Of ethnic conflicts and causes

Ethnic Conflict and Civic Life: Hindus and Muslims in India by Ashutosh Varshney; New Haven and London: Yale University Press, 2002; pages 384, &pound35.

THE publication of Ashutosh Varshney's book could not have come - very sadly - at a more apposite time. As I have been reading it, communal violence has been continuing in Ahmedabad and elsewhere in Gujarat, and Varshney's work, which is methodologically exemplary and theoretically rich, deserves to be quarried for ideas and insights into the causes and conditions of such conflict in India. In essence the argument of the book, which is immensely readable (unusually so for such a serious work of scholarship), is quite straightforward. It takes off from the striking observation that the extent and the frequency of occurrence of ethnic conflicts involving Hindus and Muslims vary enormously among Indian cities. The fact of variance, in itself, casts doubt on explanations based on competing general theories about ethnicity (those proposing arguments based on primordialism, instrumentalism, constructivism or institutionalism); and on the argument that religious violence has (in Ashis Nandy's words) "something to do with the urban-industrial vision of life and with the political process the vision lets loose". If this were true, how could it be that there is so much variation between cities? Careful analysis, drawing in part on interviews with many of the key actors who have been involved, of the historical experience of some of the cities that have seen a lot of violence and of that on the other hand of comparators which have not, suggests that it is the extent of inter-communal civic life, mainly that stimulated initially by Gandhian popular political mobilisation in the 1920s and 1930s, which accounts for the differences. This earlier politics, Varshney suggests, substantially determines whether politicians now try to mobilise support on communal lines, or not. Extensive inter-communal engagement in civic life creates a kind of 'institutionalised peace system', and it is this, rather than administrative action, that determines whether or not bloody riots take place: "It is the environment of a peaceful city that makes the police and administration perform its law-and-order functions better, irrespective of... the level of professionalism" (page 289). Quite contrary to the way in which Varshney's argument has been reported to this writer on several occasions, his work is not at all a vindication of Robert Putnam's well-known claims about 'social capital', as the cause and consequence of 'civic engagement', but rather a robust criticism of them. Whereas Putnam's argument may be described as 'society-centric' - suggesting that aspects of social relationships determine political outcomes - Varshney's, on the other hand, establishes the causal centrality of politics.

ETHNIC conflict (= 'communal conflict', in the terminology used in India) involving Hindus and Muslims is an urban not a rural phenomenon (less than four per cent of deaths owing to religious violence between 1950 and 1995 occurred in rural areas); and it has been concentrated in particular States and even more in particular cities. Eight cities are identified as being 'most riot prone', on the basis of an original data-set assembled from a careful scrutiny of The Times of India reports from 1950 to 1995 (official data series all being flawed in one way or another). These cities are Bombay (now Mumbai), Ahmedabad, Hyderabad, Meerut, Aligarh, Baroda (now Vadodara), Delhi and Calcutta (now Kolkata), though the first two stand out above the rest in terms of numbers of deaths - and the total for Ahmedabad has by now, probably, overtaken that of Bombay (now Mumbai).

Varshney's research strategy is to analyse the reasons for variance, and he enquires into the conditions that might account for the differences between certain of these cities - Ahmedabad, Hyderabad and Aligarh - and comparators that have some critical features in common but a history of a much lower incidence of ethnic violence. The paired comparisons - of Ahmedabad and Surat, Hyderabad and Lucknow, Aligarh and Calicut (now Kozhikode) - show the importance of the existence or not of the kind of civic life that brings together the members of the two religious communities. It is not 'social capital' as this has been defined by Robert Putnam in his influential books Making Democracy Work (Princeton University Press, 1993) and Bowling Alone (Simon and Schuster, 2000), that counts. In other words it is not the extent of associational life and the strength of local organisation, which are usually taken as key indicators of 'social capital', that matter - indeed strong intra-communal organisation may well be conducive to conflict - but the extent to which there is organisation that brings people of different communities together and helps to give them shared interests or identities.

In Kozhikode, in spite of the experience of the Mappilla Rebellion of the 1920s in the surrounding countryside and the successful organisation there of the Muslim League, the principal axis of social conflict has been over caste/class oppression, and the struggle against it has brought Hindus and Muslims together. In Lucknow the organisation of the chikan (embroidered textiles) industry means that rich, Hindu nationalist supporting businessmen are locked into relations of mutual dependence with masses of Muslim workers, in such a way that "if communal riots were to break out the local economy would collapse" (page 214).

And the Muslims are divided and sometimes drawn into conflict amongst themselves by the sectarian difference between the Sunni and Shia traditions. Here, importantly, though "electoral tendencies have favoured Hindu nationalism, civic patterns (deriving especially from the economy - "Lucknow's biggest mass-based civic structure is economic") have moderated the behaviour of Hindu nationalists" (page 202). In Aligarh and Hyderabad, by contrast, Muslims and Hindus tend to have discrete economic networks - there is nothing comparable with the inter-dependence that characterises the chikan industry of Lucknow - and in both cities the history of political organisation has been such that Hindus and Muslims have not been drawn together. In Hyderabad any efforts in this direction on the part of the Congress were systematically opposed by the Nizam, and in Aligarh, where a declining Muslim aristocracy confronted a rising Hindu merchant class, "the seeds of intra [as opposed to inter-] communal civic engagement were sown by the inability of the Congress Party to restructure mass politics in the 1920s and 1930s" (page 167).

The point of the comparison of Ahmedabad and Surat is rather different, and Varshney's interest in them is partly because they are cities which show such strong contrasts over time in terms of ethnic violence. Both are cities - in Mahatma Gandhi's home State - which for a long time successfully avoided serious communal violence (at the time of Partition, for instance) but which have subsequently become susceptible to it - especially so, in Ahmedabad, from 1969, and in Surat at the time of the destruction of the Babri Masjid. Varshney is able to demonstrate successfully that both cities, their economic and other differences notwithstanding, had rich inter-communal civic life, established by the cadre-based Congress Party, by Gandhian social and educational organisations, by business associations and (in Ahmedabad) by trade unions (the Textile Labour Association), that had been seriously weakened in the periods leading up to the first major outbreaks of ethnic violence. An internal Congress Party report of 1957, for example, on the state of the party organisation in Gujarat, referred to the fact that "Congress workers at the base have lost contact with the people" and called for the establishment of properly functioning local committees - explicitly "to combat the evils of communalism" (pages 268-69). But nothing was done, and it is clear, as Varshney says, that "The decline of the party as a civil body accompanied its rise to power" (page 267). Similarly, in Ahmedabad, the Textile Labour Association, had declined so far by the mid-1960s as to have lost power in municipal elections, and in 1969 textile workers killed one another on grounds of religion. In other words - the historical sequence is of course extremely important for the argument as a whole - it was not the occurrence of violence that caused the breakdown of inter-communal civic life, but demonstrably the reverse. The case also shows that 'history matters', but in a way which is somewhat at odds with Varshney's general argument about the legacy of Gandhian popular political mobilisation.

This is unquestionably a brilliant book, of immense interest. But the very 'thickness' of the analysis of each individual case rather defies the simple elegance of the general argument. For instance, though the argument about the causal centrality of politics is a powerful one - tragically confirmed in current events in Gujarat - Varshney's studies also seem to show up the importance of the character of the local economy and of the ways in which Hindus and Muslims are or are not drawn into inter-dependence with each other. For example, in an observation which is surely relevant to current events in the city, Varshney comments: "In Ahmedabad's business and professional life today, large-scale associational interaction between Hindus and Muslims simply does not exist" (page 250), but this lack of inter-communal association is shown to have been linked to the decline of the city's textile industry.

In Surat's business life, on the other hand, Hindus and Muslims are said to be "highly integrated, both at day-to-day and associational levels", and when communal peace was shattered in the city in 1992-93, violence occurred in the shantytowns but there were no deaths in the Old City - because, according to the Police Commissioner who took over during the riots, "there were prior links, especially business links, between Hindus and Muslims, and the existence of prior communication helped the police maintain peace. No such mechanisms were available in the slums" (page 260).

My point is not that Varshney's general argument is wrong, but that the explanations that he offers for each city case are in fact much more complex, and so the depth and subtlety of his analysis is in a sense betrayed by the particular theme that he draws out. The Gujarat cases show, I think, that he stresses the significance of the legacy of the politics of the 1920s and 1930s too much; while the extremely interesting vignette that he offers in his conclusion, on the way in which intelligent policing brought peace in Bhiwandi after a long history of communal violence, seems to detract from the force of his earlier conclusion (cited above) that "it is the environment of a peaceful city that makes the police and administration perform its functions better". But criticisms such as these do not detract from the interest and the importance of this book which deserves to command a wide readership.

John Harriss is Professor of Development Studies, London School of Economics.