Guarding the Boundary: Colonialism, ‘Custom,’ and the Ganesh Festival of 1890s Bombay

Ganapati Utsav

Ganesh (or, to fill out the Sanskritic final syllable, Gaṇeśa) has the most eye-catching iconography among the deities of modern mainstream Hinduism: he is a therioanthropomorphic figure, a rotund man with an elephant’s head. He is recognised with reverence and affection across India and beyond – wherever Hindu people have settled – but historically he has enjoyed a particular regard in Maharashtra. Ganesh is also popularly known as Ganapati, which means Lord of the Ganas. These are minor mythological figures, often comical, spooky, or grotesque, associated with Ganesh’s father, Shiva. In the present day Ganesh is regarded principally as a beneficent figure, a dispenser of bounty and good fortune with comical aspects of his own, most notably a certain joie de vivre. He exerts a special dominion over threshold zones in time and space, and this guardianship function is implied in another of his names: Vinayak, Lord of Obstacles. In his capacity as Vinayak he is depicted on doorways, invoked at the start of writings and performances, and addressed at the beginning of prayers even if the devotee’s principal message is directed to a different god. Altars in Hindu temples and homes seating one or more deities often have a place for an ancillary Ganesh figure at the southernmost end; the guardian is there to check any malignant forces that might approach from that inauspicious direction (Courtright 1996: 35).

It is in the Hindu month of Bhadra, Gregorian August/September, that Ganapati Utsav takes over the streets. The fourth day of the waxing moon, chaturthi, inaugurates the festival, which speakers of Marathi and Hindi often refer to in its entirety as Ganesh Chaturthi, although the public celebration lasts ten days. The culmination and climax come at the end of the bright fortnight, anant chaturdashi, when the sacred effigies of Ganesh are carried by their worshippers in procession to the Arabian Sea or other bodies of water for immersion and dissolution (visarjan).

In the present day, the urban festival follows a set of norms established with the endorsement and encouragement of Tilak and the primarily Brahmin nationalist circle that was based in Poona (nowadays spelled Pune) in the 1890s. The best way to introduce the pattern may be to work through the Marathi/Hindi terminology that has been incorporated ever since into English-language descriptions of the festival. The colourful, often extravagant images of Lord Ganesh – twenty-first-century examples are often twice as tall as a man – are known as murtis: cultic objects that receive ritual attention as symbols or vessels of gods. The question of how the idol does its ‘work’ – whether it represents the deity whose form it cites, or materially embodies it – is a live one in different schools of Hindu thought. But it is safe to say that, across the celebration’s ten-day span, the murtis are indeed recognised by many of their worshippers as efficacious stations or vessels of divine power. On the inaugural chaturthi day, they are installed in public spots in Hindu neighbourhoods and consecrated by Brahmin priests. The temporary structures that enshrine them are known as pandals, which are typically designed around topical themes that may present political critique or other sorts of religiously encoded commentary.

For the next nine days, the pandals stand as local attractions, kept open to the local community and the broader public late into the night. Augmenting their glamour as sacred spectacles are qualities of artistry and novelty. Visitors may address Lord Ganesh through the prescribed sequence of worship (puja), connecting with the deity in the ritual exchange of gazes (darshan), and food offerings (returned as the god’s blessing: prasad); they may participate in the singing of hymns (kirtan) or follow the roster of speakers who deliver sermons, narrate from Hindu mythology, or lecture on topical matters. In the early years much of the attention of the public and the authorities alike centred on the melas, the ‘crowds’ or teams of young men who serve the god and entertain spectators with coordinated dance, song, drill, and display of martial arts. On the visarjan day, the mela escorts the murti down the avenues to the site of immersion, vying with rival bands in noise, flash, and exuberance.

The planning, design, sponsorship, and production of the pandal have been organised by a neighbourhood club known as a mandal. These can be fairly informal groupings, but the participating units registered with the official city festival, which are sometimes affiliated with local temples, are known as sarvajanik, or public, mandals. The coordination of public mandals is overseen in Bombay and other cities of Maharashtra by a central authority that mediates, in turn, with the municipal authorities. In today’s Greater Mumbai, or Brihanmumbai, this body is known as the Brihanmumbai Sarvajanik Ganeshotsav Samanvay Samiti; its antecedents go back to the Ganesh Mandal of Tilak’s time. Let me stress this point about continuous state surveillance: from the festival’s inception, police and civil administrators have maintained an interest in monitoring and regulating this ritualised takeover of urban space. A contemporary development is that now, in addition to security concerns and crowd and traffic management, the authorities confront a formidable environmental challenge in the offshore disposal of an elephantine mass of polychromed idols.

To sum up: Discrete community groups across the city organise local youths and raise money from businesses; they produce spectacular floats that function as metonyms of neighbourhood identity; a period of celebration lasting several days culminates in a citywide pageant in which groups from the city’s constituent neighbourhoods parade their floats and merge on the thoroughfare in a grand procession. A religious tradition presents the occasion for an urban phenomenon in which agencies of the state, political patrons, business interests (very much including the tourism industry), and professional as well as demotic voices of cultural expression work uneasily together to open up central spaces to a release of collective libidinal energy. Masses of young, vital bodies – generally belonging to subordinated urban constituencies – exert themselves before a public gaze. Now, outlined like this, of course, the morphological portrait fits not only Ganapati Utsav – but also Carnival as celebrated in such cities as Rio de Janeiro, Port of Spain, New Orleans, and indeed all across the Caribbean and Latin America. Scholars of urban religion in other areas and periods will surely find it easy (and possibly thought provoking?) to come up with other celebrations that follow this script.

Another point of comparison for the modern, public Ganesh Festival is found closer to home. Let us consider the yearly excursions of the presiding deities of Hindu temple towns such as Puri, Bhaktapur, or Madurai. In these cases, the temple’s centrality to the town’s identity – and physical layout – reflects a historical alliance between the site’s Brahminical establishment and the kingly dynasties that used to rule the area. Normally, the murti of the deity – the guarantor of the (former) ruler’s legitimacy – must be approached for worship within the temple. On the occasion of the festival, however, either the principal murti or a secondary, traveling murti is taken out of the temple confines, mounted on a wheeled platform, and – attended by priests and other local notables – is pulled through the streets by male devotees. The conditions of darshan, the exchange of gazes, are thus reversed: from the spectacle displayed in the sanctum the deity becomes the mobile overseer of the civic domain.

At first glance, there is a closer match still for Ganapati Utsav: Durga Puja as celebrated in Bombay’s mirror city on the eastern coast of India, Calcutta (Kolkata). Here, in the erstwhile colonial capital, Hinduism’s Great Goddess in the righteous warrior persona of Durga is embodied in clay murtis, enshrined in neighbourhood pandals – often presenting highly creative and topical tableaux – and, following worship over several days, is consigned to the sea. The outstanding work of English-language scholarship on Durga Puja is Rachel Fell McDermott’s 2011 Revelry, Rivalry, and Longing for the Goddesses of Bengal. McDermott’s exposition is wide-ranging, but the most compelling strand of her argument is the historical inquiry that relates features of the present-day festival to salient observances and debates of the early 19^th century, the heyday of the East India Company’s rule. The stakes identified in this colonial encounter are very different from those of 1890s Bombay. For example, it emerges that a key spectatorship constituency for which the rich Hindu sponsors of early Durga displays competed was none other than the British officialdom of Calcutta, many of whom looked forward to the yearly Puja as a spectacular entertainment. McDermott (2011: 153–156) herself invites comparison with Ganapati Utsav, noting several points of correspondence. Again, as I reflect on the site of my interest in the light of her richly informative study, it is the points of divergence, rather than the similarities, that strike me as notable.^[2] But possibly some of the questions I will move on to present in this article – generated from my own effort to evaluate the colonial origins of practices of modern urban religion – will prove applicable to Durga Puja as well.

None of this is to deny that in the present day, the two festivals stand as complementary emblems in Indian public discourse of regional and urban Hinduisms – Bengal and Kolkata on one hand, Maharashtra and Mumbai on the other. But the larger point is that, rather than morphological or doctrinal similarities in the present, it is to the historical record that one should look for the most productive site of comparison. For there was indeed a recognised model for the Ganesh Festival, its direct inspiration and antecedent, but it comes from outside the Hindu tradition. This is the Shiʿi holiday of Ashura – or Muharram, as it is called in India after the Islamic month in which it falls – as observed, specifically, as an urban celebration in 19^th-century Bombay.

Muharram

Doctrinally speaking, Muharram is an observance of mourning. The holiday commemorates the martyrdom of Hussein, the Prophet’s grandson, at the Battle of Karbala in 680 CE. Historically and into the present day, the meaning of this event, the lessons drawn and annually reiterated, and – perhaps most germanely – the prescribed manner of enacting commemoration have been contested among diverse constituencies across the Islamic world. The lines are not necessarily drawn between rival Shiʿi and Sunni camps. In 19^th-century Bombay, the observances followed norms introduced by Sunnis from the coastal Konkan region, who were the dominant component of the city’s Muslim population through the early decades.

Jim Masselos (1982: 50) observes of the Konkanis that, at the century’s start, ‘Bombay’s Mohurrum was . . . essentially their festival’:

It had two main visible elements. First were processions when groups of men and youths representing the collectivity of various neighborhoods moved around in their own and nearby streets, lamenting but also singing and dancing. Models of the tomb of Husain and his brother Hassan were also installed in the Muslim streets. . . . The tombs or tabuts were subsequently carried in individual promenades throughout the area late on the ninth night and into the very early morning of the tenth day. On the afternoon of the tenth day all came together in a single gigantic procession. It wended its way through the Muslim parts of the city and finally went to the seaside where the tabuts were immersed.

The parallels are clear. The spectacular bamboo-and-paper tabuts – thematised around statements of topical (possibly political) import, exhibited within their home neighbourhoods for several days, and finally presented to the public at large on the way to the sea – corresponded to the Ganesh pandals. The tolis, the teams of local youths who attended the tabuts in the penultimate days of the festival, performing martial arts and generating high spirits, were a model for the melas (Kidambi 2007: 179–180; Masselos 1982: 58). The locality-, caste-, and guild-based committees that organised the tabuts’ fabrication and exhibition – and the infrastructure of support, including funding sometimes obtained through extortionary tactics – found their counterparts in neighbourhood mandals.

Masselos charts the festival’s vicissitudes. Through the 19^th century, the Muslim population grew in numbers and diversity, with Ismaili trading communities from Gujarat, North Indian Julahas, and Twelver Shiʿis from Iran being notable constituencies. When, claiming a proprietary right to Muharram, the Iranian Shiʿis sought to prescribe their own modes of observance, Konkani leaders successfully petitioned the government to keep the Iranian practices off the streets. Here is a point to which I will return: religious leaders (clients of the regime from that class often described as ‘native elites’) enlisted the British authorities as guardians of orthodoxy, charged with policing not only spatial but spiritual bounds. And over the decades, even as tabut-centred processions became the form recognised, in official parlance, as Islamic ‘custom,’ the spectacle grew in numbers and visibility. At mid-century, between processors and onlookers, the total number of people involved was estimated at a hundred thousand. ‘A figure which even if approximately correct is extraordinarily large given a city population of about half a million,’ comments Masselos (1982: 54), who goes on to state:

It was the city’s major spectacle and one which was probably observed by most inhabitants. As the Times of India noted in 1884, Bombay’s Mohurrum was “a carnival . . . the like of which for extent and eccentricity, is to be found in few other cities of the world.”

Among the carnivalesque elements were naked revellers painted in tiger stripes, and others disguised as bears, elephants, and even sannyasis, Hindu renunciants.

Or were the sannyasis, perhaps, real? Masselos’s study makes two points that need to be amplified here. One is the pervasive involvement of the colonial state. The other is the participation of Hindus, including members of Maharashtrian castes; those who found it meaningful to claim the occasion under a Sanskritic-sounding name called it Imam Jayanti.

There is no question that 19^th-century administrators saw the festival as a grave security threat. And, not to belabour the point, they also appear to have perceived a threat to two received categories within the handbook of statecraft – those terms of colonial art, Hindu and Muslim. Particularly in the aftermath of the Revolt of 1857, expressions of British anxiety over their hold on Bombay stressed the vast disparity in the manpower ‘the Muslims,’ as a collective, could demonstrably field in the city over the government’s police and military forces. It is no great stretch to theorise that for the functionaries of the Raj, the large total of Muslims among the resident population was a problem that had epistemological dimensions. Where and when and in what quantities were Muslims visible, and enumerable? Where and when would it be desirable that such numbers be visible? Perhaps yet more unsettling was a question of recognition: How could the Muslims be sorted out from the Hindus? Compounding the challenge was the legal doctrine of custom, with a peculiar push-pull dynamic at its core. It codified sets of visible and legible criteria according to which authentic religious expression could be recognised; acts or utterances seen to meet the criteria were judged matters of custom – the purview of community leaders (Muslim, Hindu, and so on) – as opposed to matters of public concern, like politics. In order to evade control by agencies of the colonial state, in other words, instantiations of religious custom had first to pass scrutiny by the state.

With reference specifically to Muharram, concerns of governmentality like these can be read into the repeated moves police officials made to define and restrict the routes on which tabuts were permitted to travel. The general idea was to close off side-alleys or residential ‘gullies’ and direct festival traffic on the broad public roads. (Out of the premodern ‘native quarter,’ one might say, and onto the rationalised modern grid, where disorder could be contained and managed.) But repeatedly, the prohibitions on use of neighbourhood lanes would be met by community leaders petitioning in the name of ‘immemorial custom.’ And thus, just as in the case of the alternative Iranian rite, the British found themselves party to a dialectic that actually defined what was, and what was not, Islamic custom: how and when and where proper Islam could manifest itself in the heart of the city. (Or on its ‘face’ or ‘skin,’ to go with my earlier figuration of the city as body.)

Given all this, is it not surprising to learn the British authorities took it upon themselves to purge the festival of its Hindu participants. In 1839, they declared a formal prohibition of Hindu groups processing with their own tabuts. And yet – it is surely worth noting – in the name of custom, there were some Muslim leaders who petitioned against this policy, too. The reaction must have surprised colonial officials, inasmuch as the spirit of the proscription was to re-inscribe Muharram within the domain of custom and thus, by definition, the concern of one religious tradition (and not the other). Notwithstanding this ruling, Hindu tabuts were observed being paraded and immersed into the 1860s and subsequently, men from Hindu communities continued to march with Muharram parties ‘even during the nineties when, with Tilak’s influence, there was some slight decline in Hindu participation’ (Masselos 1982: 57).

Masselos leaves it to others to draw the connections with Tilak’s project of enlisting Ganesh in a Hindu answer to Muharram. His study wraps up by mapping a convergence of the British perspective, at the turn of the 20^th century, with an assortment of community interests that were working to redefine norms in the name of religious custom. At the neighbourhood level, patronage of tabuts and their attendant tolis had shifted away from the old Konkani elite to rowdier, more plebeian rivals with discrete local power bases. Those Muslim spokesmen whom the British took seriously increasingly advocated interior movements as the authentically Islamic approach on the occasion of Muharram. Interior, as in an introspective or contemplative turn. But interior, too, in terms of space – away from the street and towards the spaces of the mosque and private home. By 1910, the police had at length succeeded in banning all Hindu involvement in the festival. In 1913, under the same police commissioner, a tightened regulatory regime that cracked down on neighbourhood committees had the effect of banishing tabut processions from the city altogether. Between the policing of public space on the British side and the policing of orthodoxy among the city’s Muslim elites, the famed Muharram of 19^th-century Bombay had effectively become confined to purdah.

The Tilak Argument: Politics Disguised as Religion?

Let us return to 1894 and the Ganapati Utsav. Where the case for Tilak as the architect of the festival as an urban, public form is concerned, a book written by Richard Cashman in 1975, The Myth of the Lokamanya: Tilak and Mass Politics in Maharashtra, stands as a touchstone. Cashman’s argument is strong and well-supported by research in textual sources in English and Marathi, most notably the words of Tilak and his associates as published in Kesari, his Marathi-language newspaper (the English-language organ of the same nationalist syndicate was called the Mahratta). But in this work, Tilak’s work with Ganapati Utsav comes under scrutiny as just one element within a larger study of political leadership. And Cashman not only frames his subject’s interest in the festival as instrumentalist, the author’s own analysis is focused on an evaluation of the project’s utility as a vehicle for political mobilisation – and the dissemination of Tilak’s agitations – against the British Raj. As such, Cashman (1975: 97) concludes, the festival was efficacious in the period 1894–1910 only to the extent that it helped its founder carry on the fight to other arenas: ‘The man Tilak used the god Ganesh as a stepping-stone to a bright political career.’

Use the god Tilak surely did, and yet the god was hardly used up. In her Performative Politics and the Cultures of Hinduism, Raminder Kaur does not rebut Cashman’s argument about Tilak’s agency so much as contextualise and complicate it. This 2003 monograph is a study of the Ganesh Festival as celebrated in Maharashtra from the 1890s to the present, and as a work of religious studies it draws, commendably, on multiple methodologies. The author is particularly deft in her analyses of Ganesh pandals as visual-cultural artefacts, and of the performance forms that surround them as complex spectacles whose effect is made not only visually but across the sensorium. As with Cashman, Kaur’s historical exposition centres on Poona, as opposed to Bombay. Her picture situates Tilak not as a mastermind so much as a cheerleader, who endorses and publicises the efforts of more than one community leader experimenting in the early 1890s with the organisational possibilities of the sarvajanik mandal (‘public committee’) and the spectacular properties of the pandal form.

Not even Cashman claims that the Ganapati Utsav of 1894 sprang forth from a vacuum. In a sort of reverse image of the fate of Muharram, the transformation of the Ganesh Festival in the first years of the 1890s can be seen as an outward movement into public space from behind private gates and courtyard walls. Up until that juncture (and well into it), it was chiefly Brahmins who dedicated the fourth day of the bright fortnight of Bhadra to Ganesh. The ritual attentions that centred on a temporary effigy into which the god’s presence had been invited might continue for a number of days set by the family, by the elders within the subcaste, or by strictly local convention. In fact, immersion the very next day was common, immersion being the standard method shared by Hindu communities – then as now – of decommissioning a consecrated object.

The point is not that there were set rules in place requiring a top-down intervention, but rather that in this period norms were loose. Following his arrival on chaturthi, the length of time Ganesh would remain on the earth, and the kind of spaces to be toured on his visit, were not synchronised across small-scale localities or kinship units. A report cited by Cashman (1975: 76) indicates that Ganapati murtis were already being paraded around Bombay neighbourhoods in the 1880s. But let us think back, in this connection, to the British concern with regulating Muharram traffic. There is a distinction to note between residential lanes – the narrow passages that permeate discrete neighbourhoods – and the public streets that delineate and bound them against each other. And the move from relatively more intimate to relatively more public and open spaces maps onto an expansion of social collectivity, from kinship- and caste-based communities to broader formations that encompass them.

This is where the Ganesh icon’s precolonial history comes in. The deity’s two-faced valences as a signifier of Maharashtrian caste identities – on the one hand, the region’s Brahmin communities; on the other, those peasant castes that identified themselves with the martial legacy of the Maratha Confederation – can be traced back to that polity’s glory days. In the 18^th century, much of western and central India had come under the sway of the Peshwas, the Poona-based rulers whose dynastic lineage was Brahmin but whose armies were principally recruited from agrarian caste groups of Maharashtra. And the lineage god of the Peshwas, whose image they displayed as a mark of their sovereignty, was Ganesh.^[3]

Consolidating the seasonal worship of Ganesh around a new orthopraxy, Tilak and his associates proposed, would first rally together Brahmin communities, who were an educated vanguard conscious of the not-so-distant achievements of the Peshwa regime, not to mention the recognised authorities in matters of ritual. The Maratha peasant castes would then follow their lead, recovering and reflecting on the glorious past of inter-caste solidarity even as the synchronised pattern of welcoming, felicitating, and sending Ganesh on his way would project collective hopes into the future. Father Ganapati, go before us, as the immersion-day chant goes, Come back to us soon next year. Why not? An aspirational modernity – a resurgent nation to succeed the British – inspired by the very regime the British had themselves deposed following the Anglo-Maratha Wars of not a hundred years before?

Pushing the Boundaries

Something else pivotal about 1894 must be noted. The year before had been marked in Bombay by a signal event, a Hindu-Muslim riot entirely unprecedented in scale. On a Friday, 11 August (a date that fell neatly between Muharram and Ganesh Chaturthi that year), rioting broke out following noonday prayers at a mosque. Within hours, the police found themselves obliged to call on the army for help. But if the government had reinforcements, so too did the mobs on the Hindu side; the next day the fighting escalated with the involvement of Maharashtrian labourers employed in the city’s burgeoning textile sector. Violence continued sporadically through the following week, by the end of which eighty people had been killed, 530 wounded, and 1,500 arrested for rioting, robbery, and trespass (Kidambi 2007: 118; Menon 2010: 64). What is germane to the history of Ganapati Utsav is that the event cast mastery of the downtown streets as a question of contestation between distinct native populations under the sign of religion. And not only that, but recognised within the Hindu fold as its champions were the textile workers, men from the peasant castes of the hinterland. A Gujarati-language pamphlet that circulated in the afflicted neighbourhoods later the same month praised them as true patriots, heirs of Maratha glory:

Ye be relics of Peshwai, fortunate descendants of Shri Shivaji;

Ye be no women, aye, ye are men, Bravo.

God grant you success, night and day may He keep ye happy;

Fight again in your country’s good, Bravo. (Masselos 2007: 110)

It should be reiterated that this was the first time that battle lines had been drawn across the territory of the city with one set of castes and localities coalescing under the Muslim banner, and the other, the Hindu. The energetically stated grievance that the British had seemed conciliatory to the Muslim side – in the aftermath of the riot, and even before it – animated Tilak’s strategy of winning urban space for the Hindus with a grand annual celebration (Hansen 2001: 29–30).

I suggest that the inaugural festivals should be seen together with the riot as concomitant occasions of mass mobilisation that contributed to the consolidation of rival Hindu and Muslim ‘camps’ in this period. This framing compels a reassessment of three important points. The first – revisited from my discussion of the Tilak thesis – concerns the festival’s entrenchment of an orthopraxy. It would be a specifically Maharashtrian (and Brahmin-dominated) tradition that was to provide Hinduism, in the Bombay context, with its modern public face.

The second point to review is that of the entanglement of the colonial state as the custodian of public space. In that capacity, government officials had to arbitrate questions of which aspects of Indian life were properly public and which ones private, and thus the purview of native custom. As I proposed earlier when considering the example of Muharram, the management of a great urban festival – the exhibition of religion before a public composed of diverse sectarian, caste, regional, national, and gender identities – necessarily involved the Raj, against its stated principles, in the adjudication of what was authentic religious expression and what was not. (Examples of what was not: elements deemed inflammatory, barbarous, or seditious.)^[4] Much attention has been given Tilak as the great rebel, who shrewdly saw in the Utsav a chance to stir up nationalism under the sahibs’ noses, as it were, encoded as religion. But let us state it plainly that not so much has been said about how much the practical success of the innovation lay in collaboration with the city’s administrators.

The third point is to reiterate the opposition that the early Ganesh Festivals worked to define, not against the British overlords, but the Muslim rivals, neighbours pledged to a different conception of God. If in its performance format Muharram was an inspiration, by the same token it was the competition – and competition for the involvement of precisely the sort of Maharashtrian proletarians who had been praised as street-fighting heroes in the Hindu cause. A popular theme in the songs performed by the melas of the 1894 festival was the duty of true Marathas to reject the temptation of Muharram. Cashman (1975: 78) cites some exemplary verses:

How have you forgotten Ganapati, Shiva and Maruti?

What have you gained by worshipping the tabuts? ​

What boon has Allah conferred upon you

That you have become Mussalmans today?

Do not be friendly to a religion which is alien

Do not give up your religion and be fallen.

In my reading, this is not so much a statement of a pre-existing antagonism, to be taken for granted, as a bid to consolidate such an antagonism in a rather fluid situation.^[5] In this connection, some of the pandal images Kaur (2005: 44–46) has recovered from this period present a dramatic contrast with the roly-poly Ganesh forms favoured in devotional worship at the present day. In the historical images, the god’s elephant head is juxtaposed with a lean, muscular body styled like that of a wrestler. Captured in dynamic poses with an assortment of weapons, this athletic champion triumphs over supine demon figures – enemies of dharma.

Two Questions

Let me wrap up this discussion of the research of others by formulating some original questions. Mining the textual record, historians have reviewed the statements, reports, and arguments of Tilak, associated political figures, journalists, and colonial officials. The archives have also yielded political pamphlets and even the verses of festival songs in Marathi and Gujarati – with the qualifier that material of this sort tends to have a Brahmin provenance (see Rege 2000: 202). Innovatively, Kaur has brought in visual and material culture of the period and, in the present-day frame, some ethnographic testimony. I am myself an ethnographer of religion. Given an opportunity to return to Bombay (or to Mumbai, at any rate) to advance with this inquiry, where might I look for insights that could fill out our portrait of this newly urbanised and public god?

I would propose two questions to guide fieldwork among Ganesh Festival celebrants from non-Brahmin, subaltern backgrounds. These interlocutors might not all come from Maharashtrian communities, but those who do would be likely to locate themselves within the super-caste category of Maratha, a proposition which (with the benefit of over a century’s worth of mobilisation by political interests like the Shiv Sena Party) has become even more of a quasi-national, nativist construct than it was in Tilak’s day. So the first of my questions would have to do with visualising the shift in perceptions of space for the labour migrants of the 1890s – from rural to urban. Labour migration is of course a defining feature of subaltern life in the city to this day, and insofar as it is reasonable to theorise continuities in spatial praxis across four generations, it would make sense to work with urban subalterns who retain ties with their villages of origin.

In this paper, I have already called some attention to the problem of urban public space and the publicising of religion from the point of view of the colonial government. What remains to be asked, however, is how the transposition of ritual norms from the village to the urban context would have been experienced by the rank-and-file members of the pandals and melas, especially non-Brahmin groups. And particularly suggestive in this connection are some ambiguities encompassed by the god Ganesh himself. As Vinayak, Lord of Obstacles, he is not solely (as commonly represented) a remover of obstacles. Rather, the Lord governs obstacles; he can shift them from one path and into another (Krishan 1991–92). As the custodian of the threshold, he marks and advances boundaries. In Maharashtra, as I have already observed, his cult was historically dominated by Brahmins and enjoyed the official patronage of the Brahmin Peshwas until their demise in the early 19^th century. But in the sort of village Hinduism observed among peasant castes, it could be said that Ganesh was but one among a multitude of divine personages traditionally tasked with keeping watch over the margins. Or perhaps a better way to convey the local-to-translocal, village-to-city dynamic of my interest would be to cast the elephant-headed god as a Brahmin-friendly ‘brand’ under whose label cults dedicated to local village guardians could be assimilated within a new form of metropolitan Hinduism.

The point is that there is a function that a whole class of (generally secondary) Hindu deities perform as organisers of space, enforcers of boundaries between the civilised, habitable space of the home community and enemy territory or wilderness areas. In fact, one of the best-known examples of an Indian city whose layout is configured according to a Hindu sacred geography is none other than Poona, capital of the Peshwas: its perimeter is marked by a ring of temples seating guardians at the cardinal and ordinal directions. These guardian deities are known as the Ashtavinayak, the Eight Vinayaks, each one a form of Ganesh himself. Varanasi, paramount among Hindu holy cities, has a similar cordon of guardian Ganapatis.

By contrast Bombay, whose development was directed for most of its comparatively short history by Britishers, has no such arrangement. But along with the multitude of villages it has swallowed in its process of expansion it has absorbed their formerly rural guardian shrines. And the more recent labour migrants who settle in the city, generally in unauthorised housing, consecrate their jury-rigged new neighbourhoods with tutelary deities of their own (see Elison 2018). Siddhivinayak is the name of the city’s most famous Ganesh temple – indeed, possibly its most famous Hindu temple – and in this connection it bears note that this fairly recent site is a satellite temple, an indexical extension (although much grander in form and dimensions) of Siddhatek, one of the ring temples of Poona. Thus one of the most hallowed powers to safeguard the Marathi heartland has also colonised a corner of Bombay. (What is more, the multitiered structure stands sentinel, as Rachel Dwyer [2015: 263] points out, over one of the two major motor routes into the city from the mainland.) Here I see an opening to an ethnographic intervention. Does a similar territorial logic obtain among festival celebrants as they extend the boundaries of Ganesh’s domain down the streets all the way to the sea?

If my first question is aimed at understanding how the god reconfigures urban space, the second question seeks to get at how it feels to participate in this process as a human subject. Here I would propose to centre on matters of affect, corporeality, and gender.^[6] And in so doing, I would like to recall one of the images with which I began this essay, the 19^th-century construct of ‘urban metabolism,’ the modern city as organism. On the tenth day of the festival, one of the things being exhibited on the thoroughfare was (and, of course, is) the Ganesh murti. It is an emblem of territorial sovereignty like a national flag, although in the colonial context it is distinguished by a crucial advantage: a flag is a political symbol, its display forbidden by the Raj; a murti is an artefact of religious custom, whose free expression the British are sworn to protect. But simultaneously there is another thing being exhibited. This is the juggernaut’s engine: manpower. Massed beneath the idols are bodies in vigorous motion. Historically (up to the 1980s, in fact), in the persons of the Maratha-majority textile workers, those were the muscles that powered the industrial city. Literally bearing the god aloft, they enact their service to Ganesh. In return his blessing is manifested indexically, in the vitality in their limbs and the drive in their hearts.

I have myself written elsewhere of the masculine and carnivalesque atmosphere of the contemporary Ganapati Utsav as a spectatorial experience. Something of the feeling is channelled – mediated through Bollywood glitz – in the 2012 Hindi film Agneepath. The musical number ‘Deva Shree Ganesha’ presents a vivid depiction of the idol’s traversal of gendered spaces, moving from the chawl, the working-class tenement in whose female-thronged courtyard the pandal has been installed, into the crowds of ecstatic men on the streets and ultimately to the sea, where the face-off between the hero and a rival gangster sharpens the god’s symbolism as a path-clearer for his votaries.

At the turn of the 20^th century, when the citywide ratio of approximately 1.5 men to every woman was even more skewed in working-class neighbourhoods, colonial officials identified the masculine street culture of such areas as a zone of concern (Kosambi 1986: 57). And along with liquor shops and brothels as inducements to ‘license’ – or the release of libidinal energies – there was a target that seems incongruous, at first glance: gyms. Neighbourhood spots where men assembled to train together, typically centred on wrestling pits, were popular sites of recreation and sociality for the working poor. Muslims called them taleems and Hindus akharas, and they were closely affiliated with the toli and mela teams that escorted the floats in the big urban festivals. In fact, in 1910, when the government set out to manage a Ganapati Utsav that would be as free as possible of elements deemed antisocial or seditious, the task was accomplished with a police crackdown on the city’s akharas (Kidambi 2007: 181–182, 193–194; compare Freitag 1992: 211).

Were the sahibs truly concerned about gyms as dens of proletarian license and vice? Or was the problem the opposite – discipline and health? My sense is that the perceived danger resided in a balance of these: what Ganapati Utsav (and Muharram before it) made possible was the directed channelling and public enactment of masculine vigour.^[7]

During his annual visit, Lord Ganesh is understood to spread health and prosperity across his domain, and the ideas of these blessings that circulate in contemporary popular discourse are generally quite bourgeois and anodyne. But I would want to ask whether the spell of effervescence Ganesh visits on his worshippers on the final day is understood by those worshippers as something more radical. One formulation I have in mind is possession: an intimate and volatile blessing, an infusion of special energy likely encountered by people already subjected to possession by the less high-profile guardian deities of more local cults.^[8] Another operative term might be mast, which is the sort of heightened affect that animates musical virtuosos, mystics, and others with a propensity for stepping out of the normative subject position. Pertinent here, perhaps, is that when the Hindi word mast – often spelled ‘must’ or ‘musth’ – appears in English text it does so in a specialised context: it denotes a periodical condition of high excitement affecting male elephants (see Orwell 1950).

What sorts of connections might ethnographic work reveal between physical culture, subaltern constructions of corporeal and affective states, and festival participation in the present day? And what light would then be shed on the colonial scenario, when the emergence of this vigorous guardian god in the urban crucible must have challenged more than one British official to reach for his schoolbooks, and revisit a classical conundrum. Quis custodiet ipsos custodes: Who guards the guardian?