Paul Brass spent a scholarly lifetime on the politics of north India, and the conclusion he drew from it runs against most popular explanation of communal conflict. Group identities, he argued, are not ancient inheritances that erupt when restraint weakens; they are constructed, selected from available cultural material and mobilised by elites who have interests in their existence. Nor are riots spontaneous. They are produced — organised, rehearsed and then interpreted — by identifiable actors who benefit from them.
This position, usually labelled instrumentalism or constructivism, gives Brass two lasting contributions to the sociology of Indian politics: a theory of ethnic identity formation set out in Ethnicity and Nationalism (1991), and a theory of collective violence developed in Theft of an Idol (1997) and The Production of Hindu-Muslim Violence in Contemporary India (2003), whose central concept is the institutionalised riot system.
Life and intellectual formation
Brass was an American political scientist who taught for most of his career at the University of Washington, and his field was Uttar Pradesh — studied continuously, with repeated visits, from the late 1950s onwards. That unusual longitudinal depth shaped his method. He worked from party records, factional histories, election data, police and press accounts, and long interviews with politicians and local notables, and he was as interested in how events came to be reported as in what happened.
His first book, Factional Politics in an Indian State (1965), examined the Congress Party in Uttar Pradesh and found that its politics was organised less by ideology or class than by shifting factions built on personal followings, patronage and locality. This early finding set his lifelong assumption: political outcomes in India are best explained by the strategies of organised local actors competing for power, not by the automatic expression of social categories.
The instrumentalist theory of ethnicity
Language, Religion and Politics in North India (1974) supplied the evidence and Ethnicity and Nationalism (1991) the general statement. Brass distinguishes levels: an ethnic category is simply a set of people who share some objective marker — a language, a religion, a region — of which they may be barely conscious; an ethnic community exists when that marker becomes the basis of subjective identity and organisation; a nationality or nation exists when the community makes political claims to autonomy or statehood. Nothing carries a group automatically from one level to the next.
What moves it is elite action. Cultural difference offers a repertoire of possible markers — language, script, dress, diet, sacred sites, historical memory — and political entrepreneurs select from that repertoire, emphasising the markers that best serve mobilisation and suppressing internal differences that would divide the group. Success depends on conditions: whether the chosen symbols converge on the same population, whether the group has enough internal stratification to produce leaders and enough resources to sustain organisation, and above all how the state responds — whether it concedes, represses or ignores the claim. Identity, on this account, is an outcome of political competition rather than its cause.
His case studies made the argument concrete: the Maithili movement in Bihar, which largely failed; the Punjabi Suba demand and Sikh identity formation, which succeeded; and the emergence of Muslim political separatism in the United Provinces, where he traced the political construction of a community that was internally diverse in language, sect and class. This last case produced his well-known exchange with Francis Robinson, who argued that a genuine, pre-existing Muslim community consciousness — not merely elite manufacture — lay behind the separatist movement. The Brass-Robinson debate remains the standard illustration of instrumentalism against a more primordialist reading.
Theft of an Idol and the framing of violence
Theft of an Idol (1997) begins from small incidents in eastern Uttar Pradesh — a theft, a death, a dispute over land or a woman — and asks how such events are turned into episodes of collective violence carrying communal or caste meaning. Brass's answer is that the meaning is not in the event. Every incident is immediately subject to competing accounts: the police produce one, local politicians another, rival community leaders a third, the press a fourth, each with an interest in a particular reading. What determines whether an incident becomes a riot is which narrative frame prevails and who has the capacity to make it prevail.
This yields the methodological warning that gives the book its force. Because all knowledge of a riot reaches the analyst through interested accounts, the standard explanations of Indian violence — economic competition, provocation, longstanding animosity — are often reproductions of one side's framing rather than findings. Brass also objects to the word riot itself, since it suggests a sudden, leaderless crowd and thereby conceals organisation and responsibility.
The institutionalised riot system
The Production of Hindu-Muslim Violence in Contemporary India (2003), built substantially on decades of work in Aligarh, states the concept most fully. Communal violence in India is not randomly distributed: it recurs in the same cities over decades while comparable neighbouring towns remain quiet. That pattern, Brass argues, points to standing organisation — an institutionalised riot system, a persistent network of specialists and interests that can generate violence when it is wanted.
He describes the system as operating in three phases. Preparation and rehearsal is continuous: everyday polarisation, provocative processions and slogans, rumour circuits, small confrontations that keep tension available. Activation is triggered when an incident is converted into an outrage by those he calls conversion specialists, and crowds are mobilised by known actors — party workers, criminal entrepreneurs, local strongmen — often with police inaction or complicity. Explanation follows: the violence is retrospectively interpreted by politicians, media and inquiry commissions in ways that assign blame, secure electoral advantage and normalise the event as an eruption of ancient hatred, leaving the system intact for next time. Riots, in this reading, are functional — they consolidate vote blocs, discipline minorities, remove competitors and distribute spoils — which is why they are produced.
Place in the study of Indian society
Brass belongs with the constructivist turn in the social sciences, alongside Benedict Anderson on imagined communities and Eric Hobsbawm on invented tradition, but with a fieldworker's attention to district-level politics. Within Indian studies he is one of the two indispensable authorities on communal violence, and his focus on organisation, incentive and state response supplies the political counterweight to culturalist and psychological accounts of communalism such as Ashis Nandy's. His broader work on parties, factions and Uttar Pradesh politics also made him a leading interpreter of Indian democracy's actual working.
Criticisms
Instrumentalism draws a persistent objection: it explains supply and ignores demand. If identity is manufactured by elites, why do ordinary people respond to some appeals and not others, and why do they risk their lives for them? Critics answer that pre-existing sentiment, memory and religious meaning must be doing real work, which returns some ground to primordialist and culturalist accounts, as Robinson argued and as Clifford Geertz's notion of primordial attachment implies. Related to this is the charge of reductionism — treating culture as a mere resource strips it of the affective and moral weight that participants clearly give it, and cannot account for the emotional dynamics of crowd violence.
The riot-system concept has been questioned on evidential grounds: a covert network is by nature hard to document, and demonstrating that it exists, persists and is deliberately maintained is difficult without conflating organisation with correlation. Ashutosh Varshney offered an influential alternative, arguing that variation between cities is better explained by the presence or absence of intercommunal civic and associational life, while Steven Wilkinson located the key variable in electoral incentives and the willingness of state governments to order the police to act. Both accept that violence is political but locate the mechanism elsewhere. Finally, Brass has been criticised for an excessive focus on Uttar Pradesh, which raises the question of how far the model travels to Gujarat, Maharashtra or the Northeast.
Legacy
Brass changed the questions asked about Indian communal violence. After him it is no longer respectable to describe a riot as spontaneous without asking who organised it, who profited, who framed it and why the same cities burn repeatedly. The vocabulary of production, riot specialists and institutionalised systems is now standard in the literature, in journalism and in the arguments of civil liberties inquiries, and the constructivist account of identity underlies most current work on caste, language and regional mobilisation in India.
For the UPSC answer
Brass is the strongest single authority to cite in questions on communalism and ethnic conflict, so state his thesis sharply: ethnic identity is constructed and mobilised by elites who select from available cultural markers, and communal violence is produced by an institutionalised riot system rather than erupting spontaneously. Give the three phases — preparation and rehearsal, activation by conversion specialists, and retrospective explanation — because the sequence demonstrates organisation and earns analytical credit. Use him comparatively: against primordialist readings such as Robinson's, alongside Anderson's imagined communities, and against Nandy's psychological and cultural account of communalism. Close with the standard limitation — instrumentalism explains elite supply but not popular demand — and with Varshney's civic engagement and Wilkinson's electoral incentives as the leading alternative explanations.
References & further reading
- Brass, Paul R. (1965). Factional Politics in an Indian State: The Congress Party in Uttar Pradesh. Berkeley: University of California Press.
- Brass, Paul R. (1974). Language, Religion and Politics in North India. Cambridge: Cambridge University Press.
- Brass, Paul R. (1990). The Politics of India Since Independence. Cambridge: Cambridge University Press.
- Brass, Paul R. (1991). Ethnicity and Nationalism: Theory and Comparison. New Delhi: Sage Publications.
- Brass, Paul R. (1997). Theft of an Idol: Text and Context in the Representation of Collective Violence. Princeton: Princeton University Press.
- Brass, Paul R. (2003). The Production of Hindu-Muslim Violence in Contemporary India. Seattle: University of Washington Press.