The Gujarat Pogrom of 2002
Published on: Mar 26, 2004

Paul R. Brass is professor emeritus of political science at the University of Washington. His most recent work is The Production of Hindu-Muslim Violence in Contemporary India (University of Washington Press, forthcoming 2003).

Events labelled “Hindu-Muslim riots” have been recurring features in India for three-quarters of a century or more. In northern and western India, especially, there are numerous cities and towns in which riots have become endemic. In such places, riots have, in effect, become a grisly form of dramatic production in which there are three phases: preparation/ rehearsal, activation/enactment, and explanation/interpretation.1 In these sites of endemic riot production, preparation and rehearsal are continuous activities. Activation or enactment of a large-scale riot takes place under particular circumstances, most notably in a context of intense political mobilization or electoral competition in which riots are precipitated as a device to consolidate the support of ethnic, religious, or other culturally marked groups by emphasizing the need for solidarity in face of the rival communal group. The third phase follows after the violence in a broader struggle to control the explanation or interpretation of the causes of the violence. In this phase, many other elements in society become involved, including journalists, politicians, social scientists, and public opinion generally.

At first, multiple narratives vie for primacy in controlling the explanation of violence. On the one hand, the predominant social forces attempt to insert an explanatory narrative into the prevailing discourse of order, while others seek to establish a new consensual hegemony that upsets existing power relations, that is, those which accept the violence as spontaneous, religious, mass-based, unpredictable, and impossible to prevent or control fully. This third phase is also marked by a process of blame displacement in which social scientists themselves become implicated, a process that fails to isolate effectively those most responsible for the production of violence, and instead diffuses blame widely, blurring responsibility, and thereby contributing to the perpetuation of violent productions in future, as well as the order that sustains them.

In India, all this takes place within a discourse of Hindu-Muslim hostility that denies the deliberate and purposive character of the violence by attributing it to the spontaneous reactions of ordinary Hindus and Muslims, locked in a web of mutual antagonisms said to have a long history. In the meantime, in post-Independence India, what are labelled Hindu-Muslim riots have more often than not been turned into pogroms and massacres of Muslims, in which few Hindus are killed. In fact, in sites of endemic rioting, there exist what I have called “institutionalized riot systems,” in which the organizations of militant Hindu nationalism are deeply implicated. Further, in these sites, persons can be identified who play specific roles in the preparation, enactment, and explanation of riots after the fact. Especially important are what I call the “fire tenders,” who keep Hindu-Muslim tensions alive through various inflammatory and inciting acts; “conversion specialists,” who lead and address mobs of potential rioters and give a signal to indicate if and when violence should commence; criminals and the poorest elements in society, recruited and rewarded for enacting the violence; and politicians and the vernacular media who, during the violence, and in its aftermath, draw attention away from the perpetrators of the violence by attributing it to the actions of an inflamed mass public. When successful, as it most often is, the principal beneficiaries of this process of blame displacement are the government and its political leaders, under whose watch such violence occurs. Here also, in the aftermath, social scientists become involved when they draw attention to the difficulties of “governance” in societies where interethnic and intercommunal animosities are allegedly rampant. They themselves then become implicated in a political discourse that, as Baxi has well put it, concerns itself with “the agonies of governance,” rather than with the “suffering” of the victims of misgovernance, and thereby normalizes the violence against its victims.2

These issues of images, labels, and responsibility emerged starkly once again in the months from February 27 through June 2002 in the western Indian state of Gujarat, where widespread killings, mostly of Muslims, were carried out on a scale, and with a ferocity, reminiscent of the genocidal massacres that occurred during the partition of the Punjab in 1947, and with the apparent involvement—by several eyewitness accounts—of ministers in the government itself, under the leadership of Bharatiya Janata Party (BJP) Chief Minister Narendra Modi.3 All the elements of riot production and interpretation outlined above appear in the great killings in Gujarat.

All available evidence, which is unusually well documented in the case of Gujarat, indicates beyond a shadow of doubt that the Sangh parivar(the umbrella organization of all militant Hindu organizations) was well-prepared and well-rehearsed to carry out the murderous, brutal, and sadistic attacks on Muslim men, women, and children.4 Although the precipitating incident that provided the pretext for these actions was unexpected, there was an apparent connection with local politics in the town of Godhra, the site of the incident; a clear connection with the vast movement of militant Hindu political mobilization centered around the demand for the building of a new temple to the god, Ram, in the northern Indian town of Ayodhya; and a post-pogrom intent on the part of the BJP government in Gujarat to take advantage of presumed Hindu political consolidation for electoral purposes, by calling for early state legislative assembly elections.

As in most such situations, the “causes” of the initial acts of violence appeared obscure and indeterminate, yet the process of blame displacement began before any credible facts at all emerged concerning the horrific killing of 58 persons, mostly kar sevaks (militant Hindu volunteers) returning from Ayodhya by train on February 27, burned alive in two bogeys of that train, at the railway station in Godhra. BJP leaders promptly blamed the Inter Services Intelligence unit of Pakistan (generally known as ISI) for the Godhra incidents. In the days following this and other BJP claims and accusations, however, news reports appeared that cast doubt on them and pointed to several other circumstances that opened up other interpretations. These included allegedly provocative, insulting, and lewd behavior on the part of the kar sevaks in relation to Muslim vendors at the train stations en route to Godhra, and in relation to Muslim passengers, including women. Other circumstances concerned local political rivalries between the Congress and the BJP, between rival factions in the Congress, and between local Muslim organizations. It was also revealed that the town of Godhra, with an approximately 40% Muslim, 60% Hindu population balance, has had a long history of communal riots, of which no less than thirteen were enumerated between 1947 and 1992.

But the Godhra incidents were quickly overshadowed by what followed, namely, a systematic pogrom enacted with precision and extreme brutality by persons and organizations in the institutionalized riot system5 of the RSS family of organizations, including members of the BJP government, the police, and even members of the elite Indian Administrative Service (IAS).6 This pogrom began on February 28, a day after the Godhra massacre, under the auspices of the Vishwa Hindu Parishad (VHP), which called for a state-wide bandh(closing down) to protest the killings in Godhra. The Gujarat pogrom continued until March 3, after which there was a hiatus followed by “a new round of violence” from March 15.7

Estimates of the numbers killed range from below a thousand to two thousand. Some thirty cities and towns in the state were reported to be “still under curfew” on March 27.8 Official figures provided by the Gujarat government9 also show the characteristic predominance of Muslims in the numbers killed during rioting: more than 5 Muslims to 1 Hindu, inclusive of the Hindus killed on the train at Godhra, but a ratio of 15 to 1 in the rioting that followed after Godhra. The numbers of displaced persons compelled to seek refuge in relief camps also speaks to the enormity of the cataclysm visited upon the Muslims of Gujarat: nearly 150,000 in 104 relief camps by mid-April.

Numerous features of these killings and destruction of property suggest the validity of the term pogrom and its systematic character. They include the destruction of over 500 mosques and dargahs (shrines).10 It has also been reported that many, if not most, police either stood aside or coordinated or participated in the violence against Muslims. Moreover, testifying to the high degree of preparation, the marauding mobs of killers carried lists of voters and other documents with them, which made it possible for them to identify the homes of Muslims who were to be killed and whose property was to be destroyed. Also on the riot scenes, according to eyewitnesses, were prominent BJP and VHP leaders, who moved along with the mobs of Hindu rioters. Sometimes they played the role of “conversion specialists,” addressing the mobs, after which they discreetly left, whereupon the mobs carried out their murderous attacks. As in Kanpur City in 1992 and in the great Aligarh riots of 1990-91, when attacks were made for the first time on predominantly Muslim mohallas at the outskirts of these cities,11so also in Gujarat, “Muslims were attacked even in areas where they constituted the majority.”12 Further, the killings extended to several villages in rural areas of the state. Also, as in Aligarh in 1990-91, several of the vernacular media agencies in Gujarat became, in effect, part of the institutionalized riot system of the Sangh parivar. A leading Gujarati newspaper, Sandesh, featured a front-page headline on February 28, “Avenge Blood with Blood,” above a story concerning a statement from the VHP. Sandesh and Gujarat Samachar featured many other false and incendiary stories in the following days, some of which virtually encouraged Hindus, in areas where violence had not yet spread, to kill Muslims.13

It is important to note further that the Gujarat pogrom transgressed beyond the boundaries of ordinary riots, pogroms, and massacres into the “zone of genocide.”14 In particular, the use of sexual molestation, rape, and murder of women, as well as children, including the reported case of cutting open a pregnant woman’s belly and killing the foetus, deserves note.

It is necessary to underline the implication in this pogrom not only of the BJP state government, its members, and its agents, but also that of the government of India, led by the BJP, which had the power and ultimate responsibility to stop this flagrant breakdown of law and order. Most significant was the failure to dismiss the Gujarat government, under Article 356 of the Constitution of India, for its inability or unwillingness to maintain law and order.

But members of the government of India compromised themselves and the central government in many other ways, some blatant, some subtle. Although Prime Minister Vajpayee, under pressure from the non-BJP constituents in his governing coalition, addressed the country on television on the third day of rioting “to denounce the Gujarat riots,”15 he did not visit Gujarat until thirty-six days after the Godhra massacre and the pogrom that followed it. He was then reported to have remarked “that the carnage had shamed India.”16 Aside from the fact that, in his eyes, it was India’s status in the world that was at stake, as much as or more than the plight of the victims of a state-supported pogrom, other features of his visit deserve note. Vajpayee visited Godhra first, thus expressing his solidarity with the Hindus who had been killed, the victims from the Sangh parivar. Also, he took with him on his tour of Gujarat central minister Uma Bharati, member of the VHP, whose speeches during the 1991 elections and prior to the destruction of the mosque at Ayodhya in 1992 were considered hostile to Muslims, and who was one of the most active proponents of the construction of a Hindu temple at that site.

On September 24, the “revenge and retaliation” drama was carried further in an act of “retributive terrorism,”17 this time in an assault, by two Muslim terrorists, upon the well-known Akshardham Hindu temple complex in the capital city of Gandhinagar, a mere 100 yards away from the home of Chief Minister Modi, where 37 persons were killed and 81 injured before the two gunmen were shot dead by commandos dispatched by the central government. On this occasion, contrary to their previous response to the Godhra killings and the pogrom that followed, the two most prominent BJP leaders, Prime Minister Vajpayee and Home Minister L. K. Advani, rushed immediately to the scene, where, characteristically, they blamed Pakistan for the attack, although the head of the sect that controls the temple was reported to have urged Vajpayee to “show restraint” and refused “to blame anyone.”18

The final act in the Gujarat drama is its appropriation for electoral purposes. Hopeful that Hindu sentiment in Gujarat had been consolidated by the Godhra killings and that most Hindus had been pleased by the revenge and retaliation that was taken upon Muslims in the state, the BJP government dissolved the legislative assembly on July 19 and called for elections in October. However, its call was frustrated by the Chief Election Commissioner, who noted that such early elections could not be “free and fair” under the circumstances, namely, the failure of the state government to provide protection to Muslims before the riots and rehabilitation after them, the displacement of large numbers of Muslims from their homes and their inability to return to them, and the obvious likelihood, therefore, that the Muslim population in the state would feel intimidated and might be too fearful to go to the polls.19 Elections were finally scheduled for December 12. However, the postponement of the election did not prevent the predominant leadership of the BJP, under chief minister Narendra Modi, from gaining the advantage of a polarized election contest, in which many Hindus appear to have consolidated behind the BJP, in effect registering their approval of the pogrom against the Muslims. The stakes in this contest were seen by all parties involved to have important consequences, not only for Gujarat, but for the future of the BJP-dominated coalition government in New Delhi and its commitment to a militant Hindu nationalism that scapegoats Muslims and thrives on anti-Pakistan rhetoric. A defeat for the BJP in Gujarat would have threatened both; its victory in Gujarat has encouraged militant Hindu hardliners in their political strategies that include violence against Muslims and Christians designed to promote Hindu political consolidation.20

What the BJP’s election campaign would be like was demonstrated in mid-July when the annual Hindu Jagannath rath yatra (journey by chariot) festival was allowed to proceed through the streets of the old city section of Ahmedabad, including its Muslim quarters, along a 15-kilometer route, in which the procession itself extended for three kilometers. These were followed by two other Hindu religious festivals in other parts of Gujarat, on Janmasthami (Krishna’s birthday), August 31, and the Ganesh festival on September 20 when, in Vadodara city, violence occurred in which four persons were killed as the procession passed through Muslim sections.21 Such rath yatras, though part of traditional Hindu religious processionals, here and elsewhere in India, have increasingly been dominated by militant Hindu activists, who use them to intimidate and provoke Muslims in the localities through which they proceed. On October 26, Chief Minister Narendra Modi launched an explicitly political yatra, dubbed a “Gujarat gaurav yatra” (Gujarat pride journey), with each stage of the journey beginning from a Hindu temple. When it reached Godhra, on November 10, the chief minister was reported to have “blamed the Muslims for the violence in Gujrat [sic].”22 The chief minister has also been reported, in various news media, on several occasions, to have made extremely provocative speeches, insulting Muslims in general, and mocking the plight of those Muslims displaced during the pogrom.

It appears both from reports of the campaigning as well as pre-poll interviews and the election results themselves23 that the Gujarat killings were used effectively to consolidate Hindu sentiment and voting behind the BJP, a party that was in decline in the state before the pogrom. Press reports have indicated that, despite Election Commission restrictions on direct exploitation of the Godhra and Akshardham killings, the BJP made use of slogans and videos designed to inspire fear and hatred of Muslims among Hindu voters.24

The election of December 12 resulted in an overwhelming victory for the BJP. It won 126 seats in the 182-seat Gujarat state assembly. Similar consolidation behind the Congress party was reported among Muslim voters, but it is the Hindu consolidation that matters most in a state where only nine percent of the population is Muslim. Most significant is the reported difference in the BJP’s victory in areas in which violence occurred (52 of 65 seats).25

In riot-prone India, whenever new great riots or waves of riots occur, leftist and secular writers, academics among them, commonly say that the latest wave of riots is the worst since the great Partition massacres of 1946-47. In some of the respects noted above, but especially if one takes account of all the features of this fierce outburst, it is fair enough to say as much about Gujarat in the year 2002. Others, however, proclaim a different view, taking comfort from the fact that riots did not spread from Gujarat to other parts of India, as they did in the last great wave of 1992.

But both types of statements, especially the latter, are distractions that divert our gaze from the dynamics of riot production in present-day India. The first type is useful mainly for exposing to full view the dimensions of what actually happened, and noting that yet further social and political boundaries have been transgressed. For Indians, the first image conjures up the retributive genocidal massacres of Partition in the Punjab in 1946-47, seeming to herald yet another monumental catastrophe, which will include the further weakening or disintegration of India or the obliteration of its Muslim population. If the first view maximizes the implications of such events as Gujarat in 2002, the second minimizes them. Both views have the same focus, namely, the future of India, that is, its territorial integrity, societal peace, democratic functioning, pluralism,26 and its status in a world of nation-states. But what is more important for India’s present and future in all these respects is to escape from the self-perpetuating traps of blame displacement and the complementary traps of maximizing and minimizing the significance of horrific violence. In short, it is necessary to fix responsibility and penetrate the clouds of deception, rhetoric, mystification, obscurity, and indeterminacy to uncover what can be uncovered, knowing full well that the whole truth can never be known, but that the evident actions and inaction of the perpetrators and apologists of violence, of known persons, groups, organizations, political leaders, media, academics seeking causes, and patriots seeking comfort can be identified, so that appropriate action can be taken against the perpetrators and the apologists can be discomforted.


1 The theoretical framework in this essay is elaborated more fully in Paul R. Brass, Theft of an Idol (Princeton, N.J.: Princeton University Press, 1997) and The Production of Hindu-Muslim Violence in Contemporary India (Seattle: University of Washington Press, forthcoming 2003).

2 Upendra Baxi, “The Second Gujarat Catastrophe,” Economic and Political Weekly [; hereafter EPW], August 24, 2002. See also my own remarks on the uses and misuses of the term “governance,” in Paul R. Brass, “India, Myron Weiner, and the Political Science of Development,” EPW, July 20, 2002, pp. 3031-35.

3 In order to reduce the number of footnotes in this short essay, only direct quotes from sources are cited herein. All other statements of fact concerning specific events and persons mentioned come primarily from the following sources: Frontline, issues from April 13 through November 9, 2002; various Indian newspaper reports from for the period February 28 through September 5; and the Report of a delegation of the Communist Party of India (Marxist) [CPI(M)] and the All India Democratic Women’s Association (AIDWA), which visited Gujarat March 10-13, undated, but received via e-mail forwarded to me by Clea Finkle from Subhashini Ali, a member of the team.

4 The latest confirmation of this assertion arrived after this essay was first written. A Concerned Citizens’ Tribunal, headed by a highly regarded former Supreme Court justice, Krishna Iyer, whose membership included two retired Bombay High Court judges, other former government officials, and academics, issued a 600-page report in which it was stated that the post-Godhra killings constituted “an organised crime perpetrated by the Chief Minister, Narendra Modi, and his Government.” The report was also said to have noted that the “post-Godhra violence was pre-planned and executed with ‘military precision’ by the Sangh Parivar with the State’s complicity.” It also affirmed, according to the news report, that “several members of the Modi Cabinet participated” in the violence “or instigated the mobs.” The report “also indicted the BJP-led Government at the Centre for ‘failing in its duties’”; Hindu, November 22, 2002 (online version).

5 Riaz Ahmad has found my term applicable to the Gujarat killings, in “Gujarat Violence: Meaning and Implications,” EPW, May 18, 2002.

6 Asghar Ali Engineer, “Role of Police in Gujrat [sic] Carnage,” Secular Perspective, June 16-30, 2002 and J.B. D’Souza, “Gujarat: A Civil Service Failure: How Can Credibility Be Restored?” in EPW, August 24, 2002.

7 T.K. Rajalakshmi, “Testimonies of Terror,” Frontline 19: 8 (April 13-26 2002).

8 Times of India, March 27, 2002.

9 Praveen Swami, “An Unquiet Peace,” Frontline, 19: 11 (May 25-June 7, 2002).

10 Praveen Swami, “Saffron Terror,” Frontline, 19: 6 (March 16-29, 2002). The number comes from Asghar Ali Engineer, “Can We Fight Terrorism?” Secular Perspective, October 1-15, 2002. In this and many other respects, the numbers exceed Krystallnacht—the hallmark pogrom of the twentieth century—when 267 synagogues were destroyed or damaged; see Leonidas E. Hill, “The Pogrom of November, 9-10, 1938 in Germany,” in Paul R. Brass (ed.), Riots and Pogroms (New York: New York University Press, 1996), p. 104.

11 Brass, Theft of an Idol, ch. vii and The Production of Hindu-Muslim Violence, ch. vi.

12 T. K. Rajalakshmi, “Testimonies of Terror,” Frontline, 19: 8 (April 13-26, 2002).

13 Dionne Bunsha, “Peddling Hate,” Frontline, 19: 15 (July 20 – August 02, 2002).

14 A term used by Anders Bjørn Hansen to describe the opening phases of the Punjab massacres in March, 1947, in Partition and Genocide: Manifestation of Violence in Punjab, 1937-1947 (New Delhi: India Research Press, 2002), p. 115.

15 Sukumar Muralidharan, “Appeasing the Hindu Right,” Frontline, 19: 6 (March 16-29, 2002).

16, April 4, 2002.

17 A term used by B. Raman, “A Case of Intelligence Failure?” Frontline, 19: 21 (October 12-25, 2002).

18 Dionne Bunsha, “The Modi Road Show,” Frontline, 19: 21. The RSS family of organizations, however, exercised physical restraint on this occasion, choosing not to respond with further violence against Muslims, which, at this stage, would have been counterproductive, posing the twin dangers of further postponement of the desired election and the imposition of President’s Rule.

19 Dionne Bunsha, “Boast and Bluff,” Frontline, 19: 18 (August 31 – September 13, 2002).

20 See, for example, the post-election comments of BJP leaders reported in the Hindu online edition, December 16, 2002.

21 Dionne Bunsha,”Narendra Modi’s Long Haul,” Frontline, 19:19 (September 14-27, 2002) and “Journeyman Modi,” Frontline, 19:20 (September 28-October 11, 2002).

22 Asghar Ali Engineer, “The BJP and Gujrat [sic] Elections,” Secular Perspective, November 16-30, 2002.

23 Yogendra Yadav and P. M. Patel, “Advantage BJP,” Frontline, 19: 25 (December 7-20, 2002) and Yogendra Yadav, “The Patterns and Lessons,” Frontline 19:26 (December 21, 2002-January 3, 2003).

24 For example, Dionne Bunsha reports the use of slogans arousing fear of further attacks from “Muslim terrorists,” such as the following: “You are travelling. You can be attacked.” “You are praying. You could be attacked.” “Saffron Theatrics,” Frontline, 19: 25.

25 Manas Gupta, “Landslide Win for BJP in Gujarat,” Hindu online edition, December 16, 2002. Also see Yadav and Patel, “Advantage BJP” and Yadav, “Patterns and Lessons.”

26 See, for example, Ramaswami R. Iyer, “Death of Indian Pluralism,” EPW, July 20, 2002.