in these kinds of situations, can in no way be excused. At the same
time, it is easy to see also how the police in such situations become scape-
goats for the intolerable conditions in which they must work, as a conse-
quence of the failures of the politicians and so-called planners to give even
a moment's thought to the conditions of these "ordinary" people of India,
except to make use of them for their own political advantage. Police who
must work day in and day out in such conditions also confront day in and
day out not just the neglect of their masters, but their malevolent hypocrisy,
as suggested by the police inspector himself in these additional comments
in response to my question concerning who actually perpetrates the violence
in communal riots.

The physical actions are done by persons from lower castes and classes, but
persons from middle and upper classes give shelter and guidance: Big crimi-
nals [who become involved] are close to the top lawyers, who are paid by the
politicians from the committee funds for the relief of the riot-affected. If the
 336/ Police Views of Hindu-Muslim Violence

police take a criminal to court, even for section 362,12 he is released, which sets
a bad example. They also get the support of MLAs and people in Delhi, who
pressure the government officials. The people responsible are known to the
police. Navman is the main fellow among Hindus. Shiv Hari Singhal is the lawyer
for most Muslim criminals. Manga Ram looks at a situation, then goes to the
colleges, instigates the students, and they go there.

The inspe(.ior then proceeded to give a specific example of a "big criminal"

responsible for riots in Aligarh who had "big connections" and was able to
get a false passport, with the help of a central government minister, to go to
Pakistan.
Although the inspector previously gave an account suggesting how riots
may erupt spontaneously in crowded mohallas, in these remarks it is clear
that there is in his mind a division of roles and responsibilities for the actions
undertaken during riots. These include the actual rioters, who come from
the lower castes and classes; leaders and protectors, who come from the mid-
dle and upper classes; and criminals, lawyers, politicians, and government
officials tied together in a network of influence and mutual protection. These
remarks were made in 1983 when the pervasiveness of the criminalization of
politics in India was probably less than it has been in recent years, when a
well-known mafia don of an eastern district of D.P. was made a minister in
the state government run by the BJP (in 1999), and several other members of
that government-as well as many in its predecessor governments run by other
parties-have had criminal records.
This police view is remarkable in many ways. It contains a mixture of
sense and nonsense, though more of the former than the latter. It focuses
on particular behaviors and local factors rather than upon the history or
the imagined future of Hindu-Muslim relations. It places blame generally on
socioeconomic conditions, but also pinpoints particular individuals alleged
to be fomenters of and organizers of riots, including one of the leading RSS
persons in Aligarh, the lawyer Shiv Hari Singhal, who, the inspector is sug-
gesting, is defending Muslim criminals. There is a clear implication that
Muslim criminals are paid by Hindu communalist leaders to stab persons,
including fellow Muslims, in order to set off a riot, and are then protected
by them. The police inspector's statement also, of course-like virtually all
explanations of communal and other riots-displaces blame, in this case from
the police onto the criminals protected by the big lawyers and politicians.
It appears clear to me from the whole tenor of the police inspector's remarks
that he did not have an anti-Muslim bias. I asked two further questions, the
 Police Views of Hindu-Muslim Violence /337
answers to which further bear out my impression. I asked, as I did with most
respondents for this study, specifically about the roles placed in riots by
Varshneys and Qasais. He responded that it is "true they play the most active
part because they are the strongest of the communities" in Aligarh; "there
are large numbers of both and they have the same mentality."
My second question was more direct: do Muslims always start the riots?
His response was, "most of the time." However, he added that, "since 1981,
Varshneys have taken to provoking Muslims." If my impression that the police
inspector was not biased is correct, then how could he make the statement
that most of the time Muslims do in fact start riots? He has himself pointed
to the possibility that Muslims start riots, at least since 1981, because they are
deliberately provoked by some anti-Muslim act to do so. Further, he has him-
self implied that Muslim criminals are paid and protected by the most
extreme elements among the militant Hindu nationalists for their work in
provoking riots. Here we come close to the heart of the matter of the labelling
of such large, widespread, and multifarious events as riots. Who is a Hindu
and who a Muslim in such affrays? Is a criminal, Muslim by faith, who is
paid to precipitate a riot by an extremist Hindu politician/businessman and
protected by an extremist Hindu lawyer, acting as a Muslim on behalf of his
community? The constant repetition of the words "Hindu" and "Muslim"
in these contexts, including by this writer, is both inescapable and misleading.
How can a writer write about events already classified and embedded in a com-
munal discourse without using such labels? I am here trying to sift and sort,
but in the process of sifting and sorting, I too get caught in the mesh.
I have spent much time on this single interview with an inspector of police.
A single in-depth interview does not count for much in contemporary social
science, which rests so heavily on structured questionnaires. It will certainly
not quell the doubts of social science practitioners to say that I have had
many other such interviews. 13 But I am not attempting a scientific account
of the causes of riots nor conducting a public opinion survey. I am seeking
rather to discover, through repeated interviews asking the same simple ques-
tions over and over, whether or not there is a "police view" of riots that can
be distinguished from other kinds of views of them. I can only demonstrate
here-and could demonstrate further with other interviews from other dis-
tricts that will not be covered in this book-that there was some consistency
in the accounts given to me by policemen at the lower levels of the police
hierarchy.
I will conclude this presentation of police views in Aligarh in 1983 with
another set of responses to the same questions, from an interview with a Hindu
 338/ Police Views of Hindu-Muslim Violence

station officer in another riot-prone locality, Sasni Gate. This interview, like
the previous one, has been translated from Hindi, compressed and partly para-
phrased, partly directly quoted.

PRB: How do riots get started?

so: Riots occur because of mutual misunderstanding: certain elements exploit petty
incidents of crime and give them a communal color for their own benefit. These
elements are mostly white-collar people who try to convert every incident into a
communal incident and exploit it. Then, they go into the background when it
becomes a big issue. Small incidents are exploited, then rumors are spread, espe-
cially by telephone. Then there's a stampede as people come to believe something
has happened. Then the market is dosed, a crowd forms, and the situation goes
out of control.
PRB: Who starts the riots?
so: It is not always the Muslims. The RSS is no less strong than the Muslims and as
disciplined and organized.
PRB: What about Varshney-Qasai conflicts?
so: It is partly true. Varshneys are the biggest group in the RSS and Qasais are gen-
erally Muslim League or Jamaat-i-Islami.
PRB: What are the differences between Aligarh and other V.P. towns that do not have
riots?
so: In other cities, Hindus or Muslims may be in larger numbers [that is, one or the
other predominates1or the organizations are not so organized and broadly based.
Here, also, the intellectuals are Muslims with communal feelings. Moreover, Aligarh
has always been divided on communal lines, with Hindus provoking other Hindus
[to hostility against Muslims] by telling lies and Muslims doing the same.
PRB: Who benefits from riots?
so: Businessmen, because laborers don't get paid during riots when there is curfew
and then become dependent on the bosses for help and become almost like bonded
labor; also, those industries which can keep people employed even during riots
also benefit because they can pay the laborers at low wages. Professional criminals
also benefit because they are well paid by vested interests for committing stabbings,
et cetera. Otherwise, most groups suffer.
PRB: What about the politicians?
so: Mostly, they are not around. When they are, they don't give good advice. If they
did give good advice, these situations might not tum into big riots. Some politi-
cians may exploit the situation by playing on the sentiments of the people to increase
their popularity. Then, when the DM and the police have to pay attention to them,
they become even more popular.
 Police Views of Hindu-Muslim Violence 1339
PRB: What about the general problems in these mohallas, of the alleyways and lack
of education, et cetera?
so: Naturally, because a child is born and grows up with communal riots, a riot psy-
che persists into adulthood.

Clearly, there is a close correspondence between the responses of the two

lower-level policemen to my questions, though there are some interesting vari-
ations and emendations. One emendation is the idea that riots arise out of
"mutual misunderstanding" through the manipulation of "petty incidents"
that are turned into communal questions for the benefit of others. In this
account, there are certain people from the "white-collar" classes who act, in
effect, as what I call "conversion specialists" whose function is to "try to con-
vert every incident into a communal incident": "rumors are spread, especially
by telephone."
I have myself seen how the rumor network operates, from repeated occa-
sions of sitting in interviews with BJP politicians who receive telephone calls
or visits from persons alerting them to alleged incidents, such as that a cow
has been "poisoned" or a Hindu girl has been raped by Muslims. These con-
version specialists mayor may not choose in each case to mobilize their forces
to protest against the alleged insult, assault, or injury to Hindu sensibilities,
but they remain always alert to that possibility and will do so if the political
context justifies it. But others benefit from riots also, this station officer avers,
particularly businessman and professional criminals.
The most striking aspect of the two statements by these lower-level
officials is their debunking of the whole idea that riots arise out of"communal"
differences between Hindus and Muslims. On the contrary, they say that what
are called communal riots arise out of "trivial" or "petty" incidents, out of
"misunderstandings" that are deliberately exploited by others who benefit from
giving such incidents a "communal color" and fomenting riots, from which
they derive concrete, material benefits in business, cash, or votes. Whether
or not Muslims always or do not always start the riots is not important in
their minds, because people are being provoked-and Muslims are being
deliberately provoked by particular groups in Hindu society. In a word, the
central feature of the police views cited above is that the incidents and the
contexts in which riots occur are narrowed; they are decommunalized and
decontextualized. In a sense, they are even "humanized," by which I mean
that the factors they cite arise not out of prejudice and hatred between the
two communities, but out of the general human or inhuman conditions in
which both Muslims and Hindus in Aligarh live, and by human greed.
 340 I Police Views of Hindu-Muslim Violence

Views from Above

Two further interviews with senior administrative/police officials of the dis-

trict share features with those of the two policemen analyzed in the previous
section, but differ in this one respect of contextualization. One was conducted
with the district magistrate (DM), the senior civilian officer of the district,
the other with the deputy superintendent of police, intelligence (DSP). I will
summarize first the comments of the district magistrate in response to a gen-
eral question concerning the "communal situation" in Aligarh. '4 He remarked
first that the communal situation keeps him busy all the time because of the
fact that any incident involving persons from the two communities could lead
to a communal riot. He gave several examples of the type of situation he has
had to face. But when asked for the "causes" of the communal situation, he
began with the history of the city and of the AMU in the remarks summa-
rized and paraphrased below.

The DM pointed to the history of the city and noted that Muslims had always
been very powerful here. There were riots here before Partition and many
Hindus felt the Muslims should have gone to Pakistan at Partition. In addi-
tion, AMU is a major university of India and Muslims come from all over the
world to study here. The Hindus see AMU as an eyesore, while the Muslims
consider it a question of prestige. Everyone is concerned with their own pres-
tige and dignity, and every small incident is considered to reflect upon them.
A third cause is the Hindu-Muslim balance in the city, which is about 50-50.
But basically it's because of Aligarh's history.

Thus, the DM begins by contextualizing "the communal situation" in rela-

tion to the history of the city and its relationship to the decisive moment in
modem Indian history-not Independence, but Partition-and specifically to
the place of the AMU in that history. Although our local policemen also men-
tioned the AMU, they talked about it in terms of their police work, as an alleged
site for harboring criminals and instigators of riots, not as an originating source
of the communal problem in the city. On the actual mechanics of riots, how-
ever, there was close agreement between the DM's view and that of the local
police officers, as indicated in this summary of his remarks below.

There are people here interested in creating problems, in engineering riots. It

is not for economic benefit, but for prestige. So-called good people plan the
riots, but criminals enact them. Criminals are on their payrolls. The main cul-
 Police Views of Hindu-Muslim Violence /341

prits, he acknowledged, are Navrnan and Manga Ram. Shiv Hari Singhal is
quiet now.

When I pressed the DM concerning whether or not there were similar per-
sons engineering riots among the Muslims, he also gave the names of a cou-
ple of Muslims. However, it is clear that he attributed primary responsibility
to the Hindus whose names were mentioned.
The DM, himself a man from a Scheduled Caste, concluded his analysis
by referring to popular attitudes, to what he called the "communal mental-
ity." "You can't find a 100 percent secular person in India; they are all Hindus
or Muslims. And, in crises, every Hindu is a Hindu, every Muslim a Muslim."
Although the DM, like the lower-level policemen, was, therefore, attributing
responsibility for the communal situation and Hindu-Muslim riots to pop-
ular mentalities or the popular "psyche," there is a notable difference. By call-
ing this popular mentality a "communal mentality,» he was contextualizing
it and generalizing it, placing it in the framework of a polarized population.
He was, in effect, saying that the problem is not the conditions under which
both communities live, but their identities as Hindus and Muslims. The prob-
lem in his mind is the Hindu-Muslim question.
A somewhat similar conceptualization of the communal situation in
Aligarh was presented to me in an interview with the deputy superintendent
of police, intelligence (DSP).15 He, too, stressed what he called "the histori-
cal importance of the city" because of the presence in it of the "AMU and its
minority character." He showed, as of course an intelligence officer would,
knowledge of the internal politics of the university as well as of its relations
with the people in the town.

Muslims, he said, 1lave a peculiar style ofthinking. For them, AMU is an exdu-
sive institution. A section thinks Muslim culture and style of education should
be reflected in the curriculum, but a Marxist group is equally opposed to it.
Under the present vice-chancellor, the atmosphere has changed considerably
and become very secular.
There is an emotional link between Muslims in India and Muslim coun-
tries. In AMU, 99 percent of the foreign students are from Muslim countries.

Although the then vice-chancellor received the praise of the DSP, there is
no doubt that the DSP considered the AMU a central issue in the commu-
nal situation in Aligarh. Muslims have a "peculiar" way of thinking, an exclu-
sive way of thinking, one that links them emotionally with Muslim countries.
 342/ Police Views of Hindu-Muslim Violence

Although he did not say it, this kind of statement is often followed by the
further assertion-usually implied even when it is not stated-that Muslims
lack sufficient emotional ties, loyalty, and patriotism to their own country.
However, the DSP's further remarks on the "mentality of the people" gen-
erally in Aligarh suggest that he may not have felt that way.

There is a tendency of the people of Aligarh to analyze all incidents from a com-
munal angle. So, petty crimes like robbery and stabbing can take a communal
turn. This leads to permanent hostility between Hindus and Muslims and an
atmosphere of mistrust is built up. By and large, the Muslim intelligentsia is
nationalist and secular, but the illiterate and semiliterate classes are totally com-
munal. Respectable members of the two communities are not generally
involved in communal troubles. However, a small section of the elite, a negli-
gible percentage, enlists the support of criminals to gain a political foothold.

The DSP, again like the lower-level policemen, refers to the popular men-
tality, but is more like the DM in defining it as a communal mentality that
has infected the entire Muslim lower class. This, of course, is quite the reverse
of the RSS view articulated by N avman and others, that the trouble lies more
with the intellectual Muslim classes than with the lower classes, who are, on
the contrary, infected by their elites.
From history and popular mentality, the DSP turned to the specifics of
Hindu-Muslim violence in Aligarh. When discussing the actual dynamics of
how riots begin, he referred, like the lower-level policemen, to fabricated pre-
cipitating incidents. He, too, noted that there was "a formidable criminal con-
tent among Muslims." A riot might be started even by Muslim criminals
fighting with each other. He remarked byway of example that, "in case a Mus-
lim kills another Muslim, then a Hindu also is killed to distract the admin-
istration." In other words, a potential riot situation is created to occupy the
district administration in order to distract them from focusing on a case of
murder.
One further factor emphasized by the DSP, but not by other police
officers, was his view that there was "a definite foreign hand involved." When
Indians-police officers or ordinary citi2'.ens-make this remark, they are gen-
erally referring to either the CIA or the lSI (Inter Services Intelligence) of
Pakistan, or to both. This officer's conspiratorial view of the malevolent force
of foreign influence extended, however, to the foreign media. He blamed the
"BBC and other foreign radios" for their "instant reporting of events in
Aligarh," which "only help [edl to turn the situation out of control."16 While
 Police Views of Hindu-Muslim Violence /343
the DSP, intelligence, is supposed to know about such things as the involve-
ment of "foreign hands" in internal disturbances in India, I believe his knowl-
edge was little more than a reflection of the common exaggeration of such
influences in Indian political life, a fantasy whose effect is to reduce Indian
agency and responsibility for the country's own misfortunes. In other words,
it is a further example of a particular aspect of the universal process of blame
displacement that accompanies explanations of riots everywhere.
Insofar as the media are concerned, their role, discussed in the next chap-
ter, in fact has often been to make matters worse. It is especially the case with
the scurrilous vernacular press, but the DSP chose to emphasize the "foreign
radios" instead.
It is noteworthy that in the case of all three senior police and adminis-
trative officers, including the SSP, the DM, and the DSP, rioting and "com-
munal trouble" in Aligarh are placed in broader contexts than those provided
by the inspector and station officers cited above. Although they mention some
of the same factors, the senior officers stressed history, especially the Partition
and attitudes of Hindus towards it, the importance of the AMU as a sym-
bol of difference between the two communities, Indian Muslims' contacts
with foreign countries, and foreign influences in general. In short, they con-
ceptualize, contextualize, and communalize the relations between Hindus
and Muslims in their very speech, even when they are striving to be or to
appear to be impartial. In fact, the very process of contextualizing and com-
munalizing is part of the striving to be or appear impartial: both sides are
to blame. Although they say both sides are to blame, the police at all levels
are also engaged in blame displacement. Muslim criminals, the politicians,
and the media are all implicated. Left out of account, of course, are the police
themselves.
 15 I The Role of the Media

I n the months before Independence and Partition in 1947, the watchful

British governor of the United Provinces, as Uttar Pradesh was then
called, kept reassuring Mountbatten in his fortnightly reports to the
Viceroy that the communal situation in the U.P. was manageable, if diffi-
cult. At the same time, he feared the effects on the communal situation in
U.P. of news reporting of the developing catastrophe in the neighboring state
of Punjab. He was concerned that another chain reaction might develop, such
as had happened earlier after the Great Calcutta Killings of August 1946, which
were followed by terrible riots in Noakhali, Bihar, and in western U.P. itself
at Garhmukteswar, a site holy to Hindus where approximately five hundred
Muslim men, women, and children were massacred. He blamed the Congress
press for this chain of events: "Of one thing I am certain and that is that if
the Congress Press had not been allowed to magnify Noakhali out of all rea-
son, Bihar might never have happened and after Bihar, Garhmuktesar.'"
While the governor's statement falls in the general category of blame dis-
placement, in this case diverting blame from British governance at the end
of its empire to the Congress press, it is certain that the press has also been
deeply implicated in the dynamics of riot production in post-Independence
India. The media, especially the newspapers, play important roles at all stages
in the production of riots, including the planning and rehearsal stages, the
instigation of riotous activity, and the interpretation phase. In recent years,
TV reporting on riots has also come in for some share of blame both for
inciting violence, however inadvertently, and for publication of false and
damaging news in the midst of riots. The official national TV network, Door-
dashan, for example, has been severely criticized for several items of mis-
reporting of tlle Aligarh riots of 1990-91.2
However, it is the considerably diverse array of newspapers in India that

344
 The Role of the Media /345

are most directly and repeatedly implicated in the entire process of riot pro-
duction. For analytical purposes, the press in India needs to be distinguished
by language, political orientation, sectarian identification, and locational
spread} With regard to language, there is a marked difference between the
English-language press and the vernacular press, which, in north India, means
particularly the Hindi and Urdu newspapers. With regard to political orien-
tation, several of the English-language papers as well as the vernacular news-
papers are identified with and may even be owned or directed by particular
political parties. Each of the major political parties and organizations in India
also have their own newspapers, which will not be discussed here since they
are not addressed primarily to or read by the general public. Similarly, the
sectarian papers, such as the RSS organ, the Organiser, and the Jamaat-i-Islami
paper, Radiance, whose viewpoints are predictable, partial, and directed
towards an already committed audience, will not be discussed here. Only those
that are meant for a mass audience will be considered.
In general, the degree of partiality and the use of inflammatory material
that has the effect of provoking one side to retaliatory action against the other
is most extreme among papers that are local, vernacular, and sectarian.4
Indeed, at the height of the Ramjanmabhoomi/Babri Masjid movement and
at the time of the great Aligarh Riots of 1990-91, the PueL report, in the midst
of its extensive comments on the role of the press in sustaining the riotous
momentum there, noted the "almost total communalisation of the Hindi press
in UP."5 At the other end of the spectrum, more balanced, less partial, and
less deliberately inflammatory are the national English-language press as well
as the newspapers produced by the same publishers in Hindi or Urdu as vir-
tual vernacular editions of their English counterparts. While there are
exceptions with regard to both the English-language and the independent
vernacular-language presses, it is generally the case that the English press pres-
ents a more moderate, balanced face in reporting on riots, whereas many of
the vernacular-language papers are partial to one or the other of the two reli-
gious communities (Hindus for the Hindi-language press and Muslims for
the Urdu press) and playa direct inflammatory role in the spread of riots.
Newspapers sometimes playa part in the opening phase of a riot by spread-
ing "news" that originates in the institutionalized riot system network. Thus,
for example, Banerjee notes that "the Jabalpur riots of 1961 ... were sparked
off by the news of a Hindu girl disappearing with a Muslim boy."6 In most
societies, an event such as an elopement, which happens every day, would
not be considered "news" at all. In societies sharply divided ethnically or com-
munally, it may become a matter of general interest, but it is obvious that
 346/ The Role of the Media

reporting of such an event in such societies is inflammatory and that infor-

mation on it must be provided to the press by interested parties seeking to
maintain and spread interethnic or intercommunal tension or even to delib-
erately provoke a riot. Any press reporting of such an incident, therefore, in
the north Indian context, is grounds for suspicion concerning the motives
of the press in publishing it.
But the "news" that appears in the press before and during riots is some-
times not only inappropriate but often completely false. In October 1974 in
Aligarh, when the district administration and police were working assidu-
ously to prevent the outbreak of violence in the aftermath of a quarrel between
Hindu and Muslim students at the AMU, "all important dailies from Delhi,
Agra and Lucknow carried a news item that Hindu girls inside the University
were molested and attempts were made to rape them."7 This news was utterly
fictitious and could have easily been disconfirmed by any conscientious
reporter. At this point in Aligarh, when the institutionalized riot network was
in full operation, the press became a part of it.
The press also becomes complicit in the enactment phase of riot produc-
tion by the attention it gives to statements of the riot's producers. In the midst
of the long series of riots that began in Aligarh in October 1978, during which
the principal militant Hindu figures were active-Navrnan, Shiv Hari Singhal,
B. D. Gupta, and Manga Ram-Balraj Madhok, the most extreme of the
national Jan Sangh leaders at the time, arrived in Aligarh to provide his assis-
tance. He "addressed a press conference," after which "he was quoted saying
that, since Hindus had been prevented from celebrating Diwali, Muslims must
be taught a lesson, and should be prevented from celebrating Id. »8 Here again,
we observe the local riot network in operation, assisted by the arrival on the
scene of a national figure, whose presence draws the attention of the press,
which then in effect carries out the purpose of the visit to Aligarh, namely,
to keep the riotous activity going.
Most press reporting on riots begins after the initial outbreak and is almost
always based upon "police hand outs"9 or press releases issued by the state
government. Most of the English-language newspapers will have access to
the same information at this stage and will produce more or less the same
accounts. Differences will arise only insofar as the particular newspaper has
a stringer at the riot site or an informed person upon whom it relies for infor-
mation. Lead reporters will go to the scene only in the case of larger riots.
When there is a wave of riots at a number of different sites, naturally the lead
reporters will concentrate atthe largest city or the site of the most severe riot-
ing. Even then, while the rioting and curfew are on, reporters remain highly
 The Role of the Media /347
dependent on police handouts, which in turn means, of course, dependence
on the police view of the causes of the outbreak.
Two kinds of biases and misreporting may enter at this outbreak stage.
The first concerns how it started, the second, who started it. Most press reports
will refer to a specific incident, a quarrel between a Hindu and a Muslim over
a trivial matter or a procession that ended in a communal melee. Bias enters
at this stage because of the absence of investigative reporting on the events
that preceded the precipitating incident, which, as I have demonstrated
throughout this volume, involves continuous preparation and rehearsal. In
Aligarh, the 1978 riots provide the clearest case in point. Those riots, precip-
itated during the funerary procession for the wrestler, Bhura, were preceded
by weeks of intergang fighting, Hindu communal mobilization, and direct
incitement to rioting by militant Hindu activists. Only after tlle rioting became
savage and widespread did investigative teams enter tlle scene to provide back-
ground reporting. However, the best such reporting was not done by the press,
but by civil liberties teams.
The second type of bias that occurs at this stage concerns who started the
riots. In many cases, it has been reported in Aligarh and elsewhere that Mus-
lims have begun riots by rushing out from their mosques to attack the police
or by engaging in destruction of property and arson. It is, indeed, sometimes
the case that the "first stone" is thrown by Muslims. Newspapers that report
such facts, however, may fail to mention the provocations that preceded the
disturbances. The best example here has also been mentioned earlier, namely,
the confrontations between Muslim crowds and the police duly reported in
the press, which followed upon the hourly presentation on BBC television
of the destruction of the mosque at Ayodhya that brought Muslims all over
South Asia into the streets and into such confrontations. The overall impres-
sion produced by such reporting in general confirms the prevailing beliefs
that it is always the Muslims who start the riots and the Hindus who suffer
the most from them. 10
These two biases are also linked to the more general problem raised in the
introduction to this volume, of establishing causality in the initiation and
explanation of riots. The question of explanation will be discussed below.
Here, however, it is important to note that there is a parallel between news-
paper reporting on the causes of riots and social science analyses of them.
Both types of analyses often use the term cause to refer to the precipitating
incident. In most cases that do not receive extensive coverage, all that remains
in the press, police, and government records is the precipitating incident,
which stands as the cause for all time to come. In the larger riotous events,
 348/ The Role of the Media

where investigative reporting or in-depth social science analysis may enter,

the distinction may then be made between the precipitating incident-the
proximate cause-and the "real" or underlying cause. It has been a central
methodological argument of this entire volume to demonstrate that this kind
of reporting and social science analysis is misguided l l and fails to perceive
that there is not merely a difference between proximate and underlying causes,
but a dynamic process of riot produ'-iion at work, and, in sites such as Aligarh,
an institutionalized system and network of violence production.
Once the killing begins, a third type of bias may enter, namely, the report-
ing by religion of the principal victims of rioting, the numbers killed, the
property destroyed, and the like. While the local and vernacular newspapers
sometimes do give "lurid details and community-wise breakdown of casual-
ties and loss of property, "!2 the national, English-language press generally does
not do so. Circumlocutions are often used such as references to a "particular
community" having engaged in a particular action. However, even so, it is
usually evident which community is meant by such a reference. The English-
language press, however, does not report deaths by community during riots.
Nevertheless, the English-language press has been by no means free of bias
and has sometimes reported false or misleading news in the midst of rioting. 13
The savage 1980 riots in Moradabad, near Aligarh District, which had their
repercussions in Aligarh as well, marked a turning point in press reporting
and editorial commentary. Although, as usual in post-Independence India,
it was Muslims who suffered most, press reporting, including that from
national and English-language dailies, displayed bias against Muslims.14
Particular references were made several times in previous chapters to the
reporting in the Times of India during the heyday of Girilal Jain, its editor
for many years. The paper did not generally publish false news or misreport
such matters as particular incidents at particular sites, the numbers killed, or
other specific events that could be falsified by other accounts, although it did
even go to that extent on some occasions. I5 More characteristic of Giriial Jain's
leadership, however, was the frequent publication during riots and in their
immediate aftermath of insinuating comments suggesting Muslim respon-
sibility for riots, reports of the presence of persons from Pakistan at the riot
site, reports on the more rapid increase of the Muslim than the Hindu pop-
ulation in the country, articles and editorials deploring the violence and its
consequences for the Muslims while expressing understanding at the same
time of Hindu resentments,16 and the like. Virtually all newspapers in India
today also make specific reference to the Pakistan lSI whenever Hindu-Muslim
riots occur, implying its direct or indirect involvement in them. Among the
 The Role of the Media /349
vernacular and local newspapers, publication of obviously false and inflamma-
tory items is common. The false rumor of the Medical College Hospital mur-
ders in Aligarh has been mentioned several times in this volume, but many
similar kinds of reports have been published in such newspapers at other riot
sites at other times.17
In general, the press in north India is directly involved in the spread of
rumors during riots that aids their perpetrators in recruiting and mobiliz-
ing participants, playing "a critical role in exacerbating tensions between
Hindu and Muslim communities"18 and in prolonging riotous activities. Even
though the national press does not usually resort to open partiality toward
one side or the other, or seek deliberately to exacerbate a developing riot sit-
uation, it does not act as an adequate antidote against those local papers
that do so. Thus, during the Meerut riots of 1982, the national press paid
little attention, whereas "the local Hindi press acting as the mouthpiece of
the RSS, of course, played havoc by publishing inflammatory materials against
minorities." 19
In fairness, however, at the same time it must be noted that some news-
papers and many journalists have from time to time done investigative report-
ing that has defied popular beliefs and established utterly different and more
fulsome and accurate accounts of how major riots have begun, in which
responsibility has also been pinpointed. Such has been the case, for example,
in the reports on the massacres of the Sikhs in Delhi in 1984 during the funeral
of Indira Gandhi, which provided ample evidence of the direct complicity
of "high officials of the Congress"20 and described how they recruited the
killers and provided them with transportation to the sites targeted. It remains
the case, however, that such reporting is exceptional and has not been
sufficient to overcome prevailing beliefs and stereotypes concerning the ori-
gins of Hindu-Muslim riots.
Coming now to the question of explanation and interpretation in the after-
math of riots, here is the decisive moment when the prevailing discourse of
Hindu-Muslim relations and tlle connection of tlle latter to the unity of the
country emerges full-throated. It is also the time for blame displacement and
dispersion. The local vernacular and sectarian press will, of course, blame the
other side and place the riot in the perspective of the hostility of the other and
the other's threat to the country (perceived by militant Hindus) or to secular-
ism (perceived by Muslims). The framing and blaming that comes from the
national English-language press is quite different. It produces and reproduces
the hegemonic elite discourse that is threatened both by Hindu-Muslim riot-
ing and by the communal and venomous subterranean discourses of hatred.
 350/ The Role of the Media

The structure of the elite discourse is substantial1y as follows. There is

indeed a problem of Hindu-Muslim relations that arises from the country's
history since the arrival of the Muslims in South Asia, from the division of
the country in 1947 preceded by the Muslim separatist movement, and from
the animosities that have developed between the two communities in con-
sequence thereof. This situation is, however, seen as deplorable and must be
overcome by statesmanship, secularism, and responsible behavior on the part
of the political parties. Not only in Aligarh, but in the country as a whole,
the AMU occupies a central symbolic place in this discourse. As a symbol
standing in for the Muslim community as a whole, therefore, it must disown
its alleged separatist past and declare and practice secularism, failing which
its governance must be restructured to ensure it does so. Muslims in general
must integrate fully into the mainstream of the country by abandoning their
attachment to their own Personal Law and accepting a uniform civil code.
If the leadership of the country, the political parties, the communal
organizations such as the RSS and the Jamaat-i-Islami, and the students and
faculty at the AMU are allowed to play upon the existing Hindu-Muslim prej-
udices, then it is inevitable-in the discourse of Hindu-Muslim difference-
that these prejudices will erupt in a communal conflagration. The metaphor
of fire is commonly used by the press, as well as by the politicians and by the
press when choosing its quotes from the politicians. l l Fires also may spread,
leading to conflagrations in other towns and districts. Another metaphor used
by the press is "the communal virus" that, of course, also may spread through
"contagion" from one place to another. In the early years after Independence
and during the decade-long attempt to transform the character of the AMU
from a predominantly Muslim minority institution to a more "secular" one,
elements in both the right and the left wing in the press and elsewhere saw it
as a source of this virus. Once again, the Times of India offers the clearest
example. Referring to the 1956 riots in Aligarh, that paper blamed the AMU
students for the outbreal< there and for its spread to other parts of the state.
At the same time, it conflated the problems at the university with the atti-
tudes of Muslims in the country as a whole; moving back and forth between
the AMU and the alleged communalism of its students and teachers, it raised
questions about Muslim loyalty to the country and Muslim sympathies for
Pakistan. 22
The second frame in the discourse of explanation for Hindu-Muslim riots
is interparty competition, which has also been identified in this book as a cen-
tral element in the production of such riots. However, the emphasis in the
press is different from the one presented here. I have shown that there is a
 The Role of the Media 1351

specific relationship between the production of riots and the communalization

and polarization of the electorate and suggested that the targeting of partic-
ular sites in the town of Aligarh is designed both to produce that effect and
to punish Muslims whose voting patterns either prevent the BJP from achiev-
ing dominance or endanger it. While the press sometimes provides evidence
to support these statements, when the time comes for explanation in edito-
rial and other commentary, it is usually the appeals of the parties to voters
by caste or community that are condemned, rarely the deliberate involvement
of specific political parties in riot production for their own advantage. By their
appeals, it is implied, the parties contribute to the acting out on the part of
the public of their animosities and hatreds.
Within both frames, therefore, that of the history of intercommunal rela-
tions and that of contemporary interparty conflict-which in turn are closely
linked to each other in the discourse itself-blame is dispersed in press report-
ing and commentary. It lands on a general target, such as the communal prej-
udices of the population as a whole or the attitudes of Muslims in general,
or it lands nowhere in particular.
 PART VI

Conclusion
 16/ The Persistence of Hindu-Muslim Violence
The Dynamics of Riot Production

~
. ots persist in India and have become. endemic at a multiplicity of sites
in the subcontinent. They constitute, in effect, a normal, routine aspect
f politics whose very normality and routine character are masked by
both the sincere and hypocritical comments that follow in the aftermath of
their most savage occurrences. So long as they are considered abnormal, excep-
tional, expressions of a disease that occasionally afflicts the polity, a(,:ts com-
mitted by the dregs of society drawn from the slums they inhabit, so long will
their commonness remain hidden from view. So long, then, will the political
elites, the educated, the upper-caste intellectuals, the editors ofleading news-
papers be able to reassure themselves that they live still in a basically peace-
fulland, in the world's largest democracy where such aberrations are bound
to occur in the process of India's advance from backwardness to modernity.
On the more hypocritical side lie the simplistic explanations offered by
the leaders of competing political parties, namely, that it was the other party
that was responsible. Among some parties such as the B]P, the hypocrisy rises
to an even greater height, where the national leaders of the party go to the
extent of acknowledging that riots have been promoted at the local level by
elements from their own party, and that, as 1. K. Advani said to me years ago,
"we know about" these people, implying that they would cleanse them from
the lower levels of the party as soon as practicable. The height of that
hypocrisy was, of course, reached at Ayodhya when the national leaders of
the entire RSS family, its middle-level leaders from the districts, and many
tens of thousands of their followers descended upon Ayodhya to destroy the
mosque, after which some feigned surprise and blamed excessive zeal on some
of their followers, when in fact the evidence of preplanning was over-
whelming and the personal satisfaction of all those I saw and interviewed after
the fact can only be described as joyous.

355
 356/ The Persistence of Hindu-Muslim Violence

Riots, therefore, first and foremost persist because they are unacknowl-
edged and illegitimate but well-known and accepted transgressions of rou-
tine political behavior in India. Moreover, the types of riots discussed in this
book are only one form in the repertoire of collective violent enactments that
exist widely in India, including the engagement of the Indian state in pro-
tracted civil warfare with insurrectionary groups, intracommunal sectarian
violence, intercaste riots, the spread of mafia violence and criminality in the
rural areas of north India, intergang killings, police-public confrontations of
all sorts,' and the everyday violence everywhere in India of police against cit-
izens. In one sense, therefore, Hindu-Muslim riots are but one form among
many types of persisting violence in India. In another sense, however, such
riots have a special status in that country because of the degree to which they
go to the root of the identities of the two largest categories of its population
and of that of the South Asian subcontinent as a whole, identities that are
still being formed in the very crucible of violent conflicts. In this sense, they
are linked with the two other much more violent forms of Hindu-Muslim
confrontation in South Asia, namely, the war between Muslim separatists and
the Indian state in Kashmir, and the perpetual hot-and-cold war between India
and Pakistan. Indeed, in the minds of militant Hindus especially, all three of
these sites of violence are marked upon a single grid of Hindu-Muslim con-
frontation for supremacy in the South Asian subcontinent. In their minds,
the everyday riots in towns such as Aligarh and elsewhere are part of a sin-
gle, much larger struggle.
But Hindu-Muslim riots have a distinct dynamic that shares some features
with, but differs more substantially from, insurrectionary warfare between
states and citizens or outright war between sovereign states. Those differences
cannot be spelled out here, but they can be summarized in a nutshell as
revolving around axes of unanticipated/anticipated, spontaneous/planned,
illegitimate/legitimate. Riots, not entirely unlike insurrections and external
wars, are said to break out either unexpectedly or as a consequence of a buildup
of tensions that mayor may not explode under fortuitous circumstances.
Unlike most insurrections and wars, however, they are said to be primarily
spontaneous rather than planned. They are also acknowledged by all, includ-
ing the perpetrators, as illegitimate. All who justify the violence of their own
side in riots allege that they acted only in retaliation, the aggressor's synonym
for self-defense. The whole purpose of this book has been to demonstrate the
falseness of these distinctions.
It is not my purpose to argue that there is a family resemblance between
Hindu-Muslim riots and the civil war in Kashmir or the wars between India
 The Persistence of Hindu-Muslim Violence /357

and Pakistan. Rather, these riots are another form of that conflict enacted
and reenacted in minor fracases and disputes every day in the mixed mohal-
las of towns such as Aligarh. These riots have a form, a sequencing, and a
dynamic of their own that must be masked in a self-proclaimed democracy
in ways that are not required in civil and international wars. I have tried to
penetrate the dynamic process in this book. I want to summarize here some
aspects of that dynamic, to demonstrate the functional utility of riots, and
to conclude by indicating the embeddedness of the justifications and expla-
nations for Hindu-Muslim riots in the discourse of communalism, Hindu
nationalism, and the different historical consciousnesses that reveal them-
selves in the attitudes of Hindu and Muslim protagonists.

PREPARING THE GROUND FOR RIOTS

Communal Tension
Most riots are anticipated. They are preceded by a period that is usually
described as marked by tension. Everyone knows when tension is in the air,
from ordinary citizens to the authorities of the district and, if the situation
appears to be very threatening, to the state authorities as well. The term ten-
sion has both specific and vague connotations in the English language. 2 It
implies tautness, as in a wire, strained to the utmost. It also refers specifically
to "strained relations between persons or groups" and to "uneasy suspense."3
In Aligarh, secondary accounts of the genesis of riots, and my respondents
as well, often referred to the communal tension that preceded Hindu-Muslim
riots. Such tension was sometimes said to have been precipitated by a specific
incident, sometimes by a series of incidents that led to a building up of ten-
sion over a longer period.
Several perspectives emerge from documentary and interview sources on
the relationship between communal tension and riots in Aligarh. At one
extreme is the view that communal tension is a palpable and pervasive ele-
ment in the life of the city. It is a kind of smoldering fire that can erupt into
flames over any kind of incident, however trivial, in which actions are taken
by members of one community that offend or harm a member or members
of the other community. This perspective is embedded in the view commonly
held throughout the world that riots arise out of spontaneous feelings of pas-
sion in societies where interethnic prejudices and hostilities are endemic and
have a long history.
At the other extreme is the view that communal tension and the riots
that follow in its aftermath are deliberate creations of avaricious business-
 3581 The Persistence of Hindu-Muslim Violence

men and politicians, among other types of people sometimes mentioned-

criminals and hooligans, for example. This view is often associated with the
idea that riots are not spontaneous occurrences, but deliberately planned
actions by individuals, groups, and organizations. The creation of commu-
nal tension is part of the process. Incidents are provoked, processions are
taken out, mass mobilizations are organized deliberately to offend the sen-
sibilities of the other community's members. Economic and political advan-
tages are gained both from the tension and from the violence that sometimes
follows such actions.
The perspective that seems most consistent with the results of my inves-
tigations over the years in Aligarh and elsewhere is that there are elements
both of spontaneity and planning that contribute to the creation and per-
sistence of communal tension and its development into riots. It is not that
the two extreme views are completely wrong, but that they miss the action
that takes place in intermediate space and time, between the prejudices that
exist in Aligarh society and in every multiethnic, multicultural society in the
world, on the one hand, and the undeniable element of purposive, planned
action that accompanies every riot, on the other hand. Riots do not occur
accidentally like fires from a smoldering flame nor are they meticulously
planned and coordinated from beginning to end. Rather, they are dramatic
productions, street theater performances that are meant to appear spontane-
ous, but that involve many people in a variety of roles and actions that include
inciting the interest of the audience, the dramatization and enlargement of
incidents into a fit subject for a performance, and, finally, the production of
the event.
The people of Aligarh, like devotees of theater, are kept in a state of readi-
ness for the next production through advertising of all kinds of trivial inci-
dents that hold the promise of a great drama to follow. It is never certain
whether or when the drama is going to be produced, because the production
depends on propitious circumstances, especially political circumstances such
as mass mobilizations and elections, anticipated or in progress. But the audi-
ence's interests and appetite must not be allowed to flag in the intermediate
time between productions, for each production involves audience partici-
pation. There are many bit players who must be kept ready for action and
myriad walk-on parts to be played.4 The bit players are the criminals and hooli-
gans brought in when the circumstances are propitious to start an action. The
walk-on parts are played by those people in the mohallas who are always ready
to come out for a fracas, to whom the word must be spread that an action, a
production, is about to take place. All the events that take place in interme-
 The Persistence of Hindu-Muslim Violence /359
diate space and time to maintain a state of communal tension or to increase
it when it flags are rehearsals for the production of riots.

Rumors

Virtually everyone who has written about riots has given a central place to
rumors. In fact, tensions and rumors go together. 5 Rumors keep tensions alive
in areas where riots are endemic and are an integral part of the buildup of
tension that takes place before a riot starts. During riots, rumors are used to
sustain "the momentum of violence."6 They keep "the crowds in an excited,
potential mob state."7
In most accounts, rumors are seen as the "fuel" that sets aflame the com-
bustible material of interethnic or intercommunal hostilities.s This associa-
tion between rumor and riot is also critical to the predominant view of riots
as events that arise spontaneously out of intergroup animosities. 9 Everyone
seems to agree as well concerning the function of rumors in the fomenting
of riots, namely, to communicate rapidly. Such rapid communication is espe-
cially effective when large crowds are gathered for recreational or other pur-
poses,lO but they can also "circulate at high velocity by word of mouth"ll
even without the availability of large, massed audiences. In fact, however, as
with everything else about this approach, the association between rumors and
riots gives a false impression. In this case, the false impression is that rumors
somehow appear, spread like wildfire, arouse the passions of a community,
move people into the streets into crowds that then join forces, massing into
larger and larger groups as they assemble and march towards the quarters
inhabited by other ethnic groups to retaliate for the atrocity or atrocities the
news of which has been spread by the rumors.
Among writers on riots, few have challenged this simplistic notion. Keith
is one who has done so. For one thing, he has noted, "rumours of imminent
trouble abound" in local communities when news of a riot elsewhere is spread
by the media, but, in fact, nothing at all happens in most communities, in
which the rumors fly as fast as in the areas where violence does in fact take
place.12 Although he notes that "many parts of London ... were ... pump
primed" during the 1981 disturbances by stories emanating from "the local
gossipmongers and rumour-hawkers who constructed folk discussion in the
metaphor of contagion," in most such areas there was no violence. In Keith's
view, this nonresponse to rumors is "one of the most powerful arguments
against the classification of collective violence as irrational."13 If one agrees
with Keith, as I do, then what are we to make of the universal presence of
 360/ The Persistence of Hindu-Muslim Violence

rumors that precede riots everywhere in the world and everywhere in his-
tory? If there was ever a case for a causal theory and a universal law, then
surely it lies in the role of rumors as the fuel for riots.
In Aligarh, rumors have been present before, during, and after every riot
in its history. Moreover, Pars Ram and Gardner Murphy's UNESCO study of
communal tensions in Aligarh, for which the research was done shortly after
Partition, noted that rumors abounded in the city as part of the daily life and
fears of Muslims who had been transformed by Partition into "a very inse-
cure minority group." The rumors that haunted them were that "their shops
[would1be looted and disappear at night," that they would be attacked as "cow-
killers," or that there would be "new killings" during the Hindu festival of Holi.l4
As in the case of the Russian pogroms of 1881, rumors in Aligarh in 1951 even
assigned specific dates for the beginning of the next Hindu-Muslim rioting. 15
However, rumors that reflect the fears of minority groups are not the kind
that playa role in mobilization of crowds for violent action. These are the
buzzing rumors of the marketplace. The more ominous rumors that signal
the beginnings of mobilization and presage violence are those that report on
an attack of some sort by a person of one community upon another, alleged
to have taken place somewhere in the city. A rumor, after all, is "unverified
information of uncertain origin."16 Most such rumors in Aligarh and else-
where arise out of actual or concocted incidents, often "minor incidents of
a purely private character between two individuals or two groups (one Hindu
and the other Muslim)" said to have occurred "in an out of the way place
and at an unfrequented spot by a person whose identity had not been estab-
lished by his victim."l7 But these types of rumors also often fly fast and loose
and nothing at all happens in their aft.ermath.
Yet it is also the case that an increase in the density of rumors precedes
and accompanies the start of violence. Such was the case before and during
the long series of riots that began with the stabbing of the Hindu wrestler,
Bhura, in 1978. This stabbing incident itself became the basis for a rumor that
it was an act of retaliation by Muslims for the defeat of their group of wrestlers
by the Hindu wrestlers. But other rumors soon followed, for example, that
a Hindu sweet-seller in the Babri Mandi (market) had also been stabbed by
a Muslim. Is In such times, any private quarrel may be used to provide a ker-
nel of fact to give credence to a rumor. Thus, a quarrel between two Muslims
over a money debt in Phul Chauraha led to a stabbing in November 1978, but
the rumor that was spread was that "members of different communities had
been stabbed."19
But the rumor of rumors in the history of Hindu-Muslim riots in Aligarh
 The Persistence of Hindu-Muslim Violence /361

was the totally false and concocted story of the AMU Medical Hospital mur-
ders. The PueL report on this matter, cited in detail above, characterized
this rumor as "motivated and orchestrated propaganda against the Medical
College Hospital."2o Let us recall also the salient features of this rumor, the
persons and groups with whom it was associated, and the target towards which
it was directed. It was deliberately and maliciously published in well-known
and respected Hindi newspapers; it was spread further, locally, by known per-
sons such as the Agarwal brothers, cloth merchants, and members of the RSS-
dominated Vyapar Mandal; and it was directed at the central symbol of the
Muslim community in India, the AMU. Moreover, it was accompanied by
mobilization of crowds under the leadership of known persons and groups,
including Krishna Kumar Navman and the RSS. Finally, it occurred in the
midst of rioting that had already begun. It was, therefore, an instrument in
the pursuit of violence used by known persons and groups and directed at a
specific target.
There is, therefore, little or nothing that is either arbitrary or spontaneous
about the occurrence of this rumor and others of its type. On the contrary,
as exemplified especially in the case of the AMU, certain sites become peren-
nial sources of rumors that have a specific purpose. Rumors serve the pur-
pose of mobilizing members of a community for attack or defense. Rumors
that affect the AMU generally serve the purpose of mobilizing Hindus in the
city for attack upon that living lieu de memoire of violence, confrontation, and
the Partition that many Hindus refer to as the vivisection of India. Moreover,
the rumors that revolve around AMU are repetitive: the same stories appear
over and over at times of riot.
In order, therefore, to clarify the specific relationship between riots and
rumors, I think it is necessary to distinguish between what Keith calls mere
"gossipmongers," whose rumormongering may have no effect in the mobi-
lization of crowds for violent action, and those who playa specialized role in
the spreading of rumors,21 whose activities are integral to the dynamic
process of riot-production. The latter rumors are not randomly articulated
from some general pool of blood libel accusations and are not directed ran-
domly either. In local situations such as Aligarh, rumors that are designed to
mobilize violent crowds will pinpoint a specific incident-which may be true,
false, or exaggerated-in a specific mohalla or at the AMU to arouse crowds
either to move to that place to retaliate or to take vengeance against those
seen as the blood brothers of the perpetrators of a particular atrocity. They
provide excuses for violent action that also is not random, that targets mem-
bers of a particular group in particular areas or all the members of a partic-
 362 / The Persistence of Hindu-Muslim Violence

ular group identified by lists provided to the crowds. Those screaming for
blood and revenge in the crowd are making use of slogans provided to them
as a justification for actions that serve either their interests or those of their
political organizations, or for which they are paid, partly in cash and partly
by the loot they gain-under the cover provided by the crowds so massed,
by the justification given for the violence, and by the near certainty that they
will escape prosecution.
If effective rumormongering is a specialized task in the process of riot pro-
duction, so is the squelching of rumors an important task in the containment
of riots. Suppression or denial of rumors may be undertaken by the author-
ities, the press, or responsible spokesmen from communities in potential
conflict who deliberately take on this role. 22 But it would be a mistake to imag-
ine that suppressive activities can be effective in situations where an institu-
tionalized riot system is in operation, for in such situations, the rumors are
a signal for action designed to bring out local party activists with their hench-
men and gangs who are always ready to come out for pay and loot. It is not
administrative or media action to suppress rumors that is critical, but swift
administrative and police action to break up the crowds that form and to
impose an impartial and effective curfew. If there is no such action, then the
opportunities for loot and vengeance under the cover of rumors will encour-
age others to leave their homes and join the crowds. However, it is certainly
the case that, in situations of high tension and the wild spreading of mali-
cious rumors, most people are likely to prefer to close their shops, shut their
windows, and seek to stay out of trouble. Those who participate in riots are
not aroused merely by rumors, but by the organizational and mobilizing activ-
ities of practitioners skilled in the production of crowds and riots and by the
opportunities made available for anonymous violent, destructive, and thiev-
ing actions in the crowds that are massed.
Rumors are a sign, a means of communication, a method of mobilizing
an action and, sometimes, of stopping an anticipated action by the other side,
as well as a tool used by politicians to gain support from a group that feels
threatened by the possibility of violence. Rumors are first of all signs of the
existence of a serious and potentially violent dispute. Members of opposed
sides are attuned to the slightest move attributed to the other side toward a
possible change in the status quo. Second, they warn one side of the possi-
bility of an action by the other. Third, they serve to mobilize the side so warned.
The function of this mobilization may be to attack or to demonstrate strength
so as to prevent the rumored disruption of the status quo.
In areas where riots are endemic, rumors are routine and so are the mobi-
 The Persistence of Hindu-Muslim Violence /363

lizations that follow them. The mobilizations that take place in such cir-
cumstances are of the type that give the impression of spontaneity, but it would
be wiser to consider them routinization of a practice of confrontation for the
sake of defense. Rumors sustain the tension that precipitates an unending
stream of rumors that act not simply as the means of massing mindless crowds,
but as the weapons of attack and defense in communities used to the per-
petual threat and danger of violence to members who might be caught alone,
isolated, and vulnerable to violence from members of the other community.
Such massing may also be a form of defense against anticipated police
action. 23
What then are we to make of the statement by Veena Das that "there is
no contradiction between the fact that, on the one hand, mob violence may
be highly organized and crowds provided with such instruments as voters'
lists or combustible powers, and on the other that crowds draw upon repos-
itories of unconscious images" reflected in the fast-flying rumors "that
crowds use to define themselves and their victims"? 24 Rumors certainly do
have a function of arousal. They are most effective when they draw upon
images, prejudices, and myths derived from family upbringing, socialization
in community schools, or in school textbooks in government schools, that
provide a distorted history, and political indoctrination. All these elements
function within the discourse of Hindu-Muslim difference and antagonism
that pervades contemporary Indian society, especially in northern and west-
ern India. But that discourse operates primarily as a source for explanation
and blame displacement, for the maintenance of a set of power relations in
Indian society that operates to the advantage of upper-caste Hindus and to
the disadvantage of Muslims , many among the backward castes, and the lower
castes. It justifies violent action against Muslims, explains how communal
antagonisms persist, and suggests how the predominant groups and politi-
cal organizations maintain or gain power, but it does not explain the specific
actions of violent crowds.
Rumors may reflect deep psychological fears and animosities, but too
much attention to that aspect of rumors can mislead us and has misled
countless observers of riots concerning their function. Moreover, such an
approach draws us too close to the type of explanation that focuses on mass
hatred and animosities as an explanation for riots. Rumors decidedly have
specific roles in riots and playa major role in mobilization for attack and
defense. They are important instruments in the politics of violence to which
the authorities must pay attention, but it is critical to distinguish the ordi-
nary rumors that abound in daily life, which may provide source material
 364/ The Persistence of Hindu-Muslim Violence

for psychologists and psychoanalysts, from those that have specific func-
tions that mayor may not arouse persecutory and demonizing fantasies. It
is the latter that are the material for the political and sociological analysis
of riot production.
The actions of violent crowds and the explanations for their actions con-
stitute separate, though interrelated, spheres. They depend on each other, but
they constitute different arenas. How rumors have been treated in the liter-
ature on riots is an example of the confusion that may obfuscate the difference
between the spheres. The psychological analysis of rumors belongs more in
the sphere of explanation of riots by means of blame displacement, the func-
tional analysis in the sphere of explanation of riot production as social action.
When the differences are obfuscated and the two types of explanation become
complementary, as they have been in contemporary India, they act together
to perpetuate riotous violence. When they diverge, that is, when the expla-
nations for riots reject the discursive framework into which they have been
placed and which sustains them, and instead offer alternative explanations
and ameliorative measures that replace the existing hegemonic discourse, in
this case the discourse of Hindu-Muslim antagonism, then riotous violence
may subside or take other forms.

Provocation

It is common in causal analysis of riots to make a distinction between pre-

cipitating and underlying or "real" causes. The precipitating incidents, such
as a police action against a Black citizen in an American city, are often con-
sidered to be fortuitous circumstances that may precipitate a riot because of
the simmering rage that exists among Black people against police harassment,
harm, and killing of Blacks in the past. The incident then acts as a trigger that
sets off a spontaneous conflagration arising out of that rage, the root cause.
Often, however, riots are produced as a consequence of direct provoca-
tion, of which there are myriad forms in the Indian repertoire of riot pro-
duction. Probably the most common is a universal form, the procession, which
is enacted sometimes in a fashion that has clear parallels in other societies
and cultures, sometimes with a distinctively Indian cultural flavor. Such pro-
cessions have a long history in Indian demonstrations, both those that have
been deliberately nonviolent and those that have been deliberately provoca-
tive and inciting towards violence.
The procession mayor may not be a deliberate provocation. Sometimes
it is an assertion of a right to use of public space, sometimes an assertion of
 The Persistence of Hindu-Muslim Violence /365

the strength of the group whose members have taken it out, but there are
certain types of processions that are deliberately designed to incite. In the
latter category are processions of one community that deliberately pass
through localities inhabited by another, that stop before religious places of
the other community to play music or hurl insults, that in other ways inter-
fere with the peace or even the property of the other community. Especially
provocative and/or intimidating are processions that involve displays of arms,
including in the Indian context lath is (bamboo sticks) and swords.
Among the distinctively Indian processions that have often provoked inter-
communal violence are the Ram Lila processions that occur every year in most
Hindu communities throughout north India in celebration of the god Ram,
a celebration that is also associated with actual dramatic re-creations of Ram's
life. Most such processions do not lead to Hindu-Muslim confrontations, but
many of them have done so and many have been used deliberately to pro-
voke Muslims by passing through their localities. Such religious processions
and celebrations have for over a hundred years in northern and western India
been susceptible to takeover by political leaders and groups for mobilization
of Hindus for nationalist purposes, that is, against British rule, but also to
provoke Muslims. In north India, the god Ram himself has become the polit-
ical emblem of the RSS family of organizations and Ram Lila and Ram Navami
processions are often led by or turned into provocative displays by militant
Hindu organizations, including the VHP and the Bajrang Dal. On such occa-
sions, competitive rallies of protest against or support for such processions
and the particular routes they follow may also occur. The ultimate proces-
sion for Ram was the rath yatra of L. K. Advani that left many hundreds of
Muslims dead in its wake.
Especially provocative are funerary processions in which the dead body
of a member of one community allegedly killed by a person from the other
community is carried through localities containing large concentrations of
the other. This tactic also appears to be a recent addition to the repertoire of
riot-provoking actions used by militant Hindu organizations. I have described
one such in Kanpur in December 1992. The outstanding example from
Aligarh was the funeral procession of the wrestler Bhura that precipitated
the first in the very long series of riots that began in October 1978. The funeral
processions in Kanpur and Aligarh shared the following features: forcible
removal of the dead body from the hospital by a mob, which then formed a
procession that moved through localities heavily populated by Muslims, the
shouting of provocative slogans along the way,25leading finally to the signals
and actions marking the beginning of a riot, namely, stone throwing, arson,
 366/ The Persistence of Hindu-Muslim Violence

the sound of gunshots, and the first killing(s). In both cases, also, RSS and
BJP activists were present throughout.
We have seen yet other types of actions that are alleged to be part of the
repertoire of riot provocations. The most grisly cited herein consists of the
deliberate killing of a person of a particular community, his dismemberment,
and the tossing of his body parts into a gutter to be found by persons from
one of the comm unities in order to incite a so-called retaliatory action against
the other community. I have no data that would indicate how often riots of
any scale have been attributed to such actions. What needs to be noted about
this type of action, however, is that it implies a deliberate action designed at
least to maintain intercommunal animosities at a high level or to deliberately
provoke a riot. Such an action, therefore, does contrast with other forms such
as processions, which may be designed only to assert a right, demonstrate
strength, and intimidate, but not necessarily and always to provoke a major
confrontation.

WHY DO RIOTS PERSIST?

Hindu-Muslim communal riots have been an integral part of the political

process in modern India since the 1920S. Although, since then, there has
never been an extended period of time when Hindu-Muslim riots have not
occurred somewhere in India, there have been times when they have occurred
in waves or chains that have covered large parts of the country, in the post-
Independence period notably during Partition and before and after the great
militant Hindu mobilization in the late 19805 that persisted until the destruc-
tion of the mosque at Ayodhya on December 6,1992. The history of post-
Independence riots in Aligarh has provided illustration at a single site of riots
that have been confined to the city with no apparent connection to events else-
where, those that have occurred more or less simultaneously with riots in adja-
cent areas under similar circumstances, and those that have occurred in the
midst of a great wave. In effect, Aligarh can stand for India as whole in impor-
tant respects. It is a site in which Hindu-Muslim riots may occur as a con-
sequence of local quarrels, issues, and confrontations between members of
the two communities; as a routine part of the political and electoral process
framed within the local political context; or as a part of processes of wide polit-
ical mobilization in the larger political context of the state or the country as
a whole.
One answer, therefore, to the question why riots persist in Aligarh, U.P.,
and India is simply because, despite all protestations to the contrary, they are
 The Persistence of Hindu-Muslim Violence /367

part of the general armory of weapons used by activists and interested par-
ties within both communities for personal, local, and political advantage and
by political activists and leaders to forge solidarity within their respective com-
munities in order to defeat their opponents in elections. As such, therefore,
there is in a sense nothing more to explain about riots than there is about
why there are quarrels, conflicts, and violence in general in Aligarh or else-
where in India, and in very large parts of the rest of the world, for that mat-
ter. Riots are simply there and they are there for as many reasons as there are
for nonviolent conflicts. Moreover, there are other kinds of riots that occur
regularly in India among Muslims and Hindus, notably Shia-Sunni riots
among Muslims and intercaste riots among Hindus. Once one accepts that
there are certain frameworks within which both violent and nonviolent con-
testation occurs within a particular country, the explanatory problem appears
almost to dissolve. The issue then becomes how, not why, riots take place within
a particular frame, that is, it becomes a processual issue bearing the forms
just described, more than an issue requiring causal explanation.
Still, there is variation as well as persistence in the occurrence of riots in
Aligarh, U.P., and India within the Hindu-Muslim framework. They do not
occur with the same frequency as other events, particularly scheduled events
such as elections, with which they are often associated. I have marshalled suf-
ficient evidence, I hope, to demonstrate the latter association, particularly to
show that large-scale communal riots are often staged events whose effect, if
not their deliberate intention, is to produce communal solidarity to gain elec-
toral advantage in a political context in which no other stratagem would work
so well. I want to stress once again here that, for the most part, it is riots that
produce solidarity, not electoral politics that produce riots. Those who argue
that it is "democracy," electoral participation, the spontaneous enactment of
"ancient hatreds," or the popular animosities of the masses that cause riots
are undermining the foundations of competitive political processes. When
one recognizes that it can work the other way and, on my evidence, works
that way more often than not, it takes the burden off the masses, the elec-
toral process, and "democracy," because authoritarian regimes also produce
not only riots, but massacres and genocides directed at other ethnic, tribal,
or religious groups when it is politically convenient to do so.
But I have also shown that there is variation in the temporal and spatial
occurrence of rioting within Aligarh as well as in U.P. and India as a whole.
I have examined all the causal factors that have been put forward in the lit-
erature on which I could provide evidence: personal, demographic, economic,
and political. I have argued that particular personal, demographic, and eco-
 368/ The Persistence of Hindu-Muslim Violence

nomic factors play their parts in particular riots. I have also argued that all
these factors come into play simultaneously in the larger riots. I have fur-
ther established that the larger riots always occur within a context of polit-
ical competition that they serve to intensify. Most important, I have argued
that these latter riots, especially, are productions that involve constant
rehearsal, careful timing, and specific targets of attack. From this perspec-
tive, most causal explanations of rioting of the type examined in this book,
namely, interreligious or interracial or interethnic, appear to me to be flawed
from the start. They have turned the independent variable into a dependent
variable. They have sought to explain riots rather than to understand the pur-
poses and effects of riots in interpersonal, economic, intercommunal, and
political relations.
From this latter perspective, much that has appeared mysterious and elu-
sive appears almost unproblematic. Is it surprising that, under the cover of
large-scale rioting, individuals loot and rape and take vengeance on their ene-
mies? Certainly not, but it is, on the other hand, ludicrous to consider that
riots are caused by the desire for loot, rape, and vengeance. Is it surprising
that Hindu-Muslim riots take place at sites where there are large Hindu and
Muslim populations living side by side or in close proximity? Of course not,
but tautological explanations for rioting have been common in the literature
on Hindu-Muslim rioting in India and Black-white rioting in the U.S., pre-
senting as causal findings statistical or impressionistic associations between
the relative sizes of these two types of opposed groups living in juxtaposition
to each other. Is it surprising that riots occur when authority is weak and vac-
illating and that they occur more rarely or not at all when authority is strong
and determined to suppress them? Hardly, but such associations turned into
causal explanations have provided justifications for authoritarian solutions
to perceived social problems.
Is it surprising that riots occur in filthy slums in India and Black ghettos
in the United States? Certainly not, but are the slums and the ghettos the causes
of rioting? The answer in the literature is likely to be that they are, that the
one produces economic distress that manifests itself in rioting, that the other
arises from the discrimination that produces and polices the ghettos. But, of
course, this does not work either, because not all slums and ghettos are cen-
ters of riot production. Yet, I have myself argued that slums and ghettos pro-
vide fertile grounds for riot production, not because of economic distress or
discrimination in the abstract but for two other reasons: first, they provide
a recruiting ground for specialists in crime and violence; second, commu-
nities within some slums and ghettos organize for attack and/or defense
 The Persistence of Hindu-Muslim Violence 1369

against the other community. The explanation may sound similar, but it is
not; the reversal of terms is significant because it introduces agency, purpose,
and practice.
The explanation provided for riots in this volume might bettter be
described as purposive rather than causal. I say that riots are dramatic pro-
ductions, creations ofspecific persons, groups, and parties operating through insti-
tutionalized riot networks within a discursive framework of Hindu-Muslim
communal opposition and antagonism that in tum produces specific forms of
political practice that make riots integral to the political process. Further, they
generate post hoc interpretations, analyses, and explanations that are in no
way scientific or adequate to yield satisfying causal statements, but rather them-
selves contribute to the persistence of riots. lust as drama cannot persist with-
out drama critics who, through the media, provide audiences for some types
of productions and not others, so Hindu-Muslim riots cannot persist with-
out journalistic and academic interpretations and explanations of them that
focus our gaze upon them and treat them as social problems rather than as
intolerable violent productions. However, there is a difference here between
the production of a dramatic play and the production of collective violence
because, in the latter case, the violence is treated as evidence of a social prob-
lem for which there is no feasible solution since our gaze is directed away
from the actors in the play to the nebulous realm of the causes of the exis-
tence of the drama.
It has proven difficult to answer the question, "What causes riots?" I have
argued that, in effect, it is a misdirected question. It should rather be asked,
who-individuals, organizations, groups-produces riots, how and when do
they produce them, and how is our attention diverted from questions that
could be answered to questions that cannot? But then, it is not sufficient to
say that riots are produced by riot-mongers, even if we elaborate and pro-
vide details on the processual dynamics of riot production. That would be
as if Robert Merton ended his analysis of machine politics by describing the
ways in which the machines were built by the big party bosses and how they
functioned. But Merton did not do that. In fact, the dynamics of machine
politics were well known when he wrote, so he asked a further question,
namely, how does one explain their persistence?
I want, therefore, now to present a set of more precise conclusions con-
cerning the reasons for the persistence of Hindu-Muslim riots under condi-
tions that obtain in India today. Those conclusions come under four headings:
the functional utility of riots, the role of the state, the operations of institu-
tionalized riot systems, and the role of contextualization.
 370 / The Persistence of Hindu-Muslim Violence

The Functional Utility of Riots

Riots in Aligarh, as elsewhere in U.P., provide immediate material benefits

for many persons. The beneficiaries mentioned over and over again in inter-
views include criminals, avaricious local businessmen and profiteers, and local
politicians who gain political advantage by taking the role of protectors of
the people of a particular community. Many explanations of the causes of
riots in Aligarh have been provided in which one or more of these groups
play critical roles. Marxists and other persons on the left, especially, argue
that businessmen pay criminals to start riots from which both groups gain
in precise ways.26 Criminals get not only the money from businessmen, but
the loot that becomes available during the course of a riot in which they may
act with impunity and freedom from fear of arrest. 27 Businessmen gain advan-
tage from the damage or destruction done to the property of their rivals. But
virtually all persons interviewed, of whatever ideological persuasion, give a
central place to politicians for fomenting riots in order to win the support of
members of their own community by causing trouble for the other community
and protecting or providing relief to their own.
One does not have to decide, however, whether such and such a riot or
series of riots would have taken place had criminals been apprehended or
eliminated, avaricious and dishonest businessmen arrested and prosecuted,
communal politicians denied by a responsible party leadership the nomina-
tions of their parties to contest elections. It is sufficiently clear that all these
forces are at work and that such measures would be beneficial. They cannot,
however, all be undertaken at once and some cannot be undertaken at all.
Towards the end of the long series of riots between 1978 and 1980, a zeal-
ous police officer, in complete disregard of universal standards of proper police
behavior, allegedly executed by extrajudicial means the most notorious,
mostly Muslim criminals in Aligarh, after which no major riots occurred in
the town for a decade. In some minds, the cause and the remedy are evident
from these presumed facts. But then, the riots in 1990-91 were the worst ever
seen in Aligarh, including the period 1946 to 1948. Those riots were associ-
ated with a broad Hindu mobilization in the state as a whole from which mil-
itant Hindus sought political advantage in the full knowledge that their actions
were bound to lead to violence and bloodshed, much of which was directly
fomented by their followers in distrkts such as Aligarh.
But then, what has happened to the explanation for riots in which crim-
inals play such a central role? Did a new generation of criminals arise in Aligarh
in the intervening decade or is it the political context that is decisive after all?
 The Persistence of Hindu-Muslim Violence /37l

The evidence suggests it is the latter. Criminals can be found when needed,
but they cannot produce riots on their own, as one of my informants put it.
"Without raj political force, you can't force raj riot. Goonda[s are in the] fore-
front. You can purchase them, but if there is no politics, there is no political
force behind them, they can't do [itl."28 And what about the popular men-
tality of the people, their communal psyche, about which several respondents,
especially in the police, spoke so clearly? In fact, all these aspects must be built
into a full treatment of the production of riots, their dynamics, how they begin,
and how they run their course.
We have learned enough from Aligarh to know that, while the causes of
riots must inevitably remain contestable, consensus can be reached on who
benefits from them. Our interview respondents in Aligarh disagreed funda-
mentally on their causes, but their explanations invariably involved dis-
placement of blame. And through that process of blame displacement, we
found a remarkable consensus concerning the identification of particular sets
of beneficiaries. Of course, our party was not responsible, but the other was.
Of course, our peace committee did not misuse funds for the relief of riot
victims, theirs did. Yes, of course, criminals are active in riots and paid to
instigate them, but it is done by the other side. One is led inexorably to the
unsettling conclusion, therefore, that all persons involved are guilty, though
some more than others. At one extreme, some are clearly responsible for direct
physical acts or instigation of them. At the other extreme, some are respon-
sible only to the extent that tlley refuse to see the ways in which their own
talk and action or nonaction or that of their party or group may contribute
to the persistence of communal hostility.
It is at this point that the whole political order in post-Independence north
India and many, if not most, of its leading as well as local actors-more
markedly so since the death of Nehru-become implicated in the persistence
of Hindu-Muslim riots. These riots have had concrete benefits for particu-
lar political organizations as well as larger political uses. Under the first head-
ing, it is evident that Hindu-Muslim opposition, tensions, and violence have
provided the principal justification for and the primary source of strength
for tlle political existence of some local political organizations in Aligarh and
elsewhere in north India. While the Jan Sangh, the BIP, and all the organi-
zations in the RSS family of organizations adhere to a broader ideology of
Hindutva, of Hindu nationalism, that theoretically exists independently of
Hindu-Muslim antagonisms, in practice that ideology has thrived only when
that opposition is explicitly or implicitly present. In Aligarh, that opposition
is always explicitly present because of the existence of the Aligarh Muslim
 372/ The Persistence of Hindu-Muslim Violence

University, which stands in for the Muslims of India, for Partition and the
creation of Pakistan, and for so many of the ills that afflict Indian society.
Without the presence of the Aligarh Muslim University, the Jan Sangh, the
BJP, and other local communal groups would have had greater difficulty in
establishing a strong presence in the city.
But it is not just in riot-prone cities and towns such as Aligarh that polit-
ical benefits are derived from Hindu-Muslim opposition. It has spread its
benefits even further, to the most important north Indian and all-Indian polit-
ical organizations as well as to organizations and movements within the Hindu
and Muslim communities. Hindu-Muslim opposition has had such larger
political uses since Independence. It has benefited organizations that claim
to speak for the provincial victims of communal riots, the Muslims, byaid-
ing them in the organization and consolidation of the Muslims of the coun-
try to gain political advantage, prevent the loss of special concessions such
as the preservation of the Muslim Personal Law, and preserve their cultural
institutions such as the Aligarh Muslim University itself. The fact that Muslim
political consolidation has usually backfired in post-Independence India is
another matter. The main point is that Muslim political leaders have hoped
otherwise and have played upon the real sufferings of Muslims in towns such
as Aligarh to build broader political movements. Does this then mean that
these political leaders have wanted Hindu-Muslim riots? Usually not, but their
focus has been on using them for purposes of political mobilization rather
than on stopping them.
It is not so apparent, however, how useful the persistence of Hindu-Muslim
riots has been for the secular nationalists who are supposedly their greatest
opponents. Yet, from the very foundation of the Indian state, the Partition
and the Hindu-Muslim violence that occurred during that time as well as the
intermittent violence that continued thereafter for the next fifty years have
been central to the secular nationalist justification of the need to create in
India a composite nationalism, a united people, and a strong centralized state. 29
It has been argued that the very division of India into two hostile commu-
nities, created by the machinations of the separatist Muslim League, has made
it necessary to strive ever harder to counter this hostility by recognizing their
separate existence and their right to maintain separate cultural and religious
and legal institutions recognized by the state. It is not, as in the United States,
that the state is not allowed to interfere in the religious and cultural domains
of the diverse religious communities, but that the state in India is called upon
to enforce and patronize-supposedly equally, to be sure-the religious prac-
tices, laws, and institutions of these separate communities.3°
 The Persistence of Hindu-Muslim Violence /373

But we have also seen in Aligarh that the secular parties use riots for pur-
poses of political mobilization as do the Hindu communal parties, in their
case to mobilize the Muslim community to vote for them on the grounds
that only they can provide the necessary protection to the Muslims against
their enemies. The Congress did so even during Nehru's tenure, and more
especially during Mrs. Gandhi's tenure in office until she switched her strat-
egy to pandering to the Hindu communal vote. The Janata Dal under V. p.
Singh gathered the Muslims under its wing as the Congress turned away from
them and appealed to Hindu sentiment by opening the gates of the Babri
Masjid to Hindu worship. The Samajwadi Party of Mulayam Singh Yadav also
built virtually solid support in the Muslim community in north India in the
1990S because Mulayam Singh stood forth in 1990 as the defender of the
mosque at Ayodhya and, after its destruction under a BJP government, as the
principal obstacle to the consolidation of the BJP's strength in U.P. No impor-
tant party in U.P., however, has taken the stand of the Communist Party of
India (Marxist) in West Bengal, namely, to avoid appealing on communal
grounds to either Hindus or Muslims and to instead use the full powers of
the state to ensure that communal riots simply do not happen.
There is yet another way in which Hindu-Muslim opposition and com-
munal riots serve indirectly the grander purposes of the leaders of the Indian
state. That is in the way these riots are explained, namely, as consequences
of a communal mentality exploited by the Hindu communal parties. The
cure, therefore, is, first, to change the mentality-a process that would take
decades even if anyone had any serious intention of reformulating all the
state educational curricula in India to do so-and, second, to rein in and
occasionally ban the parties and organizations of militant Hindu national-
ism. The first strategy is plainly meaningless and unworkable, the dream
mostly of sincere but powerless Delhi intellectuals. The second strategy, to
rein in or ban Hindu communal organizations, was for long itself part of
the game of manipulating communal animosities and violence to the advan-
tage of the secular parties.
But the banning of Hindu communal organizations is now out of the ques-
tion, with the rise to power of the BIP as the dominant party in a ruling coali-
tion at the Center and as the ruling party in several states of the Indian Union,
including (in 2001) U.P. With its rise to power, there has been a downplay-
ing of explicit attacks on Muslims and their leading institutions, as well as a
decline in the production of large-scale Hindu-Muslim riots and anti-
Muslim programs. No one should be deluded-although many are-into
thinking that these changes reflect any modification of the Hindutva ideol-
 374/ The Persistence of Hindu-Muslim Violence

ogy and its fundamental goals of transforming India into a Hindu nation-
state, a great military power, hegemon in its region, a member of the Security
Council of the United Nations. It is militant nationalism that drives this gov-
ernment's policies. It expressed itself most clearly in the nuclear explosions
carried out under the BJP government at Pokhran in 1998, the continuing
hostile and exaggerated rhetoric directed against Pakistan, and India's defiance
of the entire world in its refusal to sign the CTBT. In pursuit of its grand design
to achieve Great Power status in the world, the Muslims of South Asia are a
hindrance. They are seen as perpetual threats in Pakistan, in Kashmir, and
in all the so-called mini-Pakistans in the cities and towns throughout India.
They are seen as the major obstacle to the unity of the country-though in
fact their presence is essential for the creation of Hindu unity in the first place.
They must be molded into political Hindus or be disciplined, defeated, and
otherwise put in their place.

The Role of the State

How far is the Indian state as a whole or any of its federal units implicated
in the persistence of Hindu-Muslim riots in India? It is obvious that both the
central and state governments in India share responsibility for failing to pre-
vent and control riotS)l It is also evident that both the central and state gov-
ernments have sometimes acted decisively in dealing with potential riot
situations and have at other times not acted at all or have been ineffective.
Finally, it is equally clear that some state governments with large Muslim popu-
lations, including some with a history of comm unal tensions, have a(ied more
effectively than others. Governments in West Bengal, Tamil Nadu, and Kerala
have demonstrated both the will and the capacity to prevent or control Hindu-
Muslim riots.3 2 In West Bengal, which experienced massive communal vio-
lence at the time of Partition and a major communal riot in Calcutta in 1964
under Congress rule, there have been no major communal riots in the past
thirty years of Communist Party (Marxist) rule.
Insofar as the government of U.P. is concerned, there is ample evidence
of its ineffectiveness and its dereliction of duty in preventing and control-
ling riots from time to time since Independence, as well as its noncoopera-
tiveness in post-riot inquiries. There is also evidence that some governments
in U.P. have been able to act effectively when they have chosen to do so. This
has been apparent in U.P. since the mid-1990s,33 that is, since the last wave
of riots that occurred in the aftermath of the destruction of the Babri Masjid
in December 1992.
 The Persistence of Hindu-Muslim Violence /375

But no U.P. government has been willing since Independence to take any
action whatsoever to reform and professionalize the police and to investigate
the charges against the PAC of victimization of Muslims in riots. In 1978, when
the long series of riots in Aligarh began, the representation of Muslims in the
several state police forces ranged from zero to five percent. In the PAC forces,
Muslim representation was "almost negligible." One oft-repeated proposal,
therefore, has been to change the composition of the police forces and the
civil administration through changed recruitment policies, to bring about
"adequate representation of the minority community in the State and Central
Government Police Forces and also in the Civil, District administration."34
A common Hindu objection to this proposal is that the communities
are so divided that it will merely lead to the communalization of the police
forces. It is not entirely clear what lies behind this objection, however. Com-
munalization of the police forces might have one or two consequences. One
is that the police would be divided against each other and become ineffective.
The more likely fear, however, is that Muslim police might kill Hindus just
as Hindu police have killed Muslims. And, indeed, when 16 Hindus were killed
at Ayodhya by police forces that were still overwhelmingly Hindu, during the
government of M ulayam Singh Yadav in 1990, the outcry and outrage in the
Hindu population reverberated into the next election campaign and con-
tributed to the defeat of Mulayam Singh's Samajwadi Party and the victory
of the BJP.
But Mulayam Singh did not seek at that time to increase Muslim repre-
sentation in the U.P. police forces. He sought instead to increase the repre-
sentation of members of his own Yadav caste and other backward castes. In
the years since then, recruitment into the police and civil administration in
the state and their use by successive state governments have been the most
divisive issues in interparty conflicts among the three principal contending
parties in the state: the Samajwadi Party (SP) of Mulayam Singh, the BJP,
and the BSP representing the interests of the Scheduled Castes. Both the SP
and the BSP have sought to reduce upper-caste dominance in the civil and
police administrations and have fought each other as well over the appoint-
ment of backward and Scheduled Caste persons. Although figures on the caste
and communal composition of the current police and civil administrations
are not available, it is probable that Muslim representation has not been
increased significantly.35
In this context of contestation for control over the civil and police admin-
istrations, which is at the heart of a broader struggle for power in the dis-
tricts and localities of this huge state, there are several matters that need to
 376/ The Persistence of Hindu-Muslim Violence

be noted that have a bearing on the issue of whether or not increased Muslim
representation would make a difference in the prevention and control of riots.
The first is that it is unlikely to do so by itself, when the police are seen as
instruments of the party in power at the state and district levels. The assump-
tion behind the demand for increased Muslim representation is that Muslim
police will not vengefully kill innocent Muslims during riots and that, in an
integrated force, Hindus, too, will be reluctant to do so when Muslim police-
men on the force are there to observe them. But the police operate under the
orders of those who control the government, who in turn use them to harass
their political opponents, protect their supporters, and deny protection to
the latter's local rivals. It is indeed quite likely that, in a society divided by
caste and community in which the police are so used, those divisions will
also affect police work. High-caste police officers may tip off their high-caste
brethren when they have orders to arrest them for some offence against Sched-
uled Castes. Hindu police will look aside when Muslims are being attacked
by Hindu crowds. Why should not Muslims do the same? In an integrated
force, such discriminatory behavior might well lead to internal conflicts among
the police themselves.
Two matters critical to the proper functioning of any police force are rarely
mentioned in proposals to make the police somehow more impartial and more
effective in preventing and controlling riots. One concerns professionaliza-
tion, the creation of a police force trained to keep itself aloof from conflicts
among groups within society and to act impartially. Such professionalization
also, of course, requires internal incentives for proper behavior: adequate pay,
respect in society and from political superiors, and opportunities for career
advancement and promotion. Professionalization, of course, is no cure-all,
as indicated by the behavior of supposedly professionalized police forces in
industrial and postindustrial societies such as the United States from time to
time, most recently demonstrated in Los Angeles. But as an option it is cer-
tainly superior to a force dominated by educated, upper-caste persons, deeply
implicated in the everyday conflicts of society, operating with inadequate pay
at the lower levels, heavily corrupted at all levels, and offering virtually no career
incentives for any of the intermediate- and lower-level police to act differently.
But whether a professionalized force would be superior or not to the present
situation is an academic question given the political framework in which the
police operate. The politicians do not talk about professionalization of the
police because control of the police is at the center of political conflict in a
deeply and increasingly bitterly divided society. The proverbial spoils of office
 The Persistence of Hindu-Muslim Violence /377

in India include not only control over the distribution of economic resources
but control over the distribution of protection and safety. As the political strug-
gle becomes more bitter, so does the struggle for safety for oneself and one's
supporters. In the process, the dangers to all increase and so does the need
for safety in a society in which politicians increasingly carry guns, have shad-
ows (bodyguards) with them at all times, and, at higher levels, are surrounded
by armed forces and convoys of vehicles on the roads.
Yet, when all is said and done, the evidence is overwhelming that, even
with a nonprofessionalized police force, the control of which is itself a cen-
tral prize in contemporary political conflict, riots can be prevented and con-
trolled when the political will to do so exists. All political leaders in U .P. know
which districts are riot-prone, who the principal riot-mongers are in such
districts, and who are the civilian and police administrators who can be
counted upon to maintain communal peace even under trying circum-
stances in difficult districts. The people in the districts who do not like riots
also know these things. They do not condemn all the district officials and all
the police. They say that one set of officers acted partially and lor ineffectively,
another set impartially and effectively. But these police and administrative
officers cannot act impartially and effectively unless they have a clear direc-
tive from the state administration to do so. Under the political circumstances
of northern India for far too many years in the post-Independence period,
especially since the late 1960s, that political will has as often as not been absent.
On the contrary, riots have been too often treated as a normal and even a
routine aspel.."t of the political process. The advantages to be gained from allow-
ing communal conflicts to occur or the disadvantages to be incurred from
taking strong action have been too often apparent. In short, for many rea-
sons and at all levels in Indian society, from the Center to the locality, riots
have been functionally useful to far too many persons, groups, and parties.

Institutionalized Riot Systems

I have argued throughout this book that an exclusive focus on a search for
the "true" causes of riots is misdirected and itself is implicated in the per-
sistence of riots. Its implication in riot persistence arises especially from the
fact that the search for causes diverts our gaze from the dynamics of riots,
from the critical issue of how riots are produced. My emphasis here has been
on an alternative approach that focuses on the issue of persistence.
It must be stressed, first of all, that there is little spontaneous about Hindu-
 3781 The Persistence of Hindu-Muslim Violence

Muslim riots. They are, on the contrary, dramatic productions in which what
is spontaneous can occur only because the scene has been prepared with
numerous rehearsals marked by tension, rumors, and provocations, in which
the signals that an outbreak is about to occur and that the time for partici-
pation has arrived have been made clear.
Although elements of the riot ritual described by Gaborieau 36 have
remained present in the riots in Aligarh, the idea behind his schemata of a
direct connection between values and beliefs, on the one hand, and riots, on
the other hand, has to be rejec..ied. Missing in that analysis is a discussion of
the role of intermediaries between the values of the people and the riot engi-
neers. That is to say, we have to distinguish between the roles of those who
mobilize latent hostilities and those who create the incidents that make it pos-
sible for those hostilities to be mobilized; we also have to consider more care-
fully who actually mobilizes. So, let us think of the following groups.
There are, first, the communalist mobilizers, including, especially, pro-
fessionals like those associated with the former Jan Sangh, but including also
many professional politicians and activists in today's BJP, RSS, and associ-
ated organizations. They also include rabble-rousing student mischief mak-
ers, including persons at both the local Hindu degree colleges and, at times,
at the AMU as well. Also in this category are some local businessmen who
want to take advantage of riot situations to cause harm to their enemies and
rivals among Muslims.
A second category of intermediary, repeatedly mentioned in my interviews,
whose members include both paid functionaries and unpaid participants, are
criminals. The paid members are hired killers, who have two kinds of roles.
One is to provide a signal for starting a riot by stabbing a victim. The reli-
gion of the killer as well as the victim may be either Muslim or Hindu. It has
been said that even a Muslim criminal may be paid to kill another Muslim
for the purpose of starting a riot. The unpaid participants are the countless
numbers of criminals who are always ready for a riot to break out to make
money by looting. Such people may be recruited into gangs by local politi-
cians or they may act on their own. Shops owned by persons from the "tar-
geted" community are obvious sources of enrichment, but sometimes it is
necessary to storm the houses of rich men and kill all their occupants before
one can steal their possessions.
Then there are the mobilized mobs, whose composition every analyst of
riots and pogroms everywhere has faced the virtually impossible task of uncov-
ering fully. It is certain, however, that the riffraff theory of riots is faulty inso-
far as Aligarh is concerned. It is not that criminals and goondas of all sorts
 The Persistence of Hindu-Muslim Violence /379

do not get involved, as well as the riffraff or "lumpenproletariat." Indeed, the

latter are often recruited wholesale from particular categories of the popu-
lation living in compact mohallas, who come out en masse when summoned
to participate in a riot. There is no doubt either that students come out in
large numbers during riots. My own interviews in AJigarh indicate clearly, in
addition, that so-called ordinary people also come out during riots to lead
gangs of killers and arsonists and to participate themselves in killing and burn-
ing. Besides my own interviews, there are countless affidavits in which high-
caste and sometimes wealthy persons are mentioned by name by victims and
their occupations and places of business noted. But the composition of any
large-sized riotous mob can never be known fully, neither by arrest statistics
nor by survey questionnaires or random sampling or any social science tech-
nique, not even by participant observation such as has been provided in the
superb work of Bill Buford on England's football hooligansY It is certain only
that the very indefiniteness concerning the composition of riotous mobs will
continue to provide materials for the authorities to displace blame upon
riffraff, the mass of the people, the generic mob, scapegoats-even a decline
in "familyvalues"-and the like to mask their own incompetence, ineffective-
ness, and culpability.
There is, finally, another category of persons involved in riots in AJigarh,
namely, the ordinary constables and the PAC, who make it all worse. They
do so in many ways. They show partiality: standing by while Hindus attack
Muslims, as well as shooting down Muslims as soon as they leave their houses
or mass in the streets. Insofar as the reports are true-and there are too many
such reports to doubt that a great many of them must be authentic-the police
are also riot participants; they knock down the doors of people's houses, beat
up and even kill the male residents, molest the women, and even kill minor
children. They themselves rampage in the streets, striking out wildly, respond-
ing in an undisciplined manner to provocations. One must never forget, also,
when assigning responsibility for the prevention and control of riots, that
the police are agents of the state. But one must also remind oneself of the
even more unpleasant fact that the police are no different from the society
from which they come. They have the same prejudices, the same lack of dis-
cipline, the same hierarchical attitudes, the same penchant to abuse power,
the same tendem:y to corrupt behavior that exists broadly in Indian society,
unleavened by any significant degree of professionalization that would
serve-as it has for the Indian army until now, for the most part-to incul-
cate a different set of values while isolating them from society, and lack any
pay, career, or other incentives to act differently.
 380/ The Persistence of Hindu-Muslim Violence

Contextualization

There is yet a further large piece in the puzzle of riot persistence, which is the
very way in which riots are explained and contextualized. I have argued above
that large-scale riots-which provide the principal focus of this study-are
too multifarious to be subjected to conventional causal analysis. Every large-
scale riot brings out a multiplicity of persons, groups, and forces with a mul-
tiplicity of motives. Every effort to find a single cause, however broadly stated,
will fail to encompass the whole. This search for causes becomes instead a
means of blame displacement by relieving of responsibility all those left out
of the causal explanation. But it does more than that. It provides also a license
to many to loot, burn, and kill for revenge, profit, or pure fun under the cover
of broader explanations provided by the authorities, journalists, and social
scientists who attribute the criminal actions of rioters to feelings of rage against
injustice, communal prejudices, class conflicts, and so forth.
It is at this point that the much-maligned police forces-who no doubt
deserve the criticism they receive-do act differently from most of the rest
of society. Insofar as they are doing their job-and even when they are not
doing it or are misbehaving in this respect also-their task is to identify crimes
as defined in the Indian code of criminal procedure. All policemen in India-
in fact, virtually everyone in India-seem to know at least some of the vast
number of clauses in this code by heart, so well that the foreign observer never
knows what crime is being talked about unless he has memorized it as well,
because the clauses are quoted by number. An offense is not defined as mur-
der, but as a Section x crime. 38
Every person booked for a crime in India, including crimes committed
during a riot, must be booked in an FIR (First Information Report) under a
numbered clause of the criminal code. The police are also trained to inves-
tigate crimes by uncovering personal motives for them: greed, jealousy, revenge
directed at particular persons. Although, in a large-scale riot, the mere appre-
hension of a person on the basis of police or other eyewitness testimony should
be sufficient to book him, the police often think even in these situations of
the personal motivations of the perpetrator of a criminal act. Why this act
in this place at this time? The police especially pay attention to the criminal
acts that become signals for a riot to begin. They will sometimes say that it
was simply a criminal act of a certain type that was used as a pretext by others
for starting a riot.
The police cannot escape this method of identifying and analyzing crimes
even when they are behaving dishonestly. The police in north India frequently
 The Persistence of Hindu-Muslim Violence 1381

deliberately charge the wrong persons for crimes they suspect or know were
committed by others. In doing so, they act in a characteristically corrupt man-
ner. But they must still define the crime, however falsely, in relation to the
specific clauses of the criminal code. There is no clause in the criminal code
that justifies or explains away acts of looting, arson, and murder as expres-
sions of spontaneous rage and, therefore, makes them understandable, excus-
able, or unpunishable. In fact, most such acts committed during riots go
unidentified and unpunished, even when-as they almost always are-
observed by eyewitnesses and, very often, reported in affidavits and FIRs by
victims and witnesses. Nevertheless, the police have repeatedly said to me that
riots are often precipitated by criminal acts, not by spontaneous rage, and that
professional criminals roam freely during riots. The police, in my experience,
are the least inclined to offer broader contextualizations for riots. When they
do so, they do it on the basis of what they see before their eyes: filthy slums,
unhygienic conditions, narrow alleyways through which criminals may escape,
poverty, illitera<.:y, unemployment, and lack of other forms of "entertainment."
As I have already said, the police are in this case exceptional in Indian
society-and probably nearly everywhere el~e in the world as well-in their
focus on the specificities of incidents, in their desire to seek local and person-
alized explanations for them, and in their need to fit particular acts into state-
defined definitions of what constitutes crimes. In the case of large-scale
incidents such as riots or police-public confrontations, however, there is a soci-
etal tenden<.:y to seek to fit these incidents into broader contexts. Although the
form of contextualization may differ at the local and extra-local levels, the link-
age of both through modern systems of communication-even in India and
many other developing countries-contributes to a propensity towards ever-
broader forms of contextualization, especially in those cases where extra-local
persons, agencies, and authorities become involved in them or interested in
them for their own purposes. These broader forms of contextualization, even
when they appear to be high-minded in directing attention to local incidents
as examples of insidious forms of prejudice or racism or communalism,
inevitably transform, distort, and sometimes condemn, but at other times jus-
tify criminal acts.
It is difficult to contest the argument that sometimes such contextualiza-
tions are necessary to rid a society of prevailing, widespread injustice, such
as racism or discrimination against persons of a different religion, by pro-
viding special protections to particular social groups and even by defining
new state crimes or magnifying ordinary crimes by placing them in a different
category-not just killing, but killing from prejudice against a person's race
 382/ The Persistence of Hindu-Muslim Violence

or religion, for example. Thus, in the United States, the crime of murder may
be treated as either an ordinary criminal act in local courts or as a civil wrong
in federal courts, especially when it violates the rights of a person defined as
being from a minority or disadvantaged or otherwise protected group. In India,
there are ordinary crimes and crimes committed against Harijans, the only
difference between the two being the identity of the victim. However, at the
same time, it must be noted that this very process of broadening the frame-
work of explanation into which criminal and civil wrongs are placed has many
latent potential consequences that are undesirable for the peace and well-being
of society and even for the pursuit of justice. They include distortions of jus-
tice that can lead to (or justify) the phenomenon of the "backlash" against
former victims, now seen as privileged beneficiaries of a new form of dis-
crimination against members of still-dominant social categories. In India, the
backlash has taken the form of the highly distorted charge by militant Hindus
that Muslims, protected by the Indian state, have become a privileged and
pampered group in Indian society.
In discussing the phenomenon of Hindu-Muslim riots, we are clearly
faced with this question of contextualization. I have been told over and over
by police-before the mobilizations around the Ayodhya issue-that such
riots have no "concrete" basis, that they arise out of local criminal acts, that
they are placed wrongly in a broader Hindu-Muslim framework, and that
all sorts of ordinary and extraordinary criminal acts are committed under
the cover of these riots. This contextualization of Hindu-Muslim riots
derives, as noted earlier, from a broader discourse of Hindu-Muslim com-
munalism, which, however, is embedded in a divided historical conscious-
ness. Militant Hindus and separatist Muslims have agreed on the "essential"
differences between the two communities. At the same time, they have been
divided not only in how they perceive those differences to have arisen and
what needed to be done about it, but also in the very nature of their sepa-
rate historical consciousnesses.
In India, as in Sri Lanka, the two communities that are seen to be at war
or prone to intergroup violence are also associated with distinctive approaches
to history, one that Daniel characterizes as "history," the other as "heritage. "39
In India, it is the Muslims who have a true historical consciousness, based upon
their demarcation of their history as having begun at a certain time and place
and having been announced by their prophet. From that date, a vast train of
events were launched that are connected one to the other across the centuries,
events that include wars, conquests, kingdoms, empires, and chronicles of
them all, mosques and monuments whose dates and builders are known and
 The Persistence of Hindu-Muslim Violence /383

recorded, and last but by no means least, histories of the development of their
laws, punctuated from time to time by the interpretations and adaptations of
them to different times and places by their greatest scholars and holy men. 40
Most Hindus, by contrast, cannot and do not try to separate what others
consider mythology from history. They point to great monuments and tem-
pIes from the past, most of them dated by others. Their own dates for their
origins, their books, their monuments tend to be fantastic, not credible, said
to have arisen in eras that all schoolboys in the West know to have been pre-
historic, even pre-Homo sapiens. Most persons educated in Indian univer-
sities emerge believing that the length of their history and culture surpasses
that of all others, and having only contempt for countries like the United States,
which they say has "only two hundred years of history."
Whatever the differences between Hindu and Muslim approaches to their
past, it is evident that Hindus are far more absorbed in theirs than are Muslims.
They live their imagined past in the present and perceive every imagined
wrong, especially those imagined to have been done by Muslim conquerors,
as if it happened only yesterday, not 500 years before by people differently
defined and aligned in relation to each other. They blame Muslims for the
loss of their past and of the monumental evidence of their former greatness
in north India, which they believe was destroyed by Muslim generals and
rulers. Further, they claim that, while they are prepared to recognize the his-
tory of Muslim rulers, along with the Indo-Muslim monuments, art, and lit-
erature, as part of Indian history, the Muslims refuse to identify with Indian
history. Muslims, they say, see their history in India instead as part of an exter-
nal, Islamic, physical and religious conquest of the subcontinent.
These conflicting historical consciousnesses and identifications culmi-
nated in a terrifyingly precise moment in modern Indian history, that is, the
Partition, which stands for most educated Hindus-and, in northern India,
most Hindus in general-as a historical scar that not only divided the sub-
continent but defied the truth they had fought for as their rightful heritage:
the unity of India. Muslims, for their part, fougl1t for another truth invented
out of their past in India, namely, that they constituted a separate civilization
distinct from that of the Hindus, that they had always been separate, and would
have to remain so in the future. Leaving aside the question of the causes of
Partition, on which much ink has been spilt, it stands as the first catastrophe
of the historical consciousness in modern South Asia. Partition certainly arose
out of political struggles, but one of those struggles was over the past, com-
bined with. a fear of a future in which two cultures perceived as historically
distinct would not be able to live together in peace. Sayyid Ahmad Khan, in
 384/ The Persistence of Hindu-Muslim Violence

Aligarh, laid the Muslim foundation for separatism that Jinnah turned into a
political weapon. And in Aligarh itself stands the very institution that Hindus
deem to have constructed the ideology and the leadership that produced this
moment of violence and chaos, the Aligarh Muslim University. Further, the
militant Hindus claim to believe that the AMU and all the distinctive institu-
tions of the Muslims in India, even their very religious beliefs, threaten Hindu
India, India that is Hindu, with further partition, violence, and chaos.
For these Hindus, living in an imagined past, the path to the glorious
future-which rightfully belongs to India because of the greatness of its
ancient civilizations before the arrival of the Muslims and the British-is
blocked. It is blocked, on the one hand, by the remnants of that more recent
past of Muslim conquerors, empires, monuments, and mosques built upon
the ruins, real and imagined, of Hindu monuments and temples. That past
has to be rectified before Hindus can be released from its bonds to achieve
the future greatness that belongs to them. A major step in this direction was
the destruction of the mosque at Ayodhya, which, to countless Hindus, sig-
nified the beginnings of their release from "slavery."41 For some, the destruc-
tion of at least two more mosques-those in Mathura and Varanasi-and
perhaps many others may be necessary before the past can be finally rectified
and Hindus achieve full freedom at last.
On the other hand, all militant Hindus and many who are not associated
with the organizations of militant Hinduism also suffer from an obsessive
concentration on that moment when Independence was achieved and sul-
lied by Partition. They suffer from the presence in the very present of tlle evi-
dences of Partition and the imagined dangers of future partitions. In Aligarh,
the AMU stands for that presence. Elsewhere, in every major city and town
in north India, there are further symbols of that presence wherever there are
large concentrations of Muslim populations. These Muslim concentrations
are called "mini-Pakistans." These "mini-Pakistans" in turn are seen as the
centers of riot production designed to intimidate Hindus and generate more
and more Partitions, more and more violence on the Hindu body. Until the
process of historical rectification is completed and the "mini-Pakistans" are
uprooted or their residents converted into political Hindus-free, of course,
to practice their religion quietly-the concrete problems of the present can-
not be satisfactorily dealt with and India cannot achieve its rightful place in
the world. It is this mentality, inscribed in the minds of a large part of the
Hindu population of the country, produced and reproduced over many
decades, that sustains the beliefs that justify the practices that produce the
dynamic productions called Hindu-Muslim riots.
 Postscript: Aligarh and Gujarat

UTTAR PRADESH AND ALIGARH IN 2002

U tt. ar pr.adeshCU .P.), and Aligarh as well, have experienced a reduc-

tion in the incidence of Hindu-Muslim violence during the past
decade. Further, the February 2002 Legislative Assembly elections
in Aligarh City produced a result that in itself reflects a decline both in riotous
activity and in electoral communalization and polarization. Indeed, there has
not been an election result in anyway comparable to this one since the 1950S.
Three features of this election result are notable.
First, the winning candidate was Vivek Bansal, contesting on the Congress
ticket. As noted in chapter 10, Bansal-an Agarwal Hindu, a non communal
person-fought the 1993 Legislative Assembly election and came in a poor
fourth, losing his security deposit in a communalized electoral contest in which
Krishna Kumar Navman, by defeating Abdul Khaliq, won this seat for the
last time. Bansal did not contest at all in the 1996 elections, in which Abdul
Khaliq finally defeated Navman. In the 2002 election, however, Bansal won
an overwhelming plurality of the votes, 41.87 percent, compared to 20-46 per-
cent for the runner-up B]P candidate.
Second, in this election at last, Navman was replaced as the BJP candidate
by Deepak Mittal, an Agarwal Hindu who lives in a new colony, unaffected
by previous communal violence, on the outskirts of the city. His vote per-
centage was the lowest polled by a BIP Legislative Assembly candidate since
1985. In third place was the only remaining communally oriented politician
from the previous electoral contests, the sitting MLA, Abdul Khaliq, who did
not contest as the candidate of either the SP or the BSP this time, but on the
ticket of what the Election Commission of India classifies as a registered but
unrecognized party, the National Loktantrik Party. He polled 18.04 percent.
 386/ Postscript: Aligarh and Gujarat

All other Muslim candidates-including two on the tickets of the BSP and
the SP, respectively, as well as others who polled only a tiny percentage of the
votes-polled a total of 15.42 percent. Thus the total vote for all Muslim can-
didates, including Abdul Khaliq, was one-third of those polled (33.46 per-
cent), still far below the vote for the winning candidate. Moreover, the total
vote for all Hindu candidates, excluding Bansal, came to a mere 4.21 percent.
It is clear, therefore, that Bansal won this election with support from both
Hindu and Muslim voters and that the ele'-ioral results display neither a polar-
ized nor even a communalized electorate.
Third and also significant, and consistent with the findings in this book,
the turnout in this election was the lowest in the entire history of post-
Independence Legislative Assembly elections in the city. There was no com-
munal riot in Aligarh between 1995 and this election to intensify communal
solidarity and, correspondingly, there was little interest in the election con-
test. Although higl1 turnout is normally considered a favorable aspect of elec-
toral politics in a democracy, for Aligarh it appears, on the contrary, to reflect
the cleansing of an electoral process polluted for two decades by deliberate
communal provocation and routinized violence.

THE GUJARAT POGROM OF 2002

While communal violence in U.P. and in Aligarh have ebbed during the past
decade, recent events in the Western Indian state of Gujarat have surpassed
in ferocity, and probably in numbers killed, the post-Ayodhya killings that
took place in the state of U.P. a decade earlier. This book has made several
references to the images that emerge and prevail in the discourse of Hindu-
Muslim violence, images that defy all reasoned analysis of the dynamics of
such violence, how it begins, unfolds, and ends. Mention has also been made
of two of the most murderously violent events in post- Independence India,
where both the evidence of organizing and preplanning and the involvement
of known political party figures and government ministers have been noted
in the press and in official inquiries. In these two cases-the anti-Sikh vio-
lence in Delhi in 1984, and the murderous attacks on Muslims in the Bombay
killings of 1992-93-the word "pogrom" and even, at times, the words "mas-
sacre" and "genocide" have vied with the term "riots" as a summary descrip-
tive label for these events.
As I have noted in this book and elsewhere, the post hoc labelling of inci-
dents of collective violence is an important aspect of the political struggle to
 Postscript: Aligarh and Gujarat /387

gain control of their interpretation. The planners, organizers, and perpetra-

tors of collective violence want all such events to be labelled riots, expres-
sions of the spontaneous and justified feelings of people outraged by the
actions of the victims who allegedly precipitated the violence. Critics of those
alleged to have planned, organized, and perpetrated the violence seek to fix
blame upon the latter by using the term "pogrom," which implies preplan-
ning and organization by established political parties or other organizations,
often aided by the police and ministers in the government. The term "mas-
sacre" is used by those who wish to demonstrate that an event of large-scale
violence has either victimized poor, hapless, and defenseless people or
involved the sudden indiscriminate killing of a large group of people irre-
spective of age or sex. Social science cannot settle such issues by precise
definitions, since the whole project of production and criticism of collective
violence of this type involves obfuscation by bOtll sides of the differences just
adumbrated. All that the social scientist can reasonably achieve in such sit-
uations is to expose to full view, as far as possible, the agents involved in the
production of collective violence and the interests that are served by those
who seek to capture its meaning. The social scientist must then make his or
her own choice of labels.
Further, it has been noted that since the last great wave of riots that occurred
after the destruction of the Babri Masjid, of which the Bombay violence was
the most egregious instance, there was a decline-not a cessation, however-
of communal violence in India, U.P., and Aligarh. For that decline, several
reasons have been adduced, including tlle claim by the BJP that the relative
reduction in violence in the past decade proved tlle argument that the BJP-
and the RSS family of organizations (Sangh Parivar) of which the BJP is a
part-has not been to blame for such riots because it is the BJP that has been
in power during this past decade in the central government and in many states
with previous records of large-scale violence.
These issues of images, labels, and responsibility emerged starkly once again
in the months between February 27 and mid-June 2002, in the state of Gujarat,
where widespread killings, mostly of Muslims, have been carried out on a
scale and with a ferocity reminiscent of tlle genocidal massacres that occurred
during the partition of the Punjab in 1947, and with the apparent involve-
ment, according to several eyewitness accounts, of ministers in the govern-
ment itself under the leadership of BJP Chief Minister Narendra Modi.l These
events surpass, in these respects, all the riots in Aligarh described in this book,
but all the elements of riot production outlined herein are echoed in Gujarat.
 388 I Postscript: Aligarh and Gujarat

The common elements include, first, the obscurity and indeterminacy of

the "causes" of the initial acts of violence: the horrific killing on February 27
of 58 militant Hindu volunteers (kar sevaks) returning from Ayodhya by train
and burned alive in two bogeys of that train at the railway station in the town
of Godhra, in Gujarat. 2 The process of blame displacement began before any
credible facts at all had emerged about this incident. At a time when between
a million and a million and a half troops of the armed forces of India and
Pakistan were facing each other along the two countries' borders, in a standoff
that seemed to be moving dangerously close to all-out war, one that might
include the use of nuclear weapons, SIP leaders promptly blamed the lSI for
the Godhra incidents. In the days that followed these SIP claims and accu-
sations, however, news reports appeared-at first in the Washington Post and
then in Indian newspapers and in the biweekly magazine Frontline-that cast
doubt on them and pointed to several other circumstances, which opened
up other interpretations. But the Godhra incidents were quickly overshad-
owed by what followed, namely, what all available evidence points to as a sys-
tematic pogrom, enacted with precision and extreme brutality, by persons
and organizations in the institutionalized riot system of the RSS family of
organizations, members of the BJP government, the police, and even mem-
bers of the elite Indian Administrative Service} This pogrom began on
February 28, a day after the Godhra massacre, under the auspices of the VHP,
which called for a bandh (closing down) to protest the killings in Godhra.
The Gujarat pogrom continued until March 3, after which there was a hia-
tus that was followed in turn by "a new round of violence" from March 15.4
Within a week of the Godhra incidents, the press reported an official death
toll of 677. By the end of the month, the figure quoted was 783. It is certain
that these figures are too low, with the estimates of responsible observers,
including the British High Commission, ranging as high as 1,500 to 2,000. 5
Some thirty cities and towns in the state were reported to be "still under cur-
few" on March 27. 6 Fifteen or sixteen districts were reported to have been
affected bytl1e riots? though the rioting was most intense in five or six. Official
figures provided by the Gujarat government and published in Frontline also
show the characteristic predominance of Muslims in the numbers killed dur-
ing rioting: more than 5 Muslims to 1 Hindu, including the 60 Hindus killed
on the train at Godhra, but a ratio of 15 to 1 in the rioting that followed. 8 The
number of displaced persons compelled to seek refuge in relief camps also
speaks to the enormity of the cataclysm visited upon the Muslims of Gujarat:
nearly 100,000 persons in 101 relief camps by March 27, 2002,9 and nearly
150,000 in 104 relief camps two weeks later. lO
 Postscript: Aligarh and Gujarat 1389

Numerous features of these killings and this destruction of property sug-

gest the validity of the term "pogrom" and its systematic character,u Most
such features have been noted on a smaller scale in the Aligarh riots described
in this book. They include the destruction of "mosques and shrines" in
Ahmadabad, reportedly followed by the removal of the debris "by the
Ahmadabad Municipal Corporation" to remove any trace of the former
Muslim presence and even the replacement of such ruins, as in Ayodhya in
December 1992, by "makeshift" Hindu temples. 12 It has also been reported
that many if not most police either stood aside or coordinated or participated
in the violence against Muslims!3 Moreover, the marauding mobs of killers
carried lists of voters with them and were thus able to identify the homes of
Muslims who were to be killed and whose property was to be destroyed. The
mobs also carried licensing documents "and other relevant papers" from the
offices of the municipal authorities and so were able to target Muslim prop-
erties for arson attacks. 14 The attacks were accompanied by the distribution
of leaflets "in different parts of the State," calling "for an economic boycott
of Muslims."15 As in Aligarh in 1990-91, several of the vernacular media
agencies in Gujarat became, in effect, part of the institutionalized riot sys-
tem of the Sangh Parivar. 16 Reports in the English-language press and in other
English-language sources noted that, months before the enactment of these
massacres, "a leading Gujarati newspaper published an article ... naming the
restaurants in Ahmadabad owned by Muslims," which were then duly burnt
down during the pogromP
Also on the riot scenes, according to eyewitnesses, were prominent BJP
and VHP leaders who moved along with the mobs of Hindu rioters; some-
times they played the role of "conversion specialists," addressed the mobs,
and then discreetly left; after their departure, the mobs carried out their mur-
derous attacks. is Other politicians or their relations and employees were seen
allegedly encouraging the police to fire on Muslim crowds, or were them-
selves observed "directing the mobs."19 Similarly, as reported in Aligarh in
1990-91, when Scheduled Caste and otller lower-caste persons fought with
Muslims or invaded and attacked Muslim mohallas, the BJP made use in
Gujarat of tlle poorest and most deprived segments of society, persons from
the tribal communities and backward castes, to carry out much of the killing
and destruction. 20 Again, as in the great Aligarh riots of 1990-91, when attacks
were made for the first time on predominantly Muslim mohallas at the out-
skirts of the city, so also in Gujarat "Muslims were attacked even in areas where
they constituted the majority."21 Further, the killings extended to several vil-
lages in rural areas of the state. Finally, the involvement of real-estate land
 390/ Postscript: Aligarh and Gujarat

grabbers and "mafia" elements with "well-known links with ruling-party

politicians" was also reported. 22 These elements took advantage of the attacks
on slum settlements to rid them of Muslim residents and property owners.
They also exploited mob efforts to prevent residents from returning to their
neighborhoods and homes.
It is necessary as well to underline not only the implication in this pogrom
of the BJP state government, its members, and its agents but also that of the
government of India, led by the BJP, which had the power and ultimate
responsibility to stop this flagrant breakdown of law and order. Most
significant was the failure of the government of India to dismiss the Gujarat
government, under Article 356 of the Constitution of India, for its inability
or unwillingness to maintain law and order. Indeed, the situation in Gujarat
was by far the clearest case in post-Independence India for the rightful impo-
sition of that article, which has been misused countless times during the past
fifty years for inappropriate, partisan political reasons. 23
But members of the government of India compromised themselves and
the central government in many other ways, some blatant, some subtle.
Although Prime Minister Vajpayee, under pressure from the non-BJP con-
stituents in his governing coalition, addressed the country on television on
the third day of rioting "to denounce the Gujarat riots, "24 he did not visit
Gujarat until thirty-six days after the Godhra massacre and the pogrom that
followed it.25 He was then reported to have remarked "that the carnage had
shamed India." Aside from the fact that, once again, it was India's status in
the world that was at stake, as much as or more than the plight of the vic-
tims of a state-supported pogrom, other features of his visit deserve note.
Vajpayee visited Godhra first, thus expressing his solidarity with the Hindus
who had been killed, the victims from the Sangh Parivar. Also, he took with
him, on his tour of Gujarat, central minister Uma Bharati, member of the
VHP, whose speeches, during the 1991 elections and prior to the destruction
of the mosque at Ayodhya in 1992, were considered hostile to Muslims, and
who was one of the most active proponents of the construction of a Hindu
temple at that site.
In the meantime, violence continued sporadically day by day for several
months, with reports of new killings coming from Ahmadabad, the state's
metropolis, and several districts in the state. Not until May did the govern-
ment of India undertake any significant a<..iion to bring a definitive end to
the rioting. In May, the famous police chief K. P. S. Gill-no champion of
human rights, but a firm believer in maintaining law and order evenhand-
edly, a man who had presided over the final suppression of the Punjab insur-
 Postscript: Aligarh and Gujarat 1391
rection in the early 1990S and had a previous record of containing Hindu-
Muslim violence in the northeastern state of Assam-was posted to Gujarat
by Horne Minister L. K. Advani as a "security adviser" to the Gujarat chief
minister, who himself did not at all welcome the appointment. 26 Even Gill,
however, who reportedly described the situation as late as May 9,2000, as
"exceedingly bad,"27 was not provided with the quantum of force that he had
requested to end the killings promptly and decisively. Nevertheless, in keep-
ing with Gill's reputation for firmness and evenhandedness, it was reported
that "more than 200 activists of the Hindu Right" had been arrested shortly
after his arrival,28 as well as 53 persons charged with "crimes such as rape,
murder and arson at Naroda Patiya and Gulmarg Society" in Ahmadabad,
among them some "local B]P leaders."29
In riot-prone India, whenever new great riots or waves of riots occur, left-
ist and secular writers, academics among them, commonly say that the lat-
est wave of riots is the worst since the great Partition massacres of 1946-47.
In some of the respects noted above, but especially if one takes account of
all the features of this fierce outburst, it is fair enough to say as much about
Gujarat in the year 2002. It is fair enough also to note that these riots, which
took place under a majority BJP state government while the BIP was also the
dominant party in the ruling coalition of the country, make a mockery of its
claim to have freed India from riots. 30 Others, however, proclaim a different
view, taking comfort from the fact that riots did not spread from Gujarat to
other parts of India as they did in the last great wave of 1992.
But both types of statements, especially the latter type, are distractions tllat
divert our gaze from the dynamics of riot production in present-day India.
The first type of statement is useful only for the purpose of exposing to full
view the dimensions of what a'-iually happened and noting that yet further
social and political boundaries have been transgressed. For Indians, the first
image conjures up the retributive genocidal massacres of Partition in the
Punjab in 1946-47 and seems to herald yet another monumental catastrophe
that will include tlle further weakening or disintegration of India or the oblit-
eration of its Muslim population. If the first view maximizes the implications
of such events as Gujarat in 2002, the second minimizes them. Both views
have the same focus: namely, the future of India, that is, its territorial integ-
rity, societal peace, democratic functioning, and even its status in a world of
nation-states. But what is truly important for India's present and future in
all these respects is escape from the self-perpetuating traps of blame dis-
placement and the complementary traps of maximizing and minimizing the
significance of horrific violence. In short, it is necessary to fix responsibility
 392 I Postscript: Aligarh and Gujarat

and penetrate the clouds of deception, rhetoric, mystification, obscurity, and

indeterminacy to uncover what can be uncovered, knowing full well that the
whole truth can never be known, but that the evident actions and inaction
of known persons, groups, organizations, political leaders, media, academ-
ics seeking causes, and patriots seeking comfort can be uncovered, exposed,
and brought to book.
APPENDICES
 APPENDIX A

Supplementary Tables

TABLB A.l. Demographic data from 1951 census, by Muslim population per
centage (in descending order) for sensitive, riot-hit, and crime-prone mohallas

Percent Hindus Percent

Mohalla name 4 Percent Muslims and Others Scheduled Castes

Usmanpara 100.00 .00 .00

Bani IsraiJan 99.68 .32 .00
Sarai Behram Beg 86.89 13.11 .00
Turkman Gate 86.33 11.75 1.92
Atishbazan 86.32 11.91 1.77
Sheikhan 84.17 12.47 3.36
Rasalganj 81.70 10.24 8.07
Tantanpara 72.65 27.35 .00
Shah Kamal 63.79 31.38 4.83
Delhi Darwaza 63.60 25.00 11.40
Sarai Rahman 59.00 21.46 19.54
Chah Garmaya 55.43 44.57 .00
Madar Darwaza 45.54 54.46 .00
Turkman Gate 38.70 32.21 29.08
Gular Road 37.23 60.64 2.13
Madar Darwaza 31.56 53.91 14.53
Sabzi Mandi 29.41 70.59 .00
ManikChauk 24.01 75.55 .44
Madar Darwaza 23.60 76.40 .00
Rafatgan; 21.26 78.74 .00
Mamubhanja 15.42 74.40 10.17

395
 396/ Appendix A

TABLE A. 1. (continued)

Percent Hindus Percent

Mohalla name a Percent Muslims and Others Scheduled Castes

Bara Dwari 13.04 82.61 4.35

Katra 8.56 84.86 6.58
Sarai Hakim 1.91 83.96 14.13
Baniapara 1.35 98.49 .16
Raghubirpuri .00 90.91 9.09
Subhash Road .00 100.00 .00
Khaidora .00 100.00 .00
Sarrafa Bazar (Phul Chauraha) NA NA NA

aThrkman Gate and Madar Darwaza are divided among two and three separate wards,
respectively, for which census data are provided separately.
 TABLE A.2. Mohallas and other sites identified as riot-hit in major Aligarh riots, 1956-95

Centre
Mohalla IOther Site a 1956 1961 1971 1972 1978 Studyb 1979 1980 1989 1990-91 1995

Achal Talab X X
Agra Road/Chandra Talkies X X
Aligarh-Atrauli Road (bypass
linking Atrauli Road with GT Road) X X
AMU X X
Anupshahr Road X
Atishbazan X
BabriMandi X X X
Bani IsraHan X X
Baniapara X X
Banna Devi/Thana Banna Devi X X
Bara Ganuhar Ali X
Barahdwari X X X
Barain Gate (?) X
Bazar Sarai Hakim X
BhamoJa X
Bhojpur X
Chah Basanta X

(continued)
 TABtE A.2. (continued)

Centre
MohalialOther Site" 1956 1961 1971 1972 1978 Studyb 1979 1980 1989 1990-91 1995

Chah Garmaya X
Chandan Shaheed (Takia) X X
Chauraha Abdul Karim X X
Civil Lines X X X
Delhi DalWaza (including Sarak
Delhi DalWaza) X X X
Dhaurara X
Dodhpur Market X
Ektanagar X
Ghuria Bagh X
Gular Road X
Hamdardnagar X
Hathi ka Puri X
Indira Nagar X
Jagjivan Rampur X
Jaiganj X
Jama Masjid X
Jamalpur Tola X
Jangal Garhi X
Jiwangarh X
Jogipara X
KabBeg X
Kameshwar Mahadev X
Kanwariganj X
Katra X X
Khaidora X X
Kotwali Road X
Kuwarsi X
Laria X
Madar Darwaza (near Sarai Sultani) X X X X
Mahavirganj X
Mamubhanja X X X
ManikChauk X X X X
Masjid Halwaian X
Medical College Colony X
Mohammad Ali Road X
,
NagJa Mallah X
Nai Basti X
Naurangabad Road X
Pathwan Gate (?) X
Phapala X
Phul Chauraha/Sarrafa Bazar X X X X X X

(continued)
 TABU A.2. (continued)

Centre
Mohalla/Other Site U 1956 1961 1971 1972 1978 Studyh 1979 1980 1989 1990-91 1995

Purani Kotwali X X X X X
Qazipara X
,
Rafatganj X
Railway Road X
Rasalganj X X
Rasalganj Bazar X
Reori Talab (?) X
Sabzi Mandi X X X X
Saifi Colony X
Sarai Behram Beg' X X
Sarai Bhooki X X
Sarai Bibi X
SaraiHakim X X
Sarai Khirni X
SaraiMian X
SaraiPakki X
Sarai Rahman X X
Sarai Rai X
Sarai Sultani X X X X
Sasni Gate X X
Shah Jamal X
Shamshad Market X X
Sheikhan X
Sir Sayyidnagar X
Tila X
Turkman Gate X X
Upar Kot X X X
Usmanpara X
Zakaria Market X X
TOTAL SITES 7 4 4 5 27 15 4 9 6 55 4

I could not identify the correct name, the sites are marked as printed in the original source.
b Designated as such in Centre for Research ill Rural and Industrial Development (CRRW), Chandigarh, "Communal Violence and Its Impact on
Development and National Integration, ~ unpublished, undated; provided to me by Rashpal Mehrotra ill 1983.
C Includes Dahl WaJi Gill.
 402 / Appendix A

TABLE A.3. Vote shares for party candidates in their top and bottom polling
stations. 1957 Legislative Assembly elections. and demographic data for the mo-
hallas included in them. according to the 1951 census (all data in percentages)

Best/Worst Scheduled Hindus

Polling Stations Vote Share Muslims Castes and Others

INC top 5 71.98 28.39 6.54 65.06

INC bottom 5 28.31 15.33 11.94 72.73
BJStop 5 36.91 26.12 3.47 70.41
BJS bottom 5 1.22 65.34 7.49 27.18
PSPtop5 30.69 92.16 0.79 7.06
PSP bottom 5 0.91 16.05 1.69 82.27

TABLE A.4. The twentieth-century decline in Muslim

representation in the U.P. police

1935-1936 1981

Total Muslim Muslim Total Muslim Muslim

Police Police % Police Police %

Senior (gazetted)
officers 188 34 18 230 6 3
Inspectors 159 65 41 789 14 2
Sub-inspectors 1,941 864 45 8,099 348 4
Sergeants and
assistant sub-
inspectors 35 nil nil 290 22 8
Head constables,
corporals, and
constables 31,249 15,058 48 88,082 6,563 7
TOTAL 33,572 16,021 48 97,260 6,953 7

SOURCE: Steven I. Wilkinson, "The Electoral Origins of Ethnic Violence: Hindu-Muslim

Riots in India," unpublished Ph.D. dissertation, Harvard University, ch. 3, p. 31.
 APPENDIX B

Key to Maps

LOCALITIES, OHALI.AS, AND OTHER SITES OF ALIGARH,

WITH TH EIR NUMlJERED LOCATIONS ON THE MAPS

itlphahet ica I Order

Achal Sarowar 99 101

Aligarh Muslim 78 Firdous Nagar 64

Atishbazan 17 Foot 106
Babri Mandi 27 Gambhirpura 50
Badamnagar 56 Ghas ki Mandi 97
Badarbagh 66 Hamdardnagar 55
Bani lsrailan 28 Hanumanpuri 94
Baniapara 15 Indira Nagar 68
Bannadevi 86 Industrial Colony 76
Bar~hdwari 10 Inter College 98
Bar~i 47 Jaiganj 26
92 Jama Masjid 32
Bhojpur 54 Jamalpur 59
Brahmanpuri 38 Jamalpur ka Nagla 60
Chandan Shahid 93 Jangal Garhi 20

Chauraha Abdul Karim 18 Jauhar Bagh 80

Chhavani 75 81
D. S. College 40 Jwalapuri
Delhi Gate 19 Kachari
Dhaurara 100 Kale ki Sarai 22

Jain Mandir 108

Dodhpur 79 Kathpula
Dube ka Parao 8 Katra

403
 404 / Appendix B

Kelanagar 82 SaraMian 21
Khaidora 37 Sarai Nawab 2
KhirniGate 48 Sarai Pakki 43
Krishi Farm 102 Sarai Pathanan 52
Kuwarsi 83 Sarai Qazi 45
Laria 53 Sarai Rahman 89
Madar Gate 39 Sarai Rai 109
Mahavirganj 12 Sarai Sultani 42
Mahendranagar 49 Sarrafa Bazar 34
Mamubhanja 4 SasniGate 51
ManikChauk 33 Shah Jamal 69
Maulana Azad Nagar 57 Shahinshabad 67
Medical College 61 Shamshad Market 63
Medical Colony 103 Shri Varshneya College 73
Nagla Bhamola 65 Sir Sayyidnagar 104
Nagla Mallah 62 Subhash Road 7
Nagla Pala Sahibabad 95 Sudamapurl 90
Nai Basti 87 Sunhat 16
Naildgah 71 Surendranagar 72
New Abadi 58 Tamolipara 36
Pala Sahibabad 96 Tantanpara 14
Patthar Bazar 5 Turkman Gate 24
Phaphala 6 University Mosque 77
Phul Chauraha 35 UparKot 31
Purani Kotwali 29 Usmanpara 30
Qazipara 44 Ustad Sahib ka Dargah 70
Raghubirpuri 88 Vishnupuri 91
Railway Colony 110 Women's College 111

Ram Lila Ground 107

Numerical Order
Rasalganj
SabziMandi 41 Rasalganj
Samna Para 46 SaraiNawab 2
Sanichrl Painth 25 SaraiHakim 3
Sara Bibi 112 Mamubhanja 4
Sarai Bhooki 113 Patthar Bazar 5
Sarai Hakim 3 Phaphala 6
Sarai Kaba 105 Subhash Road 7
Sarai Kutub 23 Dube ka Parao 8
Sarai Lawaria 9 Sarai Lawarla 9
 Appendix B /405
Barahdwari 10 Mahendranagar 49
Katra 11 Gambhirpura 50
Mahavirganj 12 SasniGate 51
Kanwariganj 13 Sarai Pathanan 52
Tantanpara 14 Laria 53
Baniapara 15 Bhojpur 54
Sunhat 16 Hamdardnagar 55
Atishbazan 17 Badamnagar 56
Chauraha Abdul Karim 18 Maulana Azad Nagar 57
Delhi Gate 19 New Abadi 58
Jangal Garhi 20 Jamalpur 59
SaraiMian 21 Jamalpur ka Nagla 60
Kale ki Sarai 22 Medical College 61
Sarai Kutub 23 Nagla Mallah 62
Turkman Gate 24 Shamshad Market 63
Sanichri Painth 25 Firdous Nagar 64
Jaiganj 26 NagJa Bhamola 65
BabriMandi 27 Badarbagh 66
Bani IsraHan 28 Shahinshabad 67
Purani Kotwali 29 Indira Nagar 68
Usmanpara 30 Shah Jamal 69
UparKot 31 Ustad Sahib ka Dargah 70
JamaMasjid 32 Nai Idgah 71
ManikChauk 33 Surendranagar 72
Sarrafa Bazar 34 Shri Varshneya College 73
Phul Chauraha 35 Jwalapuri 74
Tamolipara 36 Chhavani 75
Khaidora 37 Industrial Colony 76
Brahmanpuri 38 University Mosque 77
Madar Gate 39 Aligarh Muslim University 78
D. S. College 40 Dodhpur 79
SabziMandi 41 Jauhar Bagh 80
Sarai Sultani 42 Jiwangarh 81
Sarai Pakki 43 KeJanagar 82
Qazipara 44 Kuwarsi 83
Sarai Qazi 45 Kachari 84
SamnaPara 46 Kathpula 85
Barai 47 Bannadevi 86
KhirniGate 48 Nai Basti 87
 406/ Appendix B

Raghubirpuri 88 Ektanagar 101

Sarai Rahman 89 KrishiFarm 102
Sudamapuri 90 Medical Colony 103
Vishnupuri 91 Sir Sayyidnagar 104
Begambagh 92 Sarai Kaba 105
Chandan Shahid 93 Foot Bridge 106
Hanumanpuri 94 Ram Lila Ground 107
Nagla Pala Sahibabad 95 Digambar Jain Mandir 108
Pala Sahibabad 96 Sarai Rai 109
Ghas ki Mandi 97 Railway Colony 110
Inter College 98 Women's College III
Achal Sarowar 99 Sarai Bibi 112
Dhaurara 100 Sarai Bhooki 113
 APPENDIX C

Riots and Interparty Competition in the Corporation

Elections: An Example from Ward No. 26

The most hotly contested ward in the city in the 1955 corporation elections was ward
no. 26. The interval between the winning candidate and the runner-up in the ward
was a mere 0.27 percent, a difference of only 7 votes. In fact, the contest was even keener
than indicated by this figure, since there were four candidates who polled within three
percentage points of each other with vote totals ranging from 501 to ;78 out of a total
of 2,;47 votes cast. The seat was won by the BJP candidate from the backward caste
of Saini. The runner-up, a Muslim, contested on the ticket of the BSP. In third place
was the SP with another Muslim candidate. The fourth-place candidate, also a
Muslim, ran as an independent.
The ward contains parts of three moluillas: Sanichri Painth, Sarai Kutub, and Turk-
man Gate (Map 2). The bulk of the voters in the ward come from the latter mohalla.
Thrkman Gate, classed as both riot-hit and crime-prone, has been a notorious cen-
ter of riot production for many years. It was at the epicenter of the rioting in 1990-91.
It also figured. importantly in the famous 1962 elections that took place in the after-
math of the 1961 riots. In those elections, the Turkman Gate area fell among the top
five polling stations in votes polled by the Republican Party candidate for the Aligarh
City Legislative Assembly seat, Dr. Bashir. It is clear, therefore, that this locality has
been at different times a center of high communal mobilization, riot production, and
intense interparty competition. It is also socially and politically fragmented, containing
a three-way caste-communal division among Hindus, Muslims, and Scheduled Castes,
and, in 199;, as just noted, a closely contested four-way party division.
The three mohallas in ward 26 lie on the southwestern edge of the old city, with
the Turkman Gate deriving its name from its location at a point where several roads
lead out of the city in different directions (Map 1). The demographic mixture in these
mohallas is uncommon in the city as a whole, but is characteristic of this section of
it, in containing such a large concentration of Scheduled Castes. Unfortunately, how-

407
 408/ Appendix C

TABLE C.l. Community and caste composition of three mohalla.~, 1951 census

Hindus and Others Muslims

Mohalla name Number Percent Number Percent

Sanichri Painth 616 91.53 51 7.58

Turkman Gate 288 32.21 346 38.70
Kutab ki Sarai 765 50.50 84 5.54
TOTAL 1,669 54.15 481 15.61

ever, it is not possible to be precise about the number of Scheduled Castes in the ward.
Insofar as the three mohallas are concerned, figures are available from the 1951 cen-
sus, but they are for the mohallas in their entirety, whereas ward 26 contains only parts
of them. The figures are, nevertheless, indicative (Table C.l).
The following features of the caste/communal composition of these mohallas are
most relevant to the analysis of the political configuration of the ward. Hindus and
others in 1951 were in the majority in the ward as a whole and in two of the mohallas,
and an overwhelming majority in one of them. Muslims were a relatively small minor-
ity in two of the mohallas, but constituted a plurality of the population of Turkman
Gate. Scheduled Castes were an insignificant minority in one mohalla, but a very sub-
stantial minority in the remaining two.
In 1951, the population of these three mohallas in their entirety numbered 3,082
persons. By 1995, the number of electors from the voters' list for those portions of
these mohallas contained within ward no. 26 was 4,418 (Table C.2). Moreover, in 1995,
Muslims constituted a majority in the ward as a whole, largely because their num-
bers had increased vastly in Turkman Gate. The ward was divided in 1995 into five
parts, three of which were for Turkman Gate exclusively. In those wards, the Muslim
percentage of registered voters ranged from 57.58 to 98.07. Non-Muslims, including
both Hindus and others and Scheduled Castes, comprised 41.81 percent of the regis-
tered voters.
Given the majority Muslim population advantage in this ward in 1995, how can
we explain both the intensity of interparty competition and the victory of the Hindu
BJP candidate here? The segment-wise distribution of the vote in this ward provides
the basis for answers to both questions. The BJP candidate, the only strong Hindu
candidate in this ward, won most of his votes in the two segments (158 and 159) in
which non-Muslims were in a majority (Table C.3). The other strong candidates, all
Muslims, polled most of their votes in the mohallas (Sarai Kutub and Turkman Gate)
where Muslims were concentrated, thereby dividing the Muslim majority and paving
the way for the victory of the BJP. In addition, the BSP candidate, though a Muslim,
 Appendix C / 409

Scheduled Castes Total

Number Percent Number Percent

6 0.89 673 100.00

260 29.08 894 100.00
666 43.96 1,515 100.00
932 30.24 3,082 100.00

running on the ticket of the party of the Scheduled Castes, polled his highest vote
share (30.62 percent) in the segment containing Sarai Kutub, which had, in 1951, a
Scheduled Caste population of 43.96 percent. Thus, there was a division in this ward
both between the Scheduled Castes and the Muslims and within the Muslim com-
munity as welL Those divisions were so great that the BJP candidate was able to win
the seat even though he polled a majority vote in only one of the five segments, in
contrast to the BSP, which won a plurality in two segments, the SP, which did so in
one, and the strong independent candidate, also in one.
What can these results tell us about the relationship between riots and interparty
competition? These elections were held in November 1995. The most proximate riot
occurred in the city a month later, but did not affect directly the mohallas in ward no.
26. So the intensity of interparty competition here bears no relation to a proximate riot.
What is most significant about interparty competition here is the degree of fragmen-
tation of the electorate, both among the three census categories and within the Muslim
category as well. Severe riots that precede elections have, as we have seen, the effect of
intensifying interparty competition in the city as a whole, but we cannot expect to see
that effect in particular mohallas, for the effect in particular localities will rather be com-
munalization and polarization of the electorate. Since no major riot preceded these elec-
tions, there was in this ward, on the contrary, a fragmentation of the non-Hindu vote
in general and the Muslim vote in particular that made possible a BIP victory.
The paradox is that the militant Hindu party, which in the past gained strength
from communalization and polarization of the vote in the city as a whole, required
precisely the opposite to win in a socially heterogeneous ward such as this. Moreover,
that situation now exists in the constituency as a whole. Since Muslims now consti-
tute a majority of the voters in the city Legislative Assembly constituency as a whole,
the HIP also now requires division of the Muslim vote rather than, or as much as,
consolidation of the Hindu vote to win the seat. That being the case, the prediction
must be that riots of the type witnessed in earlier periods in the city are not to be
expected in future under the existing electoral/political arrangements.
 410 / Appendix C

TABLB C.2. Number and percentage of registered voters by community in Ward 26,1995

Part
Number Number 0/
a/Ward Mohalia Names(s) Registered Voters

158 Sanichri Painth/Turkman Gate 858

159 Sarai Kutub/Turkman Gate 882
160 Turkman Gate 886
161 Turkman Gate 882
162 Turkrnan Gate 910
TOTAL Whole ward 4,418

TABLE C.3. Party vote shares in Ward 26, 1995 corporation elections

Segment Vote/or BIP Vote/or BSP

Number Mohalla Name(s) Number Percent Number Percent

158 Sanichri Painth/Turkrnan Gate 350 66.79 59 11.26

159 Sari Kutub/Turkrnan Gate 138 30.40 139 30.62
160 Turkman Gate 46 9.60 86 17.95
161 Turkman Gate 6 1.17 147 28.54
162 Turkman Gate 38 6.61 140 24.35
WARD TOTAL 578 22.69 571 22.41
 Appendix C / 411

Number of Number of
Registered Percent Registered Non- Percent Non-
Muslim Voters Muslim Voters Muslim Voters Muslim Voters

174 20.28 684 79.72

420 47.62 462 52.38
583 65.80 298 33.63
865 98.07 17 1.93
524 57.58 386 42.42
2,566 58.08 1,847 41.81

Community
Votefor First-Place of First-
VoteofSP Independent no. 1 Party in Place Candidate
Number Percent Number Percent Segment/Ward in Segment

46 8.78 16 3.05 BJP Hindu

84 18.50 65 14.32 BSP Muslim
68 14.20 209 43.63 IND Muslim
144 27.96 74 14.37 BSP Muslim
191 33.22 137 23.83 SP Muslim
533 20.93 501 19.67 BJP Hindu
 NOTES

1 I EXPLAINING COMMUNAl, VrOI,ENCE

1. On the distinction between riots and pogroms in relation to Hindu-Muslim vio-

lence in India, and on the more accurate description of many incidents of Hindu-
Muslim violence since Independence as anti-Muslim pogroms, see Paul R. Brass, Theft
ofan Idol: Text and Context in the Representation ofCollective Violence. (Princeton, N. J.:
Princeton University Press, 1997), pp. Iff. and passim; cf. also, among others, Amrita
Basu, "Why Local Riots are Not Simply Local: Collective Violence and the State in Bijnor,
India 1988-1993," Theory and Society 24 (1995): 35, and Gyanendra Pandey, "In Defense
of the Fragment: Writing about Hindu-Muslim Riots in India Today," in Economic
and Political Weekly [hereafter referred to as EPWj36, nos. 11 & 12 (March 1991): 559-72.
2. A point emphasized at several places in the recent, extraordinarily fine study
by Thomas Blom Hansen, The Saffron Wave: Democracy and Hindu Nationalism in
Modem India (Princeton, N.J.: Princeton University Press, 1999).
3. The term Center is generally used in India to refer to the central (Union)
Government of India, just as, in the United States, one refers to the government in
Washington, D.C., as the federal government.
4. On these two movements and their relationship to the political calculations of
and consequences for the parties and groups associated with militant Hindu nation-
alism, see Christophe Jaffrelot, The Hindu Nationalist Movement and Indian Politics:
1925 to the 19905 (London: Hurst, 1996), especially chapters sand 12.
S. The most notorious and vicious of these killings occurred in the town of
Bhagalpur in Bihar, where not just a riot occurred but massacres of many hundreds
of Muslims; see especially Indu Bharti, "Bhagalpur Riots and Bihar Government,"
EPW 34, no. 48 (December 2,1989): 2643-44.
6. Paul R. Brass, "General Elections, 1996 in Uttar Pradesh: Divisive Struggles
Influence Outcome," .EPW 32, no. 38 (September 20, 1997): 2403-23.

413
 414 / Notes to Pages 9-,16

7. Ashutosh Varshney and Steven I. Wilkinson, Hindu-Muslim RiotSl96o-93: New

Findings, Possible Remedies (New Delhi: Rajiv Gandhi Institute for Contemporary
Studies, 1996), pp. 26-27. Their definition of riot-proneness is a town that has "had
at least three communal riots, spread over at least two five-year periods in which a
minimum of 15 deaths occurred."
8. Paul R. Brass, "Introduction: Discourses of Ethnicity, Communalism, and
Violence," in Paul R. Brass (ed.), Riots and Pogroms (New York: NYU Press, 1996).
9. On the uses of the master narrative of Hindu-Muslim communalism and vio-
lence as an all-purpose explanation for disturbances of the public order in India dur-
ing British rule, see Gyanendra Pandey, The Construction ofCommunalism in Colonial
North India (Dellii: Oxford University Press, 1990).
10. See especially Doug McAdam, Political Process and the Development of Black

Insurgency, 1930-1970 (Chicago; University of Chicago Press, 1982); Charles Tilly,

"Contentious Repertoires in Great Britain, 1758-1834," Social Science History 17, no.
2 (Summer 1993): 253-79; and Sidney Tarrow, Power in Movement: Social Movements,
Collective Action and Politics (Cambridge: Cambridge University Press, 1994).
n. Brass, Theft of an Idol, pp. 9-20.
12. Journey by chariot, in this case placed on top of a Toyota.
13. Jayati Chaturvedi and Gyaneshwar Chaturvedi, "Dharma Yudh: Communal
Violence, Riots and Public Space in Ayodhya and Agra City: 1990 and 1992," in Brass,
Riots and Pogroms, p. 182.
14. Tilly and Tarrow have noted that, in every historical period of widespread
protest activity, new forms of collective action appear that are considered illegitimate,
but later become accepted and integrated into a new repertoire of accepted and legit-
imate forms. Their leading example is the industrial strike; see, for example, Sidney
Tarrow, "Cycles of Collective Action: Between Moments of Madness and the Repertoire
of Contention," Social Science History 17, no. 2 (Summer 1993): 289. As far as I know,
however, no society, even Nazi Germany or fascist Italy, however much it has prac-
ticed violence-including in the form of pogroms-has integrated violent riots into
a repertoire of accepted and legitimate forms of collective action.
15. Bill Buford, Among the Thugs (New York: W. W. Norton, 1992).
16. See Asghar Ali Engineer, "The Causes of Communal Riots in the Post-
Partition Period in India," in Asghar Ali Engineer (ed.), Communal Riots in Post-
Independence India (Hyderabad: Sangam Books, 1984), pp. 33-41; Hussain Shaheen,
"Communal Riots in the Post-Partition Period in India: A Study of Some Causes and
Remedial Measures," in Engineer, Communal Riots, pp. 165-74; Seymour Spilerman.
"The Causes of Racial Disturbances: A Comparison of Alternative Explanations,"
American Sociological Review 35, no. 4 (August 1970): 627-49.
 Notes to Pages 17-20 /415

17. Raymond J. Murphy and James M. Watson, "Ghetto Social Structure and Riot
Support: The Role of White Contact, Social Distance, and Discrimination," in Allen
D. Grimshaw (ed.), Racial Violence in the United States (Chicago: Aldine, 1969), p. 236.
According to these authors, "research findings from the riots of the 1960'S" which found
their "most obvious causes of urnest" in "the problems of poverty and discrimina-
tion," did "not exhaust our understanding of the motivations of rioters, nor do many
of the findings 'make sense' in a purely economic or discrimination framework."
18. These and many other problems in causal analysis are not effectively handled
in what has become the hegemonic statement of proper procedures for doing such
analysis in political science, namely, Gary King, Robert O. Keohane, and Sidney Verba,
Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton, N.J.:
Princeton University Press, 1994). These authors say, "We must not ask for motiva-
tions, but rather for facts." While I agree that it is futile to attempt to verify or falsify
individual motivations in the process of causal inference, it is nevertheless essential
to "ask for motivations." Not to ask for motivations assumes we do not want to know
anything about the framework that produces specific answers and the ability to pro-
duce them, including so-called factual answers concerning events. Although, in the
next sentence, they qualify their statement about never asking for motivations, it is
only to say that one may ask why someone did something only if the purpose is to
generate hypotheses, not to penetrate the framework of meaning, the discourse that
produces the statement.
19. See, for example, Ashish Banerjee, "'Comparative Curfew': Changing Dimen-
sions of Communal Politics in India," in Veena Das (ed.), Mirrors of Violence: Commu-
nities, Riots, and Survivors in South Asia (Dellii: Oxford University Press, 1990), P.54.
20. Spilerman, "The Causes of Racial Disturbances," p. 628.
21. Stanley Lieberson and Arnold R. Silverman, "The Precipitants and Underlying
Conditions of Race Riots," in Grimshaw, Racial Violence in the United States, p. 362.
22. Ashutosh Varshney, Ethnic Conflict and Civic Life: Hindus and Muslims in India
(New Haven: Yale University Press, :1.002).
23. James D. Fearon and David D. Laitin, "Explaining Interethnic Cooperation,"
in American Political Science Review 90, no. 4 (December 1996), 715-35·
24. Keith notes in connection with the Brixton riots of July 1981 that "most polit-
ical accounts or explanations of the riots were demonstrably, often openly, value-
laden"; Michael Keith, Race, Riots and Po/icing: Lore and Disorder in a Multi-Racist
Society (London: UCL Press, 1993), p. 72. A cursory reading of the press in the after-
math of the most recent great riots in America, in 1991 in Los Angeles, reveals the
same in the statements of politicians during the presidential election campaign.
25. National Advisory Commission on Civil Disorders, Report of the National
 416/ Notes to Pages 20-22
Advisory Commission on Civil Disorders (Washington, D.C.: Government Printing
Office, 1968); also known as the Kerner Commission Report.
26. Keith (Race, Riots and Polidng, p. 81) argues that the association of cause with
gUilt goes back to the original Greek term aitea, "from which the term aetiology is
derived." The Latin term causa, however, does not derive from the Greek and does
not comprehend both meanings. It is, nevertheless, clear enough that, in everyday
English, cause often implies guilt or blame, which leaves the question Keith means to
raise, as to whether or not the scientific use of "the concept of causality" in the social
sciences can escape its association "in common cultural understandings" with guilt
and blame. I believe it cannot.
27. As in the work of Neil J. Smelser, Theory of Collective Behavior (New York:
Free Press, 1962).
28. Keith, Race, Riots and Policing, p. 82. Here again, King, Keohane, and Verba,
Designing Sodal Inquiry, provide inadequate guidance. Thus, on the question of mean-
ing (p. 40), they acknowledge the necessity for qualitative research that is attuned to
cultural meanings, using Geertz's famous example of distinguishing between a "wink"
and an involuntary "twitch" of the eye. They fail to see, however, that, since actors
know the meaning of such messages, they will alter them when they want to deceive.
Further, the very methods of systematic observation that they propose for the dis-
covery of cultural meanings will nullify their meanings or universalize them, requir-
ing new codes for subterfuge or, in the issue discussed here, for blame displacement
29. That being the case, to the extent that intentionality-actual rather than objec-
tified human agency-becomes built into "causal analysis," it might be more suit-
able to characterize this form as "purposive analysis."
30. This is exemplified in the new research on the Russian pogroms in John D.
Klier and Shlomo Lambroza (eds.), Pogroms: Anti-Jewish Violence in Modem Russian
History (Cambridge: Cambridge University Press, 1992), and in books and articles pub-
lished elsewhere by the contributors to that volume. So determined are many of them
to debunk the causal theory that the Russian state was responsible for the pogroms
that they avoid the more useful task of delineating clearly the multiplicity of actors
and the roles played by them in producing the Russian pogroms in the nineteenth
century, and also end up by going too far in minimizing the role of the state author-
ities. A frustrating example is I. Michael Aronson, Troubled Waters: The Origins of the
1881 Anti-Jewish Pogroms in Russia (Pittsburgh: University of Pittsburgh Press, 1990),
which is full of archival and other documentation on the mechanics of pogrom pro-
duction in Russia that is used overwhelmingly to batter to death the theory of Russian
government involvement in the pogrom instead of constructing a coherent account
of how the pogroms were carried out.
31. See Jorge Luis Borges, Labyrinths: Selected Stories and Other Writings (New York:
 Notes to Pages 23-26 /417
New Directions, 1964), where the theme of endless, precise repetition as a form of
search for true knowledge recurs in several of the stories.
32. Robert K. Merton, "Manifest and Latent Functions," in Social Theory and Social
Structure, rev. ed. (Glencoe, III.: Free Press, 1957), p. 71.
33. Ian Litton and Jonathan Potter, "Social Representations in the Ordinary
Explanation of a 'Riot,'" European Journal of Social Psychology 15, no. 4 (October-
December 1985), p. 372.
34. The leading historians and works in this group are Pandey, Construction of
Communalism; Sandria Freitag, Collective Action and Community: Public Arenas and
the Emergence of Communalism in North India (Berkeley: University of California Press,
1989); and Ayesha Jalal, The Sole Spokesman: Jinnah, the Muslim League and the Demand
for Pakistan (Cambridge: Cambridge University Press, 1985). Each author has a dis-
tinctive perspective within the general. framework of constructivism.
35. See, especially, studies of the origins of communal consciousness and com-
munal riots in Bengal politics in the twentieth century, before and up to Partition,
particularly Joya Chatterji, Bengal Divided: Hindu Communalism and Partition,
1932-1947 (Cambridge: Cambridge University Press, 1994), and, specifically with
regard to Hindu-Muslim riots in Bengal, Suranjan Das, Communal Riots in Bengal
1905-1947 (Delhi: Oxford University Press, 1991), p. 170. Especially relevantto this study
is that Das ultimately emphasizes the extent of preplanning in the Great Calcutta
Killings of August 1946 and in others that followed it in Noakhali and Tiperra, and
the direct and indirect involvement of politicians, parties, and government ministers
in fomenting them.
36. In this group, C. A. Bayly is most prominent; see his "The Pre-History of
'Communalism'? Religious Conflict in India, 1700-1860," Modem Asian Studies 19,
no. 2 (1985), pp. 177-203. Marc Gaborieau, an anthropologist, argues for an even ear-
lier origin of communal identity and interreligious conflict; "FromAl-Beruni to Tinnall:
Idiom, Ritual and Ideology of the Hindu-Muslim Confrontation in South Asia,"
Anthropology Today 1, no. 3 (1985), pp. 7-14.
37. Most notable in this connection has been the intervention of a group of Marxist
and other historians against the movement that led to the destruction of the mosque
in Ayodhya, in which they took the position that it was a distortion of the histori-
cal facts to claim that the mosque was built upon the ruins of a Hindu temple on
the site of the god Ram's birth or that Hindus have since ancient times believed that
this very site was Ram's birthplace. See R. S. Sharma et at, Ramjanmabhumi-Baburi
Masjid: A Historian's Report to the Nation, undated b991?], mimeographed paper, and
the issue of Seminar (364 [December 1989]), titled "Mythifying History"; also rele-
vant is an article by the lawyer and journalist A. G. Noorani, "The Babri Masjid-Ram
Janmabhoomi Question," EPW 34, no. 45 (November 4-11, 1989), pp. 2461-66.
 4181 Notes to Pages 26-27

38. He heads the Centre for the Study of Society and Secularism, in Bombay
(Mumbai), from which his publications emanate. The Centre for Research in Rural
and Industrial Development (CRRID), Chandigarh, formerly headed by Rashpal
Malhotra, but now by Pramod Kumar, has also sponsored ongoing studies of com-
munal riots in India. The major publication by this organization in this connection
isPramod Kumar (ed.), Towards Understanding Communalism (Chandigarh: CRRID,
1992). This organization was also responsible a decade earlier for the most detailed
demographic study yet done in India of the distribution of riot sites by mohallas in
three western U.P. towns, including Aligarh. I make use of their study ("Communal
Violence and its Impact on Development and National Integration," unpublished,
undated) at several places below. Many other brief, but often valuable and insightful
accounts written by others of particular riots in India have been published over the
years in the pages of the EPw'
39. Asghar Ali Engineer, "An Analytical Study of the Meerut Riot," in Engineer,
Communal Riots, p. 280.
40. Asghar Ali Engineer, "Bhagalpur Riot Inquiry Commission Report-A
Comment," in Progressive Prospective 7, vol. 4 (July 1995), p. 1.
41. Beth Roy, Some Trouble with Cows: Making Sense of Social Conflict (Berkeley:
University of California Press, 1994).
42. Varslmey and Wilkinson, Hindu-Muslim Riots.
43. Varshney and Wilkinson, Hindu-Muslim Riots, p.l.
44. Aside from the issue of the utility of the paired comparison method, Varshney's
work suffers from a host of methodological problems that are revealed in a recent
article (Ashutosh Varshney, "Ethnic Conflict and Civil Society: India and Beyond,"
World Politics 53, no. 3 [Apri1200lj, pp. 362-98) heralding the imminent publication
of his book, Ethnic Conflict and Civic Life. Although he draws upon Robert Putnam's
ideas concerning the importance of civic engagement for democratic life, arguing that
interethnic civic engagement prevents interethnic violence, in Varshney's hands it is
pretty much a throwback to Arthur Bentley's arguments (though he does not refer
to Bentley) about the importance of cross-cutting cleavages and the dangers of con-
gruent cleavages between groups. In Varshney's hands, the argument becomes tau-
tological. Where there is extensive civic engagement between Hindus and Muslims,
there is peace, which amounts to saying that where there is peace, there is peace. Other
methodological problems include the following: (1) a dataset that is inherently
flawed, in which further errors were introduced in coding (a huge error was intro-
duced, for example, into the Aligarh data, to which I alerted him and which was cor-
rected in this article without acknowledgment); (2) an insistence that entire towns
and cities are appropriate units of analysis; (3) the use of his independent variable,
civic engagement, as in effect a pre-chosen cause rather than a genuine hypothesis to
 Notes to Pages 28-29 /419
be tested, with the selection of paired sites already proving the case in advance; (4)
use of the misguided distinction between remote and proximate causes; (5) a basi-
cally primordialist perspective on how riots are generated through rumors and
minor clashes that somehow "escalate," introducing a few new metaphors into this
tired approach, such as "ethnic earthquakes," among others; (6) a complete failure
to understand the workings of institutionalized riot systems, though he (inaccurately)
cites my work on the subject; (7) turning the causal chain around so that violence
itself, the dependent variable, suddenly at one point becomes the independent vari-
able; (8) a virtual freeing of the BJP and the RSS (the latter not even mentioned in
his article) from responsibility for the production of riots; (9) a complete freeing of
the police, the principal killers in most riots in India, from responsibility; and (10)
running through all this, an extraordinary faith in causal explanation and the ability
of this kind of social science research to generate full-fledged causal statements. It is
truly regrettable that such retrograde work is being brought forth at this stage in our
knowledge ofIndian politics and society, sanctified by the (mis)use of currentiyfash-
ionable methodologies in the social sciences. In a word, Varshney's work constitutes
a near-perfect example of the project of blame displacement applied to collective vio-
lence in the social sciences.
45. The case that seems to have come closest to involving an entire metropolitan
area in the United States was the New York City draft riots in July 1863, on which see
Iver Bernstein, The New York City Draft Riots: Their Significance for American Society
and Politics in the Age of the Civil War (New York: Oxford University Press, 1990),
and, for a graphic description, J. T. Headley, Pen and Pencil Sketches ofthe Great Riots:
An Illustrated History ofthe Railroad and Other GreatAmerican Riots. Including all the
Riots in the Early History of the Country (New York: E. B. Treat, 1882). In Europe, in
the nineteenth century, such an event would have been called a revolution, except
that in the New York case, it turned into a massacre of blacks by Irish.
46. Sudhir Kakar, "Some Unconscious Aspects of Ethnic Violence in India," in
Veena Das (ed.), Mirrors of Violence: Communities, Riots and Survivors in South Asia
(Delhi: Oxford University Press, 1990), PP.134-45, and Sudhir Kakar, The Colors of
Violence: Cultural Identities, Religion, and Conflict (Chicago: University of Chicago
Press, 1996).
47. Kakar, Colors of Violence, pp. 12-13.
48. Kakar, Colors of Violence, p. 16.
49. Kakar, Colors of Violence, p. 22.
50. Kakar, Colors of Violence, pp. 40-41.
51. Kakar, Colors of Violence, p. 42.
52. On the Ranchi riots of 1967, see Paul R. Brass, Language, Religion, and Politics
in North India (Cambridge: Cambridge University Press, 1974), p. 265. My analysis of
 420/ Notes to Pages 29-45

those riots formed the beginning of my own entirely different approach to the study
of riots in India.
53. See S. D. Reicher's discussion of G. Le Bon, The Crowd: A Study of the Popular
Mind (London: Ernest Benn, 1947), in "The St. Paul's Riot: An Explanation of the Limits
of Crowd Action in Terms of a Social Identity Model," European Journal of Social
Psychology 14 (1984), pp. 1-2; also, Stephen Reicher and Jonathan Potter, "Psychological
Theory as Intergroup Perspective: A Comparative Analysis of 'Scientific' and 'Lay'
Accounts of Crowd Events," Human Relations 38, no. 2 (1985), pp. 171-72 and 179.
54. Buford, Among the Thugs.
55. Kakar, Colors of Violence, p. 189 (emphasis in original).
56. Stanley J. Tambiah, Leveling Crowds: Ethnonationalist Conflicts and Collective
Violence in South Asia (Berkeley: University of California Press, 1996).
57. Brass, Riots and Progroms, pp. 12-16, and Brass, Theft of an Idol, pp. 11-20.
58. Partha, Chatterjee, The Nation and Its Fragments: Colonial and Postcolonial
Histories (Princeton, N.J.: Princeton University Press, 1993), p. 74.
59. Chatterjee, The Nation and Its Fragments, pp. 95-97.
60. Chatterjee, The Nation and Its Fragments, p. 98.
61. Chatterjee, The Nation and Its Fragments, p. 99.
62. Chatterjee, The Nation and Its Fragments, pp. 101-2.
63. Chatterjee, The Nation and Its Fragments, p. 102.
64. Pierre Nora, "From Lieux de Memoire to Realms of Memory," in Pierre Nora
(ed.), Realms of Memory; Rethinking the French Past, vol. 1: Conflicts and Divisions,
trans. by Arthur Goldhammer (New York: Columbia University Press, 1992), p. xvii:
"A lieu de memoire is any significant entity, whether material or nonmaterial in nature,
which by dint of human will or the work of time has become a symbolic element of
the memorial heritage of any community."
65. On which, see especially Paul R. Brass, Language, Religion, and Politics in North
India; Francis Robinson, Separatism Among Indian Muslims: ThePolitics of the United
Provinces' Muslims, 1860-1923 (London: Cambridge University Press, 1974); and
Mushirul Hasan, "Negotiating with Its Past and Present: The Changing Profile of
the Aligarh Muslim University," ill Mushirul Hasan (ed.), Inventing Boundaries:
Gender, Politics, and the Partition ofIndia (New Defui: Oxford University Press, 2000),
PP·135-56 .
66. Brass, Theft of an Idol, ch.7.

2 I ALIGARH

1. The census data in the next several paragraphs and in Tables 2.1 and 2.2 are derived
from the following sources: Census of India, 1951, District Population Statistics, Uttar
 Notes to Pages 46-49 /421

Pradesh, 6-Aligarh District (Allahabad: Superintendent, Printing and Stationery, 1953),

pp. 10-13; Census of India, 1951, District Census Handbook, Uttar Pradesh, 6-Aligarh
District (Allahabad: Superintendent, Printing and Stationery, 1954), pp. 182-83; Census
1961, District Census Handbook, Uttar Pradesh, 20-Aligarh District (Lucknow: Super-
intendent, Printing and Stationery, 1965), pp. 5, 10, xlviii -Iv; Census 1971, Series-21, Uttar
Pradesh, District Census Handbook, Pt. X-A: Town & Village Directory, Aligarh District,
p. 10, and Pt. X-B, Primary Census Abstract, Aligarh District, pp. 64-97; Census 198.1,
Series-22, Uttar Pradesh, District Census Handbook, Pt. XlII-A: Village & Town Directory,
District Aligarh, pp. 324-25, and Pt. XlII-B, Primary Census Abstract, District Aligarh,
pp. 120-41; Census ofIndia, 1991, SerieS-I: India, Pt. N-B (ii): Religion (Table C-9) (Delhi:
Controller of Publications, 1996), pp. 176-79, and Series-25: Pt. II-A: General Population
Tables (Delhi: Controller of Publications, 1997), pp. 100, 453-54.
2. I have incomplete voters' lists for 1975, 1984, and 1995 by mohalla, but they are
not aggregated by census categories. I have used these lists both to check the validity
of the much older 1951 census data as a source for the Hindu-Muslim population pro-
portions and to garner figures on the Hindu-Muslim population proportions in mohal-
las not included in the 1951 census.
3. Interview with head of the Political Science Department, Aligarh Muslim Uni-
versity, on December 27, 1961.
4. E. A. Mann, Boundaries and Identities: Muslims, Work and Status in Aligarh (New
Delhi: Sage, 1992), p. 34.
5. S. A. H. Haqqi, Urban Political Behavior (A Case Study of the Parliamentary
Constituency, Aligarh) (A1igarh: A1igarh Muslim University, 1978), p. 8, and Violette
Graff, "Religious Identities and Indian Politics: Elections in A1igarh, 1971-1985," in
Andre Wink (ed.), Islam, Politics and Society in South Asia (New Delhi: Manohar, 1991),
P·145·
6. Ashutosh Varshney, "Civic Life and Ethnic Conflict: Hindus and Muslims in
India," unpublished draft manuscript, February 1998, p. 168 fn.
7. Haqqi, Urban Political Behavior, p. 9.
8. Zoya Hasan, Dominance and Mobilisation: Rural Politics in Western Uttar
Pradesh, 1930-1980 (New Delhi: Sage, 1989), p. 81.
9. Graff, "Religious Identities and Indian Politics," p. 145; Haqqi, Urban Political
Behaviour, p. 9.
10. Interview with Tota ram Vidyarthi, municipal commissioner, Aligarh, defeated
Jan Sangh candidate in 1962 election, in A1igarh City, on September 16, 1962.
11. For example, the mayor of the Aligarh municipal board in 1991, an Agarwal,

was said to be a member of the RSS; interview, January 3, 1991.

12. Cf. Varshney, "Civic Life and Ethnic Conflict," p. 98. Also interview on July
19,1983: "Jan Sangh is based on the business class, Banias.... All the shopkeepers and
 4221 Notes to Pages 49-51

business men ... are ... Jan Sangh .... there is a solid vote [of the business class fori
the Jan Sangh." Many people in Aligarh continued to use the term Jan Sangh even
after the creation of the BlP primarily by former Jan Sangh members.
13. For example, in 1997, the state vice-president of the Akhil Bharatiya Vidyarthi
Parishad [All India Students Association] was an Aligarh man, Rajiv Agarwal; inter-
view with RSS members, in Aligarh, on November 21, 1997.
14. For an example of Agarwal hostility to the AMU, see Mann, Boundaries and
Identities, pp. 178-79. See also the book-length tirade against the AMU by an Agarwal
writer, Shanti S. Gupta, A.M. U.: The National Context (Aligarh: Viveka Publications,
1980). Agarwal cloth merchants also allegedly played the roles of false witnesses dur-
ing the 1991 riots (see Chapter 5), spreading malicious rumors that Muslim patients
were being killed in the AMU Medical College Hospital; People's Union for Civil Lib-
erties, "Communal Riots in Aligarh, Dec. 1990-Jan. 1991,» PUCL Bulletin (March 1991).
15. Interviews, July 22, 1983 and November 20, 1997.
16. On the 1993 elections, see Chapter 10. Bansal ultimately won the seat for the
Congress in 2002; see postscript.
17. Haqqi, Urban Political Behaviour, p. 8.
18. Interviews with Shri Niwas Sharma, MLA and President, Aligarh District
Congress Committee, on December 27, 1961, in Aligarh: April 19, 1962, in Lucknow;
and again in Aligarh on July 16, 1983.
19. E.g., on the 1996 Legislative Assembly election, see interview in Aligarh on
November 21, 1997.
20. Interview, January 3,1991.
21. Indeed, they claim that their ancestors came from Mithila and settled in Aligarh,
though they no longer have any connection with the Maithil Brahmans of Bihar; inter-
view in Aligarh, on April 2,1999. Some of the Maithels in Aligarh also take the com-
mon Maithil Brahman surname of Iha. In order to distinguish the two herein, I have
adopted the local English spelling of Maithels for those living in Aligarh.
22. Haqqi, Urban Political Behaviour, p. 8.
23. Interview, July 19, 1983.
24. Estimated from Haqqi, Urban Political Behaviour, p. 9.
25. Graff, "Religious Identities and Indian Politics," p. 146.
26. Interview with B. P. Maurya on June 28, 1963, in Chicago. Maurya claimed that
80,000 Jatavs converted to Buddhism in 1957, presumably from the whole district,
but these numbers are not reflected in any census figures for the city or the district as
a whole.
The term Achal Talab, though of Persian derivation, is commonly used to refer to
this site, also called Achal Sarowar in Hindi (derived from Sanskrit), and Achal Tank
(meaning lake or pool) in English.
 Notes to Pages 51-53 14.23
27. Haqqi, Urban Political Behaviour, p. 11.
28. See also Hasan, Dominance and Mobilisation, p. 163, on the fluctuation in the
vote of these two categories of voters in Aligarh in comparison to their support for
the Congress in the rest of the state.
29. Interview in Aligarh on November 21, 1997.
30. Interview with Hashim Kidwai, faculty member, Department of Political
Science, AMU, September 13, 1962.
31. Interview, November 21, 1997.
32. Haqqi, Urban Political Behaviour, p. 12.
33. Mann, Boundaries and Identities, p. 82.
34. Balmikis especially are said to support the BIP; inte.rview, November 21,1997.
35. Interview, January 3,1991.
36. Interview, January 3, 1991.
37. Interview, January 3, 1991.
38. Interviews (including riot victims), in Upar Kot, Aligarh, on January 3,1991.
39. Mann, Boundaries and Identities, pp. 43-44.
40. Mann, Boundaries and Identities, p. 135.
41. Haqqi, Urban Political Behaviour, pp. 39-40; Mann, Boundaries and Identities,
p. 133, has also noted "that the baradari has become increasingly politicised as a vot-
ing block."
42. Mann, Boundaries and Identities, pp. 43-44.
43. Mann, Boundaries and Identities, p. 11. Indeed, contra Ernest Gellner, she argues
(p. 75) that grand concepts ofIslarnic "renewal" and homogenization under an Islamic
identity "have little relevance for Muslims in Aligarh."
44. Mann, Boundaries and Identities, pp. 16-17.
45. Mann discusses also the distinction between Ashraf (high status) and Ajlaf
(lower status) groups, on the one hand, and zat, another kind of status ordering, on
the other hand, both of which encompass multiple baradaris; Boundaries and Identities,
pp. 38, 47, 51, 59·
46. Mann, Boundaries and Identities, pp. 48-49.
47. Mann, Boundaries and Identities, p. 34.
48. Mann, Boundaries and Identities, p. 51.
49. Thus, Qureshis were formerly known as Qasais (butchers), Ansaris as Julahas
(weavers), and Saifis as Lohars (ironworkers); Mann Boundaries and Identities, pp. 65-
66,136-37. The Saifis formerly monopolized metalworking as "skilled craftsmen in
the lock industry" of Aligarh (p. 81).
50. Interview, November 21,1997.
51. Mann, Boundaries and Identities, p. 126.
52. Haqqi, Urban Political Behaviour, p. 10.
 424 / Notes to Pages 55-61

53. Designated as such in Centre for Research in Rural and Industrial Development
(CRRID), Chandigarh, "Communal Violence and Its Impact on Development and
National Integration," unpublished, undated (provided to me by Rashpal Mehrotra
in 1983).
54. However, there are two baradaris of Ansaris, one of traditional high status,
the other deriving from the Julaha group, so it is not certain, for example, that Ansaris
at AMU and in other high-status occupations come from the Momin Ansars. More
likely, they come from the traditionally high-status Ansari baradari.
55. E. A. Mann, "Religion, Money and Status: Competition for Resources at the
Shrine of Shah Jamal, Aligarh," in Christian W. Troll (ed.), Muslim Shrines in India:
Their Character, History and Significance (Delhi: Oxford University Press, 1989), P.155.
56. Mann, "Religion, Money and Status," p. 159.
57. Mann, Boundaries and Identities, p. 158.
58. Mann, Boundaries and Identities, p. 184.
59. Mann, "Religion, Money and Status," p. 157.
60. Ashutosh Varshney, "Civic Life and Ethnic Conflict: Hindus and Muslims in
India," unpublished manuscript (February 1998), p. 168.

3 I HINDU-MUSLIM VIOLENCE IN INDIA AND ALlGARH

1. Ashutosh Varshney and Steven I. Wilkinson, Hindu-Muslim Riots 1960-93: New

Findings, Possible Remedies (New Delhi: Rajiv Gandhi Institute for Contemporary
Studies, 1996), pp. 5-6.
2. Paul R. Brass, The Politics of India since Independence, 2nd ed. (Cambridge:
Cambridge University Press, 1994), p. 240. The percentage-wise distribution of deaths
in this period by religion, for Hindus and Muslims only, is 75 percent Muslim, 25 per-
cent Hindu. Wilkinson has calculated the percentage of Muslims killed in four peri-
ods, including one pre-Independence period, as follows: 1924-27,46 percent; 1961-70,
75 percent; 1971-80, 65 percent; 1985-87, 60 percent. For some reason, the percentage
of Muslims injured is lower than for Hindus in all these periods, markedly so in all
the post-Independence riots. The explanation for this discrepancy that comes to mind
is that more Hindus than Muslims are injured in intercommunal crowd violence, but
that more Muslims are killed by police firing. Figures from Steven I. Wilkinson, "The
Electoral Origins of Ethnic Violence: Hindu-Muslim Riots in India" (unpublished
Ph.D. dissertion, Harvard University), ch. 2, p. 24> citing various sources.
3. Anonymous, "Communal Riots and Minorities," confidential mimeographed
memorandum, undated [1983?J.
4. For example, in the Bhiwandi riots of May 1970,17 Hindus, 59 Muslims, and 2
unidentified persons died. None of the Hindus, but 9 Muslims, were killed by police
 Notes to Pages 61-62 /425

firing. Muslims were fired upon on 26 occasions compared to only 4 occasions for
Hindus. The evidence, therefore, indicates clearly that Muslims were killed in dis-
proportionate numbers both by Hindu rioters and by the police. Figures from
Anonymous, "Communal Riots and Minorities"; this report and the official com-
mission reports from which the data were drawn provide ample further evidence from
most other riots in the period between 1968 and 1980 of the disproportion in police
assaults and killings of Muslims. Nor has the situation changed in the years since, as
will be discussed further below.
5. Varshney and Wilkinson, Hindu-Muslim Riots, p. 12.
6. Asghar Ali Engineer, the most important chronicler of such riots, publishes
reports on all major riots from his institute in Bombay, Centre for the Study of Society
and Secularism, which also maintains a regular journal called Secular Perspective. His
own articles, including year-end reviews, are also published frequently in the EPW.
His most recent year-end review, titled "Communal Riots, 2000," appeared in EPW
26, no. 4 (January 27 - February 2, 2001), pp. 275-79.
7. Varshney and Wilkinson, Hindu-Muslim Riots, p. 19.
8. Wilkinson, "The Electoral Origins of Ethnic Violence," chapter 2, p. 6. It must
be noted, however, that there are huge differences in the number of riots counted
depending upon the sources and the methods used. For example, the inspector-general
(IG) of police, D.P., released state government figures in 1971 that showed 924 com-
munal riots in the twenty-three years from 1948 to 1971, many times more than the
number shown by Wilkinson for the entire period from 1950 to 1993; Times of India,
January 21,1971. The Wilkinson/Varshney figures are based on reporting in the Times
of India, which is incomplete and which often collapses many discrete incidents into
one extensive "riot." The Wilkinson/Varhsney figures, in effect, do the same.
9. Ashutosh Varshney, "Civic Life and Ethnic Conflict: Hindus and Muslims in
India," unpublished draft manuscript, February 1998, p. 115.
10. Wilkinson dissertation, ch. 2, p. 17. Ashutosh Varshney's manuscript, "Civic
Life and Ethnic Conflict: Hindus and Muslims in India," using the same dataset as
Wilkinson, updated to 1995, lists one additional death in riots, bringing the estimated
total to 389.
Varshneyalso ranks all the riot-prone cities of India by the absolute numbers of
deaths. In this ranking, Bombay (1,137) stands first, Ahmadabad (1,119) second, and
AIigarh (389) third, but I have noted in the text that the AIigarh figure is inaccurate
and far too high. Nevertheless, Bombay and Ahmadabad being many times larger than
Aligarh in population, a death-per-person ranking would certainly put Aligarh, with
176 deaths, still very high in the whole of India. Citation to table from page 121 of
Varshney manuscript. Although their Aligarh data have been corrected in response
to my notification to Wilkinson of the problem, such errors are inevitable in large
 426/ Notes to Pages 62-71

datasets of this type and cast considerable doubt on their utility, especially since, as
in this case, an overcount here indicates undercounts in other parts of the database.
11. The figures are probably not accurate since the errors occurred in coding news-
paper reports that lumped together deaths from several sites for riots that occurred
during the same time period. Wilkinson and Varshney were informed of the inaccu-
racies and Wilkinson advised me that they were undertaking to correct their dataset;
personal communication from Steve Wilkinson. However, in his latest publication,
Varshney now gives a figure that is too low, rather than too high; he records 160 deaths
for 1950-95, whereas my figures are 184 plus an unknown number of deaths in the
riot of December 1992; Ashutosh Varshney, UEthnic Conflict and Civil Society: India
and Beyond," World Politics, 53, no. 3 (April 2001), p. 372.
12. Wilkinson, "The Electoral Origins of Ethnic Violence," ch. 2, p. 17.
13. Varshney, "Ethnic Conflict and Civil Society," p. 372. The population of
Greater Bombay in 1991 was 9,925,891; Ahmadabad was 3,312,216, Hyderabad 4,344,437,
Meerut 849,799, and Aligarh 480,;20. Figures from Census ofIndia, 1991, SerieS-I, India,
Part IV-B (ii): Religion (Table C-9), pp. 60, 108, 176, and 356.
14. On the multiple (mis)uses of all types of commissions in India, see Upendra
Bari, Mambrino's Helmet?: Human Rights for a Changing World (New Delhi: Har-
Anand, 1994), ch. 9 ("Sins of Commission(s),,).
15. Ratanlal Ranchhoddas, Dhirajlal Keshavlal Thakore, and Manharlal Ratanlal
Vakil, The Law of Crimes, 22nd ed. (Bombay: Bombay Law Reporter, 1971), p. 332.
16. For comparison, see the charts in Varshney and Wilkinson, Hindu-Muslim
Riots, pp. 13 and 21, which are, however, not entirely comparable, being drawn on a
yearly basis.
17. Indian Statutory Commission (hereafter referred to as ISC), volume IX,
Memorandum Submitted by the Government of the United Provinces to the Indian
Statutory Commission (London: HMSO, 1930), p. 66.
18. ISC, voL IV, Memoranda Submitted by The Government ofIndia and The India
Office to the Indian Statutory Commission, Pt. I (London: HMSO, 1930), memoran-
dum on "Communal Disorders," p. 100. I am grate.ful to Steven Wilkinson for refer-
ring me to this report.
19. Times (London), September 25 and 26, 1925.
20. ISC, "Communal Disorders," p. lll.
21. Times of India, April 12, 13, and 14, 1927.
22. Since the U.P. Legislative Assembly was in session at this time, these riots pro-
vided an occasion for a full-scale debate in the legislature, which occurred between
March 8 and 22 in response to a statement by the premier (Pandit Govind Ballabh
Pant) and an adjournment motion. The Congress was then in power with the Muslim
League as the principal opposition. The debate degenerated into an exercise in blame
 Notes to Pages 71-76/ 427
displacement in which the Muslim League leaders, including Nawabzada Liaquat Ali
Khan, who was later to become Pakistan's first prime minister, blamed the Congress
government, while the latter blamed the Muslim League and the Urdu press; U.P.
Assembly Debates, vol. IV (1-23 March 1938), pp. 460, 732-33, 800. 844-45. 965. 967-68,
1016-26, 1°36-39.
23. These citations come from the Home (Police) Box 378 File # 5004/1046
Aligarh-Riot Scheme) from the U.P. State Archives, kindly provided to me by Steven
Wilkinson.
24. United Provinces, Governor's Report, dated April 1, 1946.
25. Government of the United Provinces, Confidential Department, Fortnightly
Report for the Second Half of March 1946, dated Lucknow, Aprils, 1946.
26. Government of the United Provinces, Confidential Department, Fortnightly
Report for the First Half of April 1946, dated Lucknow, April 22, 1946.
27. United Provinces, Governor's Letter dated June 19, 1946.
28. United Provinces, Governor's Report to Lord Wavell from F. V. Wylie, April
30,1946.
29. The main continuing sites of communal violence at this time were the cities
of Allahabad and Kanpur, while the most vicious and atrocious violence occurred in
the town of Garhmuktesar in Meerut District in November 1946, where several hun-
dred men, women, and children, nearly all Muslim, were massacred.
30. Government of the United Provinces, Confidential Department, Fortnightly
Report for the First Half of June 1946, dated June 24, 1946, and Fortnightly Report for
the Second Half of June 1946, dated July 12, 1946.
31. Government of the United Provinces, Confidential Department, Fortnightly
Report for the Second Half of September 1946, dated Lucknow, October 7, 1946.
32. Governor F. V. Wylie, fortnightly letter to Lord Wavell, dated October 19, 1946.
33. Government of the United Provinces, Confidential Department, Fortnightly
Report for the First Half of November 1946, dated Lucknow, November 25, 1946.
34. Government of the United Provinces, Confidential Department, Fortnightly
Report for the Second Half of November 1946, dated Lucknow, December 10, 1946.
35. Government of the United Provinces, Confidential Department, Fortnightly
Report for the First Half of January 1947, dated Lucknow, January 29, 1947.
36. See Appendix Table A.4.
37. Wilkinson's code sheets. However, the Free Press Journal, March 6, 1950,
reported four killed.
38. Free Press Journal, March 9, 1950.
39. Pars Ram, A UNESCO Study of Sodal Tensions in Aligarh, 1950-51, ed. with an
introduction by Gardner Murphy (Ahmedabad: New Order Book Co., 1955), p. 173.
40. Wilkinson code sheets.
 428 / Notes to Pages 76-84

41. This riot appears in the Wilkinson/Times of India code sheets as one riot. The
state government, however, classed it as five discrete riots; Times of India, January 21,
1971. A brief summary of this riot may be found also in Centre for Research in Rural
and Industrial Development, "Communal Violence and its Impact on Development
and National Integration," unpublished (Chandigarh: n.d. [1983?]), p. 35.
42. This does not translate well into English, but it refers to a central government-
funded organization engaged in the production of knowledge through book
publication.
43. Times of India (Bombay), September 22, 1956.
44. Times of India (Bombay), September 7, 1956.
45. Times of India (Bombay), September 12, 1956.
46. Times of India (Bombay), September 8, 9, 10, 14, 15, 16, 18, 19, 20, and 26, 1956.
A later retrospective survey of riots in U.P. reported that rioting occurred in 12 dis-
tricts ofU.P.; Times of India, January 21,1971.
47. Times of India (Bombay), September 14, 1956.
48. Times of India (Bombay), September 10, 1956.
49. Times of India (Bombay), September 15, 1956.
50. Times of India (Bombay), September 19, 1956.
51. Times of India (Bombay), September 24, 1956.
52. Times of India (Bombay), September 17, 1956.
53. Times of India (Bombay), September 18, 1956.
54. Times of India (Bombay), September 20, 1956.
55. My information on these riots is derived principally from personal interviews
in 1961-62 in Aligarh, originally written up in Paul R. Brass, Factional Politics in an
Indian State: The Congress Party in Uttar Pradesh (Berkeley: University of California
Press, 1965), pp. 100-101. An account giving the same details as mine was prepared
years later as an official, confidential document on Aligarh riots in Anonymous, "Riots
in Aligarh," mimeographed, 7 pages (n.d., but possibly 1983).
56. Times of India, October 5, 1961.
57. Times of India, October 5, 1961.
58. Times of India, October 10, 1961.
59. Times of India, October 14, 1961.
60. Times of India, October 9, 1961.
61. Times of India, October 17, 1961.
62. Times of India, October 7, 1961.
63. Times of India, October 14, 1961.
64. Times of India, October 23 and 30, 1961.
65· Times of India, October 7, 1961.
66. These incidents are mentioned also in Graff, "Religious Identities and Indian
 Notes to Pages 84-87/429

Politics: Elections in Aligarh, 1971-1985," in Andre Wink (ed.), Islam, Politics and Society
in South Asia (New Dellii: Manohar, 1991), p. 153. Varshneyalso refers to a riot in March
1971, whose "reported cause" was the "arrest of a Hindu Nationalist student leader at
the time of elections"; Varshney, "Civic Life and Ethnic Conflict," p. 185.
67. For further details on these issues, see Paul R. Brass, Language, Religion, and
Politics in North India (London: Cambridge University Press, 1974), pp. 223-27.
68. Times of India, March 3, 1971.
69· Times of India, March 31,1971.
70. Times of India, March 9, 1971
71. Times of India, March 3, 1971.
72. Times of India, March 5, 1971.
73. The report of the Mathur Commission was not submitted to government until
1975 and was never officially released; Mukundan C. Menon and Sumanta Banerjee,
Report to the People's Union for Civil Liberties & Democratic Rights (Delhi State) on
Aligarh Riot (October 5, 1978), p. 11. Although never officially released, the report has
been published and some people have had direct or indirect access to it. Although I
have the reference for this report, I have never been able to get hold of a copy of it.
The citation is Mathur Commission Report, UP Government, 1975, Uttar Pradesh Rajya
ke Aligarh Nagar Mein 2 March 1971, and Uske Baad Huey Sampradayik UpadravoN
ke Sambandh Mein JaanchAyoj ki Report (Report of the Commission for Inquiry into
Communal Disturbances of Aligarh, U.P., on March 2, 1971, and After) (Allahabad:
Superintendent of Printing and Publishing, 1975).
74. Times of India, June 1, 1972-
75. Ninety-nine percent, according to one estimate; Times ofIndia, June 20, 1972.
76. Times of India, June 3, 1972-
77. The Muslim League in post-Independence India is but a remnant of the pre-
Independence organization. Its political strength is confined to the state of Kerala, where
it is considered a legitimate party, not an antinational, pro-Pakistan organization.
78. Times of India, June 2, 1972.
79. Times of India, June 3,1972. Later in the month, Nurul Hasan stated emphat-
ically, at a Congress meeting on the status of the AMU, "We are not going to accept
the minority character [of the AMUj"; Times of India, June 15, 1972.
80. Times of India, June 3, 1972.
81. Times of India, June ;, 1972.
82. Varshney gives the "reported cause" of this riot as "Hindu seven year old child
knocked down by a Muslim scooter driver by accident"; (Varshney, "Civic Life and
Ethnic Conflict," p. 185). Like most of his other "reported causes," this is obviously a
gross misreading of the source of the rioting that trivializes it and casts doubt on the
entire enterprise of counting riots by casual newspaper-reading.
 430 / Notes to Pages 87-89

83. Times of India, June 6, 1972.

84. Times of India, June 6 and 8, 1972.
85. Times of India, June 10, 1972.
86. However, one death is reported in Anonymous, "Riots in Aligarh." Since the
account of this riot contains other erroneous information and no details are given
concerning the death, it is possible that this information is also incorrect. This riot
nevertheless has qualified for detailed discussion above because of the extent of injuries
and the number of sites (five) from which violence was reported.
87. Times of India, June n, 1972.
88. The riots in Firozabad and Varanasi and in two other towns in U.P. were
reported in Times of India, June 17-21, 24, and 30, 1972.
89. Times of India, June 13, 1972. In fact, Hnnah did not have a dose association
with the AMU. He was not educated there. He visited the campus to make political
contacts and speeches as he did in countless places throughout the country in the late
1930S and 1940s. He did, however, have considerable success in mobilizing faculty and
students at the AMU, on which see Mushirul Hasan, "Negotiating with Its Past and
Present: The ChangingProfiJe of the Aligarh Muslim University," in Mushirul Hasan
(ed.), Inventing Boundaries: Gen.der, Politics, and the Partition of India (New Delhi:
Oxford University Press, 2000), PP.135-S6.
90. Times of India, June 20, 1972.
91. Varshney lists this sequence as three riots with their reported causes: (1) Octo-
ber, "dash between a Hindu and a Muslim wrestler," (2) November, "retribution for
previous violence," (3) December, "retribution for previous violence"; "Civic Life and
Ethnic Conflict," p. 18S. Once again, such simplistic categorization and the use of the
term cause are both utterly misleading.
92. Hindustan Times, October 17, 1978; Times of India, October 7, 1978.
93. Bhure Lal's house was located in the communally sensitive locality of Delhi
Darwaza. The Times ofIndia, October 19, 1978, report and that of Radiance (a Muslim
paper owned by the Jamaat-ul Ulama), October 22, 1978, both allege that the proces-
sionists chose deliberately to pass through the extremely communally sensitive cross-
ing of Chauraha Abdul Karim on their way, even though it was not on the direct route
to Bhure Lars house. The October 29, 1978, issue of Radiance also referred to the simul-
taneous breaking out of violence at Manik Chauk and Chauraha Abdul Karim. It is
possible, even probable, that these newspapers are feeding off each other's reports,
rather than obtaining information independently, or that they are all feeding from the
same pool of informants.
In addition to Manik Chauk, three other areas were reported to have been "the
worst~affected": Upar Kot, Rafatganj, and Phul Chauraha; Times of India, October

11,1978.
 Notes to Pages 89-94 /431
94· Times of India, October 17, 1978.
95. See Joseph S. Alter, The Wrestler's Body: Identity and Ideology in North India
(Berkeley: University of California Press, 1992).
96. Kakar notes that, even in popular parlance, the vernacular term for wrestler,
pahaiwan, also may connote goonda or hooligan. He has also noted that "Hindu-
Muslim tensions [in Hyderabadl have led the police to ban wrestling matches in the
city, since a bout between a Hindu and Muslim wrestler can easily ignite a riot between
the two communities" (p. 56). During riots, these men may meet with other wrestlers
and hooligans "on a daily basis and decide ... where the killings have to take place
and where they need to be stopped" (p. 80). Further, he remarks that the Muslim and
Hindu wrestlers have also become, in effect, icons of each "community's physical power
and martial prowess" and that they have also been "used by the politician, employ-
ing religious violence for his own purposes" (p. 85); Sudhir Kakar, The Colors of
Violence: Cultural Identities, Religion, and Conflict (Chicago: University of Chicago
Press, 1996).
97. Times of India, October 9 and 14, 1978.
98. Times of India, October 11, 1978.
99. Times of India, October 15, 1978.
100. Times of India. October 16, 1978.

101. Times of India. October 16, 1978.

102. Times of India, October 16, 1978.

103. Times of India, October 14, 1978.

104. Times of India, October 11, 1978.

105. Menon and Banerjee, Report, p. 1.

106. Menon and Banerjee, Report, pp. 1-2.
107. Menon and Banerjee, Report, p. 4.
108. Menon and Banerjee, Report, p. 5. For a similar incident of "body snatching"
by BJP/RSS politicians of a Hindu killed in Kanpur City, who had previously been
implicated in many murders of Muslims in the riots in that city in December 1992,
see Paul R. Brass, Theft of an Idol: Text and Context in the Representation of Collective
Violence (Princeton, N.J.: Princeton University Press), pp. 240ff.
109. Menon and Banerjee, Report; p. 4. The Communist (CPO newspaper, the
Patriot, used the same terms to describe the situation in Manik Chauk. Again, it is
likely these reports come from a common source or are copied from one another.
llO. Menon and Banerjee, Report, pp. 8-9.
111. Menon and Banerjee, Report, p. 9.
ll2. Menon and Banerjee, Report, p. 13.
113. These were the elections that followed the relaxation of the authoritarian
Emergency regime imposed by Indira Gandhi between 1975 and 1977 and that led to
 432/ Notes to Pages 94-97

the election of the first non-Congress government in India's post-Independence

history.
114. The rather anomalous and highly conflicted position of this man will be dis-
cussed later (see below, Chapter 13).
115. Times of India, October 20, 1978.
116. On the local divisions, see Menon and Banerjee, Report, p. 10.
117. Menon and Banerjee, Report, p. 12.
u8. Times of India, October 10, 1978.
U9· Times of India, November 7, 1978.
120. Patriot, November 9, 1978.
121. Times of India, November 21, 1978; this report said that the November 6th
riot began dose t.o Manik Chauk, but that it affected more severely the mohallas of
"Sarai Kaba and Sarai Miya [Mianj, two adjacent areas with a mixed population of
Hindus and Muslims."
122. Times of India, November 10, 1978.
123. Hindustan Times, November 10, 1978, and Times ofIndia, November 21, 1978.
124. Hindustan Times, November 11, 1978.
125. Times of India, November 12,1978.
126. Times of India, November 16, 1978.

127. Inder Malhotra reported on this matter of the failure to arrest Navman
promptly during these riots a couple of years later as follows: "During the particu-
larly nasty riots at Aligarh in 1978, when the university town was under continuous
curfew for more than two months, there was no deartlt of intelligence advice to the
authorities that a certain Mr. Navman must be arrested. But he was an important local
leader of the then ruling party, the Janata, and no one dared touch him for weeks.
Eventually, however, he was taken into custody but only after a visiting American jour-
nalist had taunted the Chief Minister: 'Who is this Mr. 'no-man' you are so afraid
of?'" Times ofIndia, August 24. 1980. (There is a play on the pronunciation of Navman's
name here, which can be pronounced either as Nuvmaan or as Naumaan.)
128. Hindustan Times, November 20, 1978, and Times ofIndia, November 22, 1978.
129. Patriot, November 26, 1978.
130. Times of India, December 16, 1978.
131. Times of India, December 31, 1978.
132. Lok Sabha Debates, Vol. 20, Nos. 12-17 (6 th Series), Dec. 5-13, 1978, Motion
Re: Situation Arising out of Recent Communal Riots in Different Parts of the Country
(New Delhi: Lok Sabha Secretariat, 1978).
133. Times of India, November 20, 1978. The specific area mentioned in this con-
nection was Mamubhanja, but it is not dear whether it is the mohalla or the entire
ward to which reference was made.
 Notes to Pages 97-103 1433

134. Times of India, November 21, 1978.

135. Times of India, November 22, 1978.
136. Interview in New Delhi on June 20, 1991; taped in English.
137. Times of India, August 30, 1980.
138. Times of India, May 4 and 15, 1979.
139. Times of India, May 11, 1979, reported the number as "about 1,000."
140. Times of India, May 10, 1979.
141. Times of India, May 10, 1979.
142. Times of India, May 11, 1979.
143. Times of India, May 11 and 12, 1979.
144. In fact, this alleged aspect of the Dadri affair, which was mentioned only once
or twice in one-line statements in t.he press at the t.ime, became the official explana-
tion. In my interview with the DSP, intelligence, in 1983, he made the following com-
ments without my having mentioned Dadri at all. "The university, which is a
predominantly Muslim area, never witnesses any violence except once in 1979, which
was not a. communal case. A group of AMU students going to Dellli was beaten up
at Dadri for alleged misbehavior with a Hindu girl. The student~ came back and, to
express their resentment at what had happened at Dadri. burnt down some shops at
Shamshad Market." Interview in Aligarh, July 30,1983.
145. Times of India, May 13, 1979.
146. Times of India, May 16, 1979.
147. Times of India, May 21, 1979.
148. Times of India, May 22, 1979.
149. Times of India, August 17, 1980.
150. Times of India, August 19, 1980. Graff, in an earlier, mimeographed version
of the article previously cited, titled "Religious Identities and Indian Politics, a Case-
Study: Aligarh (1971-1981)" (Paris: Fondation Nationale des Sciences Politiques et
Centre d'Etudes de I'Inde et de I'Asie du Sud [EHESS], n.d. [1982-83]), p. 73, reports
that the two constables were beheaded.
151. Times of India, August 21, 1980.
152. Times of India, August 24, 1980.
153. Times of India, August 2;, 1980.
154. Times of India, August 27-30, 1980.
155. Times of India, September 3, 1980.
156. Times of India, August 26,1980.
157. Times of India, August 29, 1980.
158. Times of India, August 31, 1980.
159. Times of India, September 3, 1980. Giriial Jain was one of the most influen-
tial journalists in India, as editor of the Times of India, from 1978 to 1988. The piece
 434/ Notes to Pages 103-108

from which these quotations come is a fine example of his sophistical mode of argu-
mentation. Many examples could be provided from this one piece alone, but perhaps
the most telling was his convoluted argument that a farfetched conspiratorial theory
such as Mrs. Gandhi's interpretation could not be proven wrong; therefore, it must
be given some credence.
160. Times of India, September 4, 1980.
161. Times of India, August 24, 1980.
162. Times of India, September 9, 1980; the official death toll was increased to ten
the next day when another person died of his injuries; Times of India, September 10,
1980. For the follow-up reports on Moradabad, see Times of India, September 9-12,
1980.
163. Reported to me from two sources: Graff, personal communic.ation, and inter-
view in Aligarh, July 30, 1983. Both sources confirm the killings, but I am uncertain
about the number; the figures given to me orally, but not included in my written notes,
were 41 or 49. Further information on the behavior of this SSP, including a reference
to his alleged extrajudicial mode of execution, may be found in the Indian Express,
December 6, 9, and 19, 1981.
164. Centre for Research, "Communal Violence," pp. 34-35. However, there were
major disturbances on the campus of the AMU in May and June 1981 concerning inter-
nal disputes that did not involve Hindu-Muslim relations, which were also handled
firmly by the new district administration.
165. Times ofIndia, September 11 and 12, October 17, 20, and 31, November 2, 1980.
166. Varshney's list ("Civic Life and Ethnic Conflict," p. 185), which is quite inad-
equate for these events, runs as follows. In 1980: August, "Police fires [sic] at a Muslim
protest. Rioting three times in two weeks"; September, ''Arrest of two local journal-
ists"; October, "Random stabbing"; November, "Muslims attack the Provincial
Armed Constabulary." In 1982: July, "Cause insufficiently reported."
167. Varshney, "Civic Life and Ethnic Conflict," p. 174.
168. Inte.rview in Aligarh on July 19, 1983; taped.
169. Times of India, October 9, 1988.
170. Times of India, October 12 and 15, 1988.
171. Times of India, editorial, October 15, 1988.
172. These quotes come from a signed Times of India report by two correspon-
dents reporting from Muzaffarnagar, October 13, 1988. The newspaper's editorial fol-
lowed this report two days later.
173. Centre for Research, "Communal Violence," passim.
174. Centre for Research, "Communal Violence," p. 46.
175. Times of India, October 13, 1988.
176. Atul Kohli, "From Majority to Minority Rule: Making Sense of the 'New'
 Notes to Pages 108-111 /435

Indian Politics," in Marshall M. Bouton and Philip Oldenburg (eds.), India Briefing,
1990 (Boulder: Westview Press, 1990), p. 27. Dates for these riots are from Muslim India
83 (November 1989), pp. 525-26.
177. The broadside was distributed under the auspices of the Dwitiya Rajbhasa
Urdu Virodhi Sangarsh Samiti (Action Committee against Urdu as Second State
Language).
178. Here is a sample of some of the slogans, taken from a broadside listing 52 of
them, distributed on November 20 by one Dr. Vedram Vidyarthi under the anti-AMU
title, Vishvavidyalaya ya Vishvraksh? (University or Poison Tree?) ("a tree yielding
poisonous fruit," according to McGregor's Hindi-English dictionary). Under the cat-
egory of anti-Rajiv Gandhi, there appeared such slogans (rhymed in Hindi) as the
following: (1) Ham hain Muslim bis karor. Panje ko denge maror. (We are two hun-
dred million Muslims. We will bring you down.) (2) Rajiv tune prajatantra ki hatya
ki. Tujhe bhi yad kareNga musalman. (Rajiv, you killed democracy. Muslims will think
of you also [or, will certainly remember it].) This latter slogan makes use of the very
derogatory form of the second person pronoun, used normally only for animals and
little children. The slogan also implies a death threat against Rajiv. (3) AUgarh se uthi
hai anghi. Ur jayega Rajiv Gandhi. (A storm has risen from Aligarh. It will carry off
Rajiv GandhL) (4) Rajiv Gandhi-R.S.S. tumhari. (Rajiv Gandhi-you are R.S.S.)
Under the category of anti-Hindu slogans, here is one example. Sikh Muslim Isai-
al'as meN sub bhai-bhai. (Sikh, Muslim, Christian-all are brothers in harmony.) Given
the existence at this time of intense and violent insurrectionary movements amongst
Sikhs in Punjab, Muslims in Kashmir, and Christian tribals in the northeastern part
of the country, this slogan would naturally antagonize the Hindu population and would
be considered an attack on the unity of the country, which was adopted as one of the
central campaign themes of Rajiv Gandhi and the Congress in this campaign. The
broadside also classified some slogans as traitorous, for example, the following: (1)
Prepare for jihad. Donate your blood. (In English.) (2) No Islam-no India. (In
English.) (3) Naye rashtra ki nayi rajdhani-Ayodhya. (The new capital of the new
nation-Ayodhya.) The latter is an obvious sarcastic attack on the militant Hindu
movement centered around the Ram Janmabhoomi movement in Ayodhya. (4) We
want Baba.ri Masjid. (In English.) Effigies were also seen, among which four were
described in the broadside, including one of Rajiv Gandhi hanging from a noose. All
translations mine.
179. Times of India, November 11, 1989.
180. I have only the clippings for these articles without the names of the newspa-
pers or the dates.
181. Times of India, November 16, 1989.
182. My translation from the Hindi newspaper datelined November 13, 1989.
 436/ Notes to Pages 111-118

183. It is conceivable that these posters were plastered by Hindus deliberately to

incite feelings against the Muslims and the AMU. There are several reasons for think-
ing of this possibility: (1) the inference that the three Hindu students suspended from
the AMU may have been involved with it; (2) the crudeness of the slogans themselves,
so crude that those who plastered them must have known that they would incite rage
among Hindus that might lead to violence against Muslinls and the AMU; (}) the use
of these slogans in a printed broadside for the evident purpose of inciting rage, if not
violence.
184. Centre for Research, "Communal Violence," p. 35.
185. Marc Gaborieau, "From Al-Beruni to Jinnah: Idiom, Ritual and Ideology of
the Hindu-Muslim Confrontation in South Asia," Anthropology Today 1, no. 3 (1985),
p.8.
186. Government of India, Minorities Commission, First Annual Report for the
Year Ending 31St December, 1978 (New Delhi: Government of India Press, 1979), p. 91.
187. Minorities Commission, First Annual Report, p. 92.

4I THE GREAT ALiGARH RIOTS OF DECEMBER 1990

AND JANUARY 1991

1. People's Union for Civil Liberties (PUCL), "Communal Riots in Aligarh, Dec.

1990-Jan. 1991," PUCL Bulletin (March 1991), p. 14.

2. Interview in Aligarh on January 3,1991.
3. Debasish Mukerji, "The Dance of Death: Provincial Armed Constabulary Fuels
the Communal Cauldron in Aligarh," The Week (December 23, 1990), pp. 35-37 (cover
story).
4. Interview on January 3,1991; also interviews (including riot victims) in Upar
Kot, Aligarh, on January 3, 1991.
5. PUCL Bulletin (March 1991). pp. 14-15.
6. Ashutosh Varshney, "Ethnic Conflict and Civil Society: India and Beyond,"
World Politics 53, no. 3 (April 2001), p. 381, describes the situation in an obverse way:
"Blinded by a Hindu nationalist fervor, ... the city of Aligarh plunged into horren-
dous violence.... As in the past, Aligarh's local mechanisms of peace were remark-
ably inadequate to the task of dealing with an exogenous shock." His point is that
Aligarh lacks a peace system. My point is that the Hindu residents of Aligarh are not
so much "blinded" as stimulated by a 10cally well-established network of mobilizers,
agitators, and killers, part of an institutionalized riot network or system. Furthermore.
it shows an utter lack of knowledge of the Aligarh situation to imagine that the city
is merely responding to "an exogenous shock."
7. Interview on January 3, 1991.
 Notes to Pages 118-J26 / 437
8. As one leftist Hindu AMU professor put it to me in connection with this riot,
"Aligarh is very much a part of this whole game because here is ... one thing, A1igarh
Muslim University, so you can use it as a symbol that this is a Pakistani center, anti-
national center"; interview, January 3, 1991.
9. My same informant, whom I interviewed after the 1961 riots and again after the
1990-91 riots, pointing to this aspect of mass participation in the latter, said that this
was the "first time" that there was such an identification of the middle and upper-
middle classes with violent, riotous attacks upon Muslims; interview, January 3, 1991.
10. Interview, Ianuary 3,1991.
n. PUCL Bulletin (March 1991), p. 14.
12. These included a story concerning the dematerialization of a Hindu mahant
(temple priest) after his arrest by the police, on the one hand, and the melting of "the
pupil of a senior police officer's right eye" after he "had ordered firing on the karse-
vaks in Ayodhya"; PUCL Bulletin (March 1991), p. 26.
13· Interview, January 3,1991.
14. The first Times of India report on these riots was datelined December 7 in the
December 8, 1990, edition, but it did not mention the confrontation between Muslim
youths and the police at the police picket opposite the Iama Masjid. The first specific
reference to that confrontation as the beginning of the riot came in an article in the
paper in its December 16 issue. Such dating is artificial, assuming a precise beginning
as well as a precise ending of a riot, whereas, in fact, there were many other incidents
that preceded the one on December 7, which to others signified a different beginning
for this riot.
15. PUCL Bulletin (March 1991), pp. 14-15.
16. Many of these journalists, a few of whom I have had occasion to meet over
the past few decades, are persons without professional integrity or any concern for
the discovery of truth or the exposure of falsehood. Many instances of their lack of
professional behavior could be provided from my personal experiences as well those
of others. See for, exantple, Ashutosh Varshney, "Civic Life and Etlmic Conflict: Hindus
and Muslims in India," unpublished draft manuscript, February 1998, p. 123. My own
experiences have been such that I always seek to avoid any contact with them while
in India.
17. PUCL Bulletin (March 1991), p. 27.
18. PUCL Bulletin (March 1991), pp. 20-24, and the following publications and
documents issued by the AMU, through its Public Relations Office: "Villification
Compaign [sic] AgainstAligarh Muslim University [Decembef1990 Communal Riots
at Aligarh]," printed pamphlet issued by Aligarh Muslim University; "Why the
Medical College was selected"; '1\ Victim of Rumours"; and an untitled chronology
of riot incidents that especially affected the AMU from December 7 through December
 438 / Notes to Pages 126-139

16, 1990 (all typewritten or mimeographed sheets provided to me by the AMU authori-
ties and the Public Relations Office in January 1991).
19. "Why the Medical College."
20. "Why the Medical College."
21. PUCL Bulletin (March 1991), p. 19.
22. PUCL Bulletin (March 1991), p. 20.
23. PUCL Bulletin (March 1991), p. 20.
24. According to Varshney, there are "two traders associations" in Aliga.rh, of
which one, the Vyapar Sangh, is "BJP-supported." He notes that the Vyapar Mandai
is smaller than the Vyapar Sangh and "has more than 20 percent Muslim members,"
compared to "only 4 per cent" Muslim members of the Vyapar Sangh; Varshney ms.,
"Civic Life and Ethnic Conflict," p. 164fh. To my knowledge, the only trader organ-
ization of any consequence in Aligarh is the Vyapar Mandai, which I believe is dom-
inated by Hindus with militant Hindu sympathies, including many RSS members
and BJP supporters.
25. PUCL Bulletin (March 1991), p. 23.
26. PUCL Bulletin (March 1991), p. 23.
27. Although it does not constitute the formal statement for which the PUCL team
apparently hoped, the Times ofIndia (December 16, 1990) referred to Ashok Chauhan
and Pramod Kumar as "a two-member fact-finding team to look into the incident,"
describing the former as "general secretary of the district Congress committee" and
the latter as an "industrialist." They were reported as saying: "Not only was this news
item [reporting the murders in the hospital) totally false but it symbolised the threat
which irresponsible journalism can cause the country's unity."
28. PUCL Bulletin (1991), p. 24.
29· PUCL Bulletin (1991), p. 24.

5 I THE CONTROL OF COMMUNAL CONFLICT IN ALIGARH

1. Interview, Aligarh, January 3, 1991.

2. Interview, Aligarh, on January 3,1991; also interviews (including riot victims)
in Upar Kot, Aligarh, on January 3, 1991. The reference is to the massacres of Muslims
in the 1987 riots in Meerut.
3. Times of India, December 8, 1992.
4. Times of India, December 9, 1992.
5. Times of India, December 12, 1992.
6. Times of India, December 14, 1992.
7. Times of India, December 18, 1992.
8. Times of India, December 23, 1992.
 Notes to Pages 139-150 /439
9. Times of India, December 12, 1992.
10. Times of India, December 18, 1992. In this case, m.oreover, in contrast to May

1979, there were no contradictory reports denying that kar sevaks had committed these
vi.olently abusive actions upon AMU students.
11. Times of India, December 19, 1992.
12. Times of India, December 23, 1992.
13. On the p.ost-Ay.odhya Kanpur riots, see Paul R. Brass, Theft of an Idol: Text
and Context in the Representation of Collective Violence (Princeton, N.J.: Princeton
University Press, 1997), ch. 7.
14. Times of India, March 11, 1995.
15· Times of India, March 12, 1995.
16. Times of India, March 16, 1995.
17. Asghar Ali Engineer, "Aligarh Ri.ots-An Unplanned Outburst," in Towards
Secular India 1, n.o. 2 (April-June 1995), p. 87.
18. Engineer, '~ligarh Ri.ots," p. 90.
19. Engineer, "A1igarh Ri.ots," pp. 90-91.
20. Engineer, "A1igarh Riots," pp. 91-92. Engineer's account here differs from that
.of the Times of India, noted above, that attributed this death to a stabbing.
21. Engineer, '~igarh Riots," pp. 92-93.
22. I would have liked to go inside the RAF truck, but it would have been impolitic
for me to ask, and it was anyway desirable for me and my companions t.o leave the
area as quickly as possible.

6 I THE GEOGRAPHY AND DEMOGRAPHY OF RIOTS

1. Michael Keith, Race, Riots and Policing: Lore and Disorder in a Multi-Racist Society
(London: ueL Press, 1993), p. 167.
2. The current (1997) local telephone direct.ory f.or the A1igarh Muslim University
c.ommunity, consisting primarily of university faculty and some alumni, c.ontains
appr.oximately 1,.075 names, among which I was able t.o identify .only 13.0 names .of
Hindus and .others and six uncertain: Blessing Teleplwne Directory (Aligarh: Blessing
Enterprise, 1997).
3. Map locati.ons are identified either by the name or by number, when necessary
to av.oid crowding on the maps. The names .of all numbered l.ocations on the maps
are listed in alphabetical and numerical order in the Key to Maps in Appendix B.
4. Howard F. Hirt, Aligarh, U.P., India: A Geographic Study of Urban Growth
(unpublished Ph.D. dissertation, Syracuse University, 195·5), p. 160.
5. David Lelyveld, Aligurh's First Generation: Muslim Solidarity in British India
(Princet.on, N.J.: Princet.on University Press, 1978), p. 156.
 440/ Notes to Pages 150-167

6. Hirt, Aligarh, p. 160.

7. Pierre Nora, "From Lieux de memoire to Realms of Memory," in Pierre Nora
(00.), Realms of Memory; Rethinking the French Past, vol. I: Conflicts and Divisions,
trans. by Arthur Goldhammer (New York: Columbia University Press, 1992), p. xvii;
for the definition of "lieu de memoire," see the previous citation, in Chapter 1, n. 64-
8. Lelyveld, Aligarh's First Generation, pp. 273-74.
9. Centre for Research in Rural and Industrial Development, "Communal Violence
and its Impact on Development and National Integration," unpublished (Chandigarh:
n.d. [1983?]), p. 54.
10. Hirt, Aligarh, p. 70.
11. Hirt, Aligarh, p. 70; E. A. Mann, Boundaries and Identities: Muslims, Work and
Status in Aligarh (New Delhi: Sage, 1992), pp. 27-28.
12. Hirt, Aligarh, p. 225.
13. I have seen a map showing these mohallas posted on the walls of the office of
the DM in Aligarh, but could not obtain a copy. Government of India, Minorities
Commission, First Annual Report for the Year Ending 31st December, 1978 (New Dellii:
Government of India Press, 1979), listed 12 sites as "disturbance-affected areas" that
its members visited in the aftermath of the 1978 riots.
14. This characterization of the mohallas of Aligarh comes from the report of the
Centre for Research, "Communal Violence," done after the riots of 1978-79. Of the
29, one was listed as unpopulated in 1951. This mohalla is really a four-way crossing
known as Phui Chauraha and is actually one of the most sensitive localities. Police
pickets have been stationed there continuously 24 hours a day for many, many years.
The census divides the mohallas of Turkman Gate and Madar Gate into two and three
separate units, respectively. They have been consolidated (and the population com-
position for the separate census units recalculated accordingly) on Map 4 so that the
number of mohallas shown thereon is thereby reduced to 25.
15. Mann, Boundaries and Identities, p. 33.
16. Centre for Research, "Communal Violence," pp. 22-23.
17. Mann, Boundaries and Identities, p. 34.
18. Anonymous, "Riots in Aligarh," undated mimeographed study of 7 pages,
given to me in 1983 (it is obviously the confidential report of a former district mag-
istrate of Aligarh); citation from pp. 3-4.
19. Interview on July 20, 1983.
20. Government ofIndia, Minorities Commission, FirstAnnual Report for the Year
Ending 31st December, 1978 (New Delhi: Government of India Press, 1979), pp. 74-77.
21. Centre for Research, "Communal Violence," pp. 34-35.
22. Graff, "Religious Identities and Indian Politics," mimeographed version,
 Notes to Pages 167-184 /441

pp. 67-69. The reference to the Minority Bill is to a bill in Parliament to maintain the
minority character of the AMU.
23. Interview on January 3, 1991.
24. Zoya Hasan, Dominance and Mobilisation: Rural Politics in Western Uttar
Pradesh, 1930-1980 (New Delhi: Sage, 1989), pp. 160-61.
25. Mann, Boundaries and Identities, pp. 175-76.
26. Mann, Boundaries and Identities, pp. 176-77.
27. Cited in Mann, Boundaries and Identities, pp. 177-78.
28. Interview in New Delhi on January 2, 1991; interviews in Aligarh on July 20
and 24, 1983.
29. Interview, July 20, 1983.
30. Interview with senior superintendent of police on July 21, 1983. The SSP's remark
about equality of numbers is, however, imprecise and not accurate for Manik Chauk,
where Muslims are in a minority.
31. Government of India, Minorities Commission, First Annual Report, p. 77.
32. Hasan,Dominance and Mobilisation, p. 159.
33. Mann, Boundaries and Identities, pp. 177-78.
34. The Week, December 23, 1990.
35. PUCL Bulletin (March 1991).
36. Interview, January 3, 1991.
37. Interviews in Upar Kot, January 3, 1991.
38. Interview, January 3, 1991.
39. E. A. Mann, "Religion, Money and Status: Competition for Resources at the
Shrine of Shah Jamal, Aligarh," in Christian W. Troll (ed.), Muslim Shrines in India:
Their Character, History and Significance (Delhi: Oxford University Press, 1989), p. 145.
40. Interview, January 3,1991.
41. Interview with riot victims, 1991.
42. Interview with riot victims, 1991, and interview on November 21, 1997.
43. Interview, November 21, 1997.
44. Inte.rview with riot victims, 1991.
45. PUCL Bulletin, March 1991, p. 16.
46. Times of India, December 10, 1990.
47. Interview, November 21,1997.
48. Times of India, January 20, 1991.
49. Times of India, January 2, 1991.
50. Interview with riot victims, 1991.
51. Interview in Aligarh on November 21, 1997.
52. Interview, January 3,1991.
 442 / Notes to Pages 184-191
53. PUCL Bulletin, (March 1991), p. 16.
54. Times of India, November 27 and December 16, 1990.
55. Paul R. Brass, Theft of an Idol: 1ext and Context in the Representation ofCollective
Violence (Princeton, N.J.: Princeton University Press, 1997), ch. 7.
56. Similar situations have been reported from Bhiwandi (a Bombay suburb) in
1984 and Billarsharif (in Billar) in 1981. In the former case, the population balance
was transformed by an "influx of migrant U.P. workers" to the outskirts of the city
in such a way that Muslims came to comprise 60-65 percent of the population of the
tOWl!. Although the Shiv Sena was responsible for most of the violence in Bhiwandi,
the BJP also had a stake in it, though behind the scenes. Rioters were recruited from
outlying Hindu villages where the BJP was strong to attack Muslim bastis (rough-and-
tumble settlements of poor people) in Bhiwandi. Similarly, in Biliarsharif, rioters came
from villages where "the RSS had been expanding its activities." In boili cases, the
evidence of preplanning and organization is substantial. Attempts were also made in
Bhiwandi to recruit low castes to attack Muslims, but they were not generally suc-
cessful on this occasion. See Asghar Ali Engineer, Bhiwandi-Bombay Riots: Analysis
and Documentation (Bombay: Institute of Islamic Studies, 1984), pp. 13-14; on
Biharsharif, see Asghar Ali Engineer, "Case Studies of Five Major Riots from Biliarsharif
to Pune," in Asghar Ali Engineer (ed.), Communal Riots in Post-Independence India
(Hyderabad: Sangam Books, 1984), pp. 238-46.
57. With the notable exception of Bombay.
58. In some cases, because of the enormous labor involved in counting the
Muslim names exactly, I have had to use a random number sampling meiliod to arrive
at an estimate of the number of Muslim names in the voters' list.
59. Interview with lock manufacturer in Sarai Sultani, Aligarh, on April 2, 1999;
taped in Hindi-Urdu; running translation provided by Aftab Ahmad.
60. This was the time of the Chandra Talkies riot that began outside the cinema
located in Mamubhanja, about a half kilometer from Sarai Sultani.
61. This is standard police practice in north Indian police stations, namely, denial
of requests for water from prisoners whom the police particularly dislike, followed
by the standard response to drink their own urine.
62. The term actually used in Hindi was batti, which means simply a light, but
when I questioned the translation, I was told the term was being used here for a fur-
nace. At one point, in fact, the term "blast furnace" was used. So I am not entirely
sure what kind of furnace was used here.
63. The exact dates of Mr. Punia's stay as district magistrate in Aligarh are from
May 29, 1982, to May 20, 1985. Three years as district magistrate is considered a rela-
tively long term, especially in a riot-prone district such as Aligarh. Throughout his
tenure, there was a Congress government in the state capital in Lucknow. The
 Notes to Pages 195-200 1443
Congress at that time, still dependent on the Muslim vote in U.P., did not want fur-
ther rioting in Aligarh after the years of rioting that had plagued the city. It was dear
to me during my visit to Aligarh during Punia's tenure that both he and the then SSP
were alert to any possibility of a communal disturbance and had the will and ability
to prevent any large-scale rioting.
64. And lately Christians elsewhere in India.
65. This point will be illustrated in the next chapter.

7 I THE ECONOMICS OF RIOTS

1. E. A. Mann, Boundaries and .Identities: Muslims, Work and Status in Aligarh (New
Delhi: Sage, 1992), p. 77.
2. H. R. Nevill, Aligarh: A Gazetteer, Being Volume VI of the District Gazetteers of
the United Provinces ofAgra.and Oudh (Lucknow: Govt. Branch Press, 1926), pp. 59-60;
David Lelyveld, Aligarh 's First Genera.tion: Muslim Solidarity in British India (Princeton,
N.J.: Princeton University Press. 1978). p. 149.
3. Howard F. Hift, Aligarh, U.P., India: A Geographic Study of Urban Growth
(unpublished Ph.D. dissertation, Syracuse University, 195·5), p. ISO.
4. Interview with productivity services manager, Glaxo factory, Aligarh, on
July 26, 1983.
5. Ashutosh Varshney, "Civic Life and Ethnic Conflict: Hindus and Muslims in
India" (unpublished draft manuscript, February 1998), p. 133.
6. H. R. Nevill, in the 1926 Aligarh Gazetteer, noted the importance of the "indus-
try of lock-making, for which Aligarh is famous throughout India"; p. 204.
7. Centre for Research in Rural and Industrial Development, "Communal Violence
and Its Impact on Development and National Integration" (unpublished, Chandigarh,
[1983?]), p. 28.
8. Mann, Boundaries and Identities, p. 84.
9. Interview with professor and head of the Department of Political Science,
Barahseni College, on September 12, 1962.
10. Mann, Boundaries and Identities, p. 84. The term Bania refers to a specific caste

category; it is also used as a generic term for castes that engage in business.
11. Both are Hindu castes, a great many of whose members are traders and
businessmen.
12. Mann, Boundaries and Identities, p. 85.
13. Mann, Boundaries and Identities, p. 85.
14. Mann, Boundaries and Identities, p. 86. The Centre for Research report,
"Communal Violence," pp. 46-47, describes this interdependence in considerable
detail, noting that "Kohlis [sic] are mainly involved in casting activities, Muslims are
 444 / Notes to Pages 200-203

more or less exclusively doing polishing work and in the intermediate stage," "the
assembling of the final products is done by Muslim Saifis, Maithel Brahmans, and
Jatavs," and the "suppliers are mainly ... Varshneys."
15. Mann, Boundaries and Identities, p. 86.
16. Mann, Boundaries and Identities, p. 103.
17. Mann, Boundaries and Identities, p. 124.
18. See also Centre for Research, "Communal Violence," p. 46, and Violette Graff,
"Religious Identities and Indian Politics: Elections in Aligarh, 1971-1985," in Andre Wink
(ed.), Islam, Politics and Society in South Asia (New Delhi: Manohar, 1991), p.168, tn. 3.
19. For example, Mann reports that, in the 1978 rioting, "in Sarai Kaba and Sarai
Miyan (both dominated by Muslim Qureshis), ... the Kolis, a poor and low Hindu
caste, were attacked"; Boundaries and Identities, p. 178.
20. Mann, Boundaries and Identities, pp. 130-31.
21. Mann, Boundaries and Identities, p. 79.
22. One hundred and fifteen out of 171 registered factories, according to Mann,
Boundaries and Identities, p. 80.
23. Graff, "Religious Identities and Indian Politics," p. 148.
24 This bazaar, named Russellganj (hence Rasalganj in Hindi) after a former British
collector (district magistrate), was characterized in the Gazetteer as "the principal bazar"
of the city; Nevill, Aligarh Gazetteer, p. 200. It remains so today.
25. Interview, September 12, 1962.
26. S. K. Ghosh, Communal Riots in India: Meet the Challenge Unitedly (New Delhi:
Ashish, 1987), p. 214.
27. Mann, Boundaries and Identities, p. 177.
28. Times of India, November 21, 1978: Ghosh, Communal Riots in India, p. 214;
Mann, Boundaries and Identities, p. 177.
29. Government ofIndia, Minorities Commission, First Annual Report for the Year
Ending 31st December, 1978 (New Delhi: Government of India Press, 1979), pp. 82-83.
30. Times of India, November 22, 1978.
31. Centre for Research, "Communal Violence," pp. 22-23.
32. Centre for Research, "Communal Violence," pp. 34-35·
33. Centre for Research, "Communal Violence," p. 53.
34. The Jan Sangh is the predecessor party of the BJP, which is now the party of
militant Hindu nationalism, in power at the Center (2001) and in several Indian states.
There is also a rump Jan Sangh, consisting of persons who objected to the loss of its
identity that occurred when the original Jan Sangh joined with the Janata Party in
1975 and then reorganized as the SIP when the latter party collapsed. The term Jan
Sanghi also continues to be used today as a term for Hindu communalists in politics
by older persons who remember the original Jan Sangh.
 Notes to Pages 203-209/445

35. Interview, Aligarh, July 19, 1983.

36. Interview with Colonel Zaidi, Vice-Chancellor, AMU, January 20, 1962, and
interview, July 19, 1983. Colonel Zaidi put the matter as follows: "The [pay] grades of
teachers in the colleges are lower than in the AMU. The amenities, buildings, etc. are
much better. Whereas the local colleges are shabby and squalid, the AMU looks like
a green oasis"; from notes.
37. "Why the Medical College was selected" (typewritten sheet provided by the
Public Relations Office, AMU, January 1991).
38. Interview in Aligarh, July 24, 1983.
39. See also Varshney, "Civic Life and Ethnic Conflict," p. 168fn.: "There are two
large business (vaishya) castes: the Barehsenis [sic) (12 percent), and the Agarwals (6
percent of the population). It is also believed that the latter traditionally supported
the Congress but in all probability have switched to the BJP by now. The former have
been big supporters of Hindu nationalism for a long time."
40. Interview in Aligarh, January J, 1991.
41. Government of India, Minorities Commission, First Annual Report, p. 79.
42. Centre for Research, "Communal Violence," pp. 34-35.
43. Interview, July 19, 1983.
44. See Jim Masselos, "The Bombay Riots of January 1993: The Politics of Urban
Conflagration," South Asia 17 (Special Issue, 1994),79-86, where this type of expla-
nation, among others, is discussed in connection with these riots.
45. E. A. Mann, "Religion, Money and Status: Competition for Resources at the
Shrine of Shah Jamal, Aligarh," in Christian W. Troll (ed.), Muslim Shrines in India:
Their Character, History and Significance (Delhi: Oxford University Press, 1989), p.156.
46. Graff, "Religious Identities and Indian Politics," mimeographed version,
Pp·72-74·
47. This issue of privacy, particularly for women, arises especially because of the
fact that, during summer months, many people in such neighborhoods sleep in cots
out in the open.
48. Interview, July 19, 1983.
49. Interview with station officer, Sasni Gate, Aligarh, July 24, 1983·
50. Graff, "Religious Identities and Indian Politics," published version, p. 168, fn. 3.
51. Mann, Boundaries and Identities, p. 57.
52. Mann, Boundaries and Identities, p. 85.
53. The situation changes, however, if a Muslim businessman moves to the Civil
Lines area where intercommunal relations are not charged with the same level of sus-
picion and hostility as in the old city and where, in fact, "social relations" may develop
between Hindus and Muslims in the same social stratum; Mann, Boundaries and
Identities. p. 86.
 446/ Notes to Pages 209-227

54. Mann, Boundaries and Identities, p. 107.

55. Thus, for example, Mann notes that many Qureshis, traditionally butchers,
have moved into the business of "inter-state transportation," that is, into trucking;
Mann, Boundaries and Identities, p. 127.
56. Mann, Boundaries and Identities, p. 180.
57. Pars Ram, A UNESCO Study of Social Tensions in Aligarh, 1950-1951, edited
and with an introouction by Gardner Murphy (Ahmedabad: New Order Book Co.,
1955), p. 188.
58. Varshney, "Civic Life and Ethnic Conflict," p. 164.
59. Interview, January 3,1991.
60. Mann, Boundaries and Identities, p. 81; Mohd. Izhar Ahmad, "An Over All [sic]
View of Muslim's [sic] Participation in Small Scale Industries of Aligarh District,"
mimeographed, Aligarh Muslim University (1982-83).
61. Mann, Boundaries and Identities, pp. 84-85.
62. Interview in Aligarh on November 20, 1997 (taped in English).
63. Interview, November 20, 1997.
64. I am not sure that this man's responses concerning the absence of economic
strains in Hindu-Muslim relations were honest. He seemed to want to assure me as
an American that there was no conflict in India, a kind of response that I sometimes
get when I have not established sufficient rapport with someone I have not met before.
65. Ahmad, "Small Scale Industries."
66. Interview in Upar Kot, January 3,1991 (taped in English).

8 I RIOTS AND .EI. .ECTIONS

1. Steven I. Wilkinson, "The Electoral Origins of Ethnic Violence: Hindu-Muslim

Riots in India" (unpublished Ph.D. dissertation, Harvard University, 1997), chapter
1, figure 1-3. Wilkinson believes that the causal chain is from intense electoral com-
petition to riots and discounts the counterargument "that riots cause close electoral
contests"; personal communication. Leaving out the word "cause," the evidence from
Aligarh to be presented below suggests that the sequence is oft.en the reverse, namely,
that riots precede close electoral contests.
2. Paul R. Brass, Language, Religion, and Politic.~ in North India (London: Cambridge
University Press, 1974), p. 265.
3. The atypical 1955 bye-election has been left: out of this chart.
4. The correlation coefficient between the valid votes turnout for Aligarh and the
state as a whole is a rather low .396 (N=13, p=.180).
5. Note that the larger the interval, the lower the degree of competitiveness, and
 Notes to Pages 229-248 /447

vice versa, so that a declining slope on the chart registers an increase in interparty
competitiveness.
6. It should be noted also that a communal riot did precede this election-held
sometime before July 31,1955, the exact date not having been published-but by more
than a year, in June 1954.
7. Wilkinson, "The Electoral Origins of Ethnic Violence," figure 1-6.
8. Brass, Language, Religion, and Politics in North India, p. 265.

9 I THE PRACTICE OF COMMUNAL POLITICS

1. Interview in Aligarh on January 3, 1991 (taped in English).

2. Interview, January 3, 1991.
3. The Jan Sangh, however, had been in power in coalition governments in the
period of governmental instability in the state between 1967 and 1971.
4. Interview, January 3,1991.
5. Lok and janata in Hindi are synonyms for "people."
6. Paul R. Brass, Language, Religion, and Politics in North India (Cambridge:
Cambridge University Press, 1974), p. 265.
7. Interview in Aligarh on July 24,1983 (in Hindi, from notes in English).
8. Interview with inspector of police, Bannadevi Police Station, AJigarh, on July
23,1983 (in Hindi, from notes).
9. Interview in AIigarh on July 22, 1983 [taped in English].
10. Which can be rendered nonpejoratively as non-Muslim, but literally carries a
pejorative meaning in English as an unbeliever or infidel. The latter term in English,
especially, has acquired a strongly pejorative meaning in relation to non-Muslims, even
stronger than its literal meaning of infidelity or unfaithfulness to God.
n. Interview, Iuly 24, 1983.
12. Lekin maiNne jivan parishani bahut hoti haiN; interview, Aligarh, November
21,1997 (taped in Hindi with running translation by Professor Mathur).
13. Interview, Manik Chauk, June 14,1993 (taped in Hindi with running transla-
tion by Jayati Chaturvedi).
14. Interview, November 21, 1997.
15. References to Navrnan's son's activities during riots and as "a Bajrang Dal chief"
from interview, January 3, 1991; other citations from interview, June 14, 1993.
16. Interview, June 14, 1993.
17. Interview, January 3,1991.
18. Interview with retired professor of education, Barahseni College, Aligarh, and
RSS member, in Aligarh on November 21, 1997 [taped in English].
 448/ Notes to Pages 248-267

19. Interview, January 3, 1991.

20. This term was used at the time openly by the top national leadership of the BJP,
including L. K. Advani in an interview with me on December 20, 1989, in New Delhi.
21. Interview, Aligarh, July 20, 1983 (taped in English).
22. Interview, July 22,1983; all emphases in original.
23. Police inspector, Bannadevi, interview, 1983.
24. Interview with station officer, Sasni Gate, at the Sasni Gate Police Station,
Aligarh, July 24, 1983; taped in Hindi with running translation by Pallav Kumar.
25. Interview, Iuly 24, 1983.
26. Interview on Iuly 26, 1983.
27. Inte.rview, Aligarh, July 30,1983; taped in English.
28. Interview, Aligarh, November 20, 1997; taped in English.
29. Interview, November 21, 1997; taped in English.
30. Interview with RSS members, November 21,1997; this section taped in English.
31. Interview, March 30, 1999; taped in Urdu, with running translation by Aftab
Abmad.
32. Interview, January 3,1991.
33. Interview, July 22, 1983.
34. I also observed from their dress that they were all Hindus.

10 I COMMUNAUZATION AND POLARIZATION

1. Interview with professor and head of the Department of Political Science,

Barahseni College, Aligarh, on September 12, 1962.

2. The mohallas are as follows: Ghas ki Mandi, Rangrezan, and Turkman Gate in
the old Jaiganj ward; Sabzi Mandi, Sarai Tahsil, Chauk Bundu Khan, Sunhat, Pir
Ataullah, Tila, Sheikhan, Purani Kachehri, Ghosian, and Bani Israilan in the old Kan-
wariganj ward; and Turkman Gate in the Turkman Gate ward. Of these, five can be
located precisely on the available maps in the broad Upar Kot area (Map 2). In con-
trast, Dr. Bashir's bottom five polling stations comprised the six mohallas of Sarai Raja
Ram, Sudamapuri, Begambagh, Jaiganj, Baikunthnagar, and Vishnupuri, of which
all that can be located precisely on Maps 1and 3 are in the heart of the predominantly
Hindu area, near the several Hindu-dominated colleges whose students contributed
to the crowds mobilized in the October riots.
3. See Appendix Table A.3.
4. Rae's index of fractionalization (Fe) for this election was .63 for the constituency
as a whole, but the mean for all 96 polling stations was ,50, which, in Rae's index, signifies
perfect two-party competition.
 Notes to Pages 267-291 /449
5. Part of the material in this paragraph was taken directly from Paul R. Brass,
"Caste, Caste Alliances, and Hierarchy of Values in Aligarh District," in Paul R. Brass,
Caste, Faction and Party in Indian Politics, vol. 2: Election Studies (Delhi: Chanakya,
1985), p. 228.
6. Interview with Jarrar Haider, lawyer, Congress candidate for Parliament from
Aligarh constituency, at his residence, Dodhpur, September 17, 1962.
7. Much of this paragraph was taken from Brass, "Caste, Caste Alliances, and
Hierarchy of Values in Aligarh District," pp. 228-29.
8. Interview on March 31, 1991; taped.
9. Interview, Manik Chauk, June 1.,4, 1993; taped in Hindi with running transla-
tion by Jayati Chaturvedi.
10. The problems posed for BJP dominance in this constituency by the voting pat-

tern in these mohallas are, however, somewhat complex. Had even all the votes of the
SJP candidate gone to Mohammad Sufiyan, Navrnan would still have won the seat.
The fact that Muslim votes were divided, moreover, increased Navman's margin of
victory. On the other hand, if these mohallas were eliminated from the constituency
altogether, Navman's margin of victory would have been even larger; he would have
polled 48.40 percent and Sufiyan would have polled 41.78 percent, thus increasing
Navman's margin of victory by 1.75 percent.
11. Interview on March 30, 1999; taped.
12. Interview on March 31, 1999; taped in English.

11 I COMMUNAL SOLIDARITY AND DIVISION

AT THE LOCAL LEVEl.

1. Sarai Bairagi is not listed on this map. It is located between the railway line and
the old city; Howard F. Hirt, Aligarh, U. P., India: A Geographic Study of Urban Growth
(unpublished Ph.D. dissertation, Syracuse University, 1955), p. 105.
2. I have, of course, been arguing throughout this volume that large-scale riots
are often themselves political events, though they are generally treated as if they were
societal, implying a dear separation between society and politics. The relationship is,
therefore, often in fact one between rioting, as a societal/political event, and inter-
party competition, an electoral/political event.
3. Paul R. Brass, "Caste, Caste Alliances, and Hierarchy of Values in Aligarh
District," in Paul R. Brass, Caste, Faction and Party in Indian Politics, vol. 2: Election
Studies (Delhi: Chanakya, 1985), pp. 207-79. Further evidence on these point~ is pro-
vided in Appendix C in this volume, which takes the electoral analysis down to the
even lower level of mohallas within a dty ward.
 450 / Notes to Pages 296-312

12 I THE DECLINE OF COMMUNAL VIOLENCE AND THE

TRANSFORMATION OF EI.ECTORAL COMPETITION

1. Extrapolating from the change in the total population of the city, which mul-
tiplied by 3.39 in this period.
2. They include Gambhirpuri (part), Gandhinagar, Adda Hathras, Madar Gate,
Dwarkapuri, Achal Road (but not Achal Talab), and Mahendranagar (part) in ward
15 and Gambhirpuri (part), Hanumanpuri, and Mahendranagar (part) in ward 17.
3. Interview on November 21, 1997; taped in English.
4. "Ne, vote dale nahiN, log 50sti rahe, laziness"; interview, Aligarh, November 21,
1997 [taped in Hindi with running translation by Professor Mathur].
5. Interview, November 21, 1997.
6. Interview with RSS members, in Aligarh, on November 21, 1997.
7. However, it was not until 1998 that the intensity of interparty competition declined
in the Aligarh segment of the Lok Sabha constituency. In 1996, the contest was among
the closest in the dataset, only 3.98 percentage points separating the winning candi-
date from the runner-up, making this contest the fifth most competitive among the
21 elections. Then, in 1998, the interval increased substantially to 13.76 percent.

13 I RIOT INTERPRETATION, BLAME DISPLACEMENT,

AND THE COMMUNAL DISCOURSE

1. Interview in New Delhi, June 20, 1991.

2. For the uses of the rhetoric of blame for communal and other riots in India,
with specific reference to charges and countercharges before the election of 1980, see
Paul R. Brass, Theft of an Idol: Text and Context in the Representation of Collective
Violence (Princeton, N.J.: Princeton University Press, 1997), pp. 131-32.
3. The events of September 11, 2001, in New York occurred after this was written.
Ever since, as the bodies of the victims of the attack on the World Trade Center have
been identified, the New York Times has, day by day, for every person killed, done
exactly what I have proposed here: included a photograph of each person and a sum-
mary characterization of the person in the words of one of the bereaved. The effect
of such portrayals clearly is to humanize an enormous tragedy, as opposed to dehu-
manizing it by reducing it to a body count. Such portrayals, personal experiences, and
eyewitness testimonies have sometimes been provided in films and books produced
in India, notably concerning the anti-Sikh pogrom of November 1984, in Delhi, on
which see Uma Chakravarti and Nandita Haksar, The Delhi Riots: Three Days in the
Life of a Nation (New Delhi: Lancer International, 1987).
4. Excerpts and paraphrases from inte.rview on July 20, 1983; taped in English.
 Notes to Pages 313-329 /451

5. Interview on July 22, 1983; taped in English.

6. Interview in Aligarh on July 24, 1983; in Hindi, from notes in English.
7. Interview on July 30, 1983; taped in English.
8. Interview, July 30, 1983.
9. Interview on July 26, 1983; taped in English.
10. Interview with senior superintendent of police, Aligarh, on July 21, 1983; in
English, from notes.
n. Interview with station officer, Sasni Gate, at the Sasni Gate Police Station,
Aligarh, July 24, 1983; taped in Hindi with running translation by Pallav Kumar.
12. Interview on July 22, 1983; taped in English.
13. Navrnan interview, 1983.
14. Interview, July 20, 1983.
15. SSP interview, 1983.
16. Interview with station officer, Civil Lines, on July 25, 1983; taped in English.
17. Sasni Gate station officer interview, 1983.
18. Interview with inspector of police, Bannadevi Police Station, Aligarh, on July
23,1983 [in Hindi, from notes].
19. Interview with district magistrate, Aligarh, at his residence, Aligarh, July 30,
1983; in English, from notes.
20. Conversation in Aligarh, July 26, 1983; in English, from notes.
21. However, the AMUSU (AMU Students Union) president, cited earlier, chal-
lenged the militant Hindus of the town during the 1978 riots to make a surprise visit
to any hostel or hostels of their choice in which they claimed that bombs were being
made and arms kept. They did so on one occasion, but found neither. However, he
did not claim to know about the situation in the hostels during the later riots. Interview
in New Delhi on June 20, 1991; taped in English.
22. Interview, July 22, 1983.
23. Interview with DSP, intelligence, at his office, Aligarh, July 30, 1983; in English,
from notes.
24- Interview with district magistrate, Aligarh, at his residence, Aligarh, July 30, 1983.
25. Interview with RSS members, Aligarh, November 21, 1997.

14 I POLICE VIEWS OF HINDU-MUSUM VIOI,ENCE

1. Commonly, but not universally. For a defense of the PAC as an honest, well-
trained, and highly competent force that ends up with the blame for actions ordered
by others, see the article by K. S. Sastry, "A Much Maligned Force," Times of India,
May 7, 1979. Sastry points out that the commands to the PAC forces to fire on riot-
ing crowds are actually given by police officers or the civilian administration.
 452/ Notes to Pages 330-344

2. Kya nam de diya gaya bana? Yah police action tim jo MusalmanoN ke khilaf am
tor par [kiya jata hai}. (Loose translation: What name has been given to this? This was
a police action, which is generally directed against Muslims.) Interview in Aligarh on
January 2, 1991.
3. What follows are generalizations from specific incidents that have been reported
to me in my interviews.
4. All these particular incidents and several others come from my interviews
(including of riot victims) in Upar Kot, Aligarh, on January 3,1991.
5. From an incident reported in interview in New Delhi, June 20, 1991.
6. Anonymous, "Communal Riots and Minorities," unpublished, n.d. [1980?],
p.ll.
7. Literally, youths, but meaning the lowest rank of police and armed forces recruits.
8. Anonymous, "Communal Riots and Minorities," p. 12.
9. The reader should note that the police discussed in this section are the district
and station police officers and men, not the PAC jawans, who are not accessible to
interviews from foreigners. For that matter, I have never seen any reports from Indian
scholars or newsmen of the attitudes of the PAC recruits and officers.
10. Interview with inspector of police, Bannadevi, Aligarh, on July 23, 1983.
ll. It is certain that this police inspector has not read the literature on rioting for
fun and profit.
12. That is, section 362 of the Indian criminal code, which refers to abduction.
13. For another specific example of what I am calling here "the police view" from
my interviews in Kanpur City, see Paul R. Brass, Theft of an Idol: Text and Context in
the Representation of Collective Violence (Princeton, N.J.: Princeton University Press,
1997), pp. 248-50, and the same volume, passim, for numerous accounts of police
behavior and practice in riots in five districts of north India, including Aligarh.
14. Interview with district magistrate, Aligarh, at his residence, Aligarh, July 30,
1983·
15. Interview with DSP, intelligence, at his office, Aligarh, July 30,1983.
16. This charge is not without merit. The great Bombay riots many years later in
December 1992 were referred to by some as the "BBC riots," because of its broadcast
on TV every hour on the hour of the scene of the mosque at Ayodhya being demol-
ished, which was said to have inflamed Muslims and brought them onto the streets
in cities and towns throughollt the country.

15 I THE ROLE OF THE MEDIA

1. From Governor's fortnightly report to Mountbatten, dated June 9, 1947; India

Office Library records LlP&J/sI276.
 Notes to Pages 344-348/453
2. Anonymous, "Aligarh Muslim University, A Victim of Rumours" (mimeo-
graphed sheet provided to me in January 1991).
3. Subrata Banerjee, "Communalism as a Commodity in Media Industry," in
Pramod Kumar (ed.), Towards Understanding Communalism (Chandigarh: Centre for
Research in Rural and Industrial Development, 1992), p. 385.
4. Cf. Banerjee, "Communalism as a Commodity," pp. 386-88.
5. People's Union for Civil Liberties, "Communal Riots in Aligarh, Dec. 1990-Jan.
1991," PUCL Bulletin (March 1991), pp. 13-27.
6. Ashish Banerjee, "'Comparative Curfew': Changing Dimensions of Communal
Politics in India," in Veena Das (ed.), Mirrors of Violence: Communities, Riots and
Survivors in South Asia (Delhi: Oxford University Press, 1990), p. 53. On the "inflamma-
tory role" of "the Hindi press" in the Jabalpur riot of 1961, see also Asghar Ali Engineer,
Lifting the Veil: Communal Violence and Communal Harmony in Contemporary India
(Hyderabad: Sanaam, 1995), p. 3l.
7. Anonymous, "Riots in Aligarh," undated mimeographed study of 7 pages given
to me in 1983 by a former Aligarh district magistrate, p. 6.
8. Violette Graff, "Religious Identities and Indian Politics, a Case-Study: Aligarh
(1971-1981)," mimeographed paper (Paris: Fondation Nationale des SciencesPolitiques
etCentre d'Etudes de l'Inde et de I'Asie du Sud [EHESS, n.d.]), pp. 69-71.
9. Iqbal A. Ansari, "Introduction," in Iqbal A. Ansari (ed.), Communal Riots: The
State and Law in India (New Delhi: Institute of Objective Studies, 1997), p. xv.
10. Ansari, Communal Riots, p. xv, and Rafiq Khan and Satyaprakash Mittal, "The

Hindu-Muslim Riot in Varanasi and the Role of the Police," in Asghar Ali Engineer
(ed.), Communal Riots in Post-Independence India (Hyderabad: Sangam Books, 1984),
Pp·311- 12.
n. Regrettably, the work of Varshney and Wilkinson has transferred to the study
of collective violence in India this methodologically unsound form of organizing data
on riots; see Ashutosh Varshney and Steven 1. Wilkinson, Hindu-Muslim Riots
1960-93: New Findings, Possible Remedies (New Delhi: Frank Bros. for Rajiv Gandhi
Institute for Contemporary Studies, 1996), pp. 8-9, 55, and at many other places in
their joint and separate works.
12. Subrata Banerjee, "Communalism as a Commodity," pp. 387-88.
13. People's Union for Democratic Rights, Walled City Riots: A Report on the Police
and Communal Violence in Delhi, 19-24 May, 1987 (Delhi: PUDR, 1987), pp. 33-34.
14. Satish Saberwal and Mushirul Hasan, "Moradabad Riot~, 1980: Causes and
Meanings," in Engineer, Communal Riots, p. 226.
15. Subrata Banerjee, "Communalism as a Commodity," pp. 388-89.
16. Sub rata Banerjee, "Communalism as a Commodity," p. 392, refers to a specific
article by the respected free-lance journalist and writer, Prem Shankar Jha, writing in
 4541 Notes to Pages 349-357

the aftermath of the Moradabad riots, in which he expressed "great sympathy for the
Muslims" along with anticipation of a "Hindu backlash" and a statement that "Hin~
dus of all shades and castes" felt like a "beleaguered majority in their homeland."
17. Sub rata Banerjee, "Communalism as a Commodity," pp. 388 and 394-95, pro-
vides several examples.
18. Amrita Basu, "Why Local Riots Are Not Simply Local: Collective Violence and
the State in Bijnor, India, 1988-1993," Theory and Society 24 (1995), 57-58; see also Asghar
Ali Engineer, "The Causes of Communal Riots in Post-Independence India," in
Engineer, Communal Riots, p. 36, and Khan and Mittal, "The Hindu-Muslim Riot in
Varanasi," p. 310.
19. Asghar Ali Engineer, "An Analytical Study of the Meerut Riot," in Engineer,
Communal Riots, p. 280.
20. Madhu Kishwar, "Gangster Rule: The Massacre of the Sikhs," in Religion at
the Service of Nationalism and Other Essays (Delhi: Oxford University Press, 1998),
p.21.
21. Thus, in the aftermath of the 1956 Aligarh riots, the Times of India, October
6, 1961, citing the remarks of the then Home Minister of U .P., Charan Singh, remarked
upon how the actions of "thoughtless young men" at the AMU had set "the fire of
communal frenzy.... ablaze."
22. Times of India, October 10, 1961. The Times ofIndia editors revealed their true
feelings even more dearly by their selection for publication, in their letters column a
couple of weeks later, of the following: "Before partition Aligarh University was noto-
rious for the encouragement it gave to separatist tendencies among the Muslims. In
fact a substantial part of the credit for the creation of Pakistan must go to the activ-
ities of this university. Aligarh will remain a trouble spot as long as its basic charac-
ter does not undergo a sea-change." Signed by one J. M. Kochar in Times of India,
October 23, 1961.

16 I THE PERSISTENCE OF HINDU-MUSLIM VIOLENCE

1. Several incidents of the latter type are discussed in Paul R. Brass, Theft of an
Idol: Text and Context in the Representation of Collective Violence (Princeton, N.J.:
Princeton University Press, 1997).
2. Most secondary sources and many of my interview respondents use the English
word. The word for tension in Hindi is tanav, but it did not appear often in my inter-
views. However, my informants frequently translated other expressions from Hindi
that did not contain the word tanav into the English word "tension." For example,
one of my respondents remarked that, in the town of Khurja before a riot took place,
Wahan abhi tak koi Hindu-Muslim baNdh nahiN hua (There, until then, there was
 Notes to Pages 357-359 /455
no Hindu-Muslim [hindrance, difficulty, distress, etc.]). My informant translated the
sentence as follows: "[Khurja] never had any communal tension." I sometimes intro-
duced the word "tension" into an interview myself by declaring that I was "interested
in the causes of communal tension in the city of Aligarh, and why these riots take
place."
3. The American Heritage Dictionary of the English Language.
4. The metaphor here is similar to that generated by Tilly and his colleagues, who
liken "the repertoire of collective action to a game that involves a set of basic rules
around which a considerable degree of extemporization is not only permitted but
required"; Mark Traugott, "Barricades as Repertoire: Continuities and Discontinuities
in the History of French Contention," Social Science History 17, no. 2 (Summer 1993),
P·30 9·
5. Vide I. Michael Aronson, "The Anti-Jewish Pogroms in Russia in 1991," in John
D. Klier and Shlomo Lambroza (eds.), Pogroms: Anti-Jewish Violence in Modem Russian
History (Cambridge: Cambridge University Press, 1992), p. 45, who remarks in con-
nection with the 1881 Easter pogrom in Elisavetgrad: "The start of Easter had been
greeted with tensions and rumors."
6. Suranjan Das, Communal Riots in Bengal 1905-1947 (Delhi: Oxford University
Press, 1991), p. 57.
7. The Chicago Commission on Race Relations, "The Negro in Chicago: A Study
of Race Relations and a Race Riot," in Allen D. Grimshaw (ed.), Racial Violence in the
United States (Chicago: Aldine, 1969), p. 104.
8. Sudhir Kakar, The Colors of Violence: Cultural Identities, Religion, and Conflict
(Chicago: University of Chicago Press, 1996), p. 35.
9. The association is quite pronounced in the revisionist history of the Russian
pogroms of 1881, on which see especially I. Michael Aronson, Troubled Waters: The
Origins of the 1881 Anti-Jewish Pogroms in Russia (Pittsburgh: Pittsburgh University
Press, 1990), and "The Anti-Jewish Pogroms in Russia in 1881," in Klier and Lambroza
(eds.), Pogroms. All the contributions on the 1881 pogroms in the latter volume share
this view. This revisionist school has set out to debunk the idea that the 1881 pogroms
were the result of a government-sponsored conspiracy. However, it has substituted
one simplistic view for another, the theory of uncoordinated, unplanned, spontaneous
action of the ignorant masses fed on mmors that floated naturally among them all
the time, for a theory resting on the coordinated, detailed, centralized planning of the
Tsarist state.
10. Neil J. Smelser, Theory of Collective Behavior (New York: Free Press, 1962),
P·240 •
11. Stanley J. Tambiah, "Presidential Address: Reflections on Communal Violence
in South Asia," Journal of Asian Studies 49, no. 4 (November 1990), p. 757.
 456/ Notes to Pages 359-365

12. Michael Keith, Race, Riots and Policing: Lore and Disorder in a Multi-Racist
Society (London: UCL Press, 1993), p. 63.
13. Keith, Race, Riots and Po/icing, p. 71.
14. Pars Ram, A UNESCO Study of Social Tensions in Aligarh, 1950-1951, ed. with
an introduction by Gardner Murphy (Ahmedabad: New Order Book Co., 1955), p. 15.
15. Aronson, Troubled Waters, pp. 95ff, and Ram, UNESCO Study, p. 83.
16. American Heritage Dictionary.
17. Ram, UNESCO Study, pp. 100-101.
18. Radiance, October 22, 1978.
19. Hindustan Times, November 10, 1978, and Times of India, November 21, 1978.
20. People's Union for Civil Liberties (PUCL), "Communal Riots in Aligarh, Dec.
1990-Jan. 1991," PUCL Bulletin (March 1991).
21. Tambiah, "Presidential Address," pp. 746-47.
22. Ashutosh Varshney and Steven I. Wilkinson, Hindu-Muslim Riots 1960-93:
New Findings, Possible Remedies (New Delhi: Frank Bros. for Rajiv Gandhi Institute
for Contemporary Studies, 1996), pp. 39-40. During the 1919 Chicago riots, it was noted
that "the press was responsible for giving wide dissemination to much of the
inflammatory matter in spoken rumors, though editorials calculated to allay race hatred
and help the forces of order were factors in the restoration of peace"; Chicago
Commission, "The Negro in Chicago," in Grimshaw, Racial Violence, p. 104.
23. The station officer, Sasni Gate, used an example of crowd action directed
against him to illustrate how rumor leads to massing and attack. He noted how one
of his constables encountered "some Jatavs, who were all drunk and ... had a small
altercation among themselves." The constable reported the matter back to him, the
station officer went to the scene without a supporting force, and, "as he entered the
mohalla, the rumor spread around that the police has come to arrest the people. So
they started throwing stones at him, and he got injured." There was then a "small
fight (ihagara) between the people and the police"; the SO referred to the massing
of people that occurred under these circumstances as a stampede (bhaggi). This scene,
of course, could literally come out of countless police-populace confrontations in
black neighborhoods in the United States as well. Is it a stampede or is it rational
action of communal defense of its members on the part of communities used to police
harassment?
24. Veena Das, "Introduction: Communities, Riots, Survivors-The South Asian
Experience," in Veena Das (ed.), Mirrors o/Violence: Communities, Riots and Survivors
in South Asia (Delhi: Oxford University Press, 1990), p. 28.
25. The shouting of political slogans in India is a virtual popular art form, which
itself involves a specialized role, though everybody knows how to do it. Persons move
in the middle or at the head of processions or stand on the hoods of cars, bellowing
 Notes to Pages 370-374/457

out the opening line of a statement-response sequence, to which the crowd yells its
response in unison. The person who gives the opening statement also is likely to throw
his whole body into his performance, gesticulating pronouncedly and moving his legs
in a dancelike manner. A simple example of such a statement-response slogan would
be the bellowing of a hated person's name or the name of the opposite community
itself, to which the crowd then responds, "Murdabad" (Death to). There are other stan-
dard violence-inciting slogans, such as "Khoon Ka Badia Khoon Se Lenge" (Blood
for Blood), a direct call for violent revenge, assault, and death. This slogan was cer-
tainly used during the Bhura procession. Although some of the slogans are standard,
there is a fair amount of variation and imagination in them to fit each occasion.
26. As in so many instances, there is a precise parallel here also with Russian
pogroms. Writing about the 1881 riots in Russia, Aronson remarks as follows: "The
Jews' well-to-do business competitors-merchants, industrialists, and professional
people -while not participating actively in the riots themselves, may have contrib-
uted by spreading rumors, reading antisemitic newspaper articles aloud, and even assist-
ing in the impromptll organizing of rioters on the spot, by dividing them up into groups
and sending them to different parts of town." See Aronson, "The Anti-Jewish Pogroms
in Russia in 1881," in Klier and Lambroza (eds.), Pogroms, p. 49.
27. The station officer, Sasni Gate, put it this way: "The professional criminal also
benefits from this, because-people with vested interests give ... the professional crim-
inals good amount ... for killing and stabbing at these times. So they stand to-Because
their life is dependent upon crime, this offers them ample opportunity of indulging
in crime, and making more money"; interview, July 24, 1983.
28. Interview, Aligarh, January 3,1991.
29. See Paul R. Brass, "The Strong State and the Fear of Disorder," in Francine R.
Frankel et aI., Transforming India: Social and Political Dynamics of Democracy (New
Delhi: Oxford University Press, 2000), pp. 60-88.
30. See Donald E. Smith, India as a Secular State (Princeton, N.J.: Princeton
University Press, 1963); Donald E. Smith (ed.), South Asian Politics and Religion
(Princeton, N.J.: Princeton University Press, 1966); Ved Prakash Luthera, The Concept
of the Secular State and India (Calcutta: Oxford University Press, 1964); and T. N.
Madan, Modern Myths, Locked Minds: Secularism and Fundamentalism in India
(Dellii: Oxford University Press, 1997), for this and other differences between the sec-
ular state in theory and practice in the United States and India.
31. For a thorough, comprehensive summary treatment of all the proposals and
measures recommended to government~ over the years for riot prevention and con-
trol, virtually all unheeded, see Iqbal A. Ansari, Report on Communal Riots: Prevention
and Control (New Delhi: Minorities Council, 1999).
32. On Tamil Nadu and Kerala, see Steven I. Wilkinson, "The Electoral Origins
 458/ Notes to Pages 374-383

of Ethnic Violence: Hindu-Muslim Riots in India" (unpublished Ph.D. dissertation,

Harvard University, 1997), ch. 1, p. 2. Under the leadership of Laloo Yadav, Bihar, which
has had a savage history of Hindu-Muslim violence, has also entered the list of states
that refuse to countenance Hindu-Muslim riots.
33. Lance Brennan has described the changes in the attitudes and willingness of
the state governments in U.P. to prevent and control Hindu-Muslim riots since
Independence, and has argued that the will to do so declined after Independence until
Mulayam Singh Yadav came to power in 1990, though he has also noted that the will
to do so is not in itself sufficient given the communalization of the "lower echelons
of the state apparatus" that has been taking place at the same time; "The State and
Communal Violence in UP: 1947-1992," in John McGuire, et al. (eds.), Political Violence
from Ayodhya to Behrampada (New Delhi: Sage, 1996), pp.127-42. See also Wilkinson,
"The Electoral Origins of Ethnic Violence," ch. 1, p. 32, on the change in the willing-
ness to act effectively to prevent and control riots in U.P. in the mid-1990S.
34. Government ofIndia, Minorities Commission, FirstAnnual Reportfor the Year
Ending 31St December, 1978 (New Delhi: Government of India Press, 1979), p. 94.
35. Wilkinson's figures on the change in the composition of the U.P. police force
indicate the dimensions of the problem; see Appendix Table A.4. His figures go up to
1981 only, but they would not be significantly different today.
36. Marc Gaborieau, "From AI-Beruni to Jinnah: Idiom, Ritual and Ideology of
the Hindu-Muslim Confrontation in South Asia," Anthropology Today 1, no. 3 (1985),
pp. 7-14. In contrast to Gaborieau, who sees riots as arising from the separate iden-
tities of Hindus and Muslims, van der Veer sees them as integral to the construction
of opposed identities; Peter van der Veer, "Riots and Rituals: The Construction of
Violence and Public Space in Hindu Nationalism," in Paul R. Brass (ed.), Riots and
Pogroms (New York: NYU Press, 1982), pp. 154-77.
37. Bill Buford, Among the Thugs (New York: W. W. Norton, 1992).
38. Even ordinary behavior in India is sometimes defined by reference to a clause
in the Indian Penal Code. Most famous in this regard, a clause known by everyone,
including any foreigner with any serious knowledge ofIndia, is clause 420, which refers
to cheating. Persons and acts in everyday life-from overcharging on a taxi ride to
every form of duplicity-are referred to as char sau bis (four hundred and twenty,
spoken in English as "four-twenty").
39. E. Valentine Daniel, Charred Lullabies: Chapters in an Anthropography of
Violence (Princeton, N.J.: Princeton University Press, 1996), pp. 28-60.
40. On the importance of the latter in Islamic history, see Francis Robinson,
"Islamic History as the History of Learned and Holy Men," paper presented at the
University of Washington in 1985 and as a lecture at the Centre d'Etudes de l'Inde et
de J'Asie du Sud, "Lettre d'information, no. 5, Avril 1986."
 Notes to Pages 384-389 /459

41. In my first visit to India after the destruction of the mosque, I asked many
people the same question over and over concerning that event: "Where were you
when you first heard the news, what did you feel in your heart, and what did you
do?" Several militant Hindus~BJP and RSS--replied in words to the effect that they
felt that they had at last been released from slavery.

POSTSCRIPT

1. Frontline, 19: 8 (April 13-26, 2002) points the finger directly at Chief Minister

Narendra Modi, referring to him as "apparently, the alleged mastermind of the

pogrom."
2. Frontline 19, no. 6 (March 16-29, 2002).
3. Asghar Ali Engineer, "Role of Police in Gujrat [sic] Carnage," Secular Perspective,
June 16-30, 2002.
4. Frontline 19, no. 8 (April 13-26, 2002).
5. "Upwards of 1,500 Dead," Frontline 19, no. 6 (March 16-29,2002). The figure
of 2,000 was attributed to the British High Comrnissionand announced over National
Public Radio in the U.S. and the BBC World Service during the riots.
6. Times of India, March 27, 2002.
7. The number was given as fifteen in Frontline 19, no. 8 (April 13-26, 2002) and
as sixteen in Frontline 19, no. 12 (June 8-21, 2002).
8. Frontline 19, no. 11 (May 25-June 7, 2002).
9. Times of India, March 27, 2002.
10. Frontline 19, no. 8 (Aprii13-26, 2002).

11. "An independent fact finding commission of eminent citizens"-professors

from the Iawaharlal Nehru University and the Jamia Millia in New Delhi, and for-
mer senior bureaucrats in the government, including the police, none of them
Muslims-described the Gujarat carnage as "a 'systematic and planned' pogrom
against Muslims, carried out by the state government at the instigation of the Sangh
Parivar" (Deccan Herald, Aprilu, 2002). All other fact-finding missions and com-
missions that visited Gujarat during the riots, including the National Human Rights
Commission (NHRC) and the National Commission of Minorities, both statutory
bodies of me government of India, have provided evidence that supports such
charges; see Frontline 19, no. 8 (April 13-26, 2002). It also deserves note that the NHRC
chairman is former Supreme Court Justice J. S. Verma, whose decisions in some cases
affecting the RSS family of organizations have been considered favorable to those
organizations and to the Hindutva ideology; see Gary Jeffrey Iacobsohn, The Wheel
ofLaw: India's Secuklrism in Comparative Perspective (Princeton University Press, forth-
coming), ch. 7. Several other independent organizations, such as the People's Union
 460/ Notes to Pages 389-390

for Civil Liberties, whose reports on AJigarh have also been cited in this book, all
"indicted sever[elly the Gujarat government, the State police and outfits such as the
Bajrang Dal and the Vishwa Hindu Parishad"; see Frontline 19, no. 12 (June 8-21,2002).
12. Frontline 19, no. 6 (March 16-29, 2002).
13. SifY News (www.sify.com).March7. 2002.
14. New Delhi Television (www.ndtv.com). March 12, 2002.
15. The Hindu, March 24, 2002; see also Asghar AJi Engineer, "Gujrat [sic]-An
Area of Darkness," Secular Perspective, April 16-30, 2002.
16. Frontline 19, no. 7 (March 30-April12, 2002); see also Frontline 19, no. 12 (June
8-21,2002), where the papers singled out for special mention were Sandesh and Gujarat
Samachar, along with "some local cable television channels."
17. "Report of a Delegation of the Communist Party of India (Marxist) and the
All India Democratic Women's Association {AIDWA)," The Hindu, March 24, 2002,
identifies the newspaper as Sandesh. I received this report via an e-mail message from
Subhashini AJi that was forwarded to me by Clea Finkle. In subsequent notes, the
Communist Party of India (Marxist) will be abbreviated "CPI(M)," and the docu-
ment cited here will be abbreviated as the "CPI(M)-AIDWA report."
18. The names mentioned included Gujarat Home Minister Zadaphia [Jhadapiya 1
and Revenue Minister Harin Pandya (CPI(M)-AIDWA report), VHP leader Pravin
Togadia, MLA Mayaben Kodnani, and VHP General Secretary Jaideep Patel; see
Frontline 19, no. 8 (ApruI3-16, 2002).
19. This report names, for example, Bapu Jhadapiya, brother of Gujarat Home
Minister Goverdhan Jhadapiya.
20. Asghar Ali Engineer, "BIP's Riot-Free India," Secular Perspective, March 16-31,
and "Gujrat [sic]-An Area of Darkness."
21. Frontline 19, no. 8 (April 13-26, 2002).
22. Frontline 19, no. 11 (May 25-]une 7, 2002).
23. A comment in Frontline 19, no. 11 (May 25-June 7, 2002) is noteworthy here,
supporting the argument in this volume concerning the reasons why the central and
state governments in India oft.en do not act to control riots when they occur: "No one
believes that the Union government will do the sensible thing and impose Central
rule in Gujarat; it quite simply, has no reason to do so." In other words, from the
point of view of functional-political utility, there was no advantage for the govern-
ment to act. In fact, it would not be politically sensible. There were also political rea-
sons for it not to act, the most important being that any action would have amounted
to accepting the charges against its own party government in the state and would also
have eliminated its last state bastion in the country after a series of electoral defeats
for the BJP in other states.
24. Frontline 19, no. 6 (March 16-29, 2002).
 Notes to Pages 390-391/461

25. Frontline 19, no. 8 (Aprii13-26, 2002).

26. The reasons are obvious, as reported in Frontline 19, no. 12 (June 8-21,2002):
"In the NHRC's view, the appointment of K. P. S. Gill as Security Adviser to the Chief
Minister of Gujarat, Narendra Modi, implicitly confirmed that the State had failed
to bring under control the persisting violation of the rights to life, liberty, equality
and dignity of tlte people." The VHP, of course, openly criticized the appointment
of Gill, calling it "objectionable";. see Hindustan Times (www.HindustanTimes.Com),
May 9, 2002.
27. See Hindustan Times (www.HindustanTimes.com).May8. 2002; and Sify News
(www.sify.com). May 9, 2002.
28. Frontline 19, no. 11 (May 25-June 7,2002).
29. Frontline 19, no. 13 (June 22-July 5,2002).
30. Engineer, "BJP's Riot-Free India."
 INDEX

Aaj, 119, 122, 12.4-2.6, 135 161, 249: criminals at, 322-2.4: Medical
Achal Sarowar. See Achal Talab College Hospital of, 121-22, 179, 191,
Achal Talab. 51-52, 54> 142., 144, 157, 161, 361,422; pre-Independence political
296, 403, 406 role of, 62, 71-75: professors' and other
Achal Tank. See Achal Talab faculty members' views on riots, 116,
Advani, L. K., 13, 103.116-18, 133, 135, 2.47, 133. 136, 145, 167, 169. 172., 205,254-55,
355, 365, 391, 448 298, 318-19, 321, 333; and riots, 77-82.,
Agarwal(s), 48-49, 57-58, 163, 168, 200, 205, 84-88, 90, 108~13, 126, 138-39, 141-
211,264, 285,361,385: brothers (Maltesh 42, 144, 160-61, 176-79, 316, 322- 25,
and Satish), 127-29, 361, 422 378: Students Union of, 112; symbolic
Aggarwal. See Agarwal(s) importance of, 36-37. 112., u8, 152, 215,
Agra Road, 170 327,343,350.361,384; as target of blame
Amr. See Yadav(s) by Hindus for Partition of India, 10, 35,
Akhara(s), 89-91, 168, 205 75, 88. 100,324-25, 340,343,360-61,
Aligarh City Legislative Assembly consti- 372. 384, 454. See also Rumors
tuency: boundary changes in, 156, 176, Aligarh Muslim University Students
179-80,185-86; candidates and elec- Union (AMUSU). See Aligarh Muslim
tions, 47-49, 94> 108, 134,152,205,211, University
219> 221-30, 234-39, 2.43> 247-51, 254> 256, Aligarh, riots in (by date): before 1946,
262-301,319-20,385-86,402,407,409 70-75, Of1946,76; of 1950. 76, 99: of
Aligarh College. See Mohammadan 1954,76: of 1956,76-80,112,350; of
Anglo-Oriental College 1961, 80-84, 161, 262, 228: of 1971, 84-
Aligarh Inquiry Commission, 250 85, 112., 161: of 1972, 86-89: of 1918, 89-
Aligarh Municipal Corporation, 57. 58 99,103,133,167, 169-70, 173, 203, 205-
Aligarh Muslim University (MiU), ix. 49, 6, 244, 307-8, 312, 347; of 1979, 99-101,
53,55,131,150,153,183-85,203,210-11. 133: of 1980, 101-4, 170, 348: of 1988,
2.43,2.45,251,259,268,273-74,285,307, 105-8: of 1989, 108~12; of 1990-91, 112,
309, 313, 316, 340-41, 346, 361, 434: con- 116-31,156,170 -71,173,184,189-9],
troversy over status of, 92-95, 99~102, 229, 276, 279-80, 389.
Aligarh University. See Aligarh Muslim Bani Israilan, 54, 201, 265, 403,405,448
University (AMU) Bania(s), 48-49, 104, 166, 168, 200, 203,
Amar Ujala, 119, 125, 131 205-6,210-11
America and Americans, 76, 79, 432, Bansal, Vivek. 49, 285,385
446; and political machine, 23-24; Bara Bazaar, 201
and urban riots, 18, 364, 419 Baradari, 52-56, 71, 160-61, 187, 201, 209,
AMU Hospital. See Aligarh Muslim 215, 220, 273. 280, 285, 291-92, 294-95,
University (AMU): Medical College 298
Hospital of Baraltseni Degree College, 49, 85, 93, 256
Ansari, Iqbal, 169 Barallsenis, 48-49, 204-5, 210, 264, 320.
Ansaris, 53, 55-56, 187 See also Varshney( s)
Anupshar Road, 160, 176, 183 Barauli, 244
Ataiyan,54 Bashir, Dr. See Khan, Abdul Bashir
Atishbazan, 54, 276, 403, 405 BBC (British Broadcasting Corporation),
Ayodhya, 312. 373, 375; destruction of 5,342,347
mosque at, 5, 30, 35, 137-40, 246-47, Bhagalpur, riots in, 108-9, 413
347,355,366,384, 389-90; movement, Bhagavad Gita, 77, 112
and riots associated with, 7-8, 13, 26, Bharatiya Janata Party (BJP), 6, 44, 1.06,
30, 38, 49, 61, 68, 1.05. 108-12, 116-20, 110, 138, 212, 336, 355, 372-75; and
133, 136-40, 185, 228-29, 231, 234-35, Ayodhya movement, 8, 13, 61, 118,
241,243,248,272,274,278,382,386,388 133-34,198; and communal discourse,
3U, 313, 315; and Gujarat pogrom,
Babri Mandi, 55, 96, 161, 360, 403, 405 387-91; and Krishna Kumar Navrnan,
Sabri Masjid: controversy concerning, 26, 128-29, 134, 163, 204-5, 238. 244-45,
61,105,108,118,180,198,274,345,373; 248,272,300; and Muslims, 107, 188,
destruction of, 143, 387; disturbances 247,249,280,315,319,351,371; and
associated with movement against, riots, 102-3, 140-41, 143, 179, 184-85,
192,247,296,374,387; effects of con- 219-20, 240-41, 249, 251-56, 258-59,
troversy on elections in Aligarh, 281, 279, 301, 309, 315, 317, 320, 339,351,366,
283, 285, 288-89; as object of militant 378; strength and support base of, in
Hindu hatred, 241; rath yatra to, 13 Aligarh, 10, 48-58, 139, 157, 160, 163,
Babri Masjid Action Committee, 105 177-81, 183, 185, 223-25, 234-35, 265,
Sabri Mosque. See Babri Masjid 272-77,281-84, 286-88, 293, 295-98,
Backward castes, 8, 44, 50, 134, 223, 225, 301, 385, 407-9
243,278,291,363,375,389 Bharatiya Kranti Dal (BKD), 84, 222, 225
Baltraich, 1.05 Bharatiya Vidhya Bhavan, 76-77, 79
Bahujan Samaj Party (BSP), 44, 51, 58, 160, Bhojpur, 155, 176, 179, 180-82, 276, 403,
238,256,257,273,277-78,283-84,294, 40 5
298-301,375.385,407-9 Bhopal,77
Bajrang Dal, 117-18, 130, 134. 184, 187-88, Bhura. See Bhure La!
192-93,195,246,248,365 Bhure Lal, 89-91, 92-95, 97-98, 165-66,
Bandh,101,105-6.388 168, 205-6, 307, 314, 347, 360, 365
 Index/465

Bihar, 29, 50, 61, 73, 82, 108, 116, 133, 231, Chauraha Abdul Karim, 89, 163, 168, 403,
344,413 40 5
Blame displacement, 303; as a form of Cbhavani,156
causal analysis, explanation, and inter- Chira Ghachain, 54
pretation of riots, 20, 22, 29, 308-9, CIA,342
380; as a phase in riot produ<:tion, Civil Lines, ix, 71, 75, 102, u8, 121-22, 127,
15-16,305-6,349; uses of in political 136,150,152-53,155,157,160,162,178,
and social conflict and problems, 14, 200~1

24,67,70,81-83,90-91,97,100,103, Communalism, 9, 11, 17, 25,34, 36, 39, 109,

120,130, 228, 249, 255-56, 304-1l, 316- 111,170,248-49,251,261,307,316,318,
17, 320-22,331, 336, 343-44, 351, 363- 321, 350, 357, 381-82, 413
64> 371, 379, 383, 387-88, 391 Communist Party of India (Marxist), 373
Bombay: incidence of riots in, 65; January Communists, ix, 86, 262, 268, 373, 374
1993 pogrom against Muslims of, 31, Congress, Congress (I), and Congress (R).
386; December 1990 riots in, 5, 31, See Indian National Congress
137-38,141 Contextualization, 74-75, 83, 242, 310-11,
Brahmans, 49-50,57, 267 340, 369, 381-82
Britain and the British, 45,48,72-74,150, Conversion specialists, 32-33, 252-53, 259,
315,344,384,388, 413; rule of, 5,8,25- 339,389
26,32,34-36,70,199,365 Cow protection movement, 7
Budaun, riots in, 108 Criminal(s) and criminality, 23,356,382,
Buddhism, 35, 51, 267 452; and activities during riots, 83, 90,
Buford, Bill, 29, 259, 379 93-94, 96, 134, 162-63, 252,334-35, 342,
Business in Aligarh, 152,186,192,200,205, 380-82; at Aligarh Muslim University,
209-10; castes and, 57-58, 264; impli- xvi, 126, 152, 322-24, 340; and law of
cation of in riots, 95, 133-34, 139-40, riots, 66; speciafu.ed role of in riot
167,203,206-8,215,339,357-58; rival- production, 95, 152, 167-69, 176, 252-
ries of, and riots, 97, 203, 209-16: sup- 53, 258-59, 333, 335-40, 342-43, 358,
port of for Navman and BJP, 134, 205 370-71,378-79,457. See also Goondas
Crowd(s), 69, 111, 363; behavior and
Calcutta, 73, 344, 374 violence of in riots, 17-18, 21-22, 29,
Caste, community, and political influ- 65,89,117,124,152,247,263,338,359,
ence, 56-59 363-64; confrontations of with police,
Catholics, 315 65,87,102, 117,121,123, 166, 171,173,
Central Reserve Police Force (CRPF), 100, 330,347,362,389, 424,451, 456; l1indu
105, 123, 184 and Muslim, confrontations between,
Centre for Research in Industrial 141, 143-44, 165-66, 168, 172, 258, 376,
Development, 107, 162 424: mobilization and recruitment
Chamars, 44, 50, 262, 263. See also Jatavs of, 21, 33, 81, 95, 169, 258, 360-63, 448,
Chandan Shaheed, 54 456-57
Chandra Talkies, 207, 216 CTBT (Comprehensive Test Ban Treaty),
Chauhan, Ashok, 129-31 374
 466/lndex

Curfew, and control of riots, 81, 87, 91, 1991, 7, 278-83, 294, 390; of 1993, 49,
94, 96, 100-2, 106, 113, 122, 126, 136-37, 283-85, 385; of 1996, 8, 211, 228, 256,
141-42, 162, 171, 177, 188, 190-92, 196, 298-301; of 2002, 385-86; of Aligarh
208,246,333,338,346,362,388,432 Muslim University Students Union,
80, 82, 263; and BIP, 8, 183, 205, 248-
Dadri, 99-'101, 112 49,296,301; caste and religious com-
Dargah, 56. See also Ustad Sahib ka munity voting in, 49, 51, 243-44, 287,
Dargah 289-93,298,300; municipal, 88, 157.
Das, Veena, 363 178,181,183, 286, 292, 297, 407-11; and
Degree colleges, 49,87,93, no, 127, 129, riots, 34, 43, 72, 80-81, 84-85, 101, 108,
253, 307; and demand for affiliation 112-13, 132-34, 179-81, 183, 185, 197,
with Aligarh Muslim University, 87; 219-41, 251-52, 256-58, 262-63, 279,
domination of by Rashtriya Swayam- 308, 320, 333, 358, 367,370, 375, 38S-86;
sevak Sangh (RSS), 78, 85, 95, 129, 161, turnout for, 262, 446
256; and student participation in riots, Engineer, Asghar Ali, xi, 26, 64, 142
80,85,112,144,152,378,448 Ethnic violence, primordialist perspective,
Delhi, riots in, 103. See also New Delhi on 28-29, 419
Delhi Danwza. See Delhi Gate
Delhi Gate, 54, 177, 201, 284> 403, 405 Faziabad, 105
Deputy superintendent of police (DSP), FIR (First Information Report), 121, 380
325,341-43 Firdous Nagar, 156,185
Desai, Morarji, 114 Fire tenders, 33, 255-56, 259
Dhaka,77 Furkan,Mohammad,273,294
Discourse, 197, 306; communal, 11, 15,17-
18, 24-26, 28, 32, 34-37, 39, 259, 261, Gaborieau, 113, 114, 378
307,311-21,327,337,349-51,357,363- Gali Hajjaman, 54
64> 382, 386; of national unity, 84; of Gambhirpura, 54, 403, 405
violence, 6, 258 Gandhi, Indira, 43, 84, 86, 102-3, 176, 197,
District magistrate (DM), 85, 104, 109, 307-8, 312, 349, 373, 434; assassination
123, 125, 128, 135-37. 143, 145, 184, 190- of, 7,31,223
92,197,338,340-43,440 Gandhi, Rajiv, 108-n, 132, 223, 235, 288,
Doordarshan, 126, 344 413; assassination of, 7
D. S. (Dharma Samaj) College, 49, 84-85, Gandhian, 138
129, 403, 405. See also Degree colleges Ganesh,77
Ghosi,54
East Pakistan, 26, 77 Golden Gang, 168
Elections, xv, xvii, 38, 47, 179, 197, 285-97; Goondas, 33, 431; encounter and execution
of 1957, 80, 266, 402; of 1962,44, 81-82, of, 104, 134> 316, 317-19, 370; role of
86,205,230,235,258,262-72,407;of in riots, 33, 134, 167-68, 303, 314, 371,
1971,84; of 1974, 248; of 1977, 90, 94; of 377-78
1980, 49, 294, 308;of1984,294; of 1985, Graff, Violette, xviii, 64, 166-67, 428-29,
294;Of1989,7,108-12,272-78,294;of 433-34
 Index 1467

Gujarat, 13, 61, 116; pogrom of 2002, 386- Indian Police Service, 329
92,459 Institutionalized riot systems, 15, 27, 32,
Gular Road, 275, 283 125, 258, 369
Gumti Express, 117, 122, 125-27,130,179 Interparty competition: and communal
Gupta, B. D., 93-94, 96, 167, 346 voting, 266-67, 287, 300, 448; and
riots, 32, 95, 106-7,198, 219-20, 223-
Habib, Irfan, 273 26,229-30, 239-40, 262, 290-91, 301,
Haider, Jarrar, 268, 270 317,350,407-11,449,450
HaIim, Khwaja, 183, 207, 215, 273, 276-77, Inter Services Intelligence (ISI), 342
289,294,3°8 Islam, 52, 79, 86-87, 458; and baradari, 52-
Harndardnagar,176, 178-79, 185, 299 53, 56, 423; Hindu views of, 34-35, 124,
Haqqi, S. A. H., 50, 53 245,260,310-12,316,325,327,383, 435
Hardoi,105
Harijans, 122, 184-85. See also Scheduled Jabalpur, 77, 345, 453
Caste(s) Jagjivan Rarnpur, 184
Hayat, Kaisar, 256, 298-300 Jaiganj ward, 54-55, 160, 448. See also
Hindu Mahasabha, 79, 187-88 Index of Mohallas (p. 475)
Hindutva, 7, 123, 143-44> 315, 371, 373. See Jain, 57-58, 103, 144, 189,348,403, 406
also Nationalism; Militant Hindu(s) Jainism,35
History, secular, 36 Jama Masjid: at Aligarh Muslim Univer-
Hiteshi, Madan Lal, 205, 264, 269-70,289 sity, 153; in Delhi, 96, 103,274; in Upar
Holi, 71, 74-76, 113, 360 Kot, AIigarh, 120-21, 153, 162-63, 274,
403,405
Id,346 Jarnaat -i -Islami, 103, 338, 345, 350
Ideology, secular, 315 Janlaat-ul-Ulama,77
Indian National Congress, ix-x, xiii, xvi- Jamalpur, 124, 155, 176-78, 185, 299, 403,
xvii, xix, 61, 102-3, 106, 145, 156, 190, 405
214, 243, 385; caste and communal Janata Dal, 110, 112, 133, 180, 183, 198, 225,
support bases of, 48-49, 51, 55, 205, 238, 241-42, 256-57, 273-74, 276-84,
234-39,270, 272-74,282-83,285;and 294, 298, 373
communal politics, 249-51; domi- Janata governments, 90, 100, 102, 244-45,
nance and decline of after Indepen- 308
dence, 7, 43-44,80-82, 132, 197-98, Janata party, 6, 90, 94-95,98-100,205,
229, 280; electoral support for in 222,225,234,244,254,257,259,279,
Aligarh since Independence, 45, 134, 282-83,290,294,308
221-26,263-68,276-78,287-90,292- Janata Party (SC), 222
95, 298,385; and Muslim vote bank, Jan Sangh, 6, 10, 90, 93-94> 100, 197, 248-
255-58, 373; and political competition 51, 254-55, 257,311-12,315,319-20,371-
with Muslim League before Indepen- 72; in Aligarh elections, 84, 205, 222,
dence, 71-75; and riots after Indepen- 225,234,244,248,264,266-67,269-
dence, 77, 90-91, 95, 98-100, 102-3, 70, 272, 287-89; alleged role of in riots,
109-12,133,307-9,317,344,349,374 84-85, 93, 98, 102, 104, 185, 203-4, 206,
 468/ Index

Jan Sangh (continued) Lachhman Das, 165

255. 259, 307-9, 320, 346, 378; caste Legislative Assembly (U.P.), 77, 98. See
and community support for, 48,163, (liso Aligarh City Legislative Assembly
203: electoral support for in Aligarh, constituency; Elections
84, 205,222,225,234,244,248,264, Lelyveld, David, 150
266-67, 269-70, 272, 287-89; presence Lieux de memoire, 36, 64
of in degree colleges, 49-50: views Lok Da1, 94> 222, 242, 289, 294
of on Aligarh Muslim University, 82, Lok Sabha (Parliament), 77,90,97,
87-88. See also Bharatiya Janata Party 100, 110, 133, 285, 296: and debates on
(BIP) status of Aligarh Muslim University
Jatavs. 44, 50, 51, 253, 262, 267, 298-99 (AMU), 86-87, 93, 99; election results
Jinnah, Mohammad Ali, 88, 384 in Aligarh, 87, 221-22, 234-35, 248, 251,
Jogipara, 121, 172, 194 262,267-71• 274,289-90,292,294,296;
elections for, 108, 110
Kakar, Sudhir, 28-29, 431 Los Angeles, riots in, 376
Kanwariganj ward, 54-55,157,160, Lucknow, 88, 90, 102, 112, 119, 141,300,346
448. See also Index of MohaUas
(P·475) Madar Darwaza. See Madar Gate
Kar seva, 13, 116-17, 120, 136, 139, 278-79, Madar Gate, 71, 170, 275-76. 283, 297. 440
388 Madhya Pradesh, riots in, 61, 77, 80
Kar sevaks, 13, 117, 139, 388 Mafia,336,356,390
Kartik Pumima, 111 Maharashtra, 7, 13. 51,81
Katra, 275, 403, 405 Mallendranagar, 156
Kayasthas, 49-50 Maheshwari Inter-College, 49
Kerala,374 Maheshwaris, 49
Khaidora, 55, 284, 403, 405 Mallmubanja ward, 54-55. See (llso
Khaliq, Abdul, 211, 283, 285, 298-300, 385 Mamubhanja
Khan, Abdul Bashir, 262-66, 268-69, 407, Maithel Brahman, 50, 54, 57
448 Mallall, 176, 179, 403, 405
Khan, Ahmad Loot, 235, 294 Mamubhanja, 157,161,163,403,405
Khan, Rallman Ali, 318 Manak Chowk. See Manik Chauk
Khan, Sujatullall, 207 MandaI Commission, 134, 278, 280
Khatauli, 105-6 Manga Ram, 93, 96, 253-55, 336, 341, 346
Khatri, 57, 259 Manik Chauk, 48-49, 55, 81, 89. 93, 104>
Koil, 48, 150, 176, 180, 183 161, 163-71, 203-6, 243, 246, 283-90,
Kolis, 51, 200 292-93,295,321,403,405
Koran, 245, 312, 314 Manik Chowk. See Manik Chauk
Kotwali (central police station), 106, 120, Mann, E. A., x, 47, 52-53,55-56, 162,167-
404-5 68,171,200-'1,207, 214,423.444-46
Krishna, 186, 214 Marris Road, 152. 210, 285
Kumar, Promod (Pramod), 129, 130, 131, Maulana Azad Nagar, 156, 185
438 Maurya, Buddha Priya, 262, 267-70
 Index/469

Media, 5, 24, 61, 66, 70, 117, 126, 342-51, 109,146,160-73,196,199, 201, 214, 243,
359, 362, 369, 389, 392. See also Press 252,265,275-76,283-84,286,322,333,
Medical College (at Aligarh Muslim 338,395-401,407,440: transferrred
University),126 out of Aligarh constituency, 296-97:
Medical College colony, 184 voting behavior in specific, 264-66,
Medical College Hospital. See Aligarh 274-76,281-95,402-11
Muslim University (AMU): Medical Mohammadan Anglo-Oriental College,
College Hospital of 37,150
Meerut, viii. 62, 65, 99.103. 107, 113, 132, Moradabad, 76, 101-4, 210, 333, 348, 434
137,349 Mufti Shallar (City Mufti), 137, 139, 141,
Member, Legislative Assembly (MLA), 173, 177, 181, 329
256. 313, 385 Municipal corporation. See Aligarh
Militant Hindu(s), 8, 25, 28, 35-37. 46, Municipal Corporation
56,77, 88, 101, 108, 111, u8, 142, 161, 180, Munshi, K. M., 76
185, 195, 241, 243, 278. 312, 316, 321, 325, Murphy, Gardner, 209,360
327,349,356,382,384.388: candidates Muslim League, ix, 55, 62, 71, 73, 75, 78, 86,
and parties in Aligarh elections, 47, 49, 103, 207, 338, 372
84,222-26.235-36,238.267,270-7~ Muslims: migration of from rural areas
285, 287-95, 300; groups and organi- to Aligarh City, 46, 155, 179, 181-82,
zations. 6-8, 79, 126, 144, 152, 167-68, 185-86, 282-83, 297; Personal Law
196, 200, 212. 214, 255, 292, 315, 324, 365: of, 350, 372; separatism and separatists,
ideology of. 124, 319, 331-32; parties, 29, 37, 62, 75, Ill, 320,350,356,372,382,
77,80,224-26,294,409; politics and 384; as targets of blame by Hindus for
politicians, xvi, 50, 75, 167, 200, 248: Partition of India, 10, 35, 75, 88, 100,
and riots, 51, 77, 79-80, 83, 87, 105. 119, 324-25, 340, 343, 360-61, 372, 384, 454
144,167-68,179,196,219,231,252,254, Muzaffarnagar,105-7,434
258,312, 337, 346-47, 366, 370
Militant Hinduism. See Nationalism; Nagla Mallah, 176, 179, 403, 405
Militant Hindu(s) Nagla Pala Sallibabad, 156
Mini-Pakistans, 36. 374, 384 Nagpur,77
Minorities Commission, 97, 114> 162, 166, Nai,54
170, 203, 205 Narasimha Rao, P. V., 7
Misras,54 Nationalism and nationalists, 13, 28, 35,
Mob(s), 119, 359, 379: violence and other 36,43,86,87,125,215,315,372; Hindu,
actions of in riots, 5-6, 65, 81, 84, 101, 6,36, 75, 199, 243,315-16,319,325-26,
106,109.117,120, 122, 125, 127-28, 165, 337,365; and historiography, 35-36, 70;
171-72, 178, 184, 331, 363, 365, 390. See militant Hindus and, 7, 9,34-35, 48,
also Crowd(s) 56,87, 215, 234-35, 270, 290, 295,301,
Moltallas: general description of, 160: not 319,327,357,371,373-74,384; as move-
included in the 1951 census, 155-56, ment. 35. 105. 119, 320; Muslim, xvi, 78,
173-86; sensitive, riot-hit, and crime- 156, 342; secular, 34,36, 43, 87.372. See
prone, 9, 48, 54-55, 69, 81, 89,93, 104> also Hindutva
 470/ Index

Navrnan, Krishna Kumar, 163, 204-5, 215, Pala Sahibabad, 156

234, 243-48, 251, 253-56, 290,312, 321, Parliament. See Lok Sabha
342, 385; alleged role of in riots, 51, 90- Parties, secular, 80, 133, 242, 285, 287, 295,
91,93-94,96,98,128-29,134> 167, 205, 373
215, 243-46, 253-55, 276,307-8,336,341, Partition of India (1947), 75, 268,305,
346, 361; as SJP candidate in elections, 327,350,383,420,430,454; migrations
134, 179-80, 183, 205, 211, 238, 248, 272- at the time of, 47, 55, 200; riots at the
73,277,280-83,285-86,288,297-301, time of, 5, 8, 26, 62, 67, 70, 72, 116-17,
385; support for among trading and 305,340,344,366, 372,374,387,391,
business castes, 49, 134, 248 417. See (liso Aligarh Muslim Univer-
Navman, Raj Puri, 254, 255 sity; Muslims
Nazi Germany, 196-97, 414 Pathans, 53, 55
Nehru-Gandhi fiunily, 7 Peace committees, 10, 96, 102, 139
Nehru, Jawaharlal, 6-7, 43, 47, 83, 179, 371, Peoples Union for Civil Liberties (PUCL):
373 report of on 1978 riots, 91-93; report
New Abadi, 156 of on 1990-91 riots, n6, u8, 120-24,
New Delhi, 36, 99-100, 131, 139, 186, 213, 126-31, 135-36, 171-72, 178, 184-85, 345,
300, 336, 346, 373; anti-Sikh massacres 361
in, 30, 185, 349, 386; political division Phapala,54
and power in, 7, 90, 96, 102, 112, 145, Phul Chauraha, 78, 114, 161, 165-66, 168,
240. See (liso Delhi, riots in 171, 201, 202, 360, 404-5, 440
Newspapers. See Media: Press Phool Chauraha. See Phul Chauraha
New York, 413 Pogroms, 16, 22, 24,30,32,37,39,317,
noncooperation movement(s), 12; and 328-29,373,378,413-14; anti-Jewish,
Khilafat, 8, 70 121; anti-Muslin!, 6, 10-11, 23, 30-31;
anti-Sikh, 31, 449-50; Russian anti-
P(lhalw(ln. See Wrestlers and wrestling Jewish, 360, 416, 454-55, 457. See (l/so
Paired comparison analysis, as method Gujarat
of studying riots, 19, 27-28,39, 106, Pokhran, 374
149,418 Police, ix, xvi, 17, 22, 66, 100, 106, 128, 132-
Pakistan: "agents" of, and alleged role in 34,163,169-70,192-97,212-14> 241,
riots, 77, 102-3, 342, 348; and associa- 244,246,259-60,279,306,311,317-18,
tion with Aligarh Muslim University, 346-47,362-64,371,387-90; behavior
10,37, 62, 241, 350, 372: and Indian of during particular riots in Aligarh
Muslims, 36, 77-78, 267-68,313, 336, (1961,1975,1978,1979, 1980, 1989, 1995),
340, 350; migrations to and from, 85, 90-91, 94-96, 98, 102-4, 109-10,
47,55,200; militant Hindu attitudes 141-43, 169, 263; collusion and frater-
towards, 321-22, 327; movement for nization of with Hindu rioters, 81, 123,
creation of, 8-9, 13, 35, 37, 55, 62, 71- 165-69,171,173, 196, 257, 313, 387-90;
73, 75, 88, 268; relations with India, 35, conflict and confrontation between
77,79,356-57,374,388. See also East rioters and, 10, 13. 21, 69, 100, 102, 166,
Pakistan; Mini-Pakistans 181,308,347,356.379; and extrajudicial
 Index 1471

encounters and executions, 104, 134, Qureshis, 54, 55, 107, 161, 186, 273, 280, 284,
317-18, 370: and firing during riots, 101, 292, 295, 298, 300, 320-22
109,117-19: Hindu-Muslim represen-
tation anlong, 74, 196, 375-76, 402: Racism, 381
and killing and beating of Muslims in Rae's index of fractionalization (Fe), 448
riots, 10, 52, 60-61, 65, 120-21, 166-67, Raghubirpuri, 54, 157,404,406
169-71,177-78,182,187-90: political Rallman, Habibur, 273, 277
control of, 375-n; professionalization Railway Colony, 150, 404, 406
of, 375-n; role of in Gujarat pogrom, Railway Road, 162, 163
387-90: public perceptions of, 329: Rajput(s). See Thakur(s)
views of on riot production, 24, 39, Ram (Hindu god), 8, 123, 192: and demand
170, 252-54, 321- 25, 371, 328-43, 379- to construct temple for in Ayodhya, 13,
82. See also Senior superintendent 26,105,108, 134, 274, 278; as political
of police (SSP) emblem of Rashtriya Swayamsevak
Politics, criminalization of, 336 Sangh (RSS) family, 365. See (1150 Ram
Prasad, P. S. V., 168 Janmabhoomi movement
Press: reporting of on riots, 6, 10, 14, 16, Ram Janmabhoomi movement, 106, 143,
20,39,66, 69-71,74,77-78, 81-83, 87- 240,246,248,262,294,345,417,435
88, 91-92, 95-97, 99-100, 102-5, 107, Ramjanmabhumi. See Ram Janmabhoomi
109-10,119,125-26,130,137,141,143, movement
162, 179, 181-82, 185, 196, 203, 273, 323, Ram, Lila, 13, 71, 74, 246, 365,404,406
343-51,362,386,388-89,415,426-27, Ram, Pars, 76, 209, 360
429,431,433-34,453,456-57,460: role Ramshila pujan, 26
of in riot produ<:tion, 33, 77, 118-19, Rani, Usha, 235, 294
124-25, 343-46, 349, 389 Rasalganj Road, 201
Procession(s), 109, 193,313: as provocation Rashtriya Swayamsevak Sangh (RSS),
for riots, 12-13, 26, 69-71, 74, 77-78, 85, x, 87, 107, 126-27, 197, 210-12, 240,
87,89-90,93,95,97-98,106,108,111, 248, 254, 256, 288, 290, 296, 300, 315-
114, 117, 138, 161, 166, 168-69, 250-51, 20,324-27,334,342,345,349-50,355;
347,358,364-66,430,456-57 and business community, 127, 129-
Protestants, 315 30, 139, 361: degree colleges controlled
Provincial Armed Constabulary (PAC), by, 78, 84-85, 87, 95, 129, 161; family
94,100,102,145,328-29; as alleged of organizations, 6-7, 61, 365, 371;
anti-Muslim force in riots, 92, 95,104, paramilitary volunteers of, 73-74;
119-23,137, 166-67, 169, 171-73, 177-79, and pre-Independence riots, 71: and
181-82, 188, 191, 196, 329-32, 375, 379 post-Independence riots, 84-85, 90-
Punjab, 30, 319, 344> 387, 390, 391 91,93-96,98-103,117,129,140-41,
Punjabi (Hindus), 200,259 167-68,184-85, 187-88, 243-45, 250-
51,258-59,301,307-9,311-12,315,317,
Qanungoyan, 54 319-20, 324, 336, 338, 361, 365-66, 378,
Qasai(s), 53, 170, 201, 338. See also Qureshis 387-88: support for among Hindu
Qazipara, 54, 177, 184, 404, 405 castes in Aligarh, 49-50, 168; and
 472/ Index

Rashtriya Swayamsevak Sangh (continued) 33, 113, 124> 359-64> 378, 419, 455-56.
Varshney community, 163, 204-5, 338. See also Aligarh Muslim University;
See also Sangh Parivar Muslims
Rath yatra, 13, 116-17, 119-20, 133, 135, 365
Republican Party of India (RPI), 45, 205, Sallis, 53, 55
221, 225, 243, 262, 264. 266-67, 269-72, Stlkhas, 168
407 Saleem, Mohanlmad Yunus, 84-85, 250-51
Riots: causal explanations of, 6, 10, n, 14, Samajwadi Janata Party, 279. See also
15-27,29-34,70-71,75,78,85,89,97, Samajwadi Party
98, 104, 117-18, 121, 124-25, 131, 132- Samajwadi Party (SP), 44> 50, 58, 143, 145,
34, 136, 140, 143-44, 146, 186, 198, 201, 160, 177, 181, 185, 223-24, 238, 279, 283,
209,212,214, 219, 226, 245, 261, 291, 294,298,301,373,375,385,407,409
3°5,3°8-9, 311-14> 320-21, 323, 332, Samnapara, 184
334, 336, 337-38, 340-41, 347-48, 356- Sangh Parivar, 6, 24°,387,389-90,459.
88, 392, 414-16, 418-19, 429-30, 434, See also Rashtri.ya Swayamsevak Sangh
446,453-55; curfew in control of, 81, (RSS)
87,91,94,96,100-2,106,113,122, Sarai Behram Beg, 54
126,136-37,141-42,162,171,177,188, Sarai Bibi, 54> 284, 404> 406
190-92, 196, 208, 246,333, 338,346, Sarai Ghosian, 54
362, 388, 432; deaths and killing in, Sarai Hakim, 202, 404-5
8,30,52,61-65,67,69,71,73,92-97, Sarai Hakim Takia, 184
1.03,1.05-6,108,113,117-22,126-27, Sarai Kaba, 179
134-36, 141-42, 160, 165-66, 169-70, Sarai Kale Khan, 55
173,176,178, 181, 193-97, 203, 205-6, Sarai Mian, 55, 201, 284, 404-5
214, 242, 258, 3U, 314, 318, 328, 335, 344, Sarai Nawab, 287, 404-5
348, 360, 364, 366, 379, 386-91, 413, Sarai Pathanan, 54, 404-5
417, 424-25, 430-31,434,439, 457; as Sarai Qazi, 184
dramatic productions, 15, 32, 306, 358, Sarai Rallman, 54, 404, 406
369,378; phases in the production of, Sarai Sultani, 54, 95, 120-21, 141, 143-44,
15-16, 27; prevention and control of, 170-73,176,186-97,212,214, 267, 285,
1.05, 132-46, 376, 379; riffraff theory 292-95,330,404-5
of, 378-79; role of state in,374. See Sasni Gate, 180, 189, 208,323,338, 404,
also Institutionalized riot systems; 405
Interparty competition Scheduled Caste Federation, 267
Rithambara, Sadhvi, 240 Scheduled Caste(s), 29, 43-44, 46, 48, 50-
Rumors: in Aligarh riots, 80, 112, 144, 51, 56-59, 156-57, 170, 176, 178, 180-83,
252-53, 260, 333, 338-39, 349, 456- 191-92,197,200,265-66,341,375-76;
57; of murders at Aligarh Muslim and electoral alliance with Muslims,
University Medical College Hospital, 43-45,47,51,243,269,272,279,298;
119, 124-31, 204> 241, 307, 349, 422; participation of in riots, 51, 1.07, 119,
role of in spread of riots, 1.0, 21, 30, 184-85, 330, 389; voting behavior of,
 Index/473

51,221,234> 266, 270-71, 277-80, 286- Sufiyan, Mohammad, 183, 280, 283,
87,292,297-98,300-1,375,407-9 294-95
Secularism, 26, 82, 86, 88,107,110,210,214, Sunet. See Sunhet
220,268,324,326,341-42,349-50,391 Sunhet, 54, 448
Security Council of the United Nations, Sunni Muslims, 367
374 Surendranagar, 54,404-5
Senior superintendent of police (SSP), Swatantra Bharat, u9
85,104, 125-26,134,136-37,143,145, Syedwara, 55
168,170, 190, 197, 259-60, 315, 317-18,
321-22, 328, 343, 434 Tamil Nadu, 374
Separatism. See Muslims Tantanpara, 54,201, 276,404, 405
Shah, Rajiv Ratan, 104 Thakur(s), 49-50, 54,57-58
Shahinshahabad, 156, 185 Thugs. See Goondas
Shahpara, 54, 55, 157 Tila, 54, 448
Shamshad Market, 100, 109, 160, 404-5 Times of Indm, 61, 69, 71, 76, 81-82, 87-88,
Shamsi baradan, 55-56 95, 103,105-6,109-10,138,141,143,178,
Sharma, 50, 189 181,185,348,350
Shastri, Shiv Kumar, 84> 87, 267, 269, 270, Times (London), 70
271,289 Turkman Gate ward, 54-55, 157, 160,
Sheikhan, 55, 176, 448 448. See also Index of Mohallas
Shia Muslims, 280, 367 (P·475)
Shilanyas, 108-12, 240-41, 273
Shiv Sena, 7, 13, 31, 442 Uniform civil code, 370. See also Muslims:
Sikhs, 46,185,316,349 Personal Law of
Singh, B. P., 104, 317-18 United Nations, 374
Singh, Ghanshyam, 294 United Provinces, 70, 73. 344. See also
Singh, Inder Pal, 250 Uttar Pradesh
Singh, Kalyan, 90, 139, 245, 256, 258 United States, 36, 327, 372, 376. 382-83,
Singh, Ranjan Pal, 250 413,456; riots in. 32, 419. 456; research
Singh, Vishwanath Pratap, 133, 373 and inquiries on riots in, 18. 20, 28,
Singhal, Shiv Hari, 84, 85. 93, 253. 336, 368,415
341,346 Upper caste(s): attitudes of, 28, 355;
South Asia, 79, 350, 374, 383: riots in, dominance of, xvi, 363. 375-76; and
347,356: study of riots in 26-27, 30 riots, 103, 203; voting behavior of,
Spontaneity, as explanation for riots. 50,300-1
See Riots: causal explanations of Upperkot. See Index of Mohallas
Student(s), 108, 138, 152, 211, 273, 316, (p. 475): Upar Kot
323; participation of in processions, Urdu language, 29. 77, 108-9. Ill, 160,
protests, and riots, 49, 71-72, 77-78, 242. 245, 313, 345
81, 83-84, 100-1, 109-10, 112-13, 152, Usman, Hafiz, 280
378 Usmanpara, 54, 201. 404, 405
 474/ Index

Ustad Sahib ka Dargah, 155, 177, 404 320-21, 337-38; militant Hindu senti-
Uttar Pradesh (U,P,), xv, xvii, xix, 50-51, ment among, 161, 163, 200, 204-7, 248,
134, 212, 256, 336, 402; Ayodhya move- 321
ment in, 116-19; ele<.1ions in, 133, 273, Verma, Anant Ram, 263-64, 266, 270
278-80, 291, 296; Pakistan movement Vidhan Sabha, See Legislative Assembly
in, 9-10; political parties in, 7-8, 44, Vidyarthi, Tota Ram, 205, 264, 270
49,80,94,133,250,267,272-74,296, Vishnupuri, 54, 448
373, 375; riots in, 13,37, 61-62, 68, 70- Vishwa Hindu Parishad (VHP), 49, 61,
77,80-81,88,90,98,101-3, 105,107- 105,108, 117-18, 168, 187, 193, 240-41,
8, 113, 119, 123, 131,133,135-38,199, 240, 248,256,365,388-90
279, 296, 301, 313, 328, 331, 338, 344, vyaparis (traders), 134
366-67,370,374-75,377, 385-86, Vyapar Mandai, 101, 127-30, 142, 361
See also Legislative A&o;embly
West Bengal, 373, 374
Vajpayee, Atal Bihari, 82, 390 Wilkinson, Steven, 9, 27, 61-64, 70,76,
Varshney, Ashutosh (Mayor), 48 219,231,424-27,446,453,458
Varshney, Ashutosh (Professor), 9, 27, Wrestlers and wrestling, and riots, 89-
56-57, 61-64, 210, 414-15, 418-19, 90, 92-98, 113, 144, 165, 167-68, 205,
421, 424-26, 429-30, 434, 436-38, 314,347,360,365,430-31. See also
443,445-46,453,456 Akhara(s)
Varshney, Gyan Chand, 139, 142
Varshney, Krishna Gopal, 286 Yadav, Mithilesh,142
Varshney(s): as Aligarh caste, 48, 55, 57- Yadav, Mulayam Singh, 117, 143, 181, 241,
58, 101, 127, 200, 204-5, 210, 243, 248, 246, 279-80, 298, 373, 375
264, 280, 286, 443; alleged roles of in Yadav, Surendra Mohan, 142
riots, 93, 166-68, 170,201,204, 215, 292, Yadav(s), 44, 50, 280, 321, 323, 338, 375
 INDEX OF MOHALLAS

Achal Talab, 51-52, 54,142, 144, 157, 161, Ghas ki Mandi, 448
296, 403, 406 Ghosian, 448
Anupshar Road, 160, 176, 183 Ghuria Bagh, 185
Ataiyan,54 Gular Road, 275, 283
Atishbazan, 54, 276, 403, 405
Hamdardnagar, 176, 178-79, 185, 299
Babri Mandi, 55, 96, 161, 360, 403, 405
Baikunthnagar, 448 Jagjivan Rampur, 184
Bani Israilan, 54, 201, 265, 403, 405, 448 Jaiganj, 95-96,403,405,448
Bara Bazaar, 201 Jama\pur, 124, 155, 176-78, 185, 299, 403,
Barai, 170, 189 40 5
Begambagh, 448 Jauhar Bagh, 283
Bhojpur, 155, 176, 179, 180-82, 276, 403, Iiwangarh, 283
405 Jogipara, 121, 172, 194
Brahmanpuri, 170, 276
Kanwariganj, 403, 405
Chandan Shaheed 54 Katra, 275, 403, 405
Chauk Bundu Khan, 448 Kela Nagar, 283
Chauraha Abdul Karim, 89, 163, 168, 403, Khaidora, 55, 284, 403, 405
405 Krishi Farm, 283
Chhavani,156
Chira Ghachain, 54 Madar Gate, 71, 170, 275-76, 283, 297,
440
Delhi Gate, 54, 177, 201, 284, 403, 405 Mahendranagar, 156
Mamubhanja, 157, 161, 163, 403. 405
Fudous Nagar, 156, 185 Manik Chauk, 48-49, 55, 81, 89, 93, 104,
161,163-71, 203-6, 243, 246, 283-90,
Gali Hajjaman, 54 292-93,295,321,403,405
Gambhirpura, 54, 403, 405 Maulana Azad Nagar, 156, 185

475
 476/ Index of Mohallas

Nagla Mallah. 176, 179, 403. 405 Sarai Nawab, 287, 404-5
Nagla Pala Sahlbabad, 156 Sarai Pakki. 276
New Abadi, 156 Sarai Pathanan, 54, 404-5
Sarai Qazi, 184
Pala Sahibabad, 156 Sarai Rahman, 54, 404. 406
Patthar Bazaar, 286 Sarai Rai, 170
Phapala,54 Sarai Raja RanI,448
Phul Chauraha, 78. 114, 161. 163, 165-66, Sarai Sultani, 54, 95, 120-21, 141, 143-44,
168, 171. 201, 202, 360, 404-5, 440 170-73,176,186-97,212,214, 276, 285,
Pir Ataullah, 448 292-95,330,404-5
Purani Kachehri, 448 Sarai Tahsil, 448
Sarrafa Bazar, 252
Qanungoyan, 54 Sasni Gate, 180, 189, 208, 323, 338, 404, 405
Qazipara. 54, 177, 184.404,405 Shahinshahabad, 156, 185
Shah Jamal, 56, 137, 155, 176-77, 276,
Rafatganj, 283 404-5
RagllUbirpuri. 54, 157, 404, 406 Shahpara, 54, 55, 157
Railway Colony, 150, 404, 406 Shamshad Market, 100, 109, 160, 404-5
Rangrezan, 448 Sheikhan, 55, 276, 448
Sudamapuri, 448
Sabzi Mandi, 161, 448 Sunhet, 54, 448
Samnapara, 184 Surendranagar,54,404-5
Sanichri Painth, 407 Syedwara, 55
Sarai Bairagi, 286
Sarai Barahseni, 286 Tamolipara, 276
Sarai Behram Beg, 54 Tantanpara, 54,201, 276, 404, 405
Sarai Bibi, 54, 284, 404, 406 Tila, 54> 448
Sarai Ghosian, 54 Turkman Gate, 54, 95, 170, 201, 265, 404-5,
Sarai Hakim, 202, 404-5 407-8,448
Sarai Hakim Takia, 184
Sarai Kaba, 179 Upar Kot, 87, 109, 120-21, 140, 153, 154, 155,
Sarai Kalan, 276 162, 170-71, 173, 188, 200, 201, 212,321,
Sarai Kale Khan, 55 404,405,440
Sarai Kutub, 407-8 Usmanpara, 54, 201, 404, 405
Sarai Lawaria, 277
Sarai Mian, 55, 201, 284, 404-5 Vishnupuri, 54> 448
 11111111111111111111111111
9 780295 985060