May 14, 2005

Communal Violence in India - Perspectives on Causative Factors (Part 2) by Roshni Sengupta

[ The Economic and Political Weekly, May 14, 2005
Part 1 is available at: http://communalism.blogspot.com/2005/05/communal-violence-in-india.html ]

Communal Violence in India - Perspectives on Causative Factors (Part 2)

by Roshni Sengupta

[...continued from Part 1]

Planning a Riot

Varshney, according to Paul Brass, refuses to recognise that many riots in the post-independence period have been outright pogroms. Instead, he indulges in a kind of ‘apologetics’ to the Sangh parivar.

The phrase ‘ethnic earthquake’ is, in the words of Brass, a myth about spontaneous outbreaks of communal violence. What is not a myth however is that communal riots are ‘organised’ and ‘produced’ by a network of known persons in the city or town. Most of these known persons are members of the Sangh parivar who are devoted to the cult of violence for the protection of Hinduism.17 

Despite the fact that there are ‘waves’ or ‘chains’ in the occurrence patterns of communal rioting, there has been no notable period when such violence has been absent. Paul Brass, in his book (2003) on the subject develops and demonstrates his argument as to why communal tensions are maintained, accompanied from time to time by lethal rioting, and how it is essential for the development of militant Hindu nationalism and for other organisations and individuals [Brass 2003, ‘Introduction’].

Communal riots, according to Brass, are endemic in India. The phenomenon of certain sites more prone to violence or otherwise, entering and disappearing from the list of riot-hit cities is therefore, to him somewhat flummoxing. While investigating the spatial variation in the incidence of Hindu-Muslim riots, Brass tries to classify the various issues that continue to surprise in categories of persistence, differential incidence/timing, classification/meaning and power.

The struggle over meaning, explanations and power relations requires attention to a communal discourse that has entrenched itself rather deep in the body politic of the country. Brass uses the term ‘hegemonic’ to explain the communal discourse that pervades Indian politics. This discourse, fiercely Hindu nationalistic, has been successful in corrupting history as well as memory (interaction with students, January 12, 2004).

He identifies three elements inherent in the spread of this hegemonic communal discourse. Historisation leads to the distortion of history18  and the division of history into periods where Muslims are seen as conquerors. The fundamental antagonism is thus over-emphasised. The aspect of memorialisation includes greater attention given to the dead heroes of one particular faith. For instance, dramatising and exaggerating nuances used to glorify what the Sangh parivar calls ‘the struggle for the birthplace of Lord Rama’, is essential to memorialisation.

Evidence exists to show that memorialisation leads to demonisation of the ‘other’. The Muslims, the ‘others’ in the Indian case, are portrayed in literature, cinema and other forms of human expression as ‘racially different inferior beings who invaded our country and cultural space’. Myths, lending credence to this process of demonisation, are taken recourse to in the form of the spoken word as well as the written letter. Muslims are seen as violent and a danger to Hindu women.

Some utterly preposterous myths that have made their presence felt in the last two decades or so are the invariably higher rates of population growth among the Muslims through which they are trying to take over power both at the local and the national level and the fact that most of them are ‘Pakistani agents’. Another feature of the communal discourse is ‘body symbolism’. Muslim rule is portrayed as ‘slavery’ of the Hindus. The politics of body symbolism depicts the partition in visceral terms, i e, tearing apart of the Hindu body. The Muslims thus are dangerous to the Hindu body and need to be removed before the danger can actually present itself. This explains why most post-independence riots have been outright pogroms against the Muslims.

Brass delineates three phases in the production process of riots. He compares a riot to a ‘staged drama’. Dipankar Gupta uses the term ‘picnic rioting’ to describe the manner in which Hindu mobs actually celebrate the killing of Muslims. The Gujarat riots saw the herding of saffron clad mobs into trucks and their subsequent journey into specially demarcated colonies of Muslim concentration.

The first phase therefore, is one of preparation (rehearsal) in which tensions are kept alive. Killing of a cow and the kidnap of a Hindu girl are common methods. The following phase is one of activation (enactment). The political circumstance must be right for a riot to be precipitated. The pre-election phase could be such an occasion.

The last phase of riot production is explanation or interpretation where blame displacement comes into play. Brass further argues that there exists a division of labour in the production of riots. Riot systems are institutionalised. Specific roles are assigned to persons like that of scouts or informants, rumour mongers and propagandists.

Vernacular journalists enact their parts admirably by coming out with wild and inciting stories. The recruiters are those who collect people to form mobs. The roles of the politicians has already been documented earlier in this paper. Special duties are assigned to groups. Paul Brass calls ‘fire tenders’ who go around scouting for rumours that could help in fomenting a riot. These men invariably are members of either the VHP or the Bajrang Dal. ‘Conversion specialists’ ultimately decide whether to begin a riot or not.

Spatial Spread and Mobilisation

Causal questions about the spatial spread of riots are of utmost importance. Why exactly are they produced in certain cities or towns and why not in others? Steven Wilkinson’s answer to this question has already been looked at. Brass asserts that the contexts are primarily political as mass mobilisations usually precede elections.

A sizeable number of Muslims in a particular town or city is essential for the production of riots. The demise of the Congress system has created a space that is normally filled by other political formations, which benefit from riots and their production. The BJP has emerged as such a formation that benefits disproportionally through riots. The Gujarat situation is self-evident in this regard.

Brass (2003) takes Aligarh as his case for study and periodises the history of riot production in that city. The early period belonged to the Congress, which dominated till the late 1950s. This domination began to be contested between the 1960s and the 1980s. The last period, which coincides with the current phase, is a contest between the BJP and the Janta Dal-Samajwadi Party combine.

Brass is critical of Varshney for ignoring the 1990-91 communal pogrom, which was solely the handiwork of the Sangh parivar. Both agree however, that, riots and electoral politics are closely connected to each other. They proclaim in their respective works that there exists an absence of political will to control riots and this cuts across political parties. Both further argue that the Sangh parivar are the primary sources of most communal conflict in the country since independence.

Sociologists have a penchant for examining social conflict and Imtiaz Ahmad is no different in his approach. Ahmad, one of the first scholars to propound socio-economic theories for communal conflict, views Hindu-Muslim conflict as an extension of the wider social conflict that includes inter as well as intra communal riots, caste violence and other forms of sectional upheavals. The emphasis placed on Hindu-Muslim conflict in case of social and communal violence comes but naturally considering the huge impact the various riots between the two communities have had on the Indian polity and society.19 

Ahmad argues that economic prosperity of the Muslims is a factor that precipitates endemic anger on the part of the Hindus who fear being swamped, both socially and economically by the nouveau riche Muslims. This antagonism results in riots, which spreads to other parts of the state. An argument that has been advanced by numerous scholars following Imtiaz Ahmad, the above brings to the fore the social contradictions that have given rise to many communal conflagrations in the past and are likely to do so in future.

The ghosts of Gujarat cannot be invoked here as the violence there was ‘produced’ by the government of the day. The tensions that continue to prevail, in the rural plains of the state, are somewhat akin to the argument made by Ahmad. Frequent riots in the diamond city of Surat prove his point beyond any doubt.

The democratic process, therefore, is responsible for communal conflict and the lack of it. The scholars discussed above have advanced conclusive arguments about the role of democracy and the electoral system. The kind of electoral system that gets institutionalised over time determines the frequency of communal riots. Further, the arguments made above prove that riots are produced by specialists who could be politicians or members of the majoritarian formation.

This paper attempted to put the whole issue of ‘causal factors’ of communal riots into perspective; most theories, however, are interconnected and any explanation of communal riots in India needs to consider a multiplicity of theories.


[This paper was presented at the 18th European Conference on Modern South Asian Studies at the University of Lund, Sweden, July 6-9, 2004.]
1 The full commission reports on why exactly the fire happened and who was responsible for it, have still to be presented or tabled in Parliament. Many reports suggest that the fire was started from inside the train. Thus, the complicity of the Muslims standing outside on the platform is suspect any which way.
2 The Rashtriya Swayam Sewak Sangh (RSS), the Vishwa Hindu Parishad (VHP), the Bajrang Dal and the Bharatiya Janata Party (BJP) are the principal affiliates of what is commonly known as the Sangh parivar (the Sangh family).
3 The violence in Gujarat was not concentrated in a few areas but had spread into the tribal districts of south Gujarat as well.
4 Conservative estimates put the total number of people killed at 800. Various independent commissions of inquiry and other sources talk in terms of ‘thousands’ killed and property worth crores destroyed.
5 Minorities Commission Reports, 1986-88 found that between 1985-87, 60 per cent of the 443 people killed were Muslims and that Muslims had suffered 73 per cent of the nine crores in reported property damage.
6 Wilkinson uses the term ‘ethnic’ in the broader sense to describe Hindu-Muslim riots in India. Horowitz argues, all conflicts based on ascriptive identities – race, language, religion, tribe or caste – can be called ‘ethnic’. The term ‘ethnic cleansing’ has also been used to describe the atrocities in Gujarat.
7 In India, both upper caste Hindus and Muslims live disproportionately in the urban areas. They thus, constitute the two main voting blocks in bipolar urban races.
8 A favourite strategy of Hindu nationalist leaders who calculate that they will gain electorally from polarisation around a Hindu identity is to organise unusually large religious processions that take new routes through minority neighbourhoods, to hoist the national flag over a disputed site, or take out processions to celebrate national anniversaries.
9 L K Advani is a senior BJP politician and the Union Minister of Home Affairs, government of India at the time of writing this paper.
10 State autonomy signifies the power of the administration to take independent action.
11 State capacity includes the fiscal disposition, judicial capacity and rate of transfers within the state.
12 Civic engagement leads to lower levels of ethnic violence.
13 Some examples are: business associations, professional organisations, film clubs, sports clubs, trade unions, etc.
14 Some examples are: Hindu and Muslim families visiting each other, eating together, joint participation in festivals, etc.
15 These eight cities represent a mere 18 per cent of India’s urban population even though they account for 49 per cent of all urban riot deaths.
16 In an Interaction with students and members of the faculty at the Jawaharlal Nehru University, New Delhi, India dated January 12, 2004.

17 Interaction with students and members of the faculty at the Jawaharlal Nehru University, New Delhi, India dated January 12, 2004.

18 Includes the revision of history text books used as school curricula where attempts are made to interpret history according to the political convenience of that time.
19 Excerpts from Imtiaz Ahmad’s original manuscripts.


Brass, Paul (2003): The Production of Hindu-Muslim Violence in Contemporary India, University of Washington Press, Washington, DC, US.
Jaffrelot, Christophe (2001): 'The Politics of Processions and Hindu-Muslim Riots', Atul Kohli and Amrita Basu (eds), Community Conflicts in India, OUP, India.
Varshney, Ashutosh (2002): Ethnic Conflict and Civic Life: Hindus and Muslims in India, Yale University Press, US.
Wilkinson, Steven (2002): 'Putting Gujarat in Perspective', Economic and Political Weekly, April 27.
– (2004): 'Votes and Violence: Electoral Competition and Ethnic Riots in India', Cambridge Studies in Comparative Politics Series, Cambridge University Press, New York and Cambridge, UK.