Ethnic Conflict and Civic Life: Hindus and Muslims in India

• Introduction
• Remarks
• Questions and Answers

Introduction

JOANNE MYERS: Good morning. On behalf of the Carnegie Council, I would like to welcome members, guests, and C-SPAN to our Books for Breakfast program.

This morning we are very pleased to have Ashutosh Varshney, who will be discussing his book, Ethnic Conflict and Civic Life: Hindus and Muslims in India.

The news headlines reporting on the resurgence of tensions between India's Hindu majority and its sizable Muslim minority has captured our attention in recent months. The flames of violence which whipped across Gujarat in March and April of this year seem to suggest that extreme religious conflict, which appeared to have faded of late, had returned to the political landscape.

To some, this may constitute irreconcilable differences between India's Hindus and Muslims, but, according to our guest this morning, this may be misleading, for, as Professor Varshney discovered, levels of communal violence may vary widely among Indian states. The issue thus becomes: Why do some cities in India erupt in savage confrontations while others manage to avoid bloody Hindu/Muslim clashes?

To find the answer, Professor Varshney took a different approach from other scholars and policymakers and focused his research on the makeup of civil society, rather than that of the state, for the means of conflict prevention. What he found was that in areas where there were vast networks of civic engagement - such as integrated business organizations, trade unions, political parties, and professional groups - which cross communal lines, ethnic violence seemed more controllable and the possibility for confrontations less likely to occur. His investigations led him to conclude that communally integrated associations can serve as an agent of peace, particularly by restraining those powerful politicians who would polarize Hindus and Muslims along religious lines.

Professor Varshney is currently a professor of political science and the Director of the Center for South Asian Studies at the University of Michigan. He received his Ph.D. from MIT and has taught at Harvard, Columbia, and the University of Notre Dame. Our guest has served as a consultant to Human Rights Watch, the World Bank, and the U.S. Department of State. He has received fellowships from the Social Science Research Council and the Ford Foundation.

Many of you are familiar with his writings, which have appeared in numerous journals, including World Politics, The Journal of Democracy, and The Encyclopedia of Democracy. He has also published articles in The Financial Times, The Washington Post, and The Christian Science Monitor.

Please join me in giving a very warm welcome to our guest this morning, Ashutosh Varshney.

Remarks

ASHUTOSH VARSHNEY: Thank you very much.

My book suggests an integral link between the structure of civic life - or civil society - on the one hand, and the presence or absence of ethnic violence, on the other. I illustrate these links through two arguments.

First, inter-ethnic and intra-ethnic, or inter-communal and intra-communal, civic engagement plays a very different role in times of ethnic tensions. Because they build bridges, inter-ethnic networks are agents of peace. But if communities are organized only on intra-communal or intra-ethnic lines and the interconnections across the communities are either very weak or non-existent, the possibilities of ethnic violence are quite high, given a spark.

That spark can emerge in the form of an attack on a train, the rumor of a rape, defeat of ethnic political parties in elections, desecration of a mosque or a temple, or something as apparently trivial as a Hindu boy going out with a Muslim girl, or vice-versa. Sparks can emerge at any time and in any town. The real issue is whether there are networks that can manage or extinguish those sparks. If they are not there, these sparks can become fires.

The second argument breaks down civic networks into two other types: organized and informal. I call the first associational forms of engagement and the second everyday forms of engagement.

Business associations; professional organizations of doctors, lawyers, teachers, students; film clubs, sports clubs; NGOs; trade unions; and political parties are examples of associational engagement.

Everyday forms of engagement consist of simple, routine interactions of life, such as whether families in the neighborhood from different communities visit each other, eat together, jointly participate in festivals, and allow kids to play together in the neighborhood. In my project we calculated the frequency of such everyday interaction in each city.

Both everyday and associational forms of engagement can promote peace, but the associational form is sturdier and stronger. Vigorous inter-ethnic associational life, organizational life, acts as a serious constraint over the polarizing strategies of political elites. The first argument, therefore, concerns the value of inter-ethnic civic life, and the second the greater value of organizational or associational inter-ethnic civic life.

What aspect of life are we identifying by civil society or civic life?

By "civil society" I mean that part of our life which exists between the state, on the one hand, and our families, on the other, which allows people to come together for a whole variety of public activities and which is independent, or relatively independent of the state. Civil society is not a non-political arena of our life, but a non-state arena of our life.

Moreover, in its non-state functions it can cover both social and political activities. Soccer leagues, bowling leagues, cricket clubs, playing card societies may be social not political; but trade unions, political parties, students' or teachers' organizations, can be very political and not social, as they are in India.

For the study of ethnic peace and violence, civil society is a new variable. How did this argument emerge in the first place? What was the puzzle that I was trying to sort out in the project when I started this work ten years ago?

Sooner or later, scholars of ethnic conflict are struck by a puzzling empirical regularity in their field. Despite ethnic or communal diversity, some regions, nations, towns, villages manage to remain peaceful, whereas others experience enduring patterns of violence. Similarly, some societies, after a long record of peace, explode in ways that surprise.

Variations across time and space, therefore, constituted an unresolved puzzle in the field of ethnicity and nationalism, in which this book falls. How does one account for such variations across time and space?

With very few exceptions, researchers had, until the mid-1990s or so, only focused on riots or violence or civil wars. This strategy of identifying the causes of violence by studying violence only, was methodologically flawed. Why?

Suppose that on the basis of commonalities across many cases of violence or riots you find that inter-ethnic economic rivalry, polarized party politics, and segregated neighborhoods explain that violence. Let's call inter-ethnic economic rivalry A, polarized party politics B, segregated neighborhoods C, and violence X. Essentially we are saying A, B, and C explain X in that model.

Can we be sure that our judgments are right if we follow that strategy? What if A, B, and C are also present in Y, peaceful cases, which we did not investigate in the first place by concentrating only on violence? And if A, B, and C are present both in X and Y, then A, B, and C cannot be the cause of X. It logically follows.

There could be another factor, D, which differentiates peace from violence. It will, however, be a factor that we did not discover in our research precisely because we did not study Y with X, we only studied X.

To identify then how this comparison between peace and violence would be pursued in my India project, I started first with a step-by-step, year-by-year, month-by-month investigation of all reported Hindu-Muslim riots in the country between 1950 and 1995. It is a very complicated methodology.

Two substantive results were decisive. First, the share of villages of rural India in Hindu-Muslim rioting turned out to be remarkably small. Between 1950 - when I began my research - and 1995, rural India, where 65-75 percent of Indians still live and where 85-88 percent lived in 1950, accounted for a mere 3.6 percent of deaths in communal violence. Hindu-Muslim violence in India is primarily urban.

And, second, within urban India, too, Hindu-Muslim riots are highly locally concentrated. Eight cities account for roughly 46 percent of all deaths. These eight cities as a group represent a mere 18 percent of the urban population today and only 5 percent of India's total population, urban plus rural. In other words, Hindu-Muslim rioting, the 1950-1995 data showed, was highly locally concentrated in urban India.

India's Hindu-Muslim violence, therefore, is city-specific. State and national politics may provide the context, or the sparks, but it is the local, city-level mechanisms that are activated, or not, given those sparks. To understand, therefore, the causes of communal violence, we must investigate these local mechanisms.

Following this reasoning, the project selected six cities, three of the most riot-prone - Hyderabad; Aligarh; and Handabad, which also burnt up in March, April, and May - and three peaceful cities and arranged them in three pairs. Each pair, therefore, had a city where communal violence was endemic and a city where it was rare or entirely absent. Hyderabad was compared with Lucknow; Aligarh with Calicut, and Handabad with Surat.

To ensure that we did not compare apples and oranges, roughly similar Hindu-Muslim percentages in the city population constituted the control in each comparison: Hyderabad and Lucknow, 30-34 percent Muslim; Aligarh and Calicut, 37-38 percent Muslim, the rest mostly Hindu; and Handabad and Surat, both 14 percent Muslim.

Handabad and Surat are separated only by about 150 miles. That was the most tightly controlled comparison in the project. Surat has been a peaceful city until its peace was broken after sixty-seven years in 1992-1993. Nor did it blow up during the most recent phase of communal violence in Gujarat either. Handabad had the most gruesome rioting that it has seen in long years.

So what emerged? You might ask why I controlled for demographic proportions rather than something else. Whether you look at academic arguments or political discourse, either from the Muslim side or from the Hindu side, all arguments and discourses concentrate on the point that numbers matter in India's democracy. So if Muslims are above a certain percentage, Muslim discourse is seen as threatening by Hindu activists. And Hindu activists or nationalists argue that if Muslims cross a certain threshold in a city's population - whether 15, 20, or 25 percent - they become "uppity" in that discourse. Thus, I controlled for demographic proportions.

So what argument emerged as a result of this research on six towns and what we call a large N of all Hindu-Muslim violence, riots from 1950-1995? What accounts for the difference between communal peace and violence?

The argument was not anticipated when the project began. It was not a hypothesis, but emerged during the process of research. Looking back, it sounds embarrassingly simple, but seemed very bright at the time.

The preexisting local networks of civic engagement between the two communities stand out as the single most important cause. Where such networks of engagement exist, conflict or tensions may exist, but they do not deteriorate into rioting and violence. I am, therefore, making a distinction between conflict and violence here. Where they are missing, communal identities can lead to endemic and ghastly violence.

What are the mechanisms that link civic networks, on the one hand, and violence and peace, on the other? And why is associational engagement sturdier than everyday engagement as a bulwark of peace?

One can identify two mechanisms that connect civil society and ethnic conflict.

First, routine engagement allows people to come together and form temporary organizations in times of tension. Such organizations, though short term, turn out to be highly significant. Called peace committees - or variations on that theme - which consists of members of both communities, these organizations policed neighborhoods, instead of relying on the police to do so; killed rumors, extremely important in times of tension; provided information to the local administration and facilitated communication between communities during unrest.

Such neighborhood-level organizations were virtually impossible to form in cities where everyday interaction did not cross religious lines. Sustained prior interaction allowed appropriate crisis-managing organizations to emerge.

The second mechanism allows us to understand why associational forms are sturdier than everyday forms. If vibrant organizations serving the economic, cultural, social, and political needs of the two communities exist, the support for communal peace not only tends to be strong, but an also be more solidly expressed.

Organized civic networks constrain local politicians in their strategic behavior very effectively. If politicians insist on polarizing Hindus and Muslims for the sake of electoral or political advantage, which they are likely to do if there is an electoral political gain to be made, they can tear the fabric of everyday engagement apart at the neighborhood level through the organized might of criminals and gangs.

In all violent cities in the project, a nexus of politicians and criminals was in evidence - provable social-scientifically, not provable legally. Individual complicity or culpability I cannot establish with my materials, and that is what would stand up in a court of law. But in a court of social science I can prove that without this connection you cannot have the kind of violence that you see in cities like Handabad or Barodabad or Aligarh or Hyderabad.

In some cases, you can perhaps point out which politician was involved and through which police officer and which civil servant; but, more often than not, it is virtually impossible to prove involvement. You can, however, certainly get a very good social science argument out of it.

Organized gangs could easily disturb neighborhood-level peace, often causing migration from communally heterogeneous neighborhoods to communally homogeneous neighborhoods. People moved for the sake of physical safety.

In peaceful cities, however, what might be called an institutionalized peace system exists. When organizations, such as trade unions; associations of businessmen, traders, teachers, doctors, lawyers; and at least some political parties - different from the ones that have an interest in polarizing - are integrated, when Hindus and Muslims are integrated, countervailing forces are created. Associations and organizations that would lose from a communal split fight for their turf, making not only their members aware of the dangers of communal violence, but also the public at large. Local police and civil administrations are far more effective in such circumstances. Civic organizations, for all practical purposes, become the ears and arms of the administration.

This is not an anti-state argument, but merely a suggestion of how local wings of the state and local civic organizations develop relationships of synergy, which is much more effective than if that synergy did not exist.

Unlike violent cities where rumors and skirmishes, often strategically planted and spread by those who have an interest in communal polarization and violence, are quickly transformed into riots, these skirmishes or rumors, relationships of synergy in peaceful cities, nip rumors, small clashes, and tensions in the bud. In the end, polarizing politicians either do not succeed or eventually stop trying to polarize communities by provoking and engineering communal violence.

This argument is probabilistic, not law-like. It is not a physics sort of argument. It is a social science-y argument. It should not be taken to mean that no exceptions to the generalization would exist. Indeed, pending further empirical investigation, law-like generalizations of ethnic violence may not be possible at all. As social scientists, we can only provide you an odds-based explanations.

Indeed, perhaps the best way to understand the relationship between civil society and violence is to use a materialogical analogy. If the civic edifice is inter-ethnic and associational or organizational, there is a good chance it can take ethnic earthquakes that measure quite high - seven or eight - on the Richter Scale, such as a partition, a desecration of a holy place in full public gaze, perhaps a civil war.

If, however, the civic edifice is inter-ethnic with everyday informal engagement, then earthquakes of smaller intensity - four on the Richter Scale - can bring the edifice down. Defeat of an ethnic political party in an election or police brutality in a particular city are examples.

But if civic engagement is only intra-ethnic, not inter-ethnic, small tremors, unconfirmed rumors; Hindu boys going out with Muslim girls, or vice-versa; the rumor of pigs thrown into a mosque; music before mosques; or rumors of killing of cows; victories and defeats in sports; wrestling matches between a Hindu wrestler and a Muslim wrestler, with one getting defeated, that leading to riots - this is a shock which measures only one or two on the Richter Scale. But it can unleash horrible torrents of violence. A multi-ethnic society with few interconnections across ethnic boundaries is very vulnerable to ethnic disorders and sparks.

You may want to argue at this point that I am confusing causes and consequences, that I need to establish why I am so sure that the causation was not the other way around - that is to say, did communal violence destroy the preexisting Hindu-Muslim civic networks in riot-prone towns, and that is why I did not find that? Or did the presence of such networks prevent violence from occurring? Which way do the causal arrows go?

Two chapters in the book go into this problem. For each town and civic association, the team went into the history of when it emerged, and what we know about violence at that point.

Historical research conducted in the cities demonstrated that civic networks determined the outcome in the short-to-medium run certainly, but in the long run these networks themselves were politically constructed. The 1920s were a transformative movement in India's politics because mass politics emerged under the leadership of Mahatama Gandhi. That is how far back we have to go to sort out when the civic associational order emerged.

Gandhi seized control of the freedom movement in 1920 and quietly revolutionized it by arguing that the British were unlikely to give Indians independence until the masses were involved in the freedom movement. He also argued that the battle for India's freedom had two aspects: wresting independence from the British; and conquering the "internal enemy," which meant the reconstruction of Indian society to deal with Hindu-Muslim problems, untouchability, women's, workers', and tribal rights, et cetera. He argued that the best way to do that was to build organizations and fight those battles.

It is not that India did not have Hindu-Muslim engagements before 1920, but it was mostly of the quotidian, informal sort. India's associational civic order was born with the Gandhian movement in India's freedom movement when millions of people joined in the freedom struggle and built organizations, including, famously, the Congress Party, which developed mass cadres after 1920.

Gandhi was the father of the first trade union of India, Majoor Mahajan, or TLA, Textile Labor Association. He was also the father of many other organizations - homespun societies, literary societies, lawyers' societies - in Gujarat first and then all over. The organizational effort spread all across India.

Despite Gandhi's power and the power of the Congress Party, these networks could not be built in every town and city. In local politics Hindu-Muslim cleavages had already emerged as the most significant determinant of local political fortunes. It was, therefore, difficult for Gandhi and the Congress Party to build these integrated organizations.

Where modern politics had not emerged in any big way, or where the cleavages were caste-based or within the Muslim community between Shi'ias and Sunnis, it was very easy to put together these organizations. This is how the organizational history of these cities can be written.

In other words, it is not the case that riots preceded the formation of these organizations in the 1920s. You can show that where the organizations were integrated, riots could be prevented. Where integrated organizations could not be built, a history of rioting began.

Finally, some policy implications, especially concerning the role of the state. Many book reviews have come out in the last six months here in the United States, the United Kingdom, and Asia. A point that critics repeatedly identify is that my emphasis on civil society takes me away from a full understanding of the role of the state; that I place an excessive analytical weight on civic mechanisms here.

In the India edition of the book, published two months ago, I have answered this. It was there in the book, hidden here and there, so I have pulled it out and explicitly proposed what I think of the state.

Here are four steps to think about it:

1. On the major fault lines of a polity, as the Hindu-Muslim relations are in India, the state tends to act in a politically strategic, not a legally correct or morally overwhelming, manner. This is true in much of the world. Consider the Sri Lankan state on Singalese-Tamil relations; the Malaysian state on Malay-Chinese relations; and the U.S. on race relations, though the U.S. is beginning to come out of its racial bind in the last few decades. The state should not act in this manner, but it does, combining legality, morality, and political calculations in an unpredictable way. The state in Gujarat appears to have acted in a primarily strategic, not a legally correct, manner, and has so far gotten away with it.

2. A more empirical understanding of how states function neither means that citizens should cease to criticize and pressure the state when it fails to protect lives in riots nor that they should stop trying every constitutional means of punishing the state or the functionaries of the state. But while making every attempt to bring pressure on the state, they should not bet on it to rectify its behavior any time soon. If the state corrects itself on major fault lines, it should be viewed as a happy temporary outcome of such activism, not something one can bet on. The odds are against that.

3. Working on and building integrated civic networks is a better bet than trying to pressure the state. Towns where Hindus and Muslims continue to be integrated into businesses, political parties, unions, and professional associations remain peaceful, and integrated organizational and civic life makes the local wings of the state behave better than intellectual, political and moral exhortations that it do so.

4. It follows that citizen action should take two forms: (1) while continuing to pressure the state for its dereliction of constitutional duty - one part of citizen action - it should (2) focus on building integrated civic structures. Pressuring or critiquing the state has been the primary strategy of organizations and citizen action in much of the world, not only in India. Such action is necessary but not sufficient. Citizen initiatives should follow a two-track strategy, combining pressuring the state, on the one hand, and building integrated civic structures, on the other.

Thank you very much.

JOANNE MYERS: I will open the floor to questions.

Questions and Answers

QUESTION: I was fascinated by the cases you mentioned which broke down after sixty-seven years. Was it simply that the events got higher on the Richter Scale than any structure could have coped with, or had the structures somehow decayed or deteriorated?

Secondly, you mentioned places where there were already very separate communal structures in the 1920s and that these were where it was difficult in the time of Gandhi to build the cross-community structures. You mentioned internal divisions within the communities on either side as being something that facilitated the building of cross-community structures. Are there other factors?

How discouraged should one be, because you made it sound as though it would be a great idea to have these organizational cross-community associations, but when things have got to a certain point, it's very difficult, if not impossible, to construct them?

ASHUTOSH VARSHNEY: Two excellent questions.

Why did the historically peaceful town of Surat experience ferocious rioting in the early 1990s after sixty-seven years, and why has it been peaceful ever since? There are two important factors in the model: one is the intensity of the earthquake, the intensity of the spark; and the second the presence or absence of integrated structures. It is true that the public desecration of Ayodhya Mosque, watched by, according to some estimates, 200-300 million people on TV, was a very big earthquake. An earthquake of greater intensity has indeed taken place in India - its partition of India in 1947 - but this came closest to that.

Within the city a very interesting pattern emerged. It has become the small-industry capital of India, and as such one of the fastest-growing cities in the country. The rioting took place only in the shantytowns, which are the newer parts of the city, where the migrant labor force had come down to live only recently. In the old city, where those structures were alive, not a single life was lost, even as 192 people were killed in the shantytowns between December 7th and March 10th.

So this peaceful town which experienced gruesome violence, actually supported the argument. If the shantytowns had followed the same pattern somehow, or if the political organizations and trade unions had been more active there by building these links, then perhaps the results would have been different, despite the intensity of the spark.

Can we build integrated organizations in what might be called highly inhospitable settings, where inter-community relations, inter-ethnic/inter-communal relations, have broken down completely? The research did not focus on this question, but some ideas did emerge in the process of trying to understand how to explain the difference between peace and violence.

The best I can do is to give you a brief summary of what we found in the town of Peyvandi, just outside of Bombay, which was riot-prone in the 1970s and 1980s and which has not had a single riot since 1986. It took the earthquake of the public desecration of the mosque without exploding into the horrors of violence.

In 1986 a police officer posted there wrote in his memoires: "Once I was stationed in Peyvandi, I knew that the biggest challenge of my life as a police officer had arrived." "The only way I could call myself successful," he says, "was if during my tenure there" - typically three years - "I could make sure that no rioting took place between Hindus and Muslims." He started putting together neighborhood-level committees with Hindus and Muslims.

Initially, this idea did not go down well, but he persisted and the committees did actually come into being. These meetings were not to discuss riots and violence, but rather common problems - sanitation, corruption, access to administration, electricity failures. By the time he left, he was not sure that the next police officer would persist with the same approach, but he did the same.

And in 1992-1993, when the big earthquake arrived, when the mosque was brought down, these committees were ready to keep peace in their city. There were some people who wanted to riot, but ultimately those people could not easily create mass riots if the masses are ready to take them on.

Another interesting incident in the same city took place in either March or April 2002. The head of Bajrang Dal, which is a very important Hindu nationalist organization, is quite wedded to the idea of violence. In off-the-record interviews, they proved to be absolutely committed to the idea of violence as a way to deal with Muslims. The head of Bajrang Dal was actually killed - a big spark for the town, but still no rioting took place, in April.

These organizations can be built. Notice that when Gandhi burst on the scene in 1920, you could not predict that he would be a great figure of Indian politics. When he was trying to build these organizations, he did not have the kind of charisma or reputation behind him, or Congress Party, that he developed in the 1930s. So the ballgame looking forward, ex ante, was very different. There were some local oligarchs, or local influentials, who were against the idea of integrated organizations, and Gandhi wouldn't be able to overcome all resistances.

If you put the two together, you come to the conclusion that it is indeed hard to put together integrated civic structures where they have completely broken down, but it is not an impossible task.

QUESTION: Could you please elaborate about strategic actions of a state? I can imagine two possibilities. One would be that in national versus local consideration, statesmen would act for the larger public good, even though there was local turmoil.

The other possibility is that politicians would act for electoral purposes.

ASHUTOSH VARSHNEY: It could be all of those, but in the case of Gujarat it was clearly for electoral purposes. The first level of argument has to be that the Hindu nationalist fervor that you see in the State of Gujarat is often an extreme kind. It is really the most ideologically pure form of Hindu nationalism in India. This would not be true in Delhi, nor of BJP and Hindu nationalist organizations outside Gujarat either.

One of the most shocking realizations that my research produced was the degree to which western India, the most prosperous and fastest-growing part of the country, had become communally divided, charged, and vitiated.

If you read the Hindu nationalist ideological texts, there is a consistent theme. If a Muslim takes one eye out, you have to take two eyes out, which is the only language that a Muslim understands. This is consistent through the 1920s, 19302, 1940s in any ideological text.

The Chief Minister of the state has gone through serious ideological training to come to where he is now. When scholars recalled the interviews with him in the early 1990s, they give evidence of a man truly committed to ideology.

But many also hold the view that communal polarization. Once the train had been attacked by some Muslim criminals, then retaliation was electorally good, and a retaliation that a Hindu nationalist government would either condone or ignore, or would at least passively support, if not actively support. We cannot prove that the government initiated pogroms, but we can certainly see that it looked the other way.

The hypothesis that it was the first pogrom of India has to be taken seriously. I don't think we can prove it. The strategic side there was using communal polarization for electoral purposes.

In other cases of violence, you could have strategic behavior of a different kind - for example, factional struggle within the ruling party associated with riots. You would find several examples where riots stopped, let's say in the city of Hyderabad, the day the Chief Minister resigned and a new Chief Minister belonging to a rival faction took over. There was no subtlety in the way the transition took place.

You could have factional struggle linked to riots, you could have electoral purposes, and sometimes it could be state versus national. Riots can be electorally and politically advantageous.

QUESTION: Do you see the media as an associational network? Is it neutral? Does it aggravate tensions, or does it help? What about organizations like the National Human Rights Commission, which is official, yet seems to reflect a civil society view?

My second question: The first labor union that Gandhi started was in Ahmedabad. Yet, you identify Ahmedabad as a city which is on the other scale. How would you explain that? Does it also necessitate perhaps research on the pre-1990s? You looked at it from 1950-1995. How do you figure the evolution of the associational networks before 1950?

ASHUTOSH VARSHNEY: Let me start with the second question first, because this occupied much of our attention and I devote a full chapter to the transformation of Gandhi's city from a place that had, arguably, until the 1950s the highest density of inter-communal civic engagement, to a city which after 1969 is, arguably, the most riot-prone city.

Yes, the 1950-1995 period was only for the national-level data, or the aggregate picture, but for each city we went all the way back to the turn of the century. So we investigated how Gandhi started building organizations first in Gujarat and how the idea spread.

Yes, the city of Ahmedabad's Congress Party was among the most vibrant and integrated organizations in the country, as were its unions, teacher associations, and lawyers. It was a city that inspired many of India's intellectuals, political activists, and citizen activists.

So why did a city so solidly integrated break down? The 1969 riots, incidentally, were before the current phase of rioting, the worst in independent India, and they took place in the city of Ahmedabad.

If you look at what these organizations were doing in the 1950s and 1960s, you find that, according to internal Congress Party documents, the most integrated party organization of the country by the late 1950s was in utter disarray. The Congress Party politicians were more interested - internal documents reflect this - in public offices and state patronage than in organizing and mobilizing people, and putting together or strengthening non-state structures. The Congress Party itself recognizes that.

Other documents show that by the mid-1960s it had virtually completely lost its civic vibrancy. It was only an appendage of the state by that time. Within fifteen years of independence, this is what happened to the Congress Party there.

The trade union was very integrated, with a membership of about 150,000 in a town which at that point was not more than 2 million. It was thus a huge trade union, in a textile town with a huge working class. Its internal documents showed that by the early 1960s, this massive trade union of India was also in utter disarray.

In neither case did communal violence provoke the weakening. The weakening took place for independent reasons. The Congress Party became more interested in the state than in mobilizing outside the state at the civic level. The trade union had a monopoly over labor representation, no competitive challenge, because of the laws of that state. As a result, the leadership of the trade union lost touch with the masses, and this trade union actually by the mid-1960s was no longer in power. It lost local elections.

When the spark came in 1969, it was a small spark, low on the Richter Scale. It was a rumor that could not be confirmed on September 17th and 18th, two weeks before Gandhi's centennial celebrations in his adopted home town. Rumors spread in the city that the cows of Jahanath Temple had been attacked by Muslims. This was the spark that led to five days of truly horrific rioting. Gandhians were nowhere to be seen, nor was the trade union.

Gandhians arrived after the riots had cooled down, which was very unusual. In 1947 the Gandhians, the Congress Party, and the trade union were ready for riots because of partition, but no such riots took place.

The integrated organizations were seriously weakened by the mid-1960s, as a result of which a spark led to the violent trend in Ahmedabad.

On your first question, media is a need, a civil society institution, at least wherever it is independent, as it is in India. It used to be much more dependent on the state earlier. It is becoming more and more a civil society organization or institution in India.

The media plays a horrible role in violent cities. It's very divided itself. It can print rumors as news stories. When we were doing our research there, the illegal media printed a headline story that Harvard University, Ford Foundation, and the Rajiv Gandhi Foundation, which was also supporting my work, were interested in spreading riots all over India under the instructions of the CIA.

The police, however, fully supported us and said, "If you believe the things that these rags write, you will never be able to do research." The police commissioner, the SSP of the city, stood by us, as did the district managers. So we continued, but we were interrupted because everyone was shaken by this news.

Whereas in the city of Calicut, which has twenty-six newspapers and magazines, for a city the size of about 600,000 at this point, when I went to interview one of the leading lights of the freedom struggle who was still alive in 1992, I found newspapers and photographers all around.

I said, "No, no, I came for a one-on-one interview. I don't think this will be a good interview if the press is there."

They said, "No, no. The newspapers say it's our beat to report on whoever comes to see him or interview him."

The guy said, "No, no. What I am going to tell you will not be affected by the presence of the newspaper reporters and photographers because I learned politics at a time when candor was practiced as a principle and I haven't given up on that."

So he gave me an interview. And the next day, what was the news item in the newspapers in Calicut? "A student of history interviews a maker of history." These newspapermen never believed that I was in Calicut to spread riots.

So the media plays a peaceful role, as well as a violent role. The newspapers of Gujarat in March and April were absolutely vicious. Some said that killing was good, that retaliatory brutality was good.

The issue is: is it integrated or not? The National Human Rights Commission and some other organs within the Indian State, including the judiciary, have become our spokesmen of civil society. It is correct to say that India's judiciary today represents civil society more than the state.

Within the state organs you are beginning to get self-correcting mechanisms. These are institutions, the judiciary and the National Human Rights Commission, which cannot punish right away, without full investigation, without evidence being presented, so the Gujarat government has not been punished yet. But the hypothesis that it will be punished in the future, has to be taken seriously. It may not be able to get away in the long run, but it has gotten away in the short run.

QUESTION: Do the principles you evolved in the Hindu-Muslim relationship apply also to the inter-caste conflicts in Bihar and UP as the lower castes have moved up politically?

ASHUTOSH VARSHNEY: After I did the Hindu-Muslim research, there are other researchers who started working on caste violence adopting this framework. They will give you a confirmed answer in three-to-four-to-five years. My hypothesis is that it will work, and I have some anecdotal evidence, but I can't give you a full-blown argument on caste violence.

In one sense, however, caste violence is different from Hindu-Muslim violence. In the hierarchy of the caste system, the upper castes were indisputably upper, and have been so, ritually and otherwise, for centuries. This kind of hierarchy did not apply to Hindu-Muslim relations. Violence takes place in a vertical structure, and violence in Hindu-Muslim relations is more horizontal than vertical.

In discursive hierarchy, it is not the case that all Muslims are below Hindus. All lower castes are below upper castes in the ritual hierarchy. That is being challenged now.

That it is vertical violence, as opposed to horizontal violence, might make some difference to what the conclusions will be. To watch this event on C-Span, click here.