Contentious Conversation by Charles Tilly
AYODHYA, India, long sheltered a sixteenth-century mosque, Babri Masjid, named for the first Mughal emperor, Babur. Ayodhya attracted worldwide attention on December 6, 1992, when Hindu militants destroyed Ayodhya's Muslim shrine, began construction of a Hindu temple on the same site, and launched a nationwide series of struggles that eventually produced some twelve hundred deaths (Tambiah, 1996, p. 251; Bose and Jalal, 1998, p. 228). But the campaign behind that newsworthy event began a decade earlier. During the 1980s, militant Hindu groups started demanding destruction of the mosque and erection of a temple to Ram, epic hero of the Ramayana. Just before the 1989 elections, Bharatiya Janata Party (BJP) activists transported what they called holy bricks to Ayodhya and ceremoniously laid a foundation for their temple.
The following year, President Lal Advani of the BJP took his chariot caravan on a pilgrimage (rath yatra)across northern India, threatening along the way to start building the Ram temple in Ayodhya. Advani started his pilgrimage in Somnath, fabled site of a great Hindu temple destroyed by Muslim marauders. Advani's followers had fashioned his Toyota van into a simulacrum of legendary hero Arjuna's chariot, an image familiar from Peter Brook's film Mahabharata. As the BJP caravan passed through towns and villages, Advani's chariot attracted gifts of flower petals, coconut, burning incense, sandalwood paste, and prayer from local women. Authorities arrested Advani before he could begin the last lap of his journey to Ayodhya, but not before many of his followers had preceded him to the city. When some of them broke through police barricades near the offending mosque, police fired on them, killing "scores" of BJP activists (Kakar, 1996, p. 51).
Both sides represented their actions as virtuous violence--one side as defense of public order, the other side as sacrifice for a holy cause. Hindu activists made a great pageant of cremating the victims' bodies on a nearby river bank, then returning martyrs' ashes to their homes in various parts of India. Soon the fatalities at Ayodhya became themes of widespread Hindu-Muslim-police clashes. Those conflicts intersected with higher caste students' public resistance to the national government's revival of an affirmative action program on behalf of Other Backward Classes (Tambiah, 1996, p. 249). In Hyderabad, reports Sudhir Kakar,more than a thousand miles to the south of Ayodhya, the riots began with the killing of Sardar, a Muslim auto-rickshaw driver, by two Hindus. Although the murder was later linked to a land dispute between two rival gangs, at the time of the killing it was framed in the context of rising Hindu-Muslim tensions in the city. Muslims retaliated by stabbing four Hindus in different parts of the walled city. Then Majid Khan, an influential local leader of Subzimandi who lives and flourishes in the shaded space formed by the intersection of crime and politics, was attacked with a sword by some BJP workers and the rumor spread that he had died. Muslim mobs came out into the alleys and streets of the walled city, to be followed by Hindu mobs in their areas of strength, and the 1990 riot was on. It was to last for ten weeks, claim more than three hundred lives and thousands of wounded (Kakar, 1996, p. 51).
As his remarkable Colors of Violence unfolds, Sudhir Kakar seeks explanations for Hyderabad's 1990 violence by reporting discussions with some of the principals (including Majid Khan, still very much alive), reflecting on the identities involved, and reconstructing the psychological orientations that facilitate lethal violence, tie establishes the deep grounding of ostensibly spontaneous intercommunal violence in everyday social relations and in the organization of such groups as the Hindu wrestlers-thugs-activists mobilized by local leaders including Majid Khan.
Kakar does not, however, present his work as popular history or organizational analysis. On the contrary. Kakar, a professional psychoanalyst, deliberately sets his "primordialist" account of Hindu-Muslim conflict against the "instrumentalist" accounts he rightly sees as predominating in current social-scientific explanations of ethnic and religius conflict. "There are many social scientists and political analysts," he declares,who would locate the enhancements of ethnicity (cultural identity in my terms) in a particular group not in social-psychological processes but in the competition between elites for political power and economic resources. In fact, this has been the dominant explanation for the occurrence of Hindu-Muslim riots .... Cultural identity according to this view is not a fixed or given dimension of communities but a variable one which takes form in the process of political mobilization by the elite, a mobilization which arises from the broader political and economic environment (Kakar, 1996, pp. 149-50).
At length and through finely crafted vignettes, Kakar makes his case against such instrumental interpretations of cultural identity and his case for shared forms of consciousness based on common experience, reinforced by hostile interaction, and mediated by deep psychological mechanisms. He commits himself to phenomenological reductionism. That is his version of "primordialism."
Instead of choosing between starkly opposed instrumentalist and primordialist accounts, however, students of South Asia's Hindu-Muslim conflicts, of ethnic mobilization, of nationalism, indeed of contentious politics in general can adopt a third alternative: they can recognize the conversational character of contention. They can examine such a conversation's location in continuously negotiated interchanges among specific interlocutors, its constraint and mediation by historically accumulated understandings concerning identities and relations of the parties, its incessant modification of those identities and relations, hence its crucial causal contribution to interactions that instrumentalists explain on the basis of individual or collective interests and primordialists explain on the basis of deeply grounded individual or collective sentiments.
Sudhir Sakar shares with many canny observers a tendency favored by conversational convention: a propensity to shift from shrewd interactional observation into individualistic theories about what he has observed. That propensity extends to observations of his own observation processes. Sahba, his collaborator in many any interviews, was a Muslim woman. "The Muslim pehlwans [strong men]," Sakar notes,had been open with Sahba but understandably guarded when I was also present. With Sahba they could express their bitterness and contempt for Hindus, show their pride in their role in the protection of the community from the Hindu enemy. In my presence, they became less Muslim and more inclined to express universal humanist sentiments. For instance, there was pious talk, not exactly reassuring, that if I were cut my blood would be exactly the same color as theirs. By the end of the interviews, though, all the pehlwans were perceptibly warmer. I like to believe that this opening up was because they sensed my genuine interest in them as persons rather than being due to any typical "shrink" "hm-ms," phrases, or inflections. I suspect, though, that their different--although, for my purposes, highly complementary--psychic agendas when talking to Sahba and to me were dictated by shifts in their own sense of identity. In other words, with Sahba, a Muslim, their self-representation was more in terms of a shared social identity. With me, a Hindu, once they felt reassured that the situation did not contain any threat, personal identity became more salient, influencing their self-representations accordingly (Kakar, 1996, pp. 76-77).
Sakar is describing a process in which people's deployment of identities shifts as a function of their conversation partners and modifies as a consequence of conversation itself. Similarly, Sakar's vivid descriptions of interviews and events alike teem with interactions among participants in the course of which interpersonal negotiation transforms expressed identities and behavior attached to them. His subtle observations belie the psychic gloss he gives them.
My point is not to carp at a brilliant analyst's logical inconsistencies but to underscore the conversational character of political contention. If we regard conversation as continuously negotiated communication and contention as mutual claim-making that bears significantly on the parties' interests--which is how I propose we understand the two terms for present purposes--then the two social phenomena overlap extensively. They overlap in the zone we might call contentious conversation. Conversation is contentious to the extent that it embodies mutual and contradictory claims, claims that, if realized, would significantly alter the longer-term behavior of at least one participant. Contentious conversation certainly activates visceral emotions, neurally processed cognitions, and individual anatomical performances. It also operates within limits set by historically formed conventions, with regard to collectively constituted interests, and in response to stimuli from leaders or bystanders. But it proceeds within a substantial, causally coherent realm that we cannot explain away by referring either to individual psyches or to group interests. Contentious conversation follows its own causal logic.
Beth Roy's subtle study of ostensibly communal conflict of 1954 in a Pakistani village (Bangladeshi by the time she arrived there in the 1980s) identifies the significance of that causal logic. She shows how a local scuffle among farmers working adjacent fields escalated into a full-scale alignment of self-identified Hindus against self-identified Muslims. At the start, Golam Fakir (categorically Muslim) and Kumar Tarkhania (categorically Hindu) interacted as disputants over the fact that Fakir's untethered cow had eaten Tarkhania's lentils. The series of confrontations did not begin as communal mobilization, but it approximated increasingly to classic models of Hindu/Muslim strife as it grew in geographic scope and mounted the national administrative hierarchy.
In addition to its empathetic description and astute detective work, Roy's study fascinates by its patient unpacking of complexities in actors, actions, and identities. Some Trouble With Cows (the title echoes one of the first stories about the 1954 conflicts Roy collected) centers on questions of identity:
When I consider stories of village communalism, I want to know how people saw their world, how they placed their own desires within it, and how their sense of political possibility was influenced by distant winds of change. It has become common to assert that the most intimate domestic behaviors are in fact socially constructed. Collective experience is translated into psychological reality through a web of ideas internalized as invisible assumptions about the world. To unravel the psychological realities of collective behavior, I believe we must look to shared areas of understanding and social location. For instance, group actions are formulated from the experience of identity, that is, the complex construction of an individual's location in the community and her ties with others. Similarly, the will to action is born of detailed ideologies that often are experienced as common sense or unexamined assumptions about rights and powers (Roy, 1994, p. 3).
In this introductory passage and throughout her superb reconstruction of old conflicts, Roy exhibits ambivalence between two points of view, sometimes treating identity and action as individual mental realities multiplied, sometimes locating identity and action in social relations: "an individual's location in the community and her ties with others." She thereby pinpoints a major difficulty in contemporary studies of contentious politics (see Cerulo, 1997, pp. 393-94). Analysts of social construction have called attention to the difficulty, but have failed to resolve it.
Here is the difficulty: humans live in flesh-and-blood bodies; accumulate traces of experiences in their nervous systems; organize current encounters with the world as cognitions, emotions, and intentional actions; and tell stories about themselves in which they acted deliberately and efficaciously or were blocked from doing so by uncontrolled emotion, weakness, malevolent others, bad luck, or recalcitrant nature. They tell similar stories about other people. Humans come to believe in a world full of continuous, nearly bounded, self-propelling individuals whose intentions interact with accidents and natural limits to produce all of social life. In many versions, those "natural limits" feature norms, values, and scripts inculcated and enforced by powerful others--but then internalized by self-propelling individuals. Accounts in this vein adopt phenomenological reductionism. They reduce social life to states of individual consciousness.
Closely observed, however, the same humans turn out to be interacting repeatedly with others, renegotiating who they are, adjusting the boundaries they occupy, modifying their actions in rapid response to other people's reactions, selecting among and altering available scripts, improvising new forms of joint action, speaking sentences no one has ever uttered before, yet responding predictably to their locations within webs of social relations they themselves cannot map in detail. They tell stories about themselves and others that facilitate their social interaction rather than laying out verifiable facts about individual lives. They actually live in deeply relational worlds. If social construction occurs, it happens socially, not in isolated recesses of individual minds.
The problem becomes acute in descriptions and explanations of contentious politics. Political actors typically give individualized accounts of participation in contention, although the "individuals" to which they attribute bounded, unified, continuous self-propulsion are often collective actors such as communities, classes, armies, firms, unions, interest groups, or social movements. They attach moral evaluations and responsibilities to the individuals praising or condemning them for their actions, grading their announced identities from unacceptable (for example, a mob) to laudable (for example, martyrs). Accordingly, strenuous effort in contentious politics goes into contested representations of crucial actors as worthy or unworthy, unified or fragmented, large or small, committed or uncommitted, powerful or weak, well connected or isolated, durable or evanescent, reasonable or irrational, greedy or generous.
Meticulous observation of that same effort, however, eventually tells even a naive observer what almost every combat officer, union leader, or political organizer acknowledges in private--that both public representations of political identities and other forms of participation in struggle proceed through intense coordination; contingent improvisation; tactical maneuvering; responses to signals from other participants; on-the-spot reinterpretations of what is possible, desirable, or efficacious; and strings of unexpected outcomes inciting new improvisations. Interactions among actors with shifting boundaries, internal structures, and identities turn out to permeate what in retrospect or in distant perspective analysts call actor-driven wars, strikes, rebellions, electoral campaigns, or social movements. Hence the difficulty of reconciling individualistic images with interactive realities.
South Asia's Hindu/Muslim conflicts present that difficulty acutely. As Sudhir Kakar, Beth Roy, and many other recent students of South Asia have shown, analysts have readily available single-actor characterizations of Hindus in general, of Muslims in general, and therefore of their interaction as the inevitable consequence of contact between incompatible mentalities. But relations among persons who belong to the categories "Hindu" and "Muslim" take a wide range of forms, from avoidance to cohabitation. They frequently take place without reference to religious affiliation.
As Hyderabad's bloody contention of 1990 illustrates, furthermore, many confrontations that begin under other definitions eventually activate and receive coding as conventional expressions of communal hostility. We should notice the analogy to family disputes in which available epithets, memories, and lines of fractionation only enter the struggle as it escalates or as third parties enter the fray. Such phenomena provide empirical justification for the "Instrumentalist" accounts of communal conflicts that Sudhir Kakar rightly challenges as sole explanations of the life-threatening interactions he studies. The effective response to instrumental reductionism, however, consists not of turning to phenomenological reductionism but of recognizing the conversational dynamics of such disputes.
The Hindu/Muslim conversation engages multiple interlocutors in varied settings. It therefore takes place in many modes. Majid Khan's interchanges with his counterparts who mobilize their own wrestlers-thugs-activists on behalf of Hindu causes differ greatly from the initial angry conversation between Golam Fakir and Kumar Tarkhania in the Panipur of 1954, which differ in form from exchanges between high-caste Panipur resident Mr. Ghosh and the Muslim officials who came to Panipur when the local conflict started drawing in outsiders. More important, Fakhir and Tarkhania did not initially respond to each other as representatives of competing categories, but as multiply linked poor farmers within the same community. Only as their conflict escalated did they fall into ranks of self-identified Muslims on one side, self identified Hindus on the other. Every pair of interlocutors has its own idioms and its own history, both of which frame their conversation. Which idioms they actually deploy and which histories they invoke, furthermore, varies with who else is participating. Because of learning and of constraint by relations to third parties, to be sure, conversations within similar and connected pairs share many properties. Yet we must understand that contentious conversation proceeds through incessant improvisation within limits set by the previous histories and relations of particular interlocutors.
The conversational analogy applies to a wide range of political contention. We could pursue it across other instances of ethnic and religious conflict, expressions of nationalism, electoral campaigns, revolutions, parliamentary debates, industrial conflict, and much more. Let us, however, move the discussion onto familiar ground. In his analysis of thousands of demonstrations in Marseille, Nantes, Paris, and other parts of France between 1979 and 1993, Olivier Fillieule identifies the stylized but incessant interchanges that occur among demonstrators, spectators, police, officials, and other persons involved in any demonstration. Although observers, reporters, analysts, and critics often reduce such events to attitudes and actions of the persons who occupy the street with banners, chants, and other dramatizations of their demands, detailed accounts drawn from such sources as police blotters reveal continuous streams of mutual deliberations, taunts, threats, attacks, retreats, delegations, agreements, and much more--usually reported from a single viewpoint, but always reflecting participation in communicative interchange.
Consider the 1986 testimony of an experienced commander of riot police. When asked what would happen if he received contradictory orders from the local police commissioner and his own unit's superiors, he replied that it was unlikely, but added:
At a moment like that I would probably decide for myself, as I actually once did for our buses. When we were setting ourselves up, the commissioner of the 16th [Parisian] arrondissement asked us, contrary to my view, to reinforce the street barriers with our vehicles. When I saw that the demonstrators were trying to set the barricades on fire, on my own initiative I had the vehicles moved to the middle of the bridge; my buses retreated. The other vehicles of the Parisian police that didn't retreat got burned (Fillieule, 1997, p. 257).
Torching buses and moving them back obviously constitutes a crude sort of dialogue. So do deploying shields against stonethrowers, wading into a crowd with flailing clubs, or even receiving delegations from demonstrators at a minister's office. Yet the dialogue involved is real and consequential. It engages two parties, or more. How one party responds to another affects what happens next. The conversation places unceasing improvisation within strongly defined conventions that mark the ongoing interchange as a demonstration rather than, say, a strike, a public meeting, an election rally, routine lobbying, or a coup d'etat.
As with all conversation, contentious conversation has a delightfully paradoxical property: improvisation within constraints that produce order. Demonstrators, counterdemonstrators, police, authorities, and other participants in demonstrations improvise incessantly, jockeying for surprise, effect, and strategic advantage. If they simply repeated the routines they had followed during a previous encounter, they would resemble people who utter bromides; they would cede all strategic advantage to their partners and come off as dull automata. Yet as compared with all the actions and interactions of which they are capable, they concentrate their efforts within a narrow range of symbols, utterances, and interactions. Demonstrators often march in ranks, display banners, shout slogans, and present petitions, but rarely carry machine guns, defecate in the street, strip naked, strangle spectators, sing nursery rhymes, stop to buy the day's groceries, or travel in taxis--except, of course, if they appear together in taxi drivers demanding protection from muggers. If participants in contentious conversation did not adopt recognizable idioms, they would undercut their own efforts to coordinate actions, convey messages, and influence objects of their claims.
As in less contentious forms of conversation, contentious conversation produces order by means of improvisation within constraints. At demonstration's end, marchers go off to their daily affairs, police take down their barricades, street-cleaners sweep up debris, merchants unshutter their shops, reporters file their stories, officials deliberate on how, if at all, they will respond to the day's events, and politicians chat about what it all means for their causes. In the meantime, participants have broadcast messages to their various audiences: the regime is rotten, we deserve better treatment, proposal X is an abomination, everyone should adopt proposal Y, or something else along these lines. Members of audiences, furthermore, have started to form judgments about how effectively participants performed, and how credible were their contentious claims.
We can capture the theatrical side of contention by speaking of contentious repertoires (Tarrow, 1998, chap. 2; Traugott 1995). Any pair of interlocutors has available to it a limited number of previously created performances within which the people involved can make claims. That array of performances constitutes their repertoire. Seen from a distance, the same citizen-official pair that figures in demonstrations chiefly as claimant and object of claims also has available as claim-making vehicles petitions, elections, public meetings, lobbying delegations, bureaucratic letters, and other well-established performances; those vehicles form the big clumps within their repertoire. Seen from closer at hand, the demonstration itself displays the finer grain of deploying barricades, shouting slogans, making speeches from balconies struggling for control of public spaces, and a dozen other variable elements. Just as a company of actors deploys both a set of
Moliere dramas and a hard-earned stock of two-line jokes, sword battles, pratfalls, gestures, embraces, double takes, and curtain calls, both the big claim-making routines and their fine, variable elements belong to their participants' contentious repertoire. The elements, after all, frequently recur from one kind of contentious performance to another, as when people voice the same slogans in demonstrations, pamphlets, petitions, and public meetings.
Like any other sorts of conversational forms, contentious repertoires embody history and culture. Participants and observers draw on previous experiences, incorporate readily available symbols, make selective references to shared memories, strategize as a function of what happened last time, notice the impact (if any) of their improvisations, compare notes after the fact. Repertoires matter to the course and outcome of contention for several reasons. First, they incorporate scripts that participants know to be performable, in which they know the parts and collaborative routines required, and of whose requisites and possible outcomes they share at least some awareness; all these features facilitate mobilization of participants for a new performance. Second, they draw meaning and effectiveness in part from their connection with previous iterations of the same performances--our opponents' recent meetings or our own, the received history (however mythical) of previous demonstrations, and so on. Third, they eliminate from consideration, and often from consciousness, a vast range of claim-making performances of which participants are technically capable, indeed may even undertake in other circumstances; thus demonstrators besieging their city hall leave behind the prayers, confessions, rituals, and offerings by which they regularly ask their gods for favors elsewhere. Finally, participants in contentious politics learn to grade, value, and contest the quality of performances, disputing how many people actually took part in a demonstration, arguing about how well the message got across, second-guessing the plans and strategies of the police, the mayor, or local leaders.
In the BJP's approach to Ayodhya, we saw contentious repertoires in full action. The chariot pilgrimage, the ostentatious building of temples on sacred sites, the bloody encounter between demonstrators and baton-wielding police all form recurrent performances of India's contemporary contentious conversation. The high stakes of such performances give us a salutary reminder not to take our theatrical metaphor as an indication of artificiality or triviality. In Ayodhya, Hyderabad, and elsewhere, these are deadly serious conversations.
Among other things, political identities are at stake. To continue with theatrical metaphors, claim-makers are acting out answers to the question, "Who are you?" When contentious conversation arises in the course of routine social life, the answers are often obvious; we are whoever we were before the contention began, employees of a given company, purchasers in a weekly market, worshippers at a shrine, public officials doing public business, police officers patrolling their beats. Much of the time, however, identities remain unclear until participants dramatize them. All people have multiple identities at their disposal, each one attached to a somewhat different set of social relations: neighbor, spouse, farmer, customer, tenant, schoolmate, lover, or citizen. Some available identities appear in public only intermittently, as is the case with many varieties of party affiliation, association membership, and adhesion to social movements. In these circumstances, participants in contentious conversation regularly make a point of the capacities in which they are interacting, of the identities they are activating
As in ordinary conversation, some performances actually center on the assertion of identities rather than the making of specific claims. One side says, "Recognize us as significant actors of a certain kind," while the other side accepts or contests that assertion of identity. Social movement activists often initiate performances whose central message declares that the activists and/or the constituency they claim to represent are "WUNC": worthy, united, numerous, and committed. Demonstration of worthiness varies by cultural setting, but often includes evidence of' decorous self-control and standing in the community. Unity calls for moving in time, speaking together, broadcasting the same message, and collaborating in the same actions--even when the occasion calls for displaying how wide a range of community members support a given program. Numbers matter greatly, as frequent disputes among activists, observers, rivals, and police over how many people actually participated in an action indicate. Claims of numbers, however, may also refer to supporters as indicated by association memberships, financial contributions, opinion polls, electoral results, or other quantitative signs. Commitment, finally, comes across in evidence that participants are willing to bear costs and face risks on behalf of their common cause; the costs and risks can vary from laying down your life to coming from tar away to join a collective action.
If any of the elements--worthiness, unity, numbers, and commitment--visibly falls to zero, a social movement loses credibility. Above the zero threshold, however, a high value on one element compensates a low value on another people who show strong signs of commitment (for example by taking life-threatening risks) need not be so numerous as those who merely show up for a rally. The notable presence of worthiness in the form of priests, dignitaries, prize winners, and victims can easily make up for a certain disunity of voiced aims among other participants. The relative weight of worthiness, unity, numbers, and commitment defines rather different kinds of movements.
What's going on? Social movements link two complementary activities: assertions of identity and statements of demands. The two activities' relative salience varies from one phase of social movement activity to another as well as from one kind of social movement to another. As compared with strikes, revolutions, coups d'etat, and many other forms of contentious politics, nevertheless, social movements stand out for their emphasis on identity assertion. They emphasize public assertion of identities whose possessors are worthy, unified, numerous, and committed. They do so because social movements grew up in the nineteenth century as means by which people currently excluded from political power could band together and claim that power-holders should attend to their interests, or the interests they represented. Otherwise, the actions of social movement participants signaled, they had the capacity and will to disrupt and alter routine political life. Recognition of their claimed identities as wronged workers, dispossessed peasants, or persecuted religious minorities constituted them as political actors, but also drew them into bargaining collectively with existing holders of power. That stress on identity assertion persists in social movements, especially in their earlier stages, to the present day. Social movements continue to assert the right to respect and political voice of indigenous peoples, gays, conservative Christians, unborn children, laboratory animals, and even trees--the latter three categories through the mediation of their self-designated protectors. Identity assertion in social movements has clear counterparts in everyday conversation. Effective conversation establishes in whose names and in what capacities the parties are speaking: Are we exchanging news as friends, neighbors, or business associates? Do I recognize you as a certified expert on our topic and myself as a grateful member of your lay audience? Are you speaking for your political constituency or simply on your own? Are we strangers on a train? The universal requirement for establishment of identities does not preclude error, dissimulation, contestation, or double entendre concerning the relationship among the parties. A friend can turn out to be a spy, while a sterile conversation for public consumption can carry erotic, subversive, or satirical overtones that bespeak a second relationship among the interlocutors. Conversation still proceeds on the basis of defined identities, and often centers on establishing just what those identities are.
Whether contentious or otherwise, conversational conventions and the very course of conversation exert significant influence over their participants' behavior. Established identities and their associated social relations, for example, always make a difference to the course of interaction. In the extreme case of Hindu/Muslim massacres, Sudhir Kakar shows that neighbors of' one faith often warned members of the other religious category of a coming attack, that in the heat of bloodshed Muslim households sometimes hid Hindus or vice versa, and that bands of killers ordinarily avoided blood-letting within their own neighborhoods; reduction of other people to nothing but their broadest public identities facilitated murder, while the presence of multiple ties and identities inhibited it. In the extreme reduction, extermination squads sorted potential victims by clothing or, for males, the condition of their penises, circumcision signaling Muslim identity, its absence Hindu identity.
Short of such an extreme, conversationally established identities always shape social interaction, including political contention. In Panipur, only as the struggle initiated by a cow's eating lentils came to be redefined as aligning Muslims in general against Hindus in general did armed combat among organized forces emerge as the mode of interaction. After the lethal confrontation of 1954, furthermore, most conflict took place on a smaller scale, under different auspices, with the blessing of a widely shared story about the ability of Panipur's diverse residents to live in harmony. Stories that people create in the course of contentious conversation themselves affect subsequent social interaction, both because they filter collective memory and because they build in commitments to behave consistently with those stories.
Let me state the point more strongly. Conversation in general shapes social life by altering individual and collective understandings, by creating and transforming social ties, by generating cultural materials that are then available for subsequent social interchange, and by establishing, obliterating, or shifting commitments on the part of participants. The same is true of contentious conversation, for the same reasons. In both noncontentious and contentious conversation, these processes work, both through words and through a wide variety of nonverbal interchanges--not only gestures, body language, and deployment of physical objects, but also displays of symbols, spatial shifts, altered relations to physical settings, and interventions of third parties. The crucial processes of contentious politics are not instrumental in the sense of proceeding directly from the self-centered competition of elites for political power and economic resources. They are not primordial in the sense of expressing deeply grounded individual phenomenology. They are conversational in the sense of proceeding through historically situated, culturally constrained, negotiated, consequential interchanges among multiple parties. Whatever else it requires, the explanation of political contention demands that analysts take mere speech acts seriously.
(*) I have adapted a few passages from a draft chapter of Dynamics of Contention, a book in progress for Cambridge University Press by Doug McAdam, Sidney Tarrow, and Charles Tilly. National Science Foundation Grant SBR 9601236 provided partial support for the year at the Center for Advanced Study in the Behavioral Sciences during which I wrote this paper.
References
Ashforth, Adam, The Politics of Official Discourse in Twentieth-Century South Africa (Oxford: Clarendon Press, 1990).
Beissinger, Mark R., "How Nationalisms Spread: Eastern Europe Adrift the Tides and Cycles of Nationalist Contention," Social Research 63 (1996): 97-146.
Beramendi, Justo G., Ramon Ma/z, and Xose M. Nunez, eds., Nationalism in Europe, Past and Present, 2 vols. (Santiago: Universidade de Santiago de Compostela).
Bjorn Claus, Alexander Grant, and Keith J. Stringer, eds., Nations, Nationalism and Patriotism in the European Past (Copenhagen: Academic Press, 1994).
Bjorn, Claus, Alexander Grant, and Keith J. Stringer, eds., Social and Political Identities in Western History (Copenhagen: Academic Press, 1994).
Bose, Sugata, and Ayesha Jalal, Modern South Asia: History, Culture, Political Economy (London: Routledge, 1998).
Brass, Paul R., The Politics of India Since Independence (Cambridge: Cambridge University Press, 1994).
Brubaker, Rogers, Nationalism Reframed: Nationhood and the National Question in the New Europe (Cambridge: Cambridge University Press, 1996).
Burt, Ronald S., Structural Holes: The Social Structure of Competition (Cambridge: Harvard University Press, 1992).
Calhoun, Craig, "The Problem of Identity in Collective Action" in Joan Huber, ed., Macro-Micro Linkages in Sociology (Newbury Park, Calif.: Sage, 1991).
Calhoun, Craig, "Nationalism and Ethnicity," Annual Review of Sociology 19 (1993): 211-39.
Cerulo, Karen A., "Identity Construction: New Issues, New Directions," Annual Review of Sociology 23 (1997): 385-409.
Connor, Walker, Ethnonationalism: The Quest for Understanding (Princeton: Princeton University Press, 1994).
Daniel, E. Valentine, Charred Lullabies: Chapters in an Anthropography of Violence (Princeton: Princeton University Press, 1996).
Davis, Richard H., "The Iconography of Rama's Chariot," in David Ludden, ed., Contesting the Nation: Religion, Community, and the Politics of Democracy in India (Philadelphia: University of Pennsylvania Press, 1996).
Druckman, Daniel, "Nationalism, Patriotism, and Group Loyalty--A Social Psychological Perspective," Mershon International Studies Review 38 (1994): 43-68.
Fein, Helen, "Accounting for Genocide after 1945: Theories and Some Findings," International Journal on Group Rights 1 (1993): 79-106.
Fein, Helen, "Revolutionary and Antirevolutionary Genocides: A Comparison of State Murders in Democratic Kampuchea, 1975 to 1979, and in Indonesia, 1965 to 1966," Comparative Studies in Society and History 35 (1993), 796-823.
Fillieule, Olivier, Strategies de la rue: Les manifestations en France (Paris: Presses de Sciences Po, 1997).
Freitag, Sandria B., "Contesting in Public: Colonial Legacies and Contemporary Communalism," in David Ludden, ed., Contesting the Nation: Religion, Community, and the Politics of Democracy in India (Philadelphia: University of Pennsylvania Press, 1996).
Ghosh, Amitav, In an Antique Land: History in the Guise of a Traveler's Tale (New York: Vintage, 1992).
Gurr, Ted Robert, Minorities at Risk: A Global View of Ethnopolitical Conflicts (Washington: U.S. Institute of Peace Press, 1993).
Gurr, Ted Robert, and Barbara Harff, Ethnic Conflict in World Politics (Boulder: Westview, 1994).
Harff, Barbara, and Ted Robert Gurr, "Victims of the State: Genocides, Politicides, and Group Repression since 1945," International Review of Victimology 1 (1990): 23-41.
Kakar, Sudhir, The Colors of Violence: Cultural Identities, Religion, and Conflict (Chicago: University of Chicago Press, 1996).
Lafargue, Jerome, Contestations democratiques en Afrique (Paris: Karthala and IFRA, 1996).
Lucassen, Leo, Zigeuner: Die Geschichte eines polizeilichen Ordnungsbegriffes in Deutschland, 1700-1945 (Cologne: Bohlau, 1996).
Mamdani, Mahmood, Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism (Princeton: Princeton University Press, 1996).
Marx, Anthony W., Making Race and Nation: A Comparison of the United States, South Africa, and Brazil (Cambridge: Cambridge University Press, 1998).
Ringmar, Erik, Identity, Interest and Action: A Cultural Explanation of Sweden's Intervention in the Thirty Years War (Cambridge: Cambridge University Press, 1996).
Roy, Beth, Some Trouble with Cows: Making Sense of Social Conflict (Berkeley: University of California Press, 1994).
Stoler, Ann Laura, "Rethinking Colonial Categories: European Communities and the Boundaries of Rule," Comparative Studies in Society and History 31 (1989): pp. 134-61.
Tambiah, Stanley J., Leveling Crowds: Ethnonationalist Conflicts and Collective Violence in South Asia (Berkeley: University of California Press, 1996).
Tambiah, Stanley J., "Friends, Neighbors, Enemies, Strangers: Aggressor and Victim in Civilian Ethnic Riots," Social Science and Medicine 45 (1997): 1177-88.
Tarrow, Sidney, Power in Movement (Cambridge: Cambridge University Press, 1998).
Tilly, Charles, "Police, Etat, contestation," Cahiers de la securite interieure 7 (1991): 13-18.
Tilly, Charles, Durable Inequality (Berkeley: University of California Press, 1998).
Traugott, Mark, ed., Repertoires and Cycles of Collective Action (Durham: Duke University Press, 1995).
de Waal, Alex, "Group Identity, Rationality, and the State," Critical Review 11 (1997): 279-89.
Williams, Robin, "The Sociology of Ethnic Conflicts: Comparative International Perspectives," Annual Review of Sociology 20 (1994): 49-79.
Winschafter, Elise Kimerling, Social Identity in Imperial Russia (DeKalb: Northern Illinois University Press, 1997).
By CHARLES TILLY
13617010037783X19980901