A strange fire

But,fundamentally,the Muzaffarnagar riots departed from our existing script of understanding. Over the last 15 to 20 years,ethnic conflict in different parts of the world,including India,has been studied extensively. We know a lot,but no theory could have confidently predicted the Muzaffarnagar riots. We should,of course,note that predictive accuracy is not a good way to judge social science theories. As in biology,social science theories tend to be probabilistic,indicating the odds,not proposing certainties. Just as not everybody who smokes cigarettes would get cancer,though a lot would,social science theories are better at explaining central tendencies,not departures from them. We just can’t get the precision of physics.

Nonetheless,we need to understand why the Muzaffarnagar riots were unusual and examine their larger implications. Why were these riots surprising?

First,the Muzaffarnagar riots were mostly rural. While civil wars tend to be rural,riots are primarily urban. Why this is so is aptly summarised by a modern-day classic,The Logic of Violence in Civil War,by Stathis Kalyvas. In civil wars,the insurgents generally attack the sovereignty of the state. They need to hide from the might of the state to attack it well; villages,mountains and forests provide better hideouts for planning and protection than cities. In comparison,rioters attack communities,not the sovereignty of the state. Rioters are armed with lesser weapons,and houses in urban bylanes are a good enough cover.

Second,prejudice of various sorts might be widely prevalent in Indian villages,but the countryside tends to have caste violence,not Hindu-Muslim riots. And when they do take place,rural riots tend to be small. In the period,1950-1995,for which Steven Wilkinson and I constructed a dataset,rural deaths accounted for a mere 4-8 per cent of the total deaths in Hindu-Muslim riots (goo.gl/c1MkvZ). Killing anywhere between 40-50 people,if not more,the Muzaffarnagar riots were big. India has had big rural riots before — in Bhagalpur (1989),Gujarat (2002) and a few in the Northeast. But these were exceptions to the larger reality of urban riots. The Muzaffarnagar riots belong to this exceptional category.

Third,riots also affected the city of Muzaffarnagar. Hindu-Muslim relations in Muzaffarnagar town have historically been calm. For communal violence,Aligarh and Meerut were the two worst cities in western Uttar Pradesh during 1950-95. But as I demonstrated in my book,Ethnic Conflict and Civic Life: Hindus and Muslims in India,neighbouring towns often continue to maintain their divergent characteristics. Next to the riot-prone Aligarh is the peaceful Bulandshahr; adjacent to Meerut are Saharanpur and Muzaffarnagar,both customarily tranquil. The Muzaffarnagar riots departed from the historical trend.

Fourth,while my own book was about urban,not rural,riots,there is a major argument in the field,by Wilkinson,which applies to rural settings as well. In Votes and Violence,Wilkinson argues that communal riots are unlikely to break out in a state where the government is heavily based on minority support. To keep the minority vote intact,the government would use its might to stop or contain riots. The Samajwadi Party runs UP’s government; it depends heavily on Muslim support,yet riots took place,killing a large number of Muslims. One big exception,of course,does not invalidate a probabilistic theory,but it should be noted that ever since the SP government came to power,small riots have been occurring in UP with considerable regularity. A minority-dependent government seems unwilling,or unable,to stop them.

Finally,another piece of scholarly wisdom is partly breached. At higher levels of income,riots decline in frequency and intensity. France did have riots in 2005,and the US in the 1960s and again in 1992 in Los Angeles,but such rioting is episodic. In contrast,riots can be endemic at low levels of income. At higher levels of income,communal or racial prejudice may continue,but resentments,when violent,take the form of terror or hate crimes (which are individual acts of violence),not riots. As the Indian economy grew at high rates over the last decade and as big riots also disappeared after Gujarat 2002,India seemed to follow the larger comparative script.

One big riot,again,does not undermine the larger income-based argument,but why did Muzaffarnagar burn? Might that happen again?

Press reports are normally the first drafts of history. These drafts suggest a hypothesis — and that is all we have at present. As national elections come closer,political parties might have,yet again,begun to think of riots as an attempt to create winning coalitions. During the late 1980s through 2002,communal polarisation was used to enlarge political constituencies. Is such polarisation under way in UP today? Is it of any use,especially to the SP and BJP?

The Muslim vote has always been central to the SP’s electoral performance. Press analysis suggests that the SP’s strategy is to allow riots to take place and then present itself as a saviour of Muslims. If true,this diabolical chessboard strategy can go horribly wrong on the ground. Once riots begin,they don’t necessarily stay small. If too many Muslim lives are lost,insecure Muslims are unlikely to view the SP as a protective shield against larger threats. They can easily desert the SP,blaming it for administrative failures,even malevolent intentionality. A large proportion of UP Muslims,after all,left the Congress in the late 1980s,when it could not stop riots.

What about the BJP? We know that UP is critical to the national ambitions of the BJP. The BJP under Modi would like to go from 116 seats (2009) to 180-190 (in 2014). Otherwise,it can’t attract post-election allies and form the government. Without a tripling or quadrupling of BJP seats in UP,such an increase is virtually inconceivable.

With only 10 out of 80 seats,the BJP finished fourth in UP in 2009,but both in 1996 and 1998,it had more than 50 seats. The 50-plus performance was due — considerably if not wholly — to a communal polarisation of the electorate that the Ayodhya movement had made possible.

With or without Ayodhya,is that the BJP strategy in UP again? Amit Shah’s appointment has led to widespread speculation that that is indeed the case. But it is hard to believe that while economic growth and governance will constitute Modi’s election campaign for India,there will be a reversion to the 1980s and 1990s in UP (and/ or Bihar). This is not a paradox but a flat contradiction. Communal polarisation will almost certainly be accompanied by violence; riding the communal tiger never allows full control over the dangerous beast. While the BJP may attract many Hindu votes in UP this way,it will also lose a substantial number of those attracted to Modi for non-communal,governance reasons,both in UP and beyond. The net seat gain may well turn out to be small,while the ghosts of Gujarat 2002 will return to haunt Modi’s future even more vigorously.

It is best to hope that as he wishes to scale national heights,Modi will not take a gamble of this magnitude. At present,one should believe that,if true,the BJP’s involvement in Muzaffarnagar was orchestrated not by the party principals at the top,but by their local agents. If this hypothesis is wrong,India may well be poised for a reversal of the decade-long communal peace. Self-defeating politics can,of course,prove all theories wrong,but it would be facile to assume that,given its desire for power,the BJP will devise a strategy that is likely to lead to defeat again.

The writer is Sol Goldman Professor of International Studies and the Social Sciences at Brown University,where he also directs the India Initiative at the Watson Institute. He is a contributing editor for ‘The Indian Express’ express@expressindia.com