A Pre-Partitioned City? Anti-Colonial and Communal Mohallas in Inter-War Delhi

Abstract This paper explores the mohallas of Delhi, sub-communities within the city, and asks whether Delhi was pre-partitioning before August 1947. It suggests that the mohalla was a site of political mobilisation that was systematically used by the Congress from the Civil Disobedience movement of 1930 onwards. During the early 1940s, communal voluntary associations like the Muslim League’s National Guard and the Rashtriya Swayamsevak Sangh attempted to establish representatives and training practices within mohallas. The paper concludes that the mohalla provided a space and a scale at which to view communal violence afresh, as one of the many ‘spaces before Partition’ that were reshaping (in) the 1940s.


Introduction: The spring of 1947
The main difficulty in putting down these present communal disturbances, as compared with communal disturbances in the past, is that they are linked to no specific occasion. In the past one knew that certain religious festivals were likely to result in communal outbursts but once these outbursts had been got under control, conditions usually rapidly returned to normal and remained so until some other religious festival. Now the outbreaks of communal disturbances are connected with no religious festival but are predominantly political and there is no saying how or when they will ultimately be stopped. They can be temporarily suppressed by force but that is no cure. 1 On 28 April 1947, Delhi's chief commissioner, William Christie, wrote a letter to the home secretary of the Government of India, A.E. Porter, reflecting on the violence within the capital's old city over the previous five weeks (as quoted above). Riots had been sparked by the Muslim League-sponsored 'Pakistan Day' on 23 March. The rioting subsided, but stabbings and street fights refused to die down over the following month. In the seven weeks to 12 May, the end of this particular band of violence, 29 people had been killed and seventy injured. This was in spite of the charging of 267 people under Criminal Procedure Code (CrPC) section 144 (unlawful assemblage), 618 under Indian Penal Codes (IPC) 107 (breach of the peace) and 151 (prevention of a crime), and 67 under other IPC sections.
Attempting to account for the violence which had been rocking the imperial capital, Christie drew upon both historical and temporal explanations. In the past (an historical period), 'communal disturbances' were linked to festivals (temporal recurrences) which punctured the 'normal'. In the present, such disturbances had been delinked from the cyclical temporality of religion and linked to the wild, open temporality of the 'political' for which there were no answers to the question of 'how' or 'when' the violence would stop. Force was the only answer, although police arrests and prosecutions, like the 952 charges under the CrPC and IPC, provoked as well as stemmed this violence.
Such historical and temporal explanations are borne out by the archive: of interwar tensions between communities who increasingly identified along religious lines; of the acceleration of communal violence during the latter war years; and of increasingly organised and martial violence in the years and months before Partition. 2 But within Christie's explanation lies a less regularly acknowledged geography and spatiality. For 'specific occasion' also read 'specific place'; for 'religious festival' also read 'religious procession'; for 'how and when' also read 'where' this violence might be stopped. Communal geographies did not always surface in the recollections of subaltern violence and their retrospective explanation by officials (what Guha termed their 'secondary discourse'), nor have they featured extensively in academic interpretations of that violence (our 'tertiary discourses'). 3 But policy documents, police plans, Criminal Investigation Department (CID) reports, newspaper articles and nationalist party memos ('primary discourses') repeatedly dwelled on the challenges of communal geographies (its locations, places, landscapes and environments) and communal spatialities (where they worked and where they did not, how to move people to and from sites of violence and community organisation, how to encourage them to stay, how to get them to train, how to make people die, and how to make them live).
One approach to these communal geographies would be to look synchronically across the late colonial period and to explore community relationships across a spectrum of sites and spaces: occasional sites of problematisation and controversy; routes of processions and marches; the way in which homes became politicised; and how location was imagined through the lens of religion and nationalism. The alternative approach, adopted here, is to look diachronically at a particular type of space through time, to see how it was formed by passing through various political periods, discursive regimes and material forms. The attempt here will be a spatial genealogical one, in the sense that it hopes to return to the present of the spring of 1947 through an historical excursion into the spaces that made it. 4 The form of space I will trace is that of the 'mohalla', the often walled and sometimes gated urban communities that made up much of the fabric of Old Delhi and its suburbs at the time. 5 The core question I will ask is whether Delhi's mohallas came to function as spaces for the production of communal segregation of the city and, therefore, whether they present a site through which to assess if Delhi was already 'pre-partitioning' well before August 1947. Keeping the emerging national imagination of 'two nations' in focus is an important reminder that all spaces are relational and dependent upon not only their neighbouring spaces, but also upon the small and larger scales that situate them; the study of mohallas below will acknowledge their vital bridging function between homes, cities and nations.
This analysis of mohallas is also marked out by what it is not. It is not, as the many brilliant works in the review below are, a social history of the composition of mohallas, nor is it an attempt to depict the lives of mohallas, or even of their formal political institutions. Rather, it is an attempt to explore the historical geographies of mohalla relationships between spaces, powers and bodies; the way the city became a machine for capillary circulations of increasingly communal governmentalities that encouraged subjects to view their body politic, and their own bodies, as agents of communal division. As with the broader politics of communalism, I will argue that the mohalla cannot be comprehended without understanding pre-colonial social formations, the material city, the computations of the colonial state, and the crafting of political public-private connections by the Indian National Congress. The mohalla presents an analytical device through which to negotiate this swirling mass of people, governmentalities and forces, not because it represented a separate world, but because it operated as a trans-scalar space. 6 As one of the earliest and still most vitally relevant interpretations of Bombay's (now Mumbai) 'mohollas' put it: … what is significant is how the field operates, the kind of forces which become manifest within it, and the processes of change that occur as a result. In this sense, the moholla symbolizes a greater-not a lesser-world. 7 Geographies of anti-colonialism, communalism and mohallas Arguing the case for studying mohallas does not require the construction of an a-spatial straw-man or straw-woman historian. Whilst not always at the forefront of historical analyses of anti-colonialism or communalism, the existing literature is rich with geographical analyses and inspirations. These include, just to name a few, studies of: the gendered cartographing of Indian nationalism and decolonisation; 8 the inner and outer sovereign realms of nationalism; 9 the (im)mobility of subaltern subjects and subaltern geographies; 10 sites commemorating the successes and failures of nonviolence; 11 explorations of the urban lives of nationalism; 12 territoriality and communal claims; 13 spaces of friendship and exile; 14 and the role of militant organisations in crafting communal, public spaces. 15 Mohallas feature rarely in these works, although many of their broader themes flit into the dislocated writings on mohollas, pols and paras which constitute a minor literature within urban histories of colonial India. A review of these works, focusing on those addressing Delhi, will provide a way into the history of the city and into mohalla debates, regarding their morphology, sociology and power politics. The key question for this literature, and for this paper, is how spatially and socially differentiated were the mohallas? That is, were they always a site of segregation, or a place for cosmopolitan coming together?
In terms of morphology, in 1986, Samuel Noe suggested that Delhi's mohallas were a result of the design of Shah Jahan's city around the central palace and two main axes (Chandni Chowk and Faiz Bazar). 16 Courtiers built their mansions in the remaining spaces within the city walls around which quarters organically emerged, with occupational speciality mohallas emerging later. For Noe, this resulted in ' … substantial segregation by caste, religion, family and place of origin'. 17 Noe cited Stephen Blake's doctoral thesis, which formed the basis of the latter's 1991 study of Shahjahanabad between 1639 and 1739. 18 Blake suggested that the city had been planned cognisant of ancient Hindu texts, which divided cities into gramas or padas, which were divided on the basis of caste, craft, profession or tribe. 19 The division of Mughal cities into quarters, or mahallahs, 20 administered by a kotwal (city magistrate), was explained by separating the two processes described by Noe. One type of mohalla was elite, in which a prince or noble household was built around by his, or occasionally her, followers. The second mohalla type focused on caste or craft, being headed by a chaudhuri (chief) or panchayat (caste council). These served to settle disputes, to administer status and tax, and to monitor the spatial integrity of the mohallas (its services, wells, religious buildings, information and gossip). Anticipating the changes in later periods, Blake suggested that after the British took up residency in the city from 1803, and especially after the persecution of Mughal elites and post-'Mutiny' urban demolitions from 1858 onwards, elite mohallas transitioned to craft hubs. 21 Finding the mohallas useful ways to collect tax and information, the British encouraged the walling and gating of these previously more open communities, calcifying in space the organic and evolving communities of the city (as the census would do to the population through its categories and tables).
These (social) morphological readings have been challenged by Shama Mitra Chenoy, who accused Blake of Weberian functionalism, of ignoring localities beyond the haveli (elite household) mohallas, of imagining spatial hierarchies that did not exist, and of overplaying the caste organisations of mohallas. Speaking of Delhi, she suggested: It was a well-planned city, inhabited in the seventeenth century by a population of which different communities had over a period of time, both by the natural process of evolution and syncretism, and a conscious policy adopted by the Mughal state, overcome, to a large extent, fear and distrust of each other. This, in turn, led to more inter-community involvement at socio-cultural and economic levels. 22 This interpretation had also been supported by Jamal Malik's analysis of the inscriptions and annotations of a Mughal map of Delhi as a means of exploring its social morphologies. 23 This suggested that before the colonial reshaping of Delhi, the city mostly did not separate the social means of reproduction from the economic means of production, with people living where they worked ('mahallas' were ascribed to dhobis/ washermen, scholars and shepherds), but also with ethnic or social communities (a Punjabi mahalla). The larger the area the more mixed, although mohallas with religious institutions at the core seemed to be more inwards looking. The conclusion, however, was that social and economic interactions across moahallas were more important than religious coherence within. The dissensus, therefore, is that mohallas could offer the basis for social segregation, but historically did not, necessarily, do so.
This dissensus must play out geographically too. Inland or 'up-stream' large towns and cities developed different types of mohallas to the large coastal settlements that urbanised under the influence of industry, commerce and migration. 24 But while the material form, their boundaries and the histories of these mohallas may have differed, their politicisation allows certain types of cross-reading. The inspiration for this paper's genealogical approach to Delhi's mohallas is the path-breaking work of Jim Masselos on colonial Bombay. His 1976 paper explored 'power in the Bombay "moholla"' as a way of considering the 'Indian urban Muslim', but also of refusing frames of understanding that considered scale as nested hierarchies of locality, province and nation. 25 Masselos instead insisted on viewing the moholla as a field of operation: ' … to view the local as national and in so doing to analyse the diversity of kinds of power operating within such fields'. 26 The early twentieth-century Bombay Muslim 'community' was shown to be divided along lines of doctrine, social group, language and trade. 'Mohollas' hardened these groupings, with some having up to 70 percent cohesion in census statistics: 'social exclusion was repeated in geographic segregation'. 27 A supportive community, perhaps, but here also an oppressive one, threatening innovation, organising social activities under the direction of a jamat (elder council) whilst also surveilled by the police, jobbers and influential families. Whilst never, therefore, simply the lowest rung in a nationalist ladder, increasing attempts from 1911 by the police to license and regulate mohalla activity, especially during religious processions, forced previously divided factions into unity against British influence. Yet on questions of affiliation to the Muslim League, the moholla could again fracture. The challenge Masselos set was to appreciate the moholla as a varied field of operations that reacted to and impacted upon the city and the nation. This challenge has not been well met, although mohallas and their equivalents have occasionally appeared in broader works on the modern Indian colonial city. Ratna Naidu's study of Hyderabad is one of the few to consider the morphology of mohallas, located beyond the winding primary and secondary streets of the city, but penetrated by tertiary streets which breach the mohalla interior. 28 Within a mixed city, the scale of the mohalla could be one of social homogeneity, which increased during periods of communal unrest. 29 It was the physical structure of the city that made this communal politics possible, although without a railway, like Delhi, connecting the Muslim structured city to broader commerce, the pace of change was very different to that of the capital. 30 Ian Talbot and Tahir Kamran have recently described the mohallas of colonial Lahore, clustered around famous darwazas (gates) and baghs (gardens). 31 Some were sites of cross-community friendship, forged through poetry readings or shared celebration of religious festivals. Others suggested communal clustering: centres of Muslim literature and medicine around Bhati Gate; Hindu teaching and the centre of the Ramlila festival near Shah Almi Gate; a gurdwara location dispute near Mochi Bagh. Similarly, Ahmedabad's distinctive 'pols' functioned as sites for political campaigning and electioneering, which could also serve as the basis for communal organisation. 32 Concerns about such activity led the police to increasingly intervene, not just in Bombay, but across the United Provinces (UP), as rival communal organisations became increasingly competitive for the pockets of urban poor that the mohallas promised. 33 Anindita Ghosh has also described how the layout of Calcutta (now Kolkata) retained a sense of village-like caste-ordered layouts, surviving as tollahs, tulis and paras within the modern city. 34 The para, like the mohalla, usually contained a few streets that had a residential or trading function in common: 'Thus, people belonged to certain paras by virtue of their social background-by status, trade, or regional origin-but a para could also denote a more heterogeneous settlement tied to a local feature such as a pond (pukur) or bazaar … '. 35 Regardless of origin, the units, as in Bombay and UP, became politicised. As Partha Chatterjee has it: The para became the site of a complex network of institutions catering to the educational, religious, cultural, and indeed moral life of neighbourhood residents. By the turn of the twentieth century, this nationalist cultural project, conceived as a project of self-regulation outside the colonial state's control, would be directed, at the microterritorial level of the urban locality, by a new middle class. 36 Raj Chandavarkar's posthumously published paper has reiterated the dissensus. He insisted that segregation and social control through space was easier to plan than to put into practice, and that academics had tended to falsely homogenise units like the mohalla. 37 He insisted that community togetherness often came about most in times of competition and conflict, and that there were numerous and temporally variable locations of power in the neighbourhood, demanding that we pay attention to period, place and the messy dynamics of community life, and imagination.
In Delhi itself, communal conflict had cyclically increased in the inter-war years after the transfer of the capital from Calcutta in 1911, dipping during periods of Congress-led mass movements. Communal collaboration never again reached the peaks of the early 1920s, when the Hindu reformist Swami Shraddhanand had preached to Muslim worshippers in the Jama Masjid on 4 April 1919, during the heady days of the Khilafat movement. 38  after the formal inauguration of New Delhi in 1931, solidified the sense of there being Hindu and Muslim zones in the city and led to the creation of the 1934 'Communal Riot Scheme', which distributed police pickets and patrols around the city in anticipation of, and then reaction to, a communal disturbance. 39 The police pickets explicitly aimed to protect mohallas where one community was felt to be weak or isolated. The Communal Riot Scheme viewed mohallas demographically, as spatial representations of social divisions. But what if the mohallas were being used to make and politicise these divisions, rather than to represent them? And if this was so, where had this technique, this political technology, come from? During an interview in 2001, Professor Sadiq-ur Rahman Kidwai recalled how close-knit Delhi's 1940s mohalla communities were, how quickly news spread through them, how shops offered credit to friends and families, how women who observed purdah would enjoy religious celebrations on the roofs. 40 But he also hinted at other uses of the physical infrastructure. During nationalist protests, when the local government declared a curfew, the gates to the mohalla could be locked and life could continue behind closed doors. 41 But what if the activities the gates were protecting were not non-violent and anti-colonial, but violent and communal? Can we even separate the two? As Chenoy, Malik and Ghosh confirmed above, it is difficult to use the archive to pronounce mohallas, and their equivalents, as sites of communal coming together or splitting apart. The spatial genealogy below will, instead, diachronically trace the emergence of the mohalla as a political technology of Congress, RSS and Muslim League mobilisation before returning to the spring of 1947.

Congress mohallas?
Every day the satyagrahis who make contraband salt in the city are changing their 'field of operation'. Today they have gone to Subzi Mandi, with Mr Idris as their head. It is believed that the frequent change of places has been attended by an appreciable success in their programme in so far at least as the propaganda work in favour of breaking salt laws is concerned. Enquiries at the Congress office go to show that the number of private homes in which contraband salt is made is daily increasing. 42 On 12 March 1930, Gandhi and a select band of satyagrahis (volunteers disciplined in Gandhi's philosophy of 'truth force') set off from Allahabad on the 'Salt March', reaching the sea shore at Dandi on 5 April. Here they illegally produced salt, triggering the Civil Disobedience movement. Congress' programme of non-cooperation, trialled during the early 1920s, was supplemented with more active law-breaking and public protest. Cities all over India were affected by strikes, processions, pickets, flagraisings and protests against government violence and exploitation. After the first weeks of protest, the challenge was to keep up support, to find new ways of capturing the public's imagination, and of getting people out of their homes and on to the streets. One tactic was temporal: 'National Week', for instance, began with the commemoration of the Rowlatt Disturbances of 1919 on 6 April and ended with the commemoration of the Jallianwala Bagh shootings of 1919 on 13 April. More common tactics, however, were spatial, temporally limitless, and aimed at penetrating the entire city.
On 21 April, the Hindustan Times, quoted above, reported one of the ways in which Congress was doing this. While Masselos had referred to mohallas as the 'field of operation' for a broader politics, here the same term was used to refer to a locality of the city (Subzi Mandi was a vegetable market and factory district to the north-west of the old city) in which protests were focused on a particular day. In this area, salt would be made, other laws broken, songs sung, speeches read and money collected for further Congress work. These activities would take place in streets, halls, markets and squares. But the aim of Congress protests was to change people's lives, and for Gandhi, this essentially meant penetrating the home and the family activities that took place there. 43 Switching location had helped increase salt production in the home, the Hindustan Times reported. But Congress had been building up a political and urban infrastructure for a decade that allowed it to more regularly bridge the politics of the city and the politics of the home. This bridge was the mohalla. To understand the building of this political infrastructure, we must reach back a little further, into the Non-Cooperation movement, and into the geographies of Congress' political imagination.

Constructive Non-Cooperation and the big towns
Delhi had been a national focal point for protests against the passing of the Rowlatt Act in March 1919, which proposed the continuation of emergency powers enacted during World War I, putatively to tackle the threats of terrorism and extremism. 44 Swami Shraddhanand had faced down armed soldiers near the Town Hall during public protests; members of the public had been killed as crowds gathered near the train station. Congress had capitalised on the protests, using them as a first mass experiment in satyagraha. Believing greater control would be required for future movements, Congress reorganised its structure. In December 1920, it was agreed that the party would be directed by a fifteen-member working committee, elected from the 350-member All India Congress Committee which represented provincial and district Congress committees that would, themselves, draw members from towns, villages, wards or mohallas. 45 From August 1920, Gandhi used these networks to launch the Non-Cooperation movement, focused around the boycott of foreign goods, refusal to co-operate with state institutions, and managed public protests. Each provincial or district committee organised its own agenda of protests, and this first Congress mass movement saw the emergence of many of the future stars of the nationalist movement. In Delhi one of the most prominent of these was Mohammad Asaf Ali. He was Delhi born and bred, but had trained in England as a lawyer before returning to practise in his home city. He became politicised by the Khilafat movement, protesting against the treatment of Turkey at the Paris Peace Conference of 1919, but was soon converted by Gandhi's message of non-violence and constructive work. He would work for the next 25 years as a Muslim Congress politician, becoming a member of the Constitutive Assembly before Independence in 1947 and India's first ambassador to the United States after Independence. 46 In 1921, however, Asaf Ali was a young figure trying to make his mark on a competitive political scene. One of the ways he did this was through his 73-page booklet, Constructive Non-Cooperation. 47 The booklet laid out, in five chapters, a plan for a rival but interior state within the state that could prepare Indians for self-government. The chapters concerned a grand panchayat (council), justice, peace and order, education and finance and some speculative conclusions. The scalar sovereignty of the state was accepted (linking the sovereign, the state and the subject), but a parallel institutional structure was proposed (the grand panchayat, sub-panchayats and sevaks, or voluntary workers). The panchayats would allow each locality to be 'self-reliant, vigilant and progressive, [and to] reassume the control of its own affairs and be sovereign within its own spheres'. 48 The prospectus is a richly written and imaginative political utopia, drawing upon the Gandhian rhetoric of 'Truth' and the need to discipline the awakening subaltern masses. 49 But at the heart of the book was a political infrastructure that would link the state and the subject through particular local geographies. Big towns, including Delhi, were said to have complex needs that had to be met by panch representatives, suggested at the ratio of one to every 1,000 voters. Delhi province was estimated to have a population of 450,000, of whom 100,000 were thought to be adults. If the area were divided into 25 divisions, each would have 4,000 voters (based on adult franchise) who would elect their own Arzi Panch representative, who should be a local resident. 50 Asaf Ali's hope was that the ancient tradition of panchayat councils would transfer their aura of authority to these new bodies, and that residents would acknowledge the ' … sovereignty of their Panchayat in all social and to a certain extent professional matters' in order to guarantee peace and order within each unit. 51 This would be enforced through volunteers; for Delhi's wards of 4,000 electors, 50-200 volunteers per unit were recommended. It was suggested these volunteers would wear a khaki uniform and perform twenty drills per month, in a spirit of military discipline, but not of the martial spirit (on this foreshadowing of voluntary armies, see the following sections). Though a national plan, with suggestions of how panchayats could link up to national courts, the degree of control and organisation required meant that this scheme would first be trialled in big towns. 52 Asaf Ali's vision is in part a typical Congress product of its time. It drew on the language of 'Truth' and tradition to provide an explicitly disciplinarian structure for training the masses. But it is also unusual in its urban focus, supplanting Gandhi's idolised village with the city as the training site of anti-colonial nationalism. This in part drew upon Asaf Ali's emerging influence in Delhi's political scene, but his visions of the organised city also fed directly into the organisation of Delhi during the Non-Cooperation movement itself. On 19 November 1921, a CID report detailed how Asaf Ali had been laying out plans for how Delhi would be organised: The city was divided into twelve circles. For each circle was a paid staff of one leader, two advisors and ten volunteers. These would take a census of the people and organise the spinning of cloth. This would take up to three or four weeks. Each area would then refuse to pay any taxes, would ignore the police, magistrates and other officials, but would establish their own police stations and staff. 53 Such divisions were not new, of course. The census divided Delhi into fifteen urban 'charges', while the municipality was divided into sixteen urban wards (with the addition of Delhi's five rural tehsil districts of Sonepat, Najafgarh, Mehrauli, Faridabad and Ballabgarh, Delhi district was divided into 21 areas, not too distant from Asaf Ali's proposed 25). These units were all greater than, but encompassed, that of the mohalla. While Non-Cooperation was suspended in 1922, the Congress focused on further extending their influence through constructive work. By the time the 1930 Civil Disobedience movement was launched when the 'field of operation' began to be switched between new parts of the city, it could draw on a much finergrained institutional penetration of the city.
A Congress bulletin for Delhi on 17 June 1930 stressed the need to reorganise the very way the city was conceived, including its reliance upon the colonial government for settling civil disputes. Panchayats could be organised for the latter, while 'mohalla clubs' could be established where the 'uninstructed may meet educated peoples for social intercourse. If meetings were organised in a methodological manner in each mohalla, one educated and experienced person could impart "Liberal" education to the whole mohalla in a month'. 54 By 27 August this process was well under way and it would remain a feature of Congress organising in the city throughout the 1930s. In an account summarising the month's activities, Delhi's F.K. Ansari reported to the All India Congress Committee:  Communal mohallas?
The Hindustani Seva Dal (Indian Service Corps), mentioned in Ansari's report, had been formed in 1924 by Congress as a means to increase discipline and efficiency (as anticipated by Asaf Ali's panchayat volunteers), and in 1931, it would become the official volunteer wing of Congress. 56 One of the Dal's purposes both within and beyond the period of mass movement was to stymie communal tension. But on the ground, it was also accused of aggravating Hindu-Muslim tensions, while the organisation itself drew upon the sort of mass penetration and youth-focussed drilling targets that were becoming associated with fascist movements across the globe at the time. There were also direct links between Congress and increasingly violent splinter organisations. These included Delhi's Congress Socialist Party, suspected of channelling funds to revolutionary organisations, orthodox Hindu Congress campaigners who supported Right-wing Hindu organisations associated with communal violence, and the central Congress Committee that proved ambivalent regarding targeted violence (train derailments, post-box arson, etc.) during the Quit India movement. Less directly, the Seva Dal model would, in part, inspire the efforts of more communally-minded organisations in their mass recruitment drives. While the Hindustani Seva Dal may have been an inspiration, it had been banned from 1931 to 1937. Congress did, however, revive its ward and mohalla committees during the Quit India protests from 1942, proving the continued value of this political infrastructure.
The 1937 elections had confirmed Congress as the largest party in the provinces and the 1942 protests had proven that, even without widespread Muslim support, Congress still had an unrivalled machinery for mobilising the masses. Communal organisations, many of which had refused to back Congress' position during World War II, had taken the opportunity provided by the mass incarceration of Congress leaders from 1942 to 1945 to increase their membership and support. One way in which they did this was through the city, a component of which was the penetration of the mohalla as a space of recruitment and the making of communal lives. What this suggests is that they both targeted the sort of mohallas featured in literature on nineteenth-century Delhi, conceived to be 'Hindu' or 'Muslim', whilst also contributing to a pre-partitioning of mohallas along communal lines. Brief examples will be given below from two of Delhi's most influential communal organisations, the Muslim National Guard (MNG, also known as the Muslim League National Guard or MLNG) and the Hindu RSS. The MNG had been formed in 1931 with the object of social reform and of disciplining Muslim League volunteers. The central government Intelligence Bureau circulated a note in May 1947 suggesting that the MNG had been reorganised by a sub-committee of the Muslim League in June 1939 to create an organised and trained body of men for the protection of Muslim lives, property and honour. 57 They adopted military uniforms and were reported to be trained in the use of daggers, lathis and swords, and possibly in the throwing of knives and acid. Their national membership in April 1947 was estimated to be 139,057.
In Delhi the success of the MNG had been hampered by rivalries between the new and old guards of the Muslim League, both of which were active in mobilising the local population around particular protests, but not in terms of training a dedicated band of volunteers. In 1942 it was listed as relatively inactive and by May 1943, it could list only 200-350 members in the city; its uniforms were mocked by the chief commissioner in a private report for looking 'more as if they had been borrowed from musical comedy'. 58 The local Muslim League was instructed by the All-India Muslim League to redouble its efforts; in November 1943, it held a meeting at the Anglo-Arabic College to elect the salar (commander) and naib salars (deputy chiefs) for the local Muslim National Guard. 59 By December, 150 volunteers were holding parade practice after evening prayers, but the MNG would struggle to establish itself in the city. As compared with Congress' deep city-wide organisation, by 17 February 1946, the Provincial Muslim League was still issuing calls for fortnightly meetings in each of Delhi's sixteen wards (not its over 130 mohallas). 60 On 29 September, a private meeting of 'Muslim Leaguers' was held in Masjid Khajoor in Kucha Chelan off Faiz Bazar. 61 Attendees were implored to organise MNG units 'in the mohallas' and to call meetings in mosques for this purpose. Office-bearers would then be elected for each mohalla and would organise training in physical exercises, baton fighting and parades.
CID reports showed how this push by the MNG into residential communities within the city was starting to impact upon the training of individuals, and on the proliferating reuse of buildings within and between mohallas for martial purposes. A report on 2 October showed that one Mohammed Yahya, an ex-army man living in Katra Sheikh Chand in Lal Kuan, was training local Muslims in knife-fighting on the roofs of mosques, moving inside after 9 pm. 62 Other figures, some prominent, were said to be overseeing training in mohallas throughout the city, glimpsed through occasional surveillance reports in the archive: the imam of the Jama Masjid, Sayed Ahmed, was said to be helping supervise training in the use of weapons in Masjid Machliwalan; a hat merchant from Chandni Chowk, who lived in Mohalla Bhojla Pahari, in Faiz Bazar ward, had organised 'Muslim Scouts' who were preparing for a rally; in Karol Bagh, the daily CID report for 3 October suggested a prominent Delhi Muslim lawyer had appointed a defence sub-committee in which twelve local 'Muslim Leaguers have been appointed as Commanders of the streets of the locality to arrange for the night watch whenever necessary'; 63 and a CID report on 21 October noted that the female section of the Muslim League was becoming active in Sadar Bazar, appealing to Muslim women not to purchase their requirements from Hindu shopkeepers in their mohallas. 64 The 'Pakistan Day' event on 23 March 1947, with which this paper opened, was described by the CID as being celebrated in 'Muslim mohallas'. 65 The disturbances that followed were, as Chief Commissioner Christie put it, 'temporarily suppressed by force', but the geography of the law and the city made this suppression uneven and ineffective. Section 144 of the CrPC could forbid certain acts, such as gathering in groups, but it only applied to public spaces. If drills or meetings took space in the lanes of a mohalla behind locked gates, it was not clear if the law even applied to these now quasi-private spaces. This dilemma had been made apparent much earlier by the more extensive activities of the RSS.

The Rashtriya Swayamsevak Sangh (RSS)
The RSS' history is much better known than that of the MNG: formed in Nagpur in 1925 as a vehicle for Hindu nationalism in the face of the perceived aggressive Muslim expansionism, it became an India-wide organisation in 1932, combining social reform, disciplined drills and military training. 66 The same May 1947 intelligence memo in which the MNG had been appraised suggested that the RSS was a supposedly open but actually secret organisation. 67 Its volunteers wore uniforms that closely matched those of the Indian army, they trained in lathi, sword, spear, dagger and musket use and had been trained in the use of hand grenades and acidthrowing at a training camp in Peshawar that year. Their national membership in April 1947 was estimated to be 188,571. The RSS had a deeper footprint in Delhi, being able to draw on a more stable history of Hindu Sabha and Hindu Mahasabha organising; RSS officers from Nagpur had stayed at Hindu Mahasabha Bhavan (house) in 1936 and established Delhi's first three RSS branches. 68 Occasional drills were reported in the late 1930s, and in 1942, they were listed as having 900 members, in 1943, 1,500, and by the second half of 1946, some 2,500, against the MNG's 800.
The real strength of the RSS activities in mohallas and similar semi-public spaces were its physical drills. While Gandhi's ashrams had trained bodies in the non-violence of satyagraha, the RSS shakhas (branches) focused on military drills and physical exercise. A report from a deputy superintendent of police on 23 December 1946 charted intense RSS activity in the Hauz Qazi ilaqa (police district) of the city. There, the report suggested, physical training and 'Vaidas' mantras (religious hymns) were used to 'poison the minds of young lads against Muslims'. 69 The training took place between 5.30 and 6.30 pm with instruction being given in private houses, but the exercises took place on the street. In the Subzi Mandi and Kotwali ilaqas, training took place between 6.45 am and 7.45 am in locations often listed as mohallas, or in lanes, baghs (gardens) or mandirs (temples) within particular mohallas. By 2 January 1947, the CID was tracking thirty places in the city in which between 1,160 and 1,305 RSS volunteers collected every morning for training in preparation for what many, including senior administrators, believed would be a communal civil war.
The RSS had studiously respected regulations passed in August 1940 and September 1944 using the wartime Defence of India Act restricting parading in uniform, camps and drills. On 30 September 1946, the Defence of India Act lapsed, leading to an increase in activity by the RSS and MNG, during which the RSS proved itself to have penetrated the city much more deeply than the MNG in anticipation of the regulations lapsing. Within two months, at a private meeting of the RSS on 23 November, the CID reported a call for the number of Muslim residents in a 'vicinity' dominated by Hindus to be enumerated and Hindus in Muslim-majority areas to likewise be counted; for the number of Muslims able to take part in a riot to be assessed; and for all the Hindus of Delhi to be armed. 70 As these reports reached the chief commissioner, alongside those of the increase in drills, the carrying of weapons and the wearing of uniforms, CrPC 144 was declared. Yet this worked to push RSS activity out of the public space into arguably private spaces such as high school grounds, railway colonies, dharamsalas, temples and mohallas. On 11 February 1947, a CID source reported that the sanchalaks (directors) of the nineteen RSS districts had been instructed to contact the youths in their mohallas to induce them to join, while the RSS was aiming to start branches in every mohalla. Three months later, on 5 May 1947, the CID suggested that the RSS' substantial city-wide political organisation had made significant progress towards this aim: The Vibhag Adhikaris (senior organisers) are also assisting Hari Chand (Delhi's RSS leader) in the matter by holding private meetings in every mohalla and by enrolling volunteers for the same. They have also promised to supply them arms, etc. These volunteers hold physical exercise and also learn lathi-fighting, etc., to be prepared for the next struggle. It was clear that by the early 1940s, the administration viewed the city as divided into Hindu and Muslim areas, and that these areas mapped directly on to mohallas in many cases. 73 In April 1944, the Delhi administration was attempting to appoint local representatives to mediate between the local population and the state. These representatives would be called 'mir mohallas', an invented tradition that would attempt to give the administration some voice in units that had already been penetrated by Congress and communal organisations. The mirs did not fully establish themselves before the tumult of the war victory and the granting of Independence, but the debates over appointing the mir mohallas were dominated by calculations over whether a mohalla was 'Muslim' or 'Hindu'. When three Hindu mohallas were identified within the Muslim mohalla-majority locality of Churiwalan (ward IX, to the south-west of the Jama Masjid), the city magistrate made it clear it was a shame they could not be administered alongside other Hindu mohallas under the oversight of a Hindu chaudhuri. 74 By 1939, in a press note, the chief commissioner could openly codify the city into Hindu and Muslim 'sensitive' and 'danger spots'. 75 Assessing if inhabitants imagined their mohallas in such communal terms is, of course, more difficult. When these imaginings enter the archive, it is often in times of crisis and of fraught identity politics. So, for instance, on 20 March 1947, just before the 'Pakistan Day' riots, petitions were received by Vallabhbhai Patel, the Home Member of the Interim Government in New Delhi, from the Daryaganj Mitra Mandal (community organisation). 76 It pointed out that this area, to the east of Faiz Bazar and to the south of the Mughal Palace (known now as the Fort), was a 'Hindu and Sikh locality' in need of protection. Ten days later, the Muslim inhabitants of nearby Urdu Bazar, to the west of Faiz Bazar and south of the Jama Masjid, petitioned the chief commissioner, complaining of armed Sikhs living in neighbouring mohallas brandishing their weapons (not just religious kirpans) in public.
Other petitions came from residents of mohallas throughout the city who felt themselves to be outnumbered by neighbouring mohallas of different religions. On 14 April 1947, a handwritten letter was sent by the residents of Gandhi Gala mohalla which neighboured the Muslim-majority Phatak Habash Khan area and the Fatehpuri Masjid. 77 During disturbances in November 1946, the letter claimed that Muslims from neighbouring areas had beaten 'on the doors of our mohalla' with sticks and it was feared that the Muslim National Guard would gain entry to the mohalla at night, using their uniforms to pose as policemen. The petition concluded by insisting that ' … ours is a small mohalla, surrounded on all sides by Muslims', and requesting a police picket ' … on both ends of the Mohalla to help relieve your stricken servants'. 78 It seems that a communal cartography of the city had taken hold. A petition to the home secretary on 29 April 1947 accused Muslims of being murdered in 'Hindu areas', but not the reverse. 79 It went on to list what the petitioners felt to be Hindu and Muslim areas, taking in prominent roads, shopping areas, temples and mosques, as well as mohallas.
Returning, however, to the points with which this paper opened, it is important to situate these geographical imaginations within the communal temporalities in place during 1946-47. Despite the chief commissioner's fears, the violence of March 1946 was not unstoppable and it did, after seven weeks, subside. Anecdotal evidence and passing comments in interviews remind us that the city returned remarkably quickly to a space that, simply put, had business to do. Just as Chenoy and Malik urged caution when considering the mohalla-based segregation of pre-colonial Delhi, so we should heed Masselos' suggestion that mohallas could unify and contract against outside threats, but also splinter and connect to other interest groups as required.
Mohallas were sites of violence and enmity, but also of friendship and resilience. Memories abound of small acts of cross-communal hospitality and resistance: of a hotel near Connaught Place taking in the family of a worker who was at risk where he lived; and of communist student activists bringing milk to mothers too afraid to leave the house during times of communal unrest. Violence and its appraisal leave archival traces; returns to non-violent norms rarely do.
What the mohalla offers us is an analytical viewpoint from which to observe the operation of broader forces in its particular field (of non-violence, of martial voluntary organisations, of communalism), and a geographical space through which to see how scales were interlinked (the home as a site of civil disobedience, 'Two Nations' within a city, a national capital rocked by individual violence). It is also clear that different political engagements with mohallas drew upon each other: Asaf Ali's model replicated bureaucratic wards; paramilitary groups replicated Congress mohalla committees; and, in an attempt to re-engage this scale, the government created mir mohallas.
The evidence suggests that mohallas were targeted as spaces of mobilisation by Congress and by communal organisations, and that this had fuelled the imagination of a city divided by religion through its mohallas. At the level of imagination it seems that Delhi had begun to 'pre-partition', and that these lines may well have structured Partition violence in the city, and to have directed the official mind regarding where Muslims would be allowed to stay, or not stay, in the city. 80 This diachronic approach to the mohalla can only augment studies of vernacular sources on mohalla life and investigations into heightened disputes over procession routes, shrines, mandirs, graves and mosques, which increasingly wracked the capital in the 1940s. But the hope of this paper is that, in considering the communal geographies of the inter-war years, mohallas might better be considered as a key 'space before partition' and 'a greater-not a lesser-world'. 81