Hindu-Muslim Relations in Colonial Banaras:
From the Lat-Bhairo Riots of 1809 to the ‘Gandhian’ Civil Disobedience of 1811

[Dedicated to all the Hindus, Muslims and others who have undergone,
willingly or unwillingly,
the salvific ‘punishment of Bhairava’ in Banaras]
Hop aboard the Mecca-Benares Express
Powered by groups.yahoo.com

by

Sunthar Visuvalingam
(with the collaboration of Elizabeth-Chalier Visuvalingam)

Paper submitted for publication to the

Journal of the American Academy of Religions

and to be appear in reworked and condensed form in a monograph by

Sunthar Visuvalingam (with Elizabeth-Chalier Visuvalingam)

Between Mecca and Banaras:
The Marriage of Lat Bhairava and Ghazi Miyan

Transgressive Sacrality and the Processual Approach to Religious Traditions

The Marriage and Cult of Lat Bhairava: What it means to be a Banarasi even Today

The Marriage (urs) of Ghazi Miyan: Muharram and the Sacrificial Pole (qutb) of (Indian) Islam

Divide, Rule and Unify: Religious Dualism and the Dialectics of Human Violence

Lat Bhairava, the scapegoat of the Lord of the Universe: From the Hindu-Muslim Riots of 1809 to the ‘Gandhian’ Civil Disobedience of 1811

Between Banaras and Mecca: Hierarchy, Egalitarianism and Autonomy

The Felling of the World Pillar: An Islamic Fulfillment of Vedic Cosmogony?

Hindu-Muslim Relations in Colonial Banaras:
From the Lat-Bhairo Riots of 1809 to the ‘Gandhian’ Civil Disobedience of 1811

[Dedicated to all the Hindus, Muslims and others who have undergone, willingly or unwillingly, the salvific ‘punishment of Bhairava’ in Banaras]

The 1809 Hindu-Muslim riot of Banaras—especially the manner in which it was handled by the British administration—has recently attracted the attention of political historians of colonial India. Its analysis constitutes the very first chapters, and sets the tone, of two recent works devoted to the subsequent development of communalism in India. So much scholarly discourse and intellectual energy is devoted to apportioning blame—René Girard’s scapegoat seems to have run amuck even in supposedly ‘objective’ research—that no real attempt has been made, at least to my knowledge, to address the ‘naive’ perplexity which any curious modern-day onlooker would have shared with Robinson. Armed with (often genuine) incomprehension, contemporary British reports exaggerated the magnitude of the Hindu-Muslim riots, while stressing the role of the colonial administration in quelling the riot and reconciling the aggrieved parties. This implicit politico-cultural agenda of justifying the Pax Britannia has inhibited the possibility of relating the, by all accounts, unprecedented character of the religious riot to the quasi-unanimous but peacefully organized resistance of 1811 to the colonial administration. On the other hand, notwithstanding the explicit declarations of both parties to the contrary, Sandria Freitag concludes from this concerted action by Hindus and Muslims that the preceding riot must have necessarily had purely ‘secular’ politico-economic causes. Gyan Pandey, for his part, speaks more cautiously in terms of ‘multiple and shifting identities’ but does not offer us any substantial insight into the kind of logic—whether processual, contextual, or irrational—that regulates and conditions the coexistence of such conflicting identities. Whereas his ‘subaltern’ commitment insists that it was the lowest castes who were at the forefront of the initiative to have the House-Tax repealed, Freitag’s ‘processual’ approach underlines its hierarchical ‘Hindu’ ethos, and the role of the Maharaja as the indispensable mediator between the British administration and its Indian subjects. Neither scholar has given sufficient consideration to the fact that this massive Indian affirmation of ‘civil rights’—a 100 years before Gandhi and without the unifying role of any single charismatic figure—was undertaken in defense of the traditional privileges of brahmins and fakirs. Sandwiched so neatly in between, the military-police feud—whose unpalatable role in mediating the transition from religious violence to peaceful joint-protest was at least noted by the perceptive British reporter—seems particularly embarrassing for apologists of all stripes. It nevertheless makes sense in the light of ritualized conflict in traditional societies, and even suggests a novel way of understanding the politics of divide-and-rule. Robinson’s report, which gives voice to the mutual recriminations recorded in the post-riot Hindu and Muslim memorials, is not even mentioned by Freitag and Pandey, who rather insist in speaking for—and effectively silencing—the ‘oppressed’ Indian subjects. It nevertheless offers much food for thought to those, like myself, intent on understanding the relationship between opposing religious paradigms, irrational violence and shared human values. I have cited it extensively—with the necessary corrections and comments—in the hope that the immediacy and vividness of his account will induce more Indians to reflect on these issues.

Underlying all these conceptual difficulties—which thus easily lend themselves to our own partisan politics—is an inadequate appreciation of religious differences, and of religion itself as a powerful force in shaping human perceptions and, thereby, the outer course of events. From Hindu participation in Muharram and Muslim participation in the Ramlila, Freitag, for example, concludes that in 1809 ‘Hindu’ and ‘Muslim’ as categories did not as yet exist! The scrupulous historian overlooks the obvious ‘detail’ that there were also sacred sites exclusive to each religion, that it was the Vishwanath temple and the Aurangzeb mosque that were especially targeted by the rioters. Much is made of the joint celebration of the ‘marriage of Lat Bhairava’ but no plausible explanation is offered as to why it was the Muslim weavers who desecrated and felled a pillar that they continued to claim as their own: sacrilegious acts which triggered off and provided a focus for this otherwise unprecedented riot. As a ‘historian of religions’—I do not bring to light any fresh historical documents to bear upon these crucial events—my intent is primarily hermeneutic and explanatory: can we provide an adequate conceptual framework—even a set of interlocking, if mutually irreducible, paradigms—for deciphering the course of events and the underlying motivations of the principal actors? In a book under preparation—entitled Between Mecca and Banaras: the Marriage of Lat-Bhairava and Ghazi Miyan—I use my paradigm of ‘transgressive sacrality’ (Visuvalingam 1985,1989) to derive, from the syncretic practices that provide the immediate context for the riot, a general model of Hindu-Muslim relations. This endeavor feeds into a ‘processual’ but non-reductive approach to religious traditions in general, and to their differing modes of interaction with the discourse of modernity. In reworking two of its sections here, I have restricted my efforts to showing how this model can not only account for the peculiar logic underlying these collective actions of the Banarasis, but also provide the criteria necessary for distinguishing (pre-colonial) religious conflict, as such, from (the subsequent development of) communalism proper (Bayly 1985). Above all, this essay offers fresh conceptual tools for tracking the manner in which religious values shape and constrain even ‘secular’ motivations, thereby determining the divergent ways in which a composite society responds to a common set of politico-economic pressures.

The origin myth of Bhairava, the terrifying aspect of Shiva, attests to the intimate and indissoluble link between Banaras and Bhairava (Chalier-Visuvalingam 1986,1989). After having emerged from the pillar of fiery light to violently cut off the head of Brahma, the ‘skull-bearing’ (Kapalika) Bhairava had to wander about for twelve years in order to expiate his brahmanicide. Finally he reached Banaras where the skull of Brahma, and with it the sin of Brahmanicide, fell into a tank appropriately named the ‘liberation of the skull’ (kapalamocana). Yet even after his absolution, the ‘Black’ (Kala) Bhairava remained at Kapalamocana as the ‘sin-eater’ to devour the impurities of future pilgrims to the city of final liberation. Paradoxically, Bhairava, the (ex-) criminal, also reigns as the policing magistrate (Kotwal) in Banaras, entrusted with the duty of preserving its sanctity not only by barring its access to sinners but also by punishing those who indulge in sins even within the confines of the holy city. The ‘punishment of Bhairava’ burns up the accumulated sins of seekers of liberation and is inflicted on everyone at the moment of death in this ‘great cremation-ground’ of the Hindu universe. This punishment was administered at a pillar (lat) whose stump, now called ‘Lat Bhairo,’ still stands beside the present Kapalamochana tank where it is worshipped as the phallic representation (linga) of Bhairava. The pillar was the focus of fanatical violence between Hindus and Muslims during the famous ‘Lat Bhairo riots’ of 1809, which eventually evolved by 1811 into a common political agitation directed against the excesses of colonial rule, without however resolving the question of the Lat. In the larger context of the recent revival of Hindu-Muslim claims and clashes over other holy sites in India, Lat Bhairo provides some insights into the manner in which human violence is channeled into and regulated by symbolic archetypes. Like that other holy city par excellence, Jerusalem, which is claimed alike by all the children of Abraham, Banaras lives as that inner space where the forces of modernity continue to re-enact—willingly or unwillingly—the still unresolved conflicts both between and within the religious traditions of man.

The Lat Bhairava pillar was almost completely leveled during the Banaras riots of 1809. Today this Hindu icon is a mere stump, 3 feet thick and 7 to 8 feet high, that stands stubbornly but precariously on a slightly elevated stone platform in the midst of a Muslim idgah, where the devout of both faiths continue to pray and offer their respective sacrifices. Entirely encased under protective copper sheeting installed after the riots by the District Magistrate, it is separated from the idgah only by a small enclosing brick wall which can nevertheless be overlooked by Indians of higher stature. Just outside the wall and to the north is the adjoining ‘well of Bharata’ (the youngest brother of Lord Rama), which plays a role in the local Ramlila. Towards the south of this terrace, and 5 or 6 meters below, is a large tank named Kapalamochana, a strong well-built structure with stairs and foundation of solid stone. The shrine is located on the north-eastern part of the present city on the junction of the Grand Trunk Road with the road leading to Sarnath, and is almost a mile west of the confluence of two sacred rivers, the Varana and of course the Ganga. The open praying area of the idgah is bounded on the west by a wall with the niche (mihrab) indicating the orientation (qiblah) of Mecca, so much so that some of the kneeling Muslims in the back rows could easily end up having the Lat between them and the object of their adoration. To the stump of the original Lat, which was once famous among the Hindu population both for its antiquity and for its sanctity, is normally affixed a small mask of Bhairava (see photo 1). The divinized magistrate of Banaras appears to supervise the work of the cudgel-bearing policemen who are permanently posted in the vicinity to prevent the outbreak of fresh communal tension.

The evidence points to an ‘Ashokan’ pillar, no doubt the one that Hiuen Tsang in 636 A.D. saw standing before a Buddhist stupa. The Hindu Mahatmyas refer to a ‘pillar of the great cremation-ground’ (mahashmashana-stambha) standing at the present location of the Lat. The Kashikhanda (97.64–6), which reflects the post-Islamic adaptations of the mid-14th century, speaks of the Great (Maha-) Rudra residing with his consort Uma (not in an adjoining temple but) in the pillar itself, near the ‘Lord of the Skull’ (Kapalesha), and refers to the adjacent Kapalamochana. The Kapalikas generally haunted the cremation-grounds; these radical tantric ascetics were adepts of human sacrifice. Stories still circulating among the Muslims of the surrounding area tell of the martyred warrior, Ghazi Miyan, having eradicated the regular human sacrifice at a temple of Somnath that would have existed near the confluence of the Varana and the Ganga. This may refer to the Maharudra temple, probably a Kapalika cult center, which Kubernath Sukul claims once existed around the pillar. It must have been devastated along with the rest of the city by Qutb-ud-din Aibak, chief general of Muhammad Ghori, at the end of the 12th century. The French lapidary Tavernier saw the pillar in 1665, during the reign of Aurangzeb, within the walled garden—with many sculptures and beautiful architecture—of a mosque. Its shaft, which was 32 to 35 feet high and all of one piece, terminated in a pyramid with a large sphere. Since Bhairava functioned as Sin-Eater at both the Mahashmashana-Stambha where as Kotwal he executed the ultimate punishment, and also at Kapalamochana where as Kapalin he was freed of the ultimate crime of Brahmanicide, it is perfectly logical that, in the wake of the Muslim occupation of the heart of Hindu Kashi (or Banaras), Kapalamochana had come to be (re-) identified with Lat-Bhairo, where the Kapalin remains as executioner, victim and pillar of the world.

The (Maharudra temple around the) Mahashmashana Stambha must have been the haunt of such transgressive Kapalika and Pashupata ascetics in the pre-Muslim period. Through a general evolution well attested elsewhere in North India, even in the major Bhairava temples of Banaras and Ujjain, the post-Muslim Lat seems to have been in the religious custody first of the Naths (Jogees), then the Gosains and finally the Brahmins. According to the available genealogy, the rituals were earlier performed by householders of the Gosain (Gosavi) caste. The migrations of Gosavis, who functioned as merchants and bankers in the 18th and 19th century, were connected with the sale of cattle and are intimately related to the spread of Kala Bhairava. These mendicant soldier-traders, who were the largest owners of urban property in Banaras during the late 18th century and constituted (along with other mendicants?) one fourth of the city’s inhabitants in the early 19th century, were a bulwark to the politico-religious authority of the Bhumihar Maharajas of Banaras. True to the ethos of the Bhairava cult, however, this ‘military caste’ was organized around the guru-disciple relationship and recruited its members according to ability and without any regard to caste-distinctions. Brokers between diverse social groups, they maintained the pilgrimage routes and rhythms, on the one hand, and conserved their ties with the orders of regular sadhus like the Dasnamis, on the other (Cohn, 1984; Freitag, 1989b:6,24–25). None of the Gosains once connected with this temple are now alive; they must have surrendered the care of this temple about thirty years ago due to the lack of heirs. A Committee was formed in 1958 and its direction has been always in the hands of a Brahmin. This sociological development corresponds to the progressive purification of the Hindu mode of worship from human sacrifice to bloodless vegetarian offerings, leaving it to the Muslims to perform the intermediate goat-sacrifices during their Id celebrations which the ‘brahmanized’ Hindus now find rather distasteful. The Muslim post-riot memorial observed that:

An outbreak of communal violence is especially feared during the annual ‘marriage of Lat Bhairo’ with the adjoining well on the full-moon day of Bhadrapada (August-September). Kala Bhairava is brought as a large metal mask (mukut = ‘crown’) from within the city to crown the pillar for a day. A systematic comparison with the raising and felling of a pole during Newar New Year festivals—the Indra festival of Kathmandu, celebrated the very same day, and the Bisket festival of Bhaktapur consecrated to Bhairava, who is said to have come from Banaras—reveals the vestige of a pre-Islamic royal cosmogony. Bhairava represents the Hindu king who offers himself at the (transposition of the Vedic sacrificial) stake in what is simultaneously conceived to be a sexual union. Hence the choice of this date—so inauspicious for any Hindu wedding—which marks the beginning of the fortnight reserved for the performance of funerary rituals for the manes. The Muslim memorial, which was "signed by 724 persons, 105 of whom were accounted individuals of note" (Robinson, 1877:119), records that "for some years the lower classes of Hindoos and Mussulmans have annually celebrated the marriage of the Laut, and have divided the offerings between them" (Robinson, 1877:113–4). The latter practice was still reluctantly admitted by the legal custodians of the idgah when we interviewed then in 1979 with John Irwin.

This was facilitated by their own celebration—with Hindu participation—of the tragic marriage of Ghazi Miyan on the first Sunday of the month of Jyeshtha (June). The Muslim saint was likewise represented by a wooden pole bearing his decapitated head; a stone pillar stands at the head of (the replica of) his tomb in the predominantly Muslim ward neighboring the Lat. The cult of this martyr, so popular among both Hindus and Muslims all over north India, derives from the historical figure of Salar Masud, the nephew of Mahmud of Ghazni. Killed by the Hindus while he was attempting to eradicate their sun-cult, the proselytizing Muslim paradoxically ended up incorporating it within an Islamic mold focused around his tomb at Bahraich. Among his foremost devotees are hence the untouchable Doms, whose services are indispensable for accomplishing the funerary rites of all Hindus in Banaras. Hybrid legends, still extent among the Muslim weavers of Banaras, suggest an intimate fusion between the sacrificial symbolism of the marriages of the Lat and of Ghazi Miyan. It is hardly surprising, then, that the same Muslim memorial even claimed that this pillar of the world was in fact "the structure of Feroze Shah, like the pillar (Laut) at Allahabad, Delhi and other places, and which the Hindoos state to have been erected by their own forefathers. But, be that as it may, it was not an object of their worship entitled to any great veneration like the temples of Bisseysur [Vishweshwar] and Bhyronauth [Bhairavanath]; for no account of this pillar is to be found in any of their orthodox books. The style of worship of the Hindus is this, wherever they find set up (a pillar) they call it, at the incitement of their priests, a place of their worship, and after sometime has elapsed they consider it as a place of worship of the highest sanctity" (Robinson 1877:113). It was during the reign of Feroz Shah Tughluq that the famous Arahi-Kangra mosque, the Chaukhamba and Gola ghat mosques, many others in Alavipur, and almost the entire building scheme around the Bakaria Kund were constructed, generally on the site of, and with the materials obtained from, demolished Hindu temples. The Tughluq dynasty patronized the by now already famous cult of Ghazi Miyan: Feroz Shah made the pilgrimage to Bahraich and even had a marble fort built around the tomb. This Delhi Sultan’s obsession with pre-Islamic pillars had led him to transplant several ‘Ashokan’ pillars—over great distances—to the compounds of his royal mosques with a spatial orientation that corresponds to what we see today at Lat Bhairava. This appropriation is not an innovation of Indian Islam, for it derives from preexisting Sufi representations of the ‘perfect man’ as the ‘pole’ (qutb) and Shia representations of the Imam as a pillar of light. Even the ‘marriage of Ghazi Miyan’ is, at the same time and without contradiction, an Indian transposition of the Iranian cycle of Muharram, where the mourning over the death of the martyred Husain (and Hasan) is at the same time celebrated as a (mystical) marriage (of their children). These elaborations may be ultimately traced to the assimilation of the Kaaba stone itself to a pillar—and, beyond that, to Jacob’s ‘pillar’ which is already charged with sacrificial connotations within the Jewish tradition (Visuvalingam and Chalier-Visuvalingam, 1993).

While heading for Bahraich in 1034–35, Salar Masud had dispatched a portion of his army and its retinue under Malik Afzal Alavi towards Varanasi (Sukul 1974:152–5, 1977:24–26). The invading contingent was thoroughly defeated on the northern outskirts beyond the boundary wall of the city at the site where the ‘Mosque of the Martyrs’ (Masjid Ganj-i-Shahidan) now stands near the Kashi Railway station. The Muslim civilians, with their women and children, were permitted to settle down in that area as townsmen and over the following century peacefully served the Hindu kings even as soldiers. After Qutb-ud-din Aibak had devastated the city in 1194, destroying nearly one thousand temples, the Muslim locality was renamed ‘Salarpur’ or ‘Alavipur / Alaipur’ (which today includes the two Muslim wards of Adampura and Jaitpura). Whereas their caste-fellows living in Madanpura resort to the Gyanvapi mosque, the illiterate weaver (Julaha) community of Alaipur generally congregates at the Lat idgah. Rather than immigrant Muslim weavers who were seduced by Hindu idolatry, they would rather be Hindu (or perhaps Hinduized Buddhist) castes that continued to worship the pillar even after their conversion en masse to Islam by the hard core of Ghazi Miyan’s original followers (cf. Kumar 1988:50). By leaving the aniconic Ashokan pillar standing intact before the idgah when it tore down the surrounding pantheon of Hindu idols, Islamic iconoclasm reinscribed the continuing Hindu worship within a Mecca-centered framework. Its spectrum of signification could have thus ranged from that of a divinity proper for the devotees of Bhairava to that of a mere victory monument (like the Qutb Minar) for the uncompromising legalists. The Hindu-Muslim cult of Ghazi Miyan merely confirms that Islamic proselytizing has succeeded through a judicious blend of violent imposition of symbolic (architectural) structures and syncretizing accommodation that operates on the common ground occupied by both religions.

The raising and felling of Bhairava’s linga-pole during the Bisket festival is accompanied by a ritual battle between the upper and lower halves of the city of Bhaktapur in Nepal. Cheered on by the riotous population, the hair-raising tug-of-war in order to drag the chariot of Bhairava from the center into their respective halves of the city becomes violent at the least pretext. A similar north-south conflict during the festival of Siti Nakha in Kathmandu involved deaths on both sides, and the regular sacrifice of captured prisoners to the goddess Kali. It was the model for similar battles in villages elsewhere in the Newar kingdom, which must have corresponded to an earlier dualistic tribal organization. The founding legend makes no bones about the Malla king Gunakamadeva, the reputed founder and culture-hero of Kathmandu, having instituted this custom at the behest of Skanda, the god of war, in order to destroy his enemies and to prevent his subjects from revolting. Similarly, the regular clashes between rival Hindu sects like the Shaiva Nagas and the Vaishnava Bairagis over the least pretext—such as precedence in taking their sacred bath in the Ganga—during festivals like the Kumbha Mela, reflect an underlying ritual paradigm that valorizes death as liberation. The binary pattern may well derive from archaic (pre-) Aryan institutions, but it conforms all the same to the immanent logic of human violence. Whether contained within integrative (royal) festivals, diverted to further partisan political agendas, distorted by systematic economic exploitation, exacerbated by modern racism, generalized in our age of enlightenment into intense individual competition for clambering up the ladder of ‘success’, or perverted into fissiparous outbursts of crime pure and simple, the dualistic mechanism has stubbornly persisted into our own times. It only proves that the ideology of pacifism urgently needs to be supplemented by adequate techniques for confronting, neutralizing and transmuting the innate violence that nourishes even the most refined disguises assumed by the acquisitive urges of the divided soul. The centrality of Bhairava’s symbolic role as the royal scapegoat suggests that successive rulers had (merely) exploited, and generally within certain self-imposed limits, a pre-existing socio-ritual mechanism meant rather to regulate, and provide a convenient safety-valve for, the constant and pervasive menace of self-consuming violence otherwise capable of undoing the entire community.

Even the Sunni-Shia divide, notwithstanding the crucial doctrinal differences involved, has a markedly ritualized character that feeds on and further inflames communal grievances stoked by other social factors. This was a regular scenario in pre-Safavid Iran, and it used to leave major cities devastated in its wake. After the Persian conversion to Shiism, their rhythm continued unabated, but now between opposing Shia sects like the Haydari and the Ni‘imati. Where there are no Shias in the locality with whom to reenact the battle at Karbala, the Sunnis in India celebrated Muharram by fighting among themselves. In the predominantly Hindu rural town of Bishnupur in West Bengal, for example, the entirely Sunni Muslim minority is divided into thirteen neighborhoods that jointly celebrate Muharram as an occasion of both gaiety and mourning. This reaffirmation of religious unity was nevertheless characterized by inter-locality competition for the Hindu Maharaja’s prize for the best tazia: the symbolic reenactment of Karbala with real swords and sticks resulted in injury and bleeding among the exclusively Sunni actors. Under normal conditions, in Banaras, the celebration of the festivals of Muharram and Barawafat (the anniversary of the Prophet), which have been consistently growing over the recent decades, is likewise characterized by intense but sportive competition between the neighboring Muslim wards (muhallas) themselves (Kumar 1989:158–163). If conflict erupted under any pretext between Muslims and Hindus, however, the dualistic structure would easily resurface as an expression of a fundamental opposition between the two religious communities. Due to the coincidence of Dashahara and Muharram occurring on the same day in 1821, for example, many were killed at Cuddapah in the Deccan when neither procession was willing to give way (Shurreef 1863:122). Such conflict has little to do with colonial intervention for, already in 1695 when the Italian traveler, Careri, was in the Deccan, he observed that whenever Sufi fakirs met the shaven Hindu Bairagis, they would fight desperately (Eaton 1978:238). It would be difficult to determine the respective roles of ritualized violence and religious animosity in such confrontations. Many Hindus participated fully in the Muharram festival, consumed only meat that had been sacrificed according to Islamic rites, and even disguised themselves as Muslim ascetics. If any fighting and bloodshed took place, the Hindus who had temporarily become fakirs took the part of the Muslims and fought against their own co-religionists (Shurreef 1863:122). The tenth of Muharram served as the occasion for the violent settling of private quarrels even among Muslims (Ali, 1832, vol.1:94). The Newar festivals of Bhairava provided a similar license to Hindus, even those of the same caste: established custom ruled out the subsequent seeking of redress (Chalier-Visuvalingam and Visuvalingam 1992).

The British had replaced Awadh as the national level authority in 1775; the Resident’s power at Banaras steadily increased until they finally took direct control of the city in 1784. The Bhumihars had enriched themselves and come to power in the Banaras region precisely through British land reforms that had displaced the earlier supremacy of the Rajputs and Marathas. By 1809 the then Maharaja had suffered loss of power and face through his unsuccessful decade-long agitation to free himself from the control of the East India Company. The Muslim kotwal had already lost the confidence of the Banarasis in 1803, when he had acquiesced to highhanded British attempts to impose a (phatakbandi) tax for recruiting patrols of watchmen. Unlike the ostentatious participation of the Maharaja in the public arena, the British administration held aloof from community life and exercised its authority through local intermediaries, but intruding ever more profoundly into the relations, structure and functioning of the pre-existing communities. Though the social tensions and lines of fissure generated by these far-reaching politico-economic changes are difficult to determine precisely, the rationalizing mentality introduced by the imperial British state had no doubt begun to have an insidious effect on the Banarasi ‘civic’ culture (banarasipan) which had for so long united the high and the low, both Hindus and Muslims. Freitag (1989a:1–22, 203–28; 1989b:19–52) thus rightly emphasizes the role of shared public arena activities like Muharram and Ramlila in assuring the politico-cultural unity of Banaras at the turn of the 19th century. Purely socio-historical approaches tend, however, to overlook the conflictual (and even transgressive) dimension that constitutes the essence of such religious festivals. The kotwal Bhairava, the mythical embodiment of the rule of law in the sacred city of Hinduism (as also in Kathmandu, the capital of Nepal), is also the brahmanicide par excellence. In the sacrificial paradigm that underlies the Hindu king’s unifying role vis-a-vis the diverse and competing constituents of his polity, the royal Bhairava simultaneously occupies the symbolic role of the scapegoat-victim.

Beneath the fragmentation of authority between the Company, the Maharaja and the kotwal (supposedly serving the Mughal Emperor), we may also discern the conflicting models of an English, a Hindu and a Muslim dispensation of justice. "Both in the record of the late Mughal period and in the years of the Company Raj the ‘chief registrars’ (kazis) and ‘chief executive officers’ of the towns (kotwals) often took a leading role in religious violence¼ These officers were in an ambiguous position. Their existence—particularly that of kazi—was itself a technical criterion that the land was Dar-ul-Islam. They were thus the formal guarantors of the supremacy of Islam in the locality. At the same time, however, they were justiciars of a Mughal Empire which in theory guaranteed its Hindu and Jain subjects some rights as protected subjects, and in practice gave them virtually equal rights out of equity. The tension between these two roles was intolerable so that when conflicts of a religious nature broke out the position of kazi and kotwal was always compromised on one count or another" (Bayly 1985:196). Under Muslim rule, such "disputes had been easily settled because there was no question of representation of interests; Hindus had been protected but there was no doubt that they were in a ceremonially inferior position" (Bayly 1985:198). This does not imply that the Muslim officers did not take into consideration Hindu customs and sensibilities while adjudicating disputes, especially those involving caste rules and pollution. In all likelihood, the Hindu (and Muslim) castes still retained and exercised means of settling disputes among themselves through pancayats and other community based institutions. But already from the early eighteenth century, the authority of the law officers was being whittled away by powerful Hindu landowners, and post-Aurangzeb Moghul policy was tilting towards alliances with powerful Hindu commercial interests. In Banaras, "the rise of the Bhumihar dynasty after 1739 and the growing influx of Maratha wealth and power into the city had sharply tipped political power against the Muslim law officers" (Bayly 1985:197). "Wherever Muslim states conceded effective sovereignty to Hindus or Sikhs strenuous efforts were made to stop the slaughter of sacred cattle by Muslim butchers¼ It is interesting that the only recorded acts of Maharaja Ranjit Singh which could be considered anti-Muslim were the banning of cattle slaughter throughout his domains and an order to desist from public calling to prayers in the Sikh holy city of Amritsar. His sovereignty was considered incompatible with the public assertion that Amritsar was an Islamic city or with open cow-slaughter. In Benares similarly the new Bhumihar rulers banned the slaughter of cattle within the holiest parts of the city" (Bayly 1985:187, n.36). So when in 1752 the Nawab of Awadh temporarily expelled the Bhumihar ruler from the city, "it was the kazi who led sections of the Muslim population in the demolition of the Vishwanath temple, thus setting the scene for two hundred years of sporadic local conflict." The early days of British rule saw a sharp redefinition of the already existing pressures on the local law officers due to factionalism and decentralization of power within the later Mughal empire. "It was this deliberate diminution of the social and judicial role of the kazi, mufti, and kotwal which formed an important precondition for the riots in north Indian cities in the period 1820–50. The British sought to turn these officials—the guarantors of Dar-ul-Islam—into subordinate agents of the district collectors and magistrates, thus severing the link between Muslim law and the operation of local justice and politics" (Bayly 1985:197; 1983:336). Hindu-Muslim interaction had been primarily a socio-cultural process characterized by syncretism-cum-conflict that took within its stride alternating politico-religious domination: it permitted constant re-negotiation and readjustment. British justice—which claimed, and largely tried, to be neutral—treated both religions on an ‘equal’ footing, thereby winning the respect of both Hindus and Muslims, but at the same time tending to legislate their differences into frozen barriers, and to arrest the process of mutual fecundation. Even where the colonial state did not consciously or actively embark on a policy of divide-and-rule, riots were made possible precisely by the absence of clear Hindu or Muslim domination, and by the relative impartiality of British rule. Indeed, it could be argued that the logic of divide-and-rule was as much imposed on the imperial state by its Hindu and Muslim subjects, even before it was overtly or tacitly exploited by the former to its own advantage.

Moreover, the joint participation of Hindus and Muslims in each other’s cults and festivals should not obscure the intense ideological struggle—even where peaceful and mutually accommodating—between the rival religions on the symbolic level for the heart, mind and soul of India. The Hindus could not remain oblivious of the living visual testimonies to the systematic razing of their religious architecture (c. 1660s) by Aurangzeb who had sought unsuccessfully to impose an Islamic city called ‘Muhammadabad’ upon their socio-religious center (Freitag 1989a:3). Having now lost their political supremacy in India, the Muslims, on the other hand, were not willing to submit to Hindu acculturation, not to the extent of surrendering the divergent world-view encoded into their own ritual practices. Having served to inscribe its continuing Hindu worship in the direction of Mecca, the Lat could now just as well serve to gradually reintegrate the lower classes of Muslims within the symbolic universe of Hinduism. Well before the arrival of British power, the decline of Islamic states like Bijapur were attributed by the Muslim orthodoxy to laxity in religious observances and, as a general rule, have coincided with reformist (even anti-Sufi) measures directed against syncretism and ultimately against (the symbols and institutions of) the Hindus themselves. Though the weaver community in north India reverenced the flag of Ghazi Miyan to whom they ascribed the comparatively recent conversion of their ancestors, by the early 19th century they were already beginning to abandon such syncretic, ‘un-Islamic’ practices under the growing pressure of Wahhabi reformism emanating from the Arabian peninsula. For the down-trodden castes, the stricter observation of the Islamic law and personal code (sharia) provided the means of reasserting their social status in the face of politico-economic domination by the upper classes, both Hindu and Muslim (ashraf). A parallel process of purification was also occurring among the Hindu untouchables like the Chamars, who were giving up liquor, meat, (blood-) red vegetables, etc., and demanding the abolition of caste and an end to idol worship. Despite its undisputed age-old sanctity, the now ‘brahmanized’ Lat Bhairo or Mahashmashana Stambha was largely neglected by the Hindu scriptures, no doubt because of the stigma of death and impurity associated with it. The growing Hindu-Muslim division at the turn of the 19th century was further reinforced by the attempts of the colonial administration to systematically classify and publicly record everything, thus leaving the Muslim weavers little choice but to shed their Hindu names and customs in order to gain an equal standing within the fraternity of Islam (Pandey 1990:83–90). Partly a reaction to the derogatory connotations of their appellation as ‘Julaha’ by others, the weavers now call themselves ‘Ansari’ meaning ‘Helpers’ (of the Prophet at Medina), thus crowning the tendency of Indian Muslims to see themselves as immigrants with a separate ‘biological’ ethnicity rather than as native converts (Kumar 1988:49–57, 1989:153; cf. Roy 1983:19–57, 249–53). Though such developments may be understood as already internal to the growth of an ‘Islamic consciousness’ and not necessarily the product of politico-economic rivalry with non-Muslims, it is nonetheless true that they set the social preconditions for religious conflict, especially when they are reflected in a shifting attitude to shared sacred spaces and symbols. The increasingly Meccan orientation of an otherwise Indianized Islam found its symbolic charter in the mihrab at the mosque of Aurangzeb, which had already served to firmly inscribe the continuing worship of the Lat in the direction of the Ka’ba. The resulting religious tug-of-war between the opposing symbolic universes to take complete ideological possession of the pillar conforms structurally to the ritualized conflict to drag the chariot of Bhairava into the opposing halves of the city of Bhaktapur. The mock conflict between ‘Kols’ and ‘Bhils’ at the fair held on the day before the new moon of Ashwin, fourteen days after Lat Bhairava’s marriage, has long since disappeared. Perhaps it was rendered quite unnecessary by the even greater sacrifice of battle for the world-pillar celebrated jointly by Hindus and Muslims in October 1809 under the divide-and-rule British impartiality of the latest district magistrate.

Lat Bhairava, the scapegoat of the Lord of the Universe:
The Hindu-Muslim Riots of 1809

The ‘Lat Bhairo riots’ of 1809 have played a crucial role in colonial historiography not only because of their gravity and magnitude—comparable, we are told, only to the Kanpur outbreak of 1931—but also because they are among the first to be recorded in the colonial period (Pandey, 1990:29). Though the history of Hindu-Muslim riots goes back into the pre-colonial period (17th century Gujarat, for example), and may be legitimately understood as the continuing legacy of the Islamic conquest of north India, it is noteworthy that there had been no notable outbreak of communal violence between Hindus and Muslims during the previous 100 years in Banaras, which has always remained the Mecca of Hindu orthodoxy (Pandey, 1990:26, n.6). The Muslim memorial begins, however, by declaring that "the country of Hindoostan has for many centuries been the seat of the true faith" (Robinson 1877:112), and that, because of their acquiescence to Islam and loyalty to the Emperor, the Hindus had been permitted to continue practicing their religion in their acknowledged sacred city. It also observes that the habitual slaughter of cattle for beef had not been hindered or prohibited though the province had been governed by Hindu chiefs. The Hindu memorial prefaces its narration by recalling how the Hindus had from necessity submitted with patience to Aurangzeb's imposition of mosques on the ruins of temples, beginning with Vishweshwar, because of his invincible power. It then complains how a Muslim butcher had slaughtered a cow on 9th October while the Hindus were still propitiating the manes of their ancestors. Though the culprit had been apprehended by the law officials at the insistence of the Rajputs led by Rattan Singh, the Muslim Kotwal simply expressed displeasure with the protesters and had the offender released. The Hindus were therefore determined to make a formal petition when the court reopened, after the Bharat Milap, on 20th October. The absence of clear Hindu or Muslim political domination, coupled with the relative impartiality of the British administration thus facilitated the transformation of public arena activities into occasions for overt religious confrontation. The Muslim memorial goes on to observe that for 3 years the Muharram coincided with the Hindu festival of Dashahara and for 3 years with Holi; and that trouble had been averted each time by the British authorities restraining the Hindus from celebrating Dashahara till the Muharram was over and from dancing, etc., during Holi (Robinson, 1877:112-3). The licentious Hindu festival of Holi formerly involved much indiscriminate violence on the ghats and elsewhere and it is therefore not surprising that Sherring (1868:191-5) simply attributed the origin of the riots to an unfortunate coincidence of the Hindu and Muslim calendars, which brought a mob of Holi revelers into headlong collision with a mourning procession of Muharram. In the climate of accelerated social change and disequilibrium brought about by British rule, all that was needed for the resurgence of the dualistic pattern of violence along redrawn Hindu-Muslim lines was an appropriate symbol that could condense within itself the axial issue that separated Islam and Hinduism.

Robinson (1877) has given a very detailed report using first-hand British accounts and the memorials written by the Hindus and Muslims themselves shortly after the Lat Bhairo riots, so called because the pillar was the destructive focus of the three-day carnage between Hindus and Muslims. In 1809, the conflagration was sparked off by a trivial incident: in fulfillment of a vow upon his recovery from illness, a Hindu of the Nagar caste was attempting to replace the makeshift mud dwelling of Hanuman on the contested ground between the idgah and the pillar with a permanent stone enclosure. The Muslim memorial (Robinson, 1877:114) observes that earlier legal proceedings against a similar alteration on the shared sacred space were never determined, and hence taken as a victory by the Hindus. It claims that the Hindus had even placed an image of Rama and Lakshmana into the pulpit of the idgah for worship during the Dashahara celebrations of the previous year, and that they were again guilty of the same disrespect this year. At first the Muslim weavers were content to appeal to the law-officer (kazi), and agreed not to disrupt the Bharat Milap celebrations: they would refer the dispute to the court immediately after the Dashahara holidays. But when the festival was over on 20th October, they instead held a mammoth protest meeting during their Friday prayers, in the excitement of which they polluted the Lat and its surroundings by overturning Hanuman's pedestal, uprooting the adjacent tulsi tree, and beating the pillar itself with shoes. Freitag (1989a:211, cf. 1989b:39) indicates that the Muslim weavers had continued to participate in this(?) Bharat Milap as in previous ones, which would only confirm the initially localized character of the dispute. Their objection to the repeated installation of Rama on their pulpit nevertheless suggests not only the primarily ‘civic’ character of their participation in the Ramlila, but also the primarily religious character of the now evolving confrontation. The conjuncture of events around the Lat so faithfully reflects the overlapping disposition of Hindu-Muslim sacred space between the Vishwanath temple and the Gyanvapi mosque, that Robinson's account mistakenly displaces the Muslim demonstration to this mosque within the old Vishweshwar enclosure, and shifts imperceptibly to the defilement of the pillar which is in fact quite a long way from the religious center of the modern city. Though worshipped primarily by the lower castes of Hindus, the ‘pillar of the great cremation ground’ was indeed quite central to the symbolic significance of the sacred city, and this sacrilege had occurred precisely when the Hindus, particularly the Rajputs, "were seeking justice for the slaughter of a cow" (Hindu memorial, cited in Robinson, 1877:108).

By daybreak the whole Hindu community had heard of the sacrilege: a crowd of mostly Rajputs began to assemble at the Lat, so much so that the Kotwal had to arrive on the scene with a large contingent in order to disperse them. The Muslim memorial (Robinson, 1877:115) claims that during the night some Hindus had already broken open the door of the pulpit. The Hindu memorial charges that the Kotwal, seating himself at the foot of Lat Bhairava, invited Hindu representatives to appear before him to air their grievances. On complying, Rattan Singh and three associates were simply taken to the police-station where they were beaten without any inquiry. The remaining 200 or so Hindus, and the even greater number of Muslims present, began to pelt each other and the Imambarrah and some idols were damaged in the process. The whole city was becoming restless and, apprehending the worst, the acting British magistrate, Mr. Bird, had already deployed 2 companies of sepoys to protect the Muslim places of worship. Anticipating retaliation by the Hindus, the outnumbered Muslim weavers, who were at the forefront of all these manifestations, then decided to take preemptive action by sacking the temple of the king of the gods, Vishwanath himself. Had the attempt succeeded, it would have certainly resulted in the utter annihilation of the Muslim community in Banaras. For the local dispute between Hindu and Muslim low-castes, who had been as much united as separated by the pillar, was already becoming a full-scale confrontation between Hinduism and Islam. The Hindus led by the Rajputs, whose attempts to assemble at the Lat had been thwarted by the district magistrate and the army, fell back and regrouped to bar the route of the Muslims who were advancing in a Muharrram-like procession with raised standards and crying "Hasan, Husain!". Outnumbered and beaten back by the better armed Hindus at Gai Ghat, the 7 or 8 thousand weavers retreated leaving behind their dead. To revenge their defeat, the Muslims slaughtered a cow on one of the holiest ghats, and mingled its blood with the sacred waters of the Ganga: according to Heber (1828:429, 431), the sacred well itself (Gyan Vapi or the one at the Lat?) was subjected to the same sacrilege.

Having triggered off this irreversible quid pro quo and now exulting over their short-lived triumph, the weavers simply went into hiding in their quarters. Such was the horror of the sacrilege that the Hindus, even on receiving the news, would not visit the defiled spot and kept milling around the area of the Vishweshwar temple. The Rajputs, who had already been incensed by the issue of cow-slaughter, counterattacked only on the following day.

The Rajputs had even begun to demolish the tombs around the Dargah of Fatima, the mother of the Imam Husain, and would have proceeded to do the same to the tomb of Prince Jewan Bukht, held in the highest veneration by the Muslims, had not Mr. Bird checked them in the nick of time. On re-entering the city, Bird saw that

However due to the diplomacy and firmness of the district magistrate, the rioters were eventually broken up and the city was completely in the power of the large military force by the next day, but not before some fifty mosques had been destroyed. Robinson also puts the number of dead and wounded gathered in from the streets at a hundred, which does not account for the bodies quietly removed. The subsequent Muslim memorial, which places the number of mosques destroyed at more than sixty (Robinson, 1875:116), remonstrated:

The available details on the evolution of the riots rather suggest a cathartic eruption of self-consuming violence that exploited every possible fissure in the social fabric before falling back to more normal modes of self-regulation. It was perhaps symbolically inevitable that the Lat of the ‘sin-eating’ Bhairava, who had always been the scapegoat of Vishwanath, was defiled and dismembered by the Muslims in the name of their own Lord of the Universe. The irony of divine justice demanded that the Hindus should have proceeded to desecrate and destroy the royal Jama Masjid which had once housed their own Vishweshwara. The Lord of the Universe was, after all, ultimately identical with the scapegoat Bhairava on whom he displaced his own ritual impurity, a necessary part of the sacrificial process of death and rebirth.

The Hindu-Muslim conflict over the Lat would seem at first sight to reflect, at least in part, the social tension between the low-castes of Muslim weavers and butchers, who initiated the agitation, and the higher Hindu castes grouped around the ‘aristocratic’ Rajputs. The ‘10 sects of Gosains’ became involved only at the second stage, whereas the relatively ‘secular’ Rajputs—who had once been the real mainstay of the Moghul army and provided some of its best generals—opposed the weavers’ action from the start. The reconstruction of various Hindu temples, including Vishweshwara, begun by the Rajputs under Akbar's rule was prohibited under Shah Jahan, in whose reign 76 temples under construction were demolished. Then, under Aurangzeb, the Rajputs were reduced to helpless witnesses of the transformation of Bindu Madhava, Vishweshwara and the Lat Bhairava into imposing mosques. With the advent of British administration, the Rajputs, who had been the main landed group from the 16th to the 18th centuries had lost their traditional dominance in the region to the triumvirate constituted by the Bhumihar dynasty, the merchant-bankers and the Gosains (Cohn 1987:320-42). Regaining political control of the Hindu holy places of Mathura, Prayag, Kashi and Gaya from the Muslims was central to the Maratha strategic alliances with the Nawab of Awadh in the 18th century (Bayly 1985:187). Even after Benares passed under British suzerainty, the Peshwa proposal to rebuild Vishweshwara on its old site after paying compensation to the Muslims did not succeed, not only because the local brahmins considered the mosque site to be already polluted but also because the English flatly refused. The present Vishwanath temple was rebuilt by Rani Ahilyabai of Indore in 1777: modern Banaras is largely a creation of the Marathas, who had no doubt a special pride in the Hindu religious architecture of the sacred city, to whose rebuilding they had contributed so much after the accession of Maharaja Balwant Singh in 1738 A.D (Altekar 1947:17-20,23-24; Sukul 1974:158). Freitag suggests that "the Rajputs, dependent on rural landholdings for their wealth, did not need good relations with the Muslim lower classes of the city," whereas "the Gosains laid claim to the leadership role that had become theirs in the city during the previous century" (1989b:40). This explanation, based primarily on ‘urban political economy,’ does not explain why the Gosains did not take the lead in the first place and why they, more than the Rajputs, wreaked such indiscriminate vengeance among the Muslims as a whole. Clearly, the Rajput aristocracy and the Marathas not only had the military training and temperament to take on the ‘bigoted’ Julahas, but were also duty-bound by their codes of chivalry to defend the Hindu dharma. The District Magistrate, Mr. Bird, notes that

The weavers constituted the largest segment of the numerically small Muslim community of the region, and were moreover concentrated in the urban centers. According to the Banaras Gazetteer (1909), "though they are almost certainly of Hindu extraction the Julahas are the most bigoted and aggressive of all the Musalmans, and have always taken a prominent part in the religious quarrels that have from time to time arisen in Banaras" (cited in Pandey 1990:105). It is noteworthy that the Muslims aimed primarily at religious symbols instead of attacking any particular class among the Hindu population, like the landowners or merchant corporations, that may have been subjecting them to undue exploitation. The late 18th and early 19th century witnessed a sharp increase in the demand for their goods and skills, and the weavers of Banaras, in particular, were also less affected by the subsequent socio-economic upheavals faced by their caste-fellows elsewhere in the region (Pandey 1990:72,75). The period immediately preceding the riots may have seen the reformistic (Wahhabi) wave of Islamic self-consciousness coincide with a heightened self-confidence and assertiveness conferred by recent economic prosperity (cf. Pandey 1990:96-107). Their very readiness to resist the encroachment of Hindu idolatry upon their idgah—by defiling the Lat in whose cult they had otherwise been participating—suggests that in the first place their own pillar never—or no longer—had for them the same sanctity that it had for the Hindus. It is precisely within a syncretic context—rather than among the orthodox of either faith whose practice is generally confined within their separate social domains—that the growth of such a reformistic consciousness may be expected extract its first casualties over shared spaces and symbols . The Hindu memorial (Robinson 1875:108) seems to perceive this initial desecration of the Lat not so much as a spontaneous act by the weavers but as instigated by their leaders. Fuming over the suppression of their otherwise customary cow-slaughter and subjected to repeated encroachment, the weavers and butchers had replayed the 1752 march headed by the kazi upon the Vishwanath temple, which was no doubt perceived as the supreme consecration of a reassertive Hindu polity.

Though the conflagration was sparked off by a religious dispute which eventually engulfed the whole city and polarized the population along the Hindu-Muslim divide, it remains significant nevertheless that it first assumed the form of a conflict between upper castes of Hindus and lower castes of Muslims. Hinduism and Islam, after all, embody and consecrate wholly incompatible social ideologies, the one hierarchic and the other egalitarian. The difficulty of separating religion from politics (van der Veer, 1987) is matched by the impossibility of reducing one to the other. On the one hand, the unilateral action by the otherwise ‘Hinduized’ weavers derived from perceptions which they shared with the Muslim elite. On the other hand, the initial reaction of the Rajputs, who were not particularly involved with the regular cult at the Lat, derived from its symbolic significance for the Hindu community as a whole. The Hindu ‘memorial’ of grievances presented to the British authorities after the riots was in the name of "we, all the Brahmins, Cuttries [Kshatriyas ‘aristocrats’], and persons of Byse [Vaishya ‘merchant/peasant’] and Sooder [Shudra ‘laboring’] castes¼ We, every sect of the Hindu persuasion, have emigrated from all parts of the country to this place, for our religion tells us that Casheejee (Banaras) is a spot eminent beyond all others for its religious purity and a place of worship and adoration. It is here that according to the Beyds, Poorans and Shastras, the gods have always fixed their residence" (Robinson 1875:106). It was believed by the Hindus and Muslims alike that the Lat was, and still is, slowly sinking into the ground so that, when its top became level with the ground, not only would the Hindu caste-hierarchy collapse but "all nations would be of one caste. The throwing down, therefore, of this pillar was regarded as most ominous and dangerous to Hinduism." Rev. Buyers also recorded a conversation between two brahmin soldiers guarding the prostrate pillar at the height of the riots: "Ah," said one, "we have seen what we never thought to see—Siva's Lat has its head level with the ground. We shall all be of one caste shortly. What will be our religion then?" "I suppose the Christian," answered the other; "for, after all that has passed, I am sure we shall never become Mussulmans" (Sherring 1868:192-3; cf. Heber 1828:430-1). While the weavers with their cries of "Hasan! Husain!" reenacted the apocalyptic scenario of Muharram, the Hindus in the official (Persian) version their memorial explicitly likened their own mourning over [morning after?] the felling of the pillar to that of the last day (Robinson 1875:109; cf. Freitag 1989b:40 n.61). Whatever may have been the nature, extent and composition of the hidden social tensions, they fed into a primarily religious conflict; the priorities could be easily reversed, in other contexts, with the religious idiom serving to focus and mobilize interests and grievances that are primarily politico-economic. Whatever the weightage in either direction, such ‘communal’ conflicts implied, invoked and vehicled—regardless of the subjective perceptions of the parties involved—a commitment to social order or messianic aspirations that were already encoded into the opposing religious traditions.

By the law, as then existing, the sentences passed on the offenders should have depended on the fatwa (legal decision) of the Mahommedan law officers, who would however have been obliged to release the Hindu prisoners in order to avoid meting the same punishment to the Muslim detainees. The acting British magistrate, Mr. Bird, hence protested,

That the Government eventually dispensed with the fatwa to have the trials conducted by a special court, only serves to underline the impossibility of any fundamental reconciliation so long as polytheistic Hinduism continues to define itself in terms of a caste-hierarchy which must necessarily exclude or demote the impure Muslim, and so long as monotheistic Islam continues to define itself in terms of an uncompromising iconoclasm which it must necessarily impose on all infidels. The Muslim memorial ends with an appeal for nothing less than the separation of the two communities:

The Hindu memorial makes counter-claims on the sacred sites of the city. The idgah here had no particular sanctity but was esteemed by the Muslims only because it marked the former ascendancy of Islam over the religion of the Hindus, whereas the Kapalamochana tank and (what was left of) the Lat was of the highest sanctity to the Hindus. The district magistrate (now Mr. Watson) hence proposed to hand over the whole site to the Hindus as part of an overall policy of separating the two communities—to prevent future clashes—even at the price of totally excluding one or the other at disputed sites like Vishweshwar. However, Mr. Bird opined that both Hindus and Muslims had suffered so severely that neither would molest the other. "Government adopted his counsels and no alteration whatever was made to the original position of the parties. Permission was given to both alike to repair damages, and according to their respective religious customs each purified their violated altars. The Hindus held high ceremonies, and with prayers and Ganges water the fragments of the Lat were restored to their original sanctity and reverently buried" (Robinson 1875:106). But it was not until June 1810, "when the Hindus reconsecrated their outraged shrines and the veneration paid to the original pillar was transferred to the mutilated relic, that the first riot can be said to have actually concluded" (Robinson 1875:102).

Hindu aversion and Muslim resort to religious violence

The Muslim memorial charges that "every man whom [the Hindus] met with a beard they took for a Musalman and killed" (Robinson 1875:116). By temporarily surrendering all other interests—which may otherwise bind many Hindus and Muslims more strongly to each other than to their respective co-religionists —to their traditional religious identity, the adherents of the two faiths become (often the blind and unconscious) instruments of the bloody confrontation between two opposing worldviews. To label such collective violence, which is indiscriminate in its choice of individual human targets, as ‘fanaticism’ thus does not necessarily imply that it is without meaning. Moreover, to treat such violence simply as the by-product of mismanaged disputes about sacred spaces and symbols is also to ignore that both religions have built-in and distinctive theories of violence which affect the manner and circumstances in which they resist or resort to it. As it was the Muslims (though only the weavers) who had commenced the riots, the Hindu memorial remonstrates that

The Hindu memorial adds for good measure that "the violence sustained at the hands of these short-sighted Musalmans was not once practised under the administration of the Mahommedan Emperor. It has occurred under the Government of the English Company renowned for its active goodness" (Robinson 1875:111). Though "the Hindus had exacted a very ample retribution" and it was the Muslims who suffered the highest casualties and damage, the Hindus, when all was over, "considered themselves the injured party" and pressed for satisfaction of their religious grievances, demands which the British found difficult to ignore (Robinson 1875:103-4; cf. the Muslim memorial, p.117-8). Much later, Hindu nationalists, including Gandhi, could still speak "at crucial moments of the ‘meekness’ and ‘natural cowardice’ of the Hindu and the ‘bullying nature’ of the Muslim and the need to overcome these" (Pandey 1990:12). Different land-owning groups or dominant castes were no doubt engaged in a perpetual contest for temporal power, and could easily resort to land-wars, as had the Bhumihars under Balwant Singh in usurping the prerogatives of various Rajput clans. The Gosains continued the religious traditions of warrior-mendicants, who enlisted with such incumbents, or even engaged in ritualized combat with other Hindu sects, not to mention the Muslims. There was also a fair amount of tension diffused vertically and horizontally throughout the social hierarchy in the form of inter-caste rivalry and competition. But the notion of warfare merely for the promotion of religion, such as embodied in the figure of Ghazi Miyan, was wholly foreign to the Hindus as a whole. If the all-pervasive paradigm of the brahmanical sacrifice was centered on an act of quasi-criminal violence—the death of the victim being equated rather with liberation—this was only the ideological counterpart of an equally firm commitment to non-violence in the public realm. This outlook was best demonstrated by the model behavior of the brahmins, those at the pinnacle of the caste-hierarchy and constituting the very pillar of the religious significance of the sacred city. Though it was their authority—the symbolic universe they vehicled—that was menaced by the Muslims, their primary form of resistance was through fasting and self-mortification on the ghats. What could remain mere ‘meekness’ or become unprincipled hypocrisy or lying intrigue for the corrupt some (cf. Pandey 1990:107, n.105), was for others a question of principle that could on occasion even override their formal allegiance to Hinduism. It was a Brahmin who saved the already severely wounded 11 year old child of the murdered Mutwali at the Gyanvapi mosque, a fact gratefully acknowledged by the Muslims and for which he was rewarded by the British (Robinson 1857:100 note; cf. the Muslim memorial, p.116). Before the blow could be repeated, the Brahmin rushed forward to catch the falling Muslim child in his arms, interposing his own inviolable body before the sword of the Gosain aggressor and defying anyone to strike. The child was eventually given up by its brahmin protector to the Magistrate. Ironically enough, it was the very system of values that the ‘short-sighted’ Muslims had sought to undermine by attacking the Vishwanath temple and by uprooting the Lat, that had at the same time preserved the son of the Mutwali of the Aurangzeb mosque.

violence dissolves Muslim-Hindu divide

Under the Mughal empire and various regional states, Hindu and Muslim warriors could continue to practice their respective faiths side-by-side, develop syncretic modes of worship and fight shoulder to shoulder against co-religionists enlisted by rival powers (Bayly 1985:181-84,195-6). Bayly goes on to suggest that during the rise and decline of such (warrior) states, "undisciplined mercenary soldiers isolated in an urban environment"—precisely because they were men of war-like disposition displaced and downgraded (as by the British) from their earlier roles—"could easily provide the impetus for communal violence." What is required, then, is a separation of the problematics of violence in itself from the question of religious identity, before attempting to determine their modes of articulation in a wider long-term historical context. Whereas the ‘clannish’ weavers, already notorious for their caste-solidarity, were predisposed by their Islamic faith to religious violence (Pandey 1990:102), it is noteworthy that the Hindu castes embroiled in the riots were precisely those who were labeled, along with the Muslims, as ‘fighting castes’ by the British. Human violence has a logic of its own: even the intervention of the militant Gosains, which signaled the disastrous ‘sacralization’ of the conflict, was perhaps not a complete derailment of the archaic dualistic pattern around the raising and the felling of the Bisket linga. The brahmins and higher castes had been fasting on the ghats since the evening of the 20th to mourn and protest the sacrileges at the shrines particularly the Lat and the Ganga, which meant that liberation was no longer possible in the desecrated city. When they were finally persuaded by the district magistrate to disperse on the 23rd, the Gosains and other rioters, who had been too busy slaughtering and pillaging to participate in the fast, now took their place on the ghats on the 24th morning. Bird observed that "they collected not like the Brahmins on the 23rd from religious principle, but for the purpose of obtaining concessions" which they could now extort through "the danger to be apprehended from their influence and example" at a time when the public authority naturally looked to the community leaders for support (Robinson 1875:103). Equally important, however, is that, in pursuit of their self-interest, the Gosains nevertheless found it natural to adopt a brahmanical model of ‘non-violent’ protest based on self-mortification in order to exercise moral pressure on the British administration.

Having run its quasi-apocalyptic course along a Hindu-Muslim divide, which was already fraught with future communal developments (Bayly 1985), the indiscriminate violence had now begun to cut across religious barriers and assume political overtones increasingly directed against socio-economic injustices. The police had earlier

But even as "the original disturbances marked only by shocking religious outrages had subsided in June 1810," a singular feud erupted between the military, the chief indigenous instrument of British domination and aggrandizement, and the agents of law-enforcement, namely the police: "The sepoys carried on a guerilla warfare in the streets of the city against the police, and in either body Hindus and Mussulmans were indiscriminately mingled" (Robinson 1875:92). The sepoys had not only persistently defied a magisterial order against the carrying of arms in Banaras, but also ridiculed the police for their earlier role, thus leading to a long succession of affrays in August and September 1810 (Pandey 1990:40). Though the (otherwise disciplined) resort to violence is legitimate in both these arms of government and even constitutes their raison d'être, the significant development is that the conflict was no longer over religious issues, and was not even within the control of the British administration. Already during the Lat Bhairo riots, for about 20 days in October and November 1809, the sepoys were not allowed time off to bathe, dress or prepare their food, so as "to prevent them as much as possible from communicating with the people. For this purpose they were provided with mithaie [local sweetmeats] that they might be at all times within the control and observation of their officers" (Pandey 1990:48-9, his emphasis). When a reinforcement of British troops arrived on November 21, the authorities withdrew a good many sepoys from the city but still retained, for the same reason, the entire contingent of European officers. The British civil and military officials were indeed quite concerned that, like the Hindu and Muslim police, the Indian sepoys could themselves become infected by the contagious popular violence which could have easily sought a fresh, and perhaps more legitimate, target: namely, the repressive economic order imposed by the colonial administration.

House-Tax Revolt of 1811

Even before the city had been quietened down through a partial reorganization of the city police in October 1810, "the House Tax Regulation (XV of 1810) had been extended to Banaras, and from the ashes of the sepoy-police agitation, the phoenix of riot rose in all its original strength," so much so that "the 10th of January 1811 found Banaras, as the 1st of January 1810 had found it, seething with clamorous mobs and troops holding the city" (Robinson 1875:92). In the wake of measures to ensure greater control over collective gatherings, this ill-advised House Tax was introduced in December 1810 to provide greater revenues for funding the expanded state apparatus. The Banarasis protested that this additional burden would make it impossible for them to continue with their own charitable practices of maintaining numerous traditional institutions for the widows, brahmins, mendicants and the poor who filled the sacred city. They were even willing abandon to their city for good, rather than submit to this violation of the basic principles on which it had been established (Freitag 1989b:43-50). Most unpopular both for its amount and its novelty (Heber 1828:432-6), the proposed tax provoked the whole population of Banaras and its neighborhood to sit in the passive resistance of a unanimous self-mortifying ‘strike’ (dharna). The magistrate, Bird, noted that "men of all classes and description, from the highest to the lowest, whether Mohammedans or Hindoos, Jolahirs, Raujpoots and Goshains included, were all of one mind, and engaged by oath to promote the common cause." The "religious orders" including "men of rank and respectability" exerted their full influence in favor of the agitation; Heber even wrote of religious tracts (dharmapatris) "issued by the leading Brahmins as being central to the process of mobilizing the people" (Pandey 1990:41ff.). One British official observed that "instead of appearing like a tumultuous and disorderly mob, the vast multitudes came forth in a state of perfect organization: each caste, trade and profession occupied a distinct spot of ground, and was regulated in all its acts by the orders of its own punchayet" (Pandey 1990:47; cf.197). The shrewd British did not attempt to disperse the strikers nor the subsequent (abortive) march on the Governor at Calcutta, nor did they immediately yield to the civil pressure. Once the rebellion had exhausted its momentum, however, they quietly repealed "the obnoxious tax, and thus ended a disturbance which, if it had been harshly or improperly managed, might have put all India in a flame" (Heber 1828:436). This civil disobedience, indeed from the groundroots, became "an important object lesson for Benares and other dissident towns" for "similar events occurred during 1815-17 in other urban areas when the British attempted to introduce a more general regulation for chaukidari tax" (Bayly 1983:320, 21). Even where such protest eventually took on a violent turn, as in Bareilly in 1816, the "resistance of the mob at Benares [was] in everyone's mouth" (Bayly 1983:324). The crucial development, which seems to have escaped the otherwise exemplary impartiality of the contemporary British reporters, is that the Hindus and Muslims of Banaras were now united in opposing their economic exploitation and political domination by foreigners.

Hierarchy, Egalitarianism and Autonomy:
The ‘Gandhian’ Civil Disobedience of 1811

But is it really legitimate, after all, to press the ‘nationalistic’ claim that the all-Banarasi anti-House Tax protest of 1811 was directed against the British as foreigners, when colonial historiography has constantly driven home the message that India had been forever destined to be ruled by immigrant powers beginning with the Aryans and, until only recently, by the Moghuls (Inden; Cohn *; Pandey 1990:* )? Is it not also anachronistic for a ‘subaltern’ historiography to project backwards (and forwards) into time a contemporary politico-economic egalitarianism, when the reports indicate that untouchable Hindu castes and even the Muslim weavers had wholly participated in a highly differentiated and internally hierarchized style of collective action (Freitag 1989a:205 n.4; 1989b:17 n.32)? Can one speak of ‘communalism’ as a post-colonial development when the continuities with pre-colonial religious conflicts are all too painfully obvious (Bayly 1985)? At the heart of Freitag's "processual analysis of symbolic collective behavior are the events of 1809-1811. During these two years a style and a symbolic rhetoric of protest emerged in Banaras which differed in important ways from the style of protest that evolved in other U.P. urban centers" (Freitag 1989a:205). By simultaneously stressing the ‘Hinduized’ style of the politico-cultural structures "integrated symbolically through the Maharaja" (1989a:227-8), on the one hand, and reducing religious festivals to their ‘civic’ dimension alone, on the other, Freitag allows us no insight into the manner in which Hindu paradigms had so powerfully shaped the Banarasi ethos as to determine even the non-sectarian ‘politics’ of popular anti-British protest. Moreover, there is no reason why the processual approach should be limited to the relations between communities, and with the emerging modern state. For it already defines not only the social interaction between Hinduism and Islam, prior to and independently of state intervention, but also the dynamism internal to each tradition taken separately. In this respect, it would be necessary, first of all, to contrast the differing understandings and models of community within Hinduism and Islam, and the general lines of their interaction with each other and with the ‘secular’ society promoted by the modern state; instead of simply substituting our notions about community as the pre-given ground on which they function. A careful re-reading of the manner in which the Hindu-Muslim confrontation of 1809 metamorphosed into the Indian-British confrontation of 1811, particularly of the continuities and contrasts, reveals not only the interference between a hierarchic ‘Banarasi’ and an egalitarian ‘Meccan’ model of society, but also allows for the emergence of a common principle of ‘autonomy’ whose spiritual and political dimensions have been worked out differently in the two religious traditions and in the secular discourse of modernity.

The district magistrate, W. W. Bird, who "remained in India for 31 years and led a very distinguished career, becoming Judge and Magistrate of Benares in 1814 and eventually rising to the rank of Deputy-Governor of Bengal and Senior Member of the Supreme Council" (Heitler 1972:240 n.2), was no doubt an instrument of imperial domination. But without his apparent integrity, active impartiality and personal courage in the exercise of the limited resources at his command, the loss of life and extent of damage during the 1809 riots would have probably been far greater. The undisguised appreciation of "the greatness of British rule," especially the role of local officials during the 19th century in maintaining law-and-order menaced by recurrent Hindu-Muslim tensions, which is registered by the Muslim chronicler of Mubarakpur (Pandey 1990:135-8), seems to echo the post-riot sentiments of the Banarasis. The Hindu memorial, for all its obsequy, would suggest that foreign capitalist exploitation had been compensated for—if not largely offset—by the security to life and property afforded by the imperial pax Britannia:

"Banaras was one of the fastest growing cities during the years 1750-90. It became the subcontinent's inland commercial capital¼ in 1757. It received immigrant merchant capital from the whole of north India and stood astride the growing trade route from Bengal to the Maratha territories. Yet this was also a city which benefited from the sustained agricultural performance of the rich and stable tracts which surrounded it" (Bayly 1983:104). In his analysis of structural change in the rural society around Banaras, Cohn (1987:343-421) points out that it is easier to trace the meteoric fortunes of those—like the Bhumihar clan that took over the reigns of the kingdom—who capitalized on the British land reforms through their positions as tax farmers or as subordinates in the civil service, than to determine the fate of the groups that were ‘dispossessed’ (at least economically), because the system was an open one that generated even more wealth to go around. Though the Rajput landowners, for example, lost their political dominance, they probably profited from the increased agricultural yields and, in any case, were not impoverished. Moreover, the people would tolerate excessive taxes and even ‘plunder’ if only for the policing functions of the state and its capacity to provide protection from external marauders (Bayly 1983:390-3). "As a logistical exercise, in fact, British expansion into upper India after 1800 was very much the sub-imperialism of Benares—the culmination of that process of maximising revenue and town-building which had been initiated by Mansa Ram in the 1740s" (Bayly 1983:211, cf. 210-12). In terms of urban political economy, at least, the gradual imposition of British rule on Banaras fitted into a pre-existing Indian pattern, all the more so as the English Company had accommodated the Maharaja's claims to real cultural and (at least nominal) political leadership of the diversified community. What was new then about the style of the colonial administration, that provoked the entire city to rise up in one body against the proposed House-Tax, and to persevere in such a prolonged act of civil disobedience?

The Mughal Emperors and the Nawabs of Awadh participated in, and expended a considerable portion of their resources on, public arena activities and courtly ceremonies which amounted to ostentatious displays of the manner in which they incorporated the body politic of their subjects. Reciprocally, the people enjoyed the liberty, if not to revolt, at least to publicly undergo self-inflicted deprivation and suffering in order to attract the attention of a paternalistic ruler to their legitimate grievances. The Indian(-ized) rulers, even in their excesses, had remained bound to their subjects in a system of mutual obligations which could be set aside only at their own risk. "A ruler's power was also constrained by the religious context. A ruler needed to be legitimate in the eyes of the powerful Islamic elites of writers, jurists and service communities. He had to offer them patronage and also sustain the life of the community of the faithful. More subtly, no regime could hope to become a dynasty unless it adjusted itself to the predilections of the mass of its Hindu subjects" (Bayly 1983:115-6). Even the greed and ambitions of the nouveau riche, like that of the ruling clan, was soon channeled into the traditional Hindu pattern that had shaped the policies and lifestyle of even the low-caste ‘upstarts’ and ‘robber-barons’ who had founded earlier empires (cf. Cohn 1987:339). Though the Company's insatiable greed for revenue took the form of administrative measures that encroached ever more deeply into areas of Indian life that had till now remained immune to state intervention and control, the British authorities preferred to exercise their ever more pervasive powers from behind the scenes. They were willing to deal individually with ‘respectable natives’ but insisted on remaining invisible in the public arena activities through which the Indians expressed and reiterated their social bonds and complex system of values (Freitag 1989a:215, 1989b:53). In the aftermath of the 1809 riots, the Gosains thus refused to budge from the ghats until the Acting Magistrate felt "compelled to go in person to remove them" (cited in Robinson 1877:103), just as he had felt obliged the day before to (con-)descend to the ghats in order to urge the Brahmins to disperse and to allay the apprehensions of the people by breaking their self-imposed fast. When the protesters of 1811, now all the Banarsis, submitted on January 23 that "they were willing to disperse, providing [the Acting Magistrate] came to them [in the field where they were residing] in person to request it" (cited in Dharampal 1971:30), W. W. Bird, who was already indignant in 1809, now flatly refused. If his condescendence was perhaps largely compensated for then by his prestigious role as the impartial arbitrator between the rival claims of Hindus and Muslims, he now risked compromising the supposedly absolute authority of British ‘justice’ by conceding to the combined moral pressure of all the Banarasis. Though the military basis of political power was at the state's command, he was obliged to turn to the Maharaja to lend it the minimum of moral legitimacy (Freitag 1989a:219-20).

Banaras had been a ‘mughalizing’ city in the early 18th century and reflected cultural patterns that continued to be fostered by the Nawab’s court at Awadh. Hence the strong ties established early in the career of the present Bhumihar dynasty—a landlord family which had served as tax officials for the Nawab but had become virtually independent by 1750—with Muslim lower-caste groups like the weavers. The triumvirate of powerholders—royal clan, merchant-bankers and Gosains—patronized an innovated and grandiose form of the Ramlila, centered on the symbolic identification of the Bhumihar Maharaja with Lord Rama, in which the rival landowning and commercial groups like the Rajputs and the Marathas also participated. Even the Muslims joined in celebrating it, more as a civic festivity around the unifying figure of the king offered as an aesthetic spectacle in the public arena at Ramnagar and Nati Imli (Freitag, 1989a:1–22, 203–28; 1989b:19–52). During the previous 100 years of Muslim rule, the primary religious focus of the Hindu community had been already shifting from the ascetic-cum-brahmanical temple-cult of Shiva to the household and devotional cult of Rama. The integrative royal festival par excellence was now no longer the cosmogonic marriage of the god Shiva-Bhairava but the reenactment of the exemplary life of the divine hero of Ramcaritmanas.

The Maharaja, who was identified with Shiva and whose Bhumihar ancestor had imposed the ban on cow-slaughter in the first place, may have justifiably been expected to champion the cause of Hinduism during the abortive 1809 attack by the Julahas on the Vishwanath temple. He had however not only remained aloof, but had disarmed his men voluntarily and removed them from the city during the violent period. "In the aftermath of the riot, he worked to restore peace, treating with the British administrators, on the one hand, and working with the leadership of various muhallas, on the other" (Freitag 1989a:212; 1989b:41). Had he intervened on behalf of the Hindus, he would have jeopardized the economic symbiosis between Hindu merchants and Muslim weavers, that contributed to the prosperity of the city and indirectly buttressed his own authority. He would have also provided a pretext for the British to not only curtail his already diminished powers but to thereby further extend their administrative hold over the life of the city. Over and above the dictates of self-interest, the Maharaja was striving, at least before the eyes of the public, to live up to the traditional role of an integrative (rather than exclusive) Hindu ruler. The cultural prestige of this role-model was such that even the leader of the Marathas, who was otherwise anxious to fill the role of Hindu avenger, vacillated and sought intermittently to pretend to this style of exemplary leadership—much to the exasperation of the Magistrate—by intervening with the Banarsi crowds at the behest of the British (Freitag 198b:41 n.65). The anti-House Tax protest now presented the Maharaja with the difficult choice of either surrendering his authority over the Banarasis, or further humiliation by challenging the British. Though the Maharaja proceeded in person, with all the distinctions of his rank, to the assembled people and used his prestige to convince the Banarasis to abandon their agitation, he thus "did not see his role simply as an underling and supporter of British authority but rather as mediator in a dispute. It was largely through the influence of the Raja that peace had been restored. His position had become pivotal" (Heitler 243-4). Though he brought fifty of the leading protesters to [the Acting Magistrate] to acknowledge their offenses [of disturbing the peace!]¼ the Rajah himself interceded in their behalf and solicited [the Acting Magistrate] to endeavor to procure both for themselves and for the subject of their complaint, the indulgence of the Government" (Bird, cited in Dharampal 1971:32-33). The Maharaja's continuing pleas with the Magistrate on behalf of his subjects were "buttressed by expressions of ongoing public discontent; indeed, householders greeted the assessors with such ‘sullen silence’ that the government could never actually collect the tax" (Freitag 1989b:50 n.87). Freitag's conclusions are worth quoting in full: "Indeed, the figure of the Raja proved essential in focusing and expressing the continued cultural integration of Banaras. On the one hand, he could represent symbolically the culture shared by upper and lower classes, by Hindus and Muslims, and by Gosains, merchant, and Rajput alike. On the other hand, he provided the only symbol of authority on which both Banarsis and British could agree—the only figure who could operate both in public arenas, expressing the reformulated system of self-rule emerging in Banaras, and as an intermediary judged legitimate by the British administration" (Freitag 1989b:50).

It was not so much that the economic burden of the House-Tax was considered excessive but, as Bird himself came to realize, that it constituted a serious encroachment upon the local autonomy and privacy of the Banarsis (Heitler 1972:252-5; cf. Freitag, 1989:*). Though it amounted to a substantial sum of money and was collected from each householder, the earlier phatakbandi tax was an entirely local operation organized within each muhalla and without the direct involvement of any British official. The night-watchmen were selected by the inhabitants themselves; the British magistrates merely provided a pro forma approval for the appointments. Hence "even after the government agreed to remit the phatakbandi tax and exempt the houses of the poor and religious institutions from the operation of the house-tax—steps which considerably reduced the economic burden of the tax—the people continued to resist" (Heitler 1972:253). Not only had such an innovation absolutely no sanction in both Hindu and Islamic legal custom; unlike the various other taxes which the people had accepted, it licensed the penetration of government agents into the intimacy of private homes in order to assess their value. By retaining the earlier arrangements, and allowing each householder a deduction from the house tax equivalent to the amount of phatakbandi paid, the final compromise allowed the muhallas to conserve a considerable measure of freedom in deciding its own affairs. Each muhalla had, in fact, originally presented its own petition. The sources seem to indicate that "the population of each [muhalla] was primarily made up of one caste" and that the "organizational base of the Varanasi hartal were the various caste [pancayats] of the city" (Heitler 1972:257,256). In any case, "the plural organization of living quarter, caste and occupation existed as part of a wider universe of political and religious notions" and thousands of fellow caste-members from the surrounding countryside were mobilized by the Lohars, Kunbis and Khorees. "There was no question of the city being a world of its own; it was the cultural centre of most rural groups and the political centre of the dominant landholders" (Bayly 1983:320,321). What was thus at stake was not so much the economic exploitation in itself, but the manner in which it was being used to confiscate the relative autonomy enjoyed by each caste in managing its own affairs, and in negotiating and regulating its relations with other castes.

Indeed, the British administration was contesting not only the legitimacy of the Indian claims to immunity from state intervention in certain areas of community life. They were contesting also, and especially, the legitimacy of their circumventing the approved channels of the new system in order to press those demands (Freitag 1989b:47-48; Pandey 1990:46-8,53). Though an illegal "disturbance" in the eyes of the British, "the protesters in Varanasi in 1810 considered the hartal to be a technique of great antiquity and sanctified by tradition. They made this point clearly in their final petition to the Governor-General" (Heitler 1972:249). The practice in the Banaras region is dated as far back as Duncan’s period of residency (1787-1795); the efficacy of similar actions by a particular caste of agricultural laborers or urban artisans in response to some petty insult or extortion suffered by one of its members is reported for the same period even from South India. More than a time-worn method of protest, the unanimous cross-caste resort to civil disobedience was constitutive and expressive of the very liberties that the Indian social order was attempting to defend.

The British encroachment on local institutions of self-rule was nevertheless facilitated, and even partly imposed, by the use—rather, abuse—of such liberties by one community in order to infringe upon and violate the prerogatives of its neighbor (cf. Freitag 1989a:214-15). Just as the pancayat of the Julahas became the occasion to take the unilateral collective action of desecrating the Lat, the Rajput leader "considered this a dispute in which the public authority had no business to interfere. That as the injury had been offered to the Hindus at large it was for them alone to determine the measure of their revenge and unite in the common resolution of inflicting it." The Acting Magistrate hence diagnosed the threat to law and order as arising precisely from the "abuse of that privilege which the Natives have been permitted to enjoy, of assembling among themselves to deliberate on matters of common interest. I found it expedient to prohibit all assemblies of this nature without previous application to the police" (Bird, cited by Pandey 1990:47; by Freitag 1989a:214-5, 1989b:48). Similarly, the failure of the police to remain impartial and prevent the initial damage to the Lat, resulted in British schemes to recompose and reorganize the force by hiring more Hindus to counterbalance the overwhelming number of Muslim police. Though Bird's original proposal that the position of kotwal be completely abolished was not implemented, much of his responsibilities were still reassigned to the Magistrate himself. Given also the difficulty—if not impossibility—of arbitrating in a religious riot by determining the exact roles of apprehended participants, the administration decided that certain persons would be "brought to trial not for the outrages each had committed against the other, but one common offence against the peace¼ as an offence against the State" (cited in Freitag 1989a:214; 1989b:43). Freitag points out that such a trial, of persons possessing "considerable respectability," was recognized even by the British as an unprecedented act by the state. The ‘natural leaders’ who were eventually singled out as appropriate symbolic representatives for their entire communities were precisely those who had indulged in or instigated the ‘communal’ violence. Thus, not only did the Muslims and the Hindus unwittingly impose, to their own detriment, the logic of divide-and-rule upon the colonial state; by undermining the moral authority upon which their civil liberties were based, their religious disputes—especially the resort to violence—led inexorably to the progressive confiscation of whatever precious little was still left of community self-rule.

The total commitment of all social groups to non-violent resistance, both as a conscious tactical decision and as a matter of principle, even in the absence of any single charismatic leader like Mahatma Gandhi a century later (Heitler 1972:239,244), is abundant proof that, beyond the immediate politico-economic issues, even the Muslim weavers, who only recently had not hesitated to take up arms in the name of their religion, were still defending a socio-cultural order that was fundamentally Hindu. Though similar anti-house tax protests through the following years in other urban centers of north India took the success of Banaras as model and incentive, they seem to have sometimes degenerated into violent revolt. Such was the case with the one in and around Bareilly in 1816, which was speedily suppressed by the British before it spread (Bayly 1983:323-28). Given their numbers and unity, the Banarasis could have easily overpowered their colonial overlords, and the British commandant was ever fearful that "the Rajputs, Gosains, ‘Muslims’ and other ‘fighting cast[e]s’ [sic] might take up arms, especially if the blood of Brahmans or other ‘religious orders’ were spilt" (Pandey 1990:42; cf. Heitler 1972:248). Whatever part considerations of possible long-term consequences and even of the larger benefits of British rule may have played, this ‘sudden’ conversion to non-violence is only the culmination of the process which led the Gosains and other ‘fighting castes’ to imitate the Brahmans and to take their place on the ghats after the riots of 1809. Traditions of vegetarianism and non-violence, observed by the Sufis already in the Middle East, were reinforced in India through their contacts with more orthodox currents of Hindu ascetics. The voluntary self-denials and bodily mortification through which the protesters expected to exercise moral pressure on British political power derive from not only from the Hindu dread of responsibility for injuring Brahmans (cf. Cohn 1987:353), but also from respect for the power of tapas shared by Hindu ascetics with their Muslim counterparts (Heitler 1972:245 n.20). There is indeed a continuity between the ritual purity of the brahmin and the self-denial of the renouncer, which accounts for the immunity of both and for their authority over the Hindus and even the Muslims. "An oath was taken to ensure the commitment and solidarity of the group. The entire protest had religious overtones (often articulated in terms of the defense of Brahmins and fakirs). Yet support for the protest crossed religious lines and there is evidence of Hindu-Muslim unity" (Heitler 1972:244). The religious sanction involved is captured in the colorful note added by Mill to Bird's account: "the very thieves refrained from the exercise of their vocation although all the shops and houses were left without protection—the people deserting the city in a body" (cited in Pandey 1990:41). There were more than 40,000 Brahmins living on charity in Benares city alone about 1810 accounting for 17-20 per cent of the population; a census of c. 1799 shows just under 20 per cent of all houses occupied by such Brahmins living on charity (Bayly 1983:126). The social, and even economic, life of Banaras had long been structured by a reigning ideology based on charity to brahmins. Not only was the mass political action informed by their principle of non-violence, the protesters were perfectly aware that in defending their own ‘civic’ rights they were also defending the religious guarantors of these values.

The "key to the issues which brought merchant and townsmen into conflict with early colonial rulers" lay almost always in "some perceived threat to the moral economy of the merchant family¼ It is notable that the only times when Hindustan's bazaars closed down altogether during our period were when cattle, Brahmins and the purity of the River Ganges seemed under threat" (Bayly 1983:392-3). The religio-cultural determinations of Indian commercial activities served to resist, reduce and buffer the social impact of the thoroughgoing economic mentality impinging from without through administrative and fiscal measures introduced by the British East India Company (Bayly 1983:229-62). Indeed, though the changes accompanying early colonial rule further consolidated the emergence of a unified north Indian merchant class, what is really striking is the persistence, vitality and, in some respects, even revival of its Hindu ethos (Bayly 1983:369-93). It was for moral rather than for economic reasons that the merchant community of Banaras had resisted the British House Tax, whereas the defectors among them had been prompted precisely by the hope of material and political rewards. The (inviolability of the) sacred cow was the most appropriate symbol for mobilizing the caste-society for it not only embodied all the religious values invested in the brahmin, but also encoded the powerful hold of this sacrificial ideology upon the moral economy of the entire Hindu community.

The anti-House Tax protest of 1811 was no doubt an all-Banarasi affair in which its diverse communities stood united in defense of their respective autonomies that interlocked to constitute and guarantee their urban self-rule. Though British attempts to selectively rally or coerce the leaders of the protest, like the powerful leader of the Maratha community (Amrut Rao), did not prove effective, whatever relative success they met with was rather among the local elite (Heitler 1972:251-2). One Baba Juman Das who is identified as "the principal merchant of Benares" kept his shop open in defiance of the ban imposed by his fellow citizens. When urban protesters coopted their caste-fellows from the rural areas, Bird managed to convince some of the landowners that the participation of their tenants was detrimental to their interests as well. They were persuaded to forcibly remove the tenants they controlled from the encampment and take them back to their estates. The first to take this initiative (and responsible for keeping one grain bazaar open) was one "Babu Sheonarian Singh" who belonged to a family with a proven record of loyalty to the British at the expense of a series of Indian causes. He was typical of the extent to which some of the ‘respectable natives,’ who had heaped their fortunes through their British connections, compromised their social allegiance to their larger community out of purely economic motives. On the other hand, even after government offered to exempt the houses of the poor, the latter not only continued to resist but also prevented the landowners, who would otherwise have had the most to lose financially, from ceding to the British strategy. Pandey points out that the Rajputs, who had championed the ‘Hindu’ cause in 1809, did not participate so enthusiastically in this popular uprising against the British administration; many Rajput landowners even assisted the colonial authorities in their attempts to disperse the crowds. "And while many of the ‘leading native inhabitants’ and ‘religious orders’ of Banaras were certainly involved in the anti-house tax agitation, they appear to have conceded the leadership, at least in the initial stages of the protest, to artisans, skilled workers and other sections of the lower classes" (Pandey 1990:42). The blacksmiths (Lohar), in particular, took the initiative of halting the production and repair of implements of cultivation and of harvest (which was fast approaching) so as to ensure the cooperation of even the landowners. This ‘blackmail’ of the higher Hindu castes by the supposedly ‘oppressed’ lower castes was not merely economic but also religious in nature, for Bird notes that "during the 26th, the dead bodies were actually cast neglected into the Ganges because the proper people could not be prevailed upon to administer the customary Rites" (cited in Pandey 1990:41). Nowhere perhaps is the ritual solidarity—involving a host of other low-caste specialists, like barbers and Mahabrahmins, amounting to possibly 2-3 per cent of the total population (Bayly 1983:126)—between the brahmin priests and the lowest untouchables, like the Doms, more evident than in the perpetual rite of cremation which gives its central significance to Banaras. Whereas it has become fashionable to decry brahmanical ideology for imposing the stigma of untouchability, which is the logical consequence of the caste-order, we are faced with the paradox of these lower castes themselves championing the cause of the brahmin ‘gods on earth’ in the face of possible defection on the part of the so-called ‘dominant’ castes!

"Lower castes among the Hindus—Kohars and Kahars, Koeris and Kurmis, Mallahs and Doms, and above all the Lohars—took an impressive and prominent part in the defense of the interests of the local community (i.e. the community of Banaras, Muslim as well as Hindu) on the imposition of the new house-tax" (Pandey 1990:197). If such a demonstration of self-discipline and caste- or corporation-based autonomy, in the very process of defending the principles of self-rule, could nevertheless find explicit articulation in terms of the defense of Brahmins and ascetics, it is certainly time to ask whether it was the underlying religious ideology that had assured such a degree of auto-determination in the first place. Relations of hierarchical interdependence—an ‘institutionalized inequality’ which is never devoid of tension, competition and even violence—run the constant risk, however, of degrading into relations of domination. The regular reenactment of festivals and other forms of ritual in the public arena—based directly or indirectly on the structural paradigm of the brahmanical sacrifice, itself centered on an act of (at least symbolic) violence—was meant precisely to minimize and offset the effects of entropy already inherent in the social ‘eco-system’ and constantly impinging upon and eroding it from without. Despite all the new pressures, dislocations and disequilibrium introduced into the social system by British administrative measures over the preceding half a century, what the Banarsis—even the Hindu untouchables and the low caste Muslim weavers—were defending against the British was a hierarchized system of mutual dependence that nevertheless ensured a certain degree of group (and probably even personal) autonomy to their respective lifestyles. This explains why specific caste-interests, when threatened, could react by reaffirming and enforcing inter-caste solidarity through a distinctively religious idiom formulated in terms of defending the privileges of brahmins and fakirs. Logically, homo hierarchicus could, and did, go even further by defending the system of values and social relations as a whole, even at the cost of forgoing (short-term) material or political gains at the expense of other (even dominant) castes. What is indeed surprising is that this holistic ‘moral economy’ was so ingrained in Indian (rural) attitudes that, long after such relations of caste inter-dependence had slided irremediably under colonial rule into relations of class domination and sheer exploitation, the Indian peasant revolt of 1919-1922 in Awadh could have still hesitated to assume the role of the ‘proletariat’ in order to dispossess their (often absentee) landlord oppressors (Pandey 1988:264-5). If Gandhi could so successfully ‘authorize’ and impose his policy of ‘trusteeship’ by the privileged rich, and of Indian class-solidarity, in forcing the British to quit—thereby avoiding the terrible costs in human life and in rebuilding