Criminal tribes and concepts of the citizen in India

India's Constitution, inaugurated in 1950, remains the largest document of its type, with one of the world's most extensive and intricate statements of the ‘fundamental rights’ of the citizen. The Constitution, however, also grew out of colonial frameworks: the Government of India Act 1935, which in turn was shaped by the system of ‘dyarchy’ that operated under the prior Government of India Act 1919. The latter not only partitioned the subjects of legislation on the basis of a temporal concept of political ‘responsibility’, it also physically excluded regions in an ‘uncivilized’ state from its jurisdiction. These spatial exclusions contributed to the idea of the ‘margin’ both within India's unchartered interior and its geographical limits, such as the Andamans. Here, the notion of terra nullius (land belonging to no-one) was an idea that established the principle that the expansion of constitutional rights could be related to processes of further colonization.Footnote ¹⁷ By extension, exclusions also took place via the differential application of constitutional rights and formal (that is, nationality) as well as substantive (concerned with legal rights) citizenship values which marked some parts of India's population as putative inhabitants of wild and uncivilized frontiers, both in terms of itineracy/habitation and culture. ‘Tribal’ communities, especially those explicitly marked by law as existing outside the boundaries of social convention such as ‘nomadic’ tribes, therefore challenged the boundaries of constitutional rights in different ways.

Scholars have reflected on a bifurcation within substantive citizenship rights in India, particularly around the relationship between universalist concepts of the citizen, produced in different ways by both liberal Nehruvian and ethno-nationalist Hindu right conceptions, on the one hand,Footnote ¹⁸ and an idea of citizens’ rights that relate to differential treatments (associated with the first law minister B. R. Ambedkar) on the basis of ethnic group or caste, on the other.Footnote ¹⁹ These two strands of rights, however, were not necessarily mutually incompatible.Footnote ²⁰ In effect, ‘differential rights’ led (over time) to affirmative action for certain categories of disadvantaged citizen as outlined within the Directive Principles of State Policy and Fundamental Rights. Yet this grew out of an older colonial notion of differentiation around caste and community privilege in which basic fundamental rights were absent. These older colonial reservations around caste were not based on social justice but on notions of esprit de corps and configured in legal structures that separated out the domain of the family from that of the public sphere, allowing the colonial system to forego social reform. Longer colonial processes of consociationalism in the state therefore shaped these frameworks,Footnote ²¹ and caste-based claims for bureaucratic employment or political representation formed the background.Footnote ²²

Crucial to the working of ‘group’ or ‘differential’ rights just after independence was the idea of the citizen who was marked by modern definitions of ‘caste’ and/or ‘tribe’—identities forged largely through strategies from an earlier era. Some scholars therefore have described the act of ethnic group definition as a historical process that was flexible and reflexive, involving a dialogue between the state and the object of its gaze, transforming the social and political meaning of such ethnicities.Footnote ²³ The category of the criminal tribe, however, is not well described by this framework, since it involved the uneven superimposition of both modern (and early modern) concepts of ‘criminality’ on a broad array of highly differentiated pre-existing ethnicities. In many cases, these identities were already associated with law breaking.Footnote ²⁴ But, as we will see below, the cultures of illegality that such definitions implied hardly lent themselves to a straightforward strategy of community-based rights claims.

Unlike Scheduled Castes or Dalits and, more recently, Backward Castes, groups described as criminal tribes were structurally hindered in negotiating strategies of ethnic mobilization. Any concept of disadvantage they might publicize in this milieu would always imply a heritage of law breaking, and therefore voluntary social marginality. Yet, as we will see in the third section of this article, there were means by which community heritage, including that of law breaking, could be mobilized to demand certain rights, which anticipated the more explicit negotiation of the group-rights framework from the 1970s. That these have not been recognized in most accounts of criminal tribes is, itself, a function of the colonial record. The key documents setting out late colonial definition of caste by region bypass such histories: the 1931 Census detailed the range of caste organizations that were, among other projects, re-representing their status and demography in response to the census commissioner, but in most regions such terms of negotiation for criminal tribes were absent.Footnote ²⁵ Most groups defined as criminal tribes were exotic footnotes in ethnographies of social custom in caste panchayats; curiosities of the official gaze, viewed as custodians of highly internalized and ‘pre-modern’ community practices, these communities were often represented as potentially dangerous marginals. Rather than being marked by political, social, or reformist organizations, groups of related criminal tribes were, in these documents, marked by everyday customs and bodily practices.Footnote ²⁶

There was another important logic to why group rights were unsuited to the criminal tribes. The ‘group’ itself, compared to those defined only by social disadvantage (as was seen in the Scheduled Caste and, latterly, Dalit movements) was highly contingent, spatially uncertain, and often chimerical. Despite apparent detailed ethnographic scrutiny, it was very difficult to arrive at a clear or consistent estimate of the population of criminal tribes in the Bombay Presidency. This uncertainty was due partly to demographic inaccuracies and partly to defining which communities might be described as such, which changed over time and administrative space. The demographic recording of criminal tribes was an example of colonial biopolitics par excellence. Its features are worth tracing for a sense of the powerful faultlines in the demographic experiment itself, which had significant effects on civic mobilization later on. For most of the period between 1930 and 1952, between 21 and 28 ‘tribes’ were notified in the Bombay Presidency, amounting to a population that varied between 875,500 and 1,175,469 (of a total Bombay Province population of between 13 and 15 million), but anywhere between 2 per cent and 7 per cent of this population were directly brought under the provisions of the CTA.Footnote ²⁷ There was therefore a considerable margin for change and redefinition within the structure of marking ‘criminals’ which has implications for how we might view administrators’ own attachment to the notion of ‘hereditary’ criminality.

Further uncertainties arose around movement and space. The communities ‘named’ in Bombay within this loose demography included groups defined as ‘nomadic’, and a dynamic of their movement was that they could not be easily enumerated by district, region, or state—Bauriah, Mang Garudi, Marwar Baori, Oudhia, Pardhi, Sansia, Waddar, and others, including Berads, Bhamptas (Rajput and Takari), Bauria, Bhars, Pasis, Bhils, Chapparbands, Dharalas, Futgudis, Kaikadis, Kammis, Katbus, Kathodis, Kolis, Lamanis, Minas, Ramoshis, and Tadvis—did not necessarily have fixed territorial populations due to movement and sometimes eviction.Footnote ²⁸ Indeed, the names themselves changed rapidly over time and space.Footnote ²⁹ Equally, the number of ‘settlements’ or open prisons for those directly controlled under the CTA varied over time, numbering, for example, 17 in 1928 to 12 in 1946.Footnote ³⁰ Settlement populations ranged from 12,861 in 1924Footnote ³¹ to 5,622 in 1933Footnote ³² and 4,402 in 1946.Footnote ³³ They could spring up with the opening of new settlements, such as that focused on Chharas in Ahmedabad, which added around 1,100 to the total settlement population in one year.Footnote ³⁴ But while there was no fixity to official definitions of so-called hereditary criminality, the stigma of communities once notified as ‘criminal’, or being part of a larger/related ethnic category that were defined as such, was palpable for generations. In many ways, generational continuity in the idea of ‘hereditary’ criminality was therefore often a function of colonial ethnography, especially as officials would often attempt to clearly define certain ‘sections’ of tribes as ‘criminal’ and other sections as ‘non-criminal’.Footnote ³⁵ This uncertainty led to a variation in civic strategies from the 1930s, which were inextricably linked to (and limited by) the contingent and fluctuating colonial project itself.

Moreover, these difficulties in precisely defining the criminal tribes saw their gradual exclusion as a distinct, if ill-defined, category in the development of the group rights framework from the 1910s.Footnote ³⁶ One of the most vexed debates at the successive conferences that aimed to bring India closer to ‘responsible government’ during the interwar period was representation for the ‘depressed classes’. The problem, however, lay in who exactly this term encompassed. As stated by the 1932 Franchise Committee, ‘the term “depressed classes” has been used to cover various classes of people such as criminal and wandering tribes, aboriginal tribes, untouchables and sometimes other backward and economically poor classes’.Footnote ³⁷ Importantly, in at least some of its incarnations, the depressed classes had included the criminal tribes as a separate category.Footnote ³⁸ Although the category itself, as noted above, was time and space contingent, in most provinces the communities who fell within the ‘criminal and wandering tribes’ were defined by their ‘nomadic and thieving habits’.Footnote ³⁹ The 1930s, though, saw a closer alignment of the depressed classes with ritual Untouchability—defined by ‘pollution by touch or approach’.Footnote ⁴⁰ This shift can be traced in large part to B. R. Ambedkar's project of creating Dalits as a separate political minority.Footnote ⁴¹ In order to establish political legibility within the vast array of other ‘depressed’ groups, clear boundaries and definitions needed to be drawn. On 23 October 1928, Ambedkar gave evidence before the Simon Commission, asking that body to grant depressed classes separate representation and reservations in the new Constitution. Ambedkar's estimate of the depressed classes’ population was relatively high, but he made it clear that he did not include criminal tribes.Footnote ⁴² He told the Commission that he considered criminal tribes to be among the most oppressed, alongside Dalits. However, when asked whether they too should be given special representation, he stated, ‘I do not think it would be possible to allow them the privilege of adult suffrage.’Footnote ⁴³ As expressed by Hari Singh Gaur during the debates, ‘depressed classes’ and ‘Untouchables’ became practically synonymous.Footnote ⁴⁴ In 1954, while reflecting on his stance, Ambedkar referred to the ‘mangarudis for instance, who were criminal tribes but were not Untouchables in the technical sense of the word; they were practically outside the pale of society and yet were not untouchables’.Footnote ⁴⁵ It was this tension which ultimately excluded the criminal tribes from the development of group rights in the late colonial period. As Ambedkar put it, political representation needed ‘something more precise, more definite’.Footnote ⁴⁶

As the boundaries of the depressed classes narrowed, no alternative categorization was suggested for the criminal tribes. In 1930, a Depressed Classes and Aboriginal Tribes Committee was presided over by pivotal administrator-expert and writer on ‘vagrancy’ O. H. B. Starte in Bombay. It recommended the use of the term ‘backward classes’, which would include three groups: ‘the depressed classes’, meaning the Untouchables only; ‘the aboriginal and hill tribes’, and ‘other backward classes’, which could include criminal tribes.Footnote ⁴⁷ Tellingly, however, in the first Administration Report by the newly constituted Backward Class Department for 1931–33, there was not a single mention of the criminal tribes.Footnote ⁴⁸ It was from this point, then, that the Bombay government eroded a separate criminal tribe category in favour of their assimilation within more the economically determined ‘backward classes’—a point to which we return in the fourth section.

The difficulties faced by criminal tribes in negotiating the late colonial ethnographic state had particular effects once questions of citizenship rights became more central to Indian politics from the 1930s. The Indian National Congress had set out its ‘fundamental rights of the citizen’ in the Karachi Resolution in 1931. After the elections of 1937, the party was able to articulate and shape some of the resolution's principles at the level of provincial government, in the eight provinces where Congress ministries were formed. But it was not until the establishment of the Constituent Assembly, and the role of Ambedkar within it as first law minister, that a tension between competing citizenship visions, namely, universal (non-marked) and group-based (or ‘differentiated’) rights, fully emerged.Footnote ⁴⁹ This tension was never resolved, both in terms of how far caste and community should be valorized as a basis for rights, but also how categories of ‘marked’ communities are defined and what their specific categories of rights are. Groups and communities who had been defined at some point as criminal tribes faced specific difficulties not only in relation to basic citizenship rights but also those of differentiated rights. Group-based rights contain an inherent tension, whereby individuals must negotiate their position as part of an ascriptive community, yet still commit to the idea of a civic community.Footnote ⁵⁰ This was an irreconcilable difference for criminal tribes, as their group identity itself was in direct contestation of the idea of civic culture. The crux of the matter was that any potential ‘group’ recognition that they might aim to mobilize would automatically affect the citizenship rights of others, since the rights and freedoms of other citizens were putatively threatened by their ‘criminality’, which itself was the key marker of their historic social disadvantage.

Criminal tribes, therefore, were not easily mobilized by either universalistic or differentiated (for example, Ambedkarite or Adivasi) forms of rights claims, nor did they necessarily attempt to do so. As the next section explores, the unstable and highly variable application of the CTA on the ground itself rendered problematic any collective political mobilization against it, in terms of citizenship movements. Yet, as the Act began to be scrutinized in detail, with a view to its repeal—and crucially in the context of emergent citizenship models—its meanings came to be generalized for those who were (or were potentially) subject to its provisions. It was the detailed and contingent discussion of its repeal, then, that came to define how easily the soon to be ex-criminal tribes—the range of very different communities affected by its provisions—might enjoy citizenship rights. Equally, the forms of civic organization and strategies that particular communities mobilized in the criminal tribe settlements from the late 1930s shadowed changing constitutional generalizations and debates about legislative reform. Importantly, these nascent forms of community mobilization laid the foundations for more direct negotiations of citizenship values in the 1950s and, later, group-rights claims for reservations from the 1970s.

Criminal tribes and civic responsibility

The late 1930s to the early 1950s were a period in which the status of criminal tribes was the subject of new debates around political rights, institution building, and regimes of citizenship. On the one hand, the evident discrepancies in the administration and ideology of the CTA led to an official enquiry into its potential reform. On the other, was a pervasive sense that these communities were incapable of civic responsibility and thus not yet ready to enjoy the rights of the citizen. Accompanying these ‘rights’ were what the drafters of the 1931 Karachi Resolution and Part IVA (51A) of the 1950 Constitution described as ‘Fundamental Duties’. In this period, when such rights were being settled and discussed, certain putative categories of citizen were seen as potentially forgoing the enjoyment of such rights, by virtue of being unable to respect the duties and responsibilities that went with them. For example, section (i) of the Fundamental Duties stipulated the protection of ‘public property’ and the duty to ‘abjure violence’. Yet, theft and violent crime, if not seen as ‘hereditary’ behaviours, were still viewed as largely immutable community cultures. Key to reformative efforts in the Bombay Presidency, then, was the question of how to incorporate communities considered inimical to civic culture into an increasingly liberal, democratic polity.

A central outcome of this process was an increasing sense, both by legislators and community activists, of the general predicament faced by criminal tribes as a specific group. The politics of identity transformation within communities thus related to the points where such communities and ‘the state’ interacted around occupation and movement. Principally, this lay partly in the fragile logic of the CTA from an earlier period and, crucially, the debates surrounding it. In the 1910s and 1920s, even as the Act was being amended to implement increasingly draconian controls of settlement and repeat offence,Footnote ⁵¹ those administering it on the ground were in no doubt about the problems of its implementation. Many officials found the idea of group criminality distasteful and had done so for decades.Footnote ⁵² Debate about its local efficacy therefore had a long legacy. The CTA, many realized, was applied in a haphazard, often arbitrary, and reactive way.Footnote ⁵³ For example, in the Kaira district in Bombay Province in 1919, a so-called criminal tribe of Dharalas made up around 40 per cent of the total population, even though only a small proportion of the total community were ever found to be involved in criminal activity, leading to the complaint that ‘the view that Dharalas are, as a class, addicted to crime, appears to be a misreading of the statistics’.Footnote ⁵⁴ Officers in the district also pointed out the means by which the hazri, or roll call, had become a new means by which headmen exercised power and authority over this particular community, even to the extent of extorting money and making threats.Footnote ⁵⁵ This focus on the Dharalas was important in later reform efforts in the late 1930s, as we will see below, linking the failed logic of colonial administration of the CTA to the views of the emergent national leadership.

This problem was not just confined to specific districts where there were relatively large, albeit internally differentiated, populations such as the Dharalas. The policy of ‘notification’ was riddled with inconsistency and uncertainty. For example, in late 1924 in the Satara district of Bombay Province, the local government estimated that despite the collective population of Kaikadis numbering just 726, only one person had ever been convicted. It was therefore not clear why they had been notified under the CTA at all. Satara's deputy superintendent of police was dismayed to find ‘that many Kaikadis of other districts have criminal tendencies is true, but this cannot be said of the Satara Kaikadis’. Therefore, in Satara their registration (which went right back to 1912) was cancelled.Footnote ⁵⁶ If the physical and static identification was not difficult enough, given the logic that particular ‘tribes’ were habitually criminal, further problems existed around conversion and change in status. In January 1919, a group of Mangs in Satara requested exemption from roll call because they had converted to Christianity and were, therefore, theoretically no longer able to take part in caste proceedings.Footnote ⁵⁷

Looked at from the perspective of an entire province, it was clear that the interlaced problems of tribal identification, regional differences, and movement rendered the Act either obsolete or a licence for arbitrary repression in many places. The inspector general of police in Bombay pointed out in 1925 that policies often revolved around the will of individual officers: ‘The degree of activity displayed in different districts in the administration of the Act has depended largely on the interest taken by the District Superintendent of Police (SP) and the extent to which he has interested his District Magistrate in the matter.’Footnote ⁵⁸ The result of this was that, ‘Different policies and different methods have been evolved in dealing with the same tribe in different districts and much confusion and waste of energy appears to me to have taken place.’Footnote ⁵⁹ Differential responsibility (or claimed lack of responsibility) for ‘wandering tribes’ also characterized the workings of the CTA in collaboration with the princely states.Footnote ⁶⁰

Two administrative problems undermined the working of the CTA: first, spatial/regional differentiation, and, secondly, that of defining the specific ethnicities of communities. As we will see in the next section, it was these two themes that became part of the liberation narrative articulated by inmates of some of the larger settlements in Bombay Province. Yet, despite continually entering debates about the administration of the CTA throughout the 1910s and 1920s,Footnote ⁶¹ neither of these problems was fully addressed in the period when India's late colonial Congress regimes (elected in 1937 under provincial autonomy), often hand in hand with ‘backward classes’ officers, attempted to reform the system. Both problems, too, were ultimately carried into debates leading to the Habitual Offenders legislation from the late 1940s. Throughout the drawn-out process of reform and repeal surrounding the CTA, older assumptions about caste and the spatial management of criminal law and control continued to dominate. Juxtaposed with these administrative and ideological dynamics, therefore, was the question for India's emergent national leadership of how the targets of the CTA could be transformed into responsible Indian citizens.

The first place in which we can see this playing out was in the Bombay Congress Government's own Criminal Tribes Act Committee of 1938–39 and the events leading up to it. This Committee, chaired by the then home minister for Bombay K. M. Munshi was attended by high levels of expectation from criminal tribes in Bombay settlements. It was tasked to investigate necessary changes in policy and the existence of ‘grievances’. In its opening chapters it proceeded to list all of the defined criminal tribes of the province, with the same modus operandi contained in police ethnographies used by the colonial state. Bhamptas were described as experts in disguise and passing themselves off as Marathas and mendicants; Kaikadis committed house dacoities in groups; Mangs were cattle lifters who also used poisoning and armed themselves with particular kinds of weapons; among Mang Garudis, women were more criminal than men. To frustrate the police, Mang Garudi women, it was written, had been known to ‘strip off their clothes and stand naked before them’ or to ‘seize a child by its legs and threaten to dash its brains out on the ground’; Thankankar Pardhis made and repaired grindstones as a pretext to spy out localities and were ‘expert cheats’; Sansi gangs settled outside villages before they attacked.Footnote ⁶² The report's authors cited the case of the Dharalas of Kaira district as an example of the immense and varied difficulties in administering the Act. Yet, they did not consider that the logic of tribe-focused group criminality was therefore thrown into question, but rather the administrative difficulties of ‘tribal’ recognition. ‘We would have no objection,’ the report concluded, ‘if well defined classes with not too large numbers can be brought under the machinery of the Act on a very clear case being made out against them as regards their abnormal criminality.’Footnote ⁶³

This apparent acquiescence in ideas about group and habitual criminality on the basis of ‘tribe’ was given a new kind of significance as the authors of the report began to envisage that these communities could be incorporated into a liberal democratic polity. Integral to this project was the reconciliation of two problems: first, the existence of a criminal law system in which normal civic rights were suspended or withheld on the basis of ethnicity, and, secondly, the extent to which those who had been subject to this system could be designated as Indian citizens, when their supposed lifestyles arguably hindered their ability to exercise civic rights responsibly. This was presented most starkly as the report considered a specific moment when internees in the Umedpur criminal tribe settlement in Sholapur involved themselves in mill strikes growing out of the Red Flag Union actions in the city in 1937–38. In the official account, the Union deliberately targeted the settlement during the government enquiry with their mobilizing rhetoric that compared the fencing around the settlement with the fencing around India imposed by British imperialism.Footnote ⁶⁴ For the report writers, ‘… such influences and propaganda are not conducive to the main purpose of settlement life, as its discipline becomes irksome at a stage when the settlers have not yet been “nursed” to a non criminal life. The criminal tribes by heredity and temperament fall an easy prey to any irresponsible agitation which does not impose self control.’Footnote ⁶⁵

This view that the inmates at the Sholapur settlement were not yet ready for civic responsibilities, being unable fully to recognize and therefore exercise their democratic rights, was shared by Munshi. In a speech on 13 January 1938, he argued that the criminal tribes

In reality, a well-organized movement did develop from within the settlement, as we will see in the third section, and put forward a range of specific demands relating to the Commission of Enquiry, set out under a defined leadership.Footnote ⁶⁷

Munshi's concept of civic rights and responsibilities, and their effective absence in these communities, arose from a more widespread ambivalence surrounding the definition and targets of social reform. The Bombay Congress had never advocated a programme of reform or rehabilitation for criminal tribes before the formation of the 1938 government. Even where they advocated the ‘uplift’, as they put it, of ‘depressed classes’, social reformers focused on the removal of presupposed cultural habits as a means of moving more towards a majoritarian (often Brahmanical) ideal.Footnote ⁶⁸ These projects never included communities defined as criminal tribes. Some ‘Scheduled Caste’ leaders in the Bombay Presidency recognized the idea of a coherent, self-defining group of criminal tribes in many regions. But they, too, were ill disposed, not least for strategic reasons, to consider criminal tribes as a separate category: Ambedkar's refusal to acknowledge them as part of the ‘depressed classes’ movement is a case in point.

Munshi's stance on the (lack of) civic culture among criminal tribes is also partly explained by the historical milieu of the Congress Party in Bombay Province. Both B. G. Tilak and G. K. Gokhale had expressed support for the continuation of the CTA and, in the case of Tilak, the exclusion of criminal tribes from special consideration in military recruitment. In opposing the recruitment of Berads and Ramoshis into the Indian Army, the latter argued that ‘Thieving is their occupation. To receive beating is their occupation. They will not stand the discipline of the army.’Footnote ⁶⁹ Although the Congress governments in Bombay and the United Provinces were ostensibly committed to progressive social reform, the voices of cultural conservatism were strong.Footnote ⁷⁰ In the same year that the Committee undertook its work in 1938, Munshi founded the Bharatiya Vidya Bhavan, which focused on Sanskrit research and had as one of its objectives a promotion of ‘faith’ in the culture of the land, in particular its epics and the Bhagavad Gita. Not surprisingly, his response to the work of Christian missionary organizations in the criminal tribe settlements was highly reactive. Rather than encouraging the promotion of community-specific religious cultures, Munshi stressed the need for a Hindu Kritankar and Muslim moulvi once a week and the banning of bible classes and Sunday schools.Footnote ⁷¹

Implicit in the 1939 report and Munshi's views was the idea that not only were the criminal tribes not ready for release, but they were also incapable of civic responsibility. Indeed, any alternative claim of rights that they might propose could only ever be viewed not just as a challenge to mainstream ideas of the rights-bearing Indian citizen (as came to be the case with Dalits), but actually a direct threat to them, in everyday terms of presumed criminality. This tension can be illustrated by the debates over the Bombay Police Act Amendment Bill in 1938. When the Bill's restrictive provisions were called into question by some members of the Legislative Assembly, Ambedkar responded: ‘there are occasions when, in order to protect the liberty of the large mass of the people, the liberty of the hooligans, the criminal sections in the society, can be suspended’.Footnote ⁷² A notion of this threat to the general ‘rights’ of society at large was repeatedly illustrated in debates about where to settle criminal tribes and was particularly sharp in larger cities. For example, in the summer of 1929, the Bombay government proposed setting up a criminal tribe settlement and ‘free colony’ in the Worli Chawls area of Bombay City. However, the commissioner of police, objected on the basis that the chawls were situated near Worli Marine Drive where much money had been spent to provide building sites for ‘better class residences’: ‘it is obvious that the better classes will not care to take up their residence in that part of the city’ if criminal tribes were to move there.Footnote ⁷³

Key here too was the idea of the public good. In the quantitative calculations underpinning the maintenance of settlements in the 1920s, as viewed by O. H. B. Starte in budgetary discussions surrounding the costs of maintaining settlements, custody was viewed in stark terms as a quantifiable public good that could be measured by the reduced effect of free criminal tribes on civil society:

For Starte, ‘settlement work was a paying proposition for the public’, which had quantifiable results at multiple levels, sometimes connected to criminal tribe cultures themselves. The amount of ‘false coinage’ in circulation, Starte argued, had dropped since the Chapparband community had been systematically ‘reformed’ in settlements.Footnote ⁷⁵

In addition was the long-standing colonial conceit that connected characteristics of indigeneity to an intrinsic inability to understand or enjoy civic rights. In the early government debates about the uses of reformation in the criminal tribes settlements, Starte argued that criminal tribes ‘had to be governed from a personal point of view … he tends to bring all the problems to one man. He will not easily learn the art of petition sending and of lobbying government—they may be a curse to him.’Footnote ⁷⁶ From the late 1920s, however, organizations and associations among criminal tribes in western/northwestern India did begin to mobilize and exercise civic consciousness. These instances (as explored below) were often uncoordinated and locally rooted, but they revealed an essential moment of proto-citizenship claim-making. By mobilizing around their specific identity as criminal tribes, and the attendant denial of civic rights, these organizations foregrounded the ‘criminal tribe’ as a particular type of political subject.

Civic mobilization and mobilizing criminality

One of the abiding narratives of the contemporary Denotified and Nomadic Tribe—or ‘DNT’—movement is the many-guised account that criminal tribe freedom was ‘delayed’ by five years following independence.Footnote ⁷⁸ Celebrations of ‘independence’ for those associated with the movement do not take place on 14–15 August, with reference to 1947, but on the occasion of Vimukti Divas on 31 August, with reference to 1952, when the legal status of the communities changed.Footnote ⁷⁹ In fact, criminal tribes in Bombay Province were denotified in 1949, not 1952, suggesting that a larger and more generalized narrative of DNT identity has gradually transformed, over time, processes of memorialization. The creation of the universal category of DNT is itself therefore part of this memorializing project and its subsidiary characteristics—the narrative of how a ‘people’ were denied justice in relation to a larger national movement of liberation, but were architects of something autonomous of it. It is also the product of a process of redefinition of these communities that took place over a longer period of negotiation, and which has been galvanized by the likes of Ganesh Devy and Mahasweta Devi in the formation of the DNT Rights Action Group (RAG) from 1998. Some of the key themes of the DNT RAG's agenda around civic consciousness, justice, and citizenship can be traced back to the late 1930s in western/northwestern India, and especially to the Munshi Report of 1939. Although clearly deployed for contemporary political ends, the key message is that criminal tribes were active agents in their eventual freedom, able to determine and define their own narratives of liberation, and not simply passive recipients of the repeal of the CTA.

As explored above, the debates on civic responsibility from the late 1930s increasingly positioned the criminal tribes as an identifiable group which (as for the Dalit movement) faced a specific predicament. In important respects, though, this was quite unlike the Dalit movement in which ‘organic’ intellectuals were able to posit convincingly an idea of Acchuts (Untouchables) as original inhabitants.Footnote ⁸⁰ Equally, they were unable to collectively articulate a clear ‘Adivasi’ identity in terms of indigeneity, isolation, and primitivism.Footnote ⁸¹ In certain regions, communities marked as criminal tribes did adhere to such characteristics, or fell within the scope of Dalit movements, but their internal heterogeneity and regional variation precluded the formation of a shared political identity in these terms. Unlike many early Dalit movements, groups marked as criminal tribes could not articulate autonomous identities within a framework that reconfigured Hindu religious culture in opposition to Brahmanism, since their ‘stigma’ was not principally related to ritual status. It was, however, inherently connected to historical narratives of movement, occupation, and settlement. And the latter were ultimately projects that posited the idea of the community in relation to the powers of the state—powers to define certain occupations and ways of life as ‘illegal’ and which imposed forms of settlement on the itinerant. In other words, the (largely hidden) community activists mobilized civic identities which valorized, on the one hand, a past occupational status around community ‘skills’, and, on the other, a past of free movement and spatial inter-community interaction that complicated geographical boundaries.

Within the official and public record of the late colonial period, there was no recognition of political/community organizations among criminal tribes. Yet, in direct interviews with 32 individuals across Maharashtra and through more careful archival work, we discovered that this was often a function of archival practices in the compiling of data for official circulation to government officers.Footnote ⁸² The formal administration reports on the working of settlements in Bombay in the 1930s, produced each year, contained sub-headings such as ‘health and education’, ‘sports’, ‘Boy Scouts and Guides’, and the operations of small cooperatives and labour activities. Alongside these mundane reform activities were tabulated figures of offenders by ‘tribe/caste’, numbers of absconders, or figures released to the ‘free colony’, but nothing on political agitation, lobbying, or civic movements.Footnote ⁸³ However, although absent from the reports, government discussions around the opening and closing of settlements around Sholapur certainly suggested that officers already had experience of inmates getting involved in labour disputes. For example, the Kalyanpur Settlement was closed down in 1931 and residents sent to Umedpur, for fear of the effect of being so close to the city and its labour troubles in the late 1920s and early 1930s.Footnote ⁸⁴

Community mobilization at the end of the decade was more obvious but still not properly documented in official reports. The 1939 report made only passing reference to the role of criminal tribes in the Red Flag Union strikes in Sholapur.Footnote ⁸⁵ Yet, in police reports, the names of specific criminal tribe leaders who had taken an instrumental role in mobilizing settlers for the strike are mentioned. Settlement inmates certainly had powerful motivations for mobilizing against labour exploitation, given the use of criminal tribe labour in projects such as Nira in other parts of the province.Footnote ⁸⁶ In police reports, Rhising Khanda Bhat, Garhya Shabhu Bhat, Mahadu Gurappa Bhat, and Tippa Krishna Kaikadi were all named as leading rabble rousers. Reports on speeches included direct mention of how settlers would be encouraged to ‘take the lead in any future strikes’.Footnote ⁸⁷ Insubordination in the settlement, too, had led to the transfer of three other leaders—Shesya Gundi Kaikadi, Bhima Ambaji, and Tuka Parsu—on 10 November 1937, and as they were being transported, other settlers threw stones at their transport. When those allegedly involved in the commotion were arrested, they claimed that they had been framed by the settlement management for attempting to organize a ‘reception committee’ for the arrival of the premier of the Bombay Presidency.Footnote ⁸⁸

It was during the late 1930s, too, that criminal tribes began to mobilize more concretely against the CTA. The existence of a criminal tribe-led strike/protest (andolan) in 1938 in Sholapur was discussed in debates on the Bombay Habitual Offenders’ Act 1947 by Shivbishalsing Harpalsing, representing the Railway Union. The movement had a clear set of aims, the central one being the repeal of the CTA.Footnote ⁸⁹ The mobilization for demands that specifically related to their status as criminal tribes, as opposed to a certain category of exploited labourers, represented a significant shift. Although these demands were far from the group-rights claims that emerged in the 1970s, they revealed an early form of civic consciousness that related specifically to their status as criminal tribes. These demands for the repeal of the CTA, or exemption from it, were frequently couched in terms of formal civic responsibility. In northwestern India, groups like the Adi Hindu Conference and the All-India Shraddhanand Dalit Uddhar Sabha had lobbied for the exclusion of certain—although not all—communities from the CTA. They emphasized the contribution of groups like the Bawarias and Harnies to the war effort and their ‘law abiding habits’.Footnote ⁹⁰ By the mid-1940s, such narratives had been appropriated by individual communities who petitioned the government for their removal from the list of criminal tribes. When members of the Sansi community in Karnal lobbied the district magistrate, for instance, they made sure to emphasize that, ‘We are almost all voters of the P.L.A. [Punjab Legislative Assembly].’Footnote ⁹¹ And Bawaria petitioners cited the jail returns of previous years to legitimize their argument that they were the ‘most law abiding and peaceful citizens of the State’.Footnote ⁹² At the same time, then, as the Congress government in Bombay was debating reform of the CTA, criminal tribes across a far-flung region had begun to mobilize using similar motifs of civic responsibility.

Interviews with a community elder who grew up in the Sholapur settlement corroborated and enriched the archival evidence of organized direct action, and challenged the Congress ministers’ dim view of criminal tribes’ ability to exercise civic responsibility. How this activity is remembered by ex-members of the settlement is now part of a larger discourse of the anti-colonial struggle—a distinctive narrative of ‘freedom fighting’ which forms a repeated theme in the political strategies of the contemporary DNT movement. As we will see below, this anti-colonialism among criminal tribes is also formed from usage of community ‘law breaking’ that inverts notions of criminality as socially productive and draws on anti-state discourses of the interwar period. Revealing an indigenized notion of civic responsibility, these strategies employed anti-colonialism as a means of celebrating illegality and criminality.

It was clear that during 1938–39 the leadership in the Sholapur settlement had framed specific demands that were neither purely political nor simply related to the actions of the Union. One set related to the strike itself, but further demands were connected to, on the one hand, release from the settlements under the new Congress government, and, on the other, the development of specific settlement facilities and welfare: a temple, better sanitation, and infrastructure in the area of the settlement and free colony. Inmates had also set up their own committees to respond to the work of the Red Flag Union. Bhimrao Jadav, a teenager at the time of the strike, articulated this in terms of a broader anti-colonial struggle:

The organizations were neither exclusively nor directly political, however. In the mode of public manifestations of popular performative anti-colonialism in the streets of Bombay, but crucially via specialist ‘professional skills’ of the community, Jadav described the use of theatre as a means of expressing frustration and solidarity in the face of the local administration and settlement management:

There was a tension between these performances—which aimed at publicity for the settlement organization's serious claims for civic and political responsibility—and, at the same time, the promotion of this via ‘traditional’ community skills, including those that might have been seen as ‘criminal’. In a vein that was common to popular anti-colonial movements in cities across North India, this was a protest that revolved around the idea of reclaiming cultural space.Footnote ⁹⁵ It also involved the inversion of consolidated ethnic ‘traits’ of communities themselves as a means of extending forms of protest and appropriation of majoritarian Hinduism.Footnote ⁹⁶ A common theme in European discussion of criminal tribes’ ‘traditional occupations’ was that they were, not unlike the handloom, obsolete in the face of reform and ‘modernization’ brought by colonialism.Footnote ⁹⁷ Jadav pointed out that alongside the theatrical performances in the settlement, ‘sometimes people would show the tricks they used in thieving. The women from the Kanjar Bhat community would dance before the officer, and the Chapparband community would show how they make duplicate coins.’Footnote ⁹⁸

The local Congress organization's response was less to the traditionally civic-minded Settlement Sewa Sangh than to the anti-colonial possibilities of these skills. As with the Red Flag Union, the local Congress co-opted members of the settlement as foot soldiers into its movements—especially those requiring direct action: ‘In the year of 1942, our group in the Settlement were asked to destroy the mills and thermal power centre with a bomb. But somehow it did not happen,’ Jadav explained.Footnote ⁹⁹ As the basis of their involvement in Quit India, settlers even thought it might possible to get a ticket for one of the constituencies in the Sholapur region for the 1946 elections: ‘Me and Nagarkar went to Dr. Antrolikar (Congress chairman in Sholapur) but that time he described us as “rats”. He said “you are not capable of contesting the election”.’Footnote ¹⁰⁰ Jadav's emphasis on criminal tribes as participants in the freedom movement is therefore connected to some of the main ‘moments’ of national anti-colonial protest—the late 1930s strikes in industrial cities and the 1942 Quit India movement.

On the other hand, Jadav's narrative deliberately evoked the unique position of criminal tribes as ‘law breakers’, as figures legitimately ‘fighting’ or ‘robbing’ an illegitimate state, especially via ‘brave’ or life-risking activities. In this connection, some of our interviewees made reference to the famous Bhantu rebel-bandit Sultana Dhaku.Footnote ¹⁰¹ The latter carried out dacoities in the United Provinces during the period of Gandhi's ‘Non-Cooperation’ movement. Official accounts relate how the gang of ‘Bhantus have informers and friends in every village from about Kashipur and Jaspur … the jungle dweller, be he forest guard or cowherd, zemindar or cultivator, is still ready to supply them with food and to give them news of the movements of the police.’Footnote ¹⁰² They also used the criminal tribe settlement itself as a base for operations, inverting the functions of colonial control.Footnote ¹⁰³ Figures like Dhaku formed part of a larger theme in the historical memory of the anti-colonial functions of the settlement: Avinash Gaikwad, a descendant of a member of a criminal tribe hostel in Baramati, argued that,

Western India had its own Sultana Dhakus—figures made popular in subsequent folklore by opposition to the administration or police. Mirkhan Sultan Khan Barouch, for instance, who assaulted revenue officers, escaped to Baroda state and, in April 1922, recruited ‘tribes’ of Minas.Footnote ¹⁰⁵ Or Laxminarayan Jharwal, who led large protests in the streets of Jaipur and burnt effigies of the CTA during the 1940s.Footnote ¹⁰⁶ Such narratives have been important in creating an internal, community hagiography that allow certain groups of ex-criminal/denotified tribes to identify and celebrate individuals who struggled against the British. In Chharanagar, Ahmedabad, one of our interviewees who was related to a member of the Ahmedabad settlement reported that,

Contemporary DNT activists thus position the criminal tribes as forgotten ‘warriors’ in the freedom movement:

The emphasis on recognition, or a lack thereof, has been key in the formation of cross-communal DNT movements since the 1970s.Footnote ¹⁰⁹ Not only has their historic contribution to the freedom struggle been overlooked by society at large, goes the narrative, but their invisibility has been reproduced within citizenship frameworks. Recognition, therefore, is twofold: reclaiming community prestige by reinterpreting their ‘criminality’ in terms of freedom fighting, and utilizing this to define the parameters of their difference for group rights.

The contemporary movement stands in contrast to the earlier instances of civic mobilization during the late colonial period in its efforts to attain group-differentiated citizenship rights. Clearly, it has also inherited certain narrative strategies based in community histories of movement that were first articulated by the nascent and individualized demands for rights from the 1930s. On the one hand, movement was a powerful idea in relation to direct and local opposition to the state. A number of our interviewees who had spent time in criminal tribe settlements, whether traditionally defined as ‘nomadic’ or not, argued that it was the lifestyle of wandering and movement that marked independence. By extension, in the 1940s and 1950s, movement and dispersal became, too, a claim for special status in relation to the state. For Ladooben, who had been confined variously in Ludaki jail, Ahmedabad settlement, and Ambernath settlement, the experience of movement, travel, absconding from settlements, and the links with kin in Saurashtra was a matter of community pride.Footnote ¹¹⁰ On the other, movement was also the basis for community histories which established a status of both respectability and regional power. Most ‘caste’ organizations from the early twentieth century narrated a transition in status from a past, regionally fixed association with Brahmanical or Rajput status to one of a ‘pariah’.Footnote ¹¹¹ Freedom of movement for many criminal tribes, for example, Chharas and Sansis, was also closely related to the trans-regional links between different communities marked as ‘criminal’ and the idea of a common Rajput ancestry, which itself was lost because of migration or movement. Here, then, was a claim of state misrecognition of the community's historical importance. Ketananand, in a similar vein to the Sansi elder at the beginning of this article, described the relationship between space, identity, and political mobilization:

This historic mythologizing of a fall from grace, as it were, from a glorious past of Rajput respectability to degradation and eventual classification as criminals forged a distinct political narrative, one which was articulated across several regions and divergent communities, and for multiple aims. From the 1930s, individual communities adopted these narratives to negotiate their problematic incorporation within the evolving frameworks of citizenship rights. In their petition to the government, the Bawarias cited earlier contended that ‘they once lived in Marwar as Chohan, Rauthor, Puar, Bhatti and Scongrey Rajputs but were driven out of their ancestral province after the fall of Maharana Pratap’.Footnote ¹¹³ Before independence, then, such civic strategies were largely utilized to contest their status as ‘criminals’. After 1947, and especially after the founding of the Constitution in 1950, communities employed their mythological pasts to directly engage with the promise of universal citizenship rights. Lobbying the government for exemption from the CTA was taken up in earnest between 1947–52 as criminal tribe activists recognized the tension between a liberated India and their lack of freedom. In a petition to the Home Ministry in 1949, migrants from Pakistan in Karnal district, East Punjab, lamented that:

In the context of independence, greater emphasis was placed upon their active fight against invading Muslim marauders. Denied the possibility of civic responsibility in the present day, as they were considered ill-suited to the duties and obligations such rights bestowed, these narratives articulated their historic contributions to society, namely in the form of a symbolic sacrifice of their ‘home-land’ and consequent degradation. In the vein of nationalizing histories, such a narrative posited the criminal tribes as actors within a heroic landscape of struggle against invaders.

Reflection on movement and migration also formed a way of connecting the lifestyles of ‘pariah’ migrants to the national rehabilitation of partition migrants in the late 1940s. A further repeated refrain from those recollecting the period of independence was the tension between post-partition rehabilitation for migrants from Pakistan and the failures of the Indian government to incorporate criminal tribes within its welfare regimes. Ketananand told us that:

Implicit in such reflections was the idea of the temporary ‘encampment’ for migrants, akin to the criminal tribe settlement, but which would eventually result in rehousing and full integration into regimes of citizenship. In contrast, criminal tribes faced a more permanent form of ‘encampment’, both spatially, in terms of their continued physical placement within settlements, and politically, as the government failed to incorporate them within its citizenship regime. Recognition of this tension is not merely an element of the contemporary DNT movement's historical memorialization, though. After the founding of the Constitution, this failure was seized upon by members of the criminal tribes who increasingly positioned their claims in relation to principles of citizenship. In Delhi, the local Sansi Mahasabha lobbied the State Congress Committee in the early 1950s to place the repeal of the CTA on their agenda.Footnote ¹¹⁶ Recognizing their new-found voting power, the Sabha threatened, ‘we will not exercise our right to vote’ unless the Act was repealed before the next elections. In addition to their new democratic powers, the freedoms supposedly promised by their status both as citizens of independent India and also, from 31 August 1952, as ex-criminal tribes—and therefore as theoretical recipients of full citizenship rights—were deployed for political ends. In 1954, for instance, Joginder Singh and Bhagwan Singh sent a petition to Delhi's chief commissioner after his visit to the Kasturba Nagar Colony:

In the early 1950s, the demands being made were still far from the cross-communal claims for group-differentiated rights that emerged from the 1970s, or that we see in the contemporary DNT movement today. Often, bar the larger settlements such as Sholapur, individual community identities or associations remained the primary means of mobilizing for group recognition. Yet, a more cohesive political identity, predicated upon their status as criminal tribes and, especially, as not-yet liberated citizens in a free India, was emerging. The earlier repeal of the CTA in Madras in 1947 and Bombay in 1949 was seized upon by activists in other regions, such as Delhi, which had yet to do so.Footnote ¹¹⁸ Similar forms of strategies were adopted by a heterogeneous and geographically diffuse set of communities across the western/northwestern region. These drew upon the longer, historic narratives of movement, occupation, and criminality which were evident in the anti-colonial protest of the late colonial period but more explicitly engaged with the idea of a political or civic culture.

The nomadic citizen and ‘rehabilitation’

Just as the criminal tribes recognized the tension between a free India and their ‘delayed’ freedom, the government too had to reconcile the question over their problematic incorporation within the liberal democratic state. Central to this was the decision not to simply repeal the CTA but to replace it with a new regime of controls on habitual offenders. The early state-level Habitual Offenders legislation passed in Bombay in 1947 and Madras in 1948 were implemented, effectively, on the basis of continuity of administration and commitment to control, as envisioned by the 1939 Munshi Report. This legislation, containing many of the same provisions regarding control of movement and settlements as the CTA, formed the basis for discussions surrounding a centrally enacted All-India repeal Bill and possible Habitual Offenders Act (hereafter HOA). These debates presupposed that the objects of the legislation would be problematic recipients of civic rights. Compared to the context of the state-level observations in the Munshi Report, the Fundamental Rights of the citizen had been recognized (or were about to be recognized) in the Constitution of independent India. This brought urgency to the subject of substantive citizenship rights, especially as its themes were being debated and defined around partition migrants. Most importantly, the promise of such rights had to be squared with regionally specific forms of penal control that maintained some features of the CTA.

In 1949 The Bombay Chronicle reported on the imminent repeal of the CTA in the province:

G. D. Tapase, the minister for backward classes in Bombay, was responsible for cutting the barbed wire enclosing the Sholapur settlement. More than mere spectacle, however, his ‘liberation’ of the criminal tribes symbolized a different kind of shift in their status. In 1947, Bombay Province had enacted its own Habitual Offenders Act which, in theory, suggested the incorporation of all ex-criminal tribes with ‘backward classes’ categorization.Footnote ¹²⁰ Because of the communities’ real status within this putative category, and owing to the fate of ‘backward class’ recognition in the 1950s,Footnote ¹²¹ this remained theoretical. In 1951, the Government of India published the report of its own Criminal Tribes Act Enquiry Committee (1949–50), which recommended the replacement of the CTA with central legislation along the lines of the existing provincial Habitual Offenders legislation. The report was replete with references to CTA's iniquitous nature:

Yet, in acting on the recommendations of the Committee, the Government of India was faced with two interrelated problems. First, how was it possible to square the evident injustice of the CTA with the fact that society at large, and the administration and police in particular, favoured the continued direct control and surveillance of specific communities on the basis of their ethnicities and the supposed cultures of criminality that went with them? Secondly, given that the HOA would have to accommodate, according to commentators, aspects of how the CTA had been administered in different ways in different regions, how could it be modelled as a central piece of legislation?

The first problem was demonstrated clearly in the reactions and objections of a number of individuals and agencies, including the chief commissioner of Delhi and members of the Punjab government, to the suggestions of a rapid move towards a more liberal HOA. These objectors were concerned that the new legislation might not incorporate, in substance, what they saw as the essential provisions and rules of the CTA. In response to concerns from the Punjab government about the proposed pace of repeal, the Home Ministry gave assurances that there was ‘no reason why those members of “criminal tribes” who are habitual offenders should be free from all restrictions even for a day’.Footnote ¹²³ Certainly, the authors of the report thought it possible to demographically quantify tribes by state. In addition, Section 16 of the Madras Act, the Government of India pointed out, provided, in certain cases, orders passed under the CTA would continue to subsist and would be deemed to have been issued under that Act. There was also no reason, they stated, that the settlements should be closed down. Both the Bombay and the Madras legislation, in Sections 15 and 8 respectively, provided for the establishment of settlements.Footnote ¹²⁴

The objections of Delhi's chief commissioner were addressed by again stressing the similarities between the existing Habitual Offenders legislation and the CTA. In both Bombay and Madras—states considered to be the most ‘enlightened’ in terms of penal policy towards criminal tribes and whose acts would form the basis of the proposed central HOA—the legislation contained many points of continuity from the CTA. In Madras in 1951, for example, 4,300 members of notified criminal tribes were still retained under the new act.Footnote ¹²⁵ Under its provisions, a ‘habitual offender’—so classified after three offences—was required to notify his or her place of residence under section 5, which was directly analogous to section 10 of the CTA. They also had to report a change of residence under section 6, which was analogous to section 11 of the CTA. They were also potentially liable to detention in industrial and reformatory settlements upon committing fourth offences under section 8, which was analogous to section 16 of the CTA.Footnote ¹²⁶

Perhaps the most important parameters of this question, however, focused on identifying the ‘unreformable subject’—one who could not be identified clearly as an individual rights bearer, but who belonged to a community whose tradition involved movement that could be defined as ‘vagrancy’. This was seen in the Bombay Legislative Assembly debates on the passage of the Bombay Habitual Offenders Act in 1947. On 8 October of that year, Morarji Desai presented the new bill, pointing out that the Munshi Report had recommended the replacement of the CTA with legislation along the lines of the Restriction of Habitual Offenders (Punjab) Act of 1918. Desai's views shadowed that of the central Congress organization: ‘… we are emphatically not of the opinion that the so-called criminal tribes are inherently criminal in the sense either that their criminality is necessarily hereditary or that no amount of attempts at improving them can alter their habits’.Footnote ¹²⁷ However, Desai did believe that

For these reasons, Desai proposed that it was necessary to continue to ‘restrict movement’ and to ‘maintain some settlements’. Desai's argument stretched further: the criminality of criminal tribes was inherently linked to traditions of nomadism and movement:

The second problem of how to create a central HOA also reinforced the sense that this legislation was a direct continuation from the CTA. Moreover, its justification as such was based on a constitutional device which firmly linked it to the CTA's stress on ‘vagrant and nomadic tribes’. The competence of the central government to repeal the CTA was based on item 3 (preventive detention) and, more importantly, item 15 (vagrancy, nomadic, and migratory tribes) of the Concurrent List of the Constitution. The Ministry of Law took the view that the CTA was concerned with nomadic tribes and was therefore relatable to entry 15. This was also supported by the fact that criminal tribes were recognized in the Government of India Act of 1935 as a separate subject for legislation, and that the CTA 1924 was held by the reforms office to be relatable to this entry. This gave the central government competence to repeal and enact.Footnote ¹³⁰ Overall, these legal arguments conceptually linked the CTA to the newly proposed HOA, via the notion that both essentially dealt with vagrancy arising from nomadism and movement.

Recent work on the state in India has explored how the superstructure of legislative enactment, administrative rules, and even constitutional ‘rights’ are part of an imaginative or discursive edifice, in reference to which most Indians struggle with day-to-day parochial bureaucratic engagement.Footnote ¹³¹ Given the underlying relationship between the CTA and the raft of Habitual Offender legislation enacted in the early 1950s, both at the level of ideology and provision, the new legislation could effectively become as close to the old as officers desired it to be. Beyond this legislative link—but also because of it—the CTA could be kept alive through police manuals and the usual extra-constitutional forms of control and punishment. Ultimately then, the rhetoric of a new, enlightened penology was somewhat empty, given the nature of how the state's coercive control often transgressed formal systems of punishment.Footnote ¹³² As we can see, running through even the highest level discussions around the proposed HOA, too, there was a clear discrepancy between the intention of the new legislation and its concrete administration on the ground. At one level, this led to a noticeable delay in the expected implementation of the new legislation on the basis of individual criminality (or the continuation of a presupposition of collective/hereditary criminality). In response to repeated complaints from individuals about this, the Government of India reported that:

Letters of complaint poured into the state and central governments that police departments continued to use the same methods of control and policing as in the colonial period. In January 1952, for example, one complainant who had migrated from Multan in Pakistan to India was challenged by the Rohtak police in 1950 simply because he was identified as belonging to a criminal tribe.Footnote ¹³⁴ Another appellant from Ludhiana in January 1952 complained that the deputy commissioner for criminal tribes aimed to ‘have some hold over and links with the released criminal tribes members’ by roping ‘as many of them within the purview of the Habitual Offender's Act as possible’.Footnote ¹³⁵ The contemporary arguments of DNT activists of a ‘delayed independence’ can therefore be found directly in the archive.

These features of the hangover of the CTA were an early sign of a much longer term tendency in the interpretation of the provincial Habitual Offenders legislation. The Bombay State rules accompanying the Habitual Offenders Act, as set out in 1949, still granted considerable authority and autonomy to village-level police patels in terms of the issue of passes for movement restriction.Footnote ¹³⁶ This was in spite of the, by then, widespread recognition of local police corruption in the administration of the CTA.Footnote ¹³⁷ A police officer recruited in the 1960s described in an interview how methods arising from the police manual continued to identify forms of criminality in connection with the modus operandi of criminal tribes.Footnote ¹³⁸ In the police manuals that he showed us from the early 1960s, the term ‘criminal tribe’ was still used. This explains how the newly denotified communities still experienced the same kind of penal system, despite the ostensible change in law regarding registration and group/ethnically defined criminality. As we have seen in the sections above, the CTA was always differentially administered on the basis of local knowledge and preference, particularly around the uncertain view of ethnic identity.

There were other factors at play in how these continuities in criminal law affected potential rights to state welfare. It was clear that the legislative and political associations between the CTA and the proposed HOA were recognized by governments as part of a larger project in which the soon-to-be ex-criminal tribes might enjoy potential rights to state relief and rehabilitation as Indian citizens. There were several organizations and government projects to provide grants and subsidies to ‘denotified’ communities in the early 1950s in some parts of India. They developed, for example, educational institutions, scholarships, midday meals, and employment exchanges. However, this was not applied uniformly and was certainly insufficient, to the extent that questions were repeatedly addressed in the Lok Sabha to ask ‘what was being done’ in the area of rehabilitation.Footnote ¹³⁹ Some help was provided by the state governments. In Delhi, the government allotted funds to the Ex-Criminal Tribes Welfare Board which conducted rehabilitative work.Footnote ¹⁴⁰ But funds were not distributed evenly, nor with consideration of the specific problems faced by the communities. In 1954, residents of the Kasturba Nagar colony complained to the chief commissioner that the Welfare Board, rather than facilitating their rehabilitation, was instead ‘always trying by fair means or foul to make us quarrel with each other’.Footnote ¹⁴¹ And in 1958, the same year the Board was dissolved, Kartar Chand of the Andha Moghal colony accused it of embezzlement.Footnote ¹⁴² Recognition, albeit unevenly applied, of their distinct status within welfare schemes, then, failed to translate into a differentiated political category. Official documents from the period pointed out that some denotified communities would benefit from the normal policies set aside for Scheduled Castes and Scheduled Tribes. But many were not included in those categories. For those, there would be ‘inclusions in the lists of Backward Classes in due course’.Footnote ¹⁴³

The rhetoric surrounding rehabilitation and welfare clearly foregrounded similar themes of political maturity or readiness, the problems for ‘society’, and the ‘Indian citizen’ in dealing with unreformed communities allegedly prone to criminal activity, as found in the 1930s. This was one of the clear contexts in which those ‘fit’ for denotification were separated from those who, along the lines of ethnicity and cultural practice, were still considered explicitly ‘unfit’ for the full enjoyment of rights. For example, the report of the Criminal Tribes Act Enquiry Committee, set up in the United Provinces in 1947, made this differentiation clear in its suggestion that the existing ‘gipsies’ currently registered under the CTA should be treated differently to most other communities. The report found that ‘… [as] on scrutiny [they] are not yet found fit for complete freedom we propose that provision should be made under the HOA for their restriction to Settlements and industrial or agricultural colonies … They don't have, yet, “moral anchorage”.’Footnote ¹⁴⁴

Crucially, for legal and administrative officers whose opinions were sought on the passage of a central HOA, this meant that certain groups of citizens could not be subject to the same rules of penal policy as others. For the chief commissioner of Delhi in 1951, the main targets of the new legislation would continue to be ex-criminal tribes, for whom special powers of detention would ideally need to be incorporated into any replacement legislation. Once again, it was the identification of traditions of movement that cemented the argument. He pointed out that existing rules of restriction of movement outside the old CTA, such as those contained in the Code of Criminal Procedure, could take a long time to implement: ‘a possible consequence could be that, since most members of the criminal tribes are nomadic in character and origin, a person proceeded against might jump his bail and disappear for good before an order of restriction could be made final or effective’.Footnote ¹⁴⁵ Secondly, these were still considered people who did not abide by the laws to which ‘normal’ citizens were subject and to whom special penal procedures would still need to be applied: ‘Most of the criminal tribes do not have a bias to crime against the person, but against property, and the offences are notoriously theft, burglary, and cattle lifting. The last has always presented a problem and they have been known to march cattle to long and unknown destinations through forced and quick marches.’Footnote ¹⁴⁶ He concluded that ‘The problem is essentially one of moral and material rehabilitation.’Footnote ¹⁴⁷ At state levels, too, efforts were made to ‘interpret’ the Habitual Offenders legislation to make it easier to identify ex-criminal tribes rather than those who had committed multiple offences. Although opposed by the High Court and Sessions Judiciary, the Home Department in Bombay pointed out the insufficiency of habitual offender definitions in the Bombay Jail Manual in setting out the number of offences for such a categorization. However, the Home Department proposed that a ‘first offence’ could be the result of habitual behaviour, if the probation officer of the court could prove this, not in the strict legal sense of the evidence, but in the moral sense.Footnote ¹⁴⁸

Until such time as ‘moral and material rehabilitation’ was theoretically achieved, denotified communities could still be subject to executive rather than judicial actions, which in some cases had dubious constitutional validity. Critically, the limits to their civic freedoms, especially around ‘movement’, were directly related to ‘public interests’ as defined by the Constitution. This tension was evident in the Constituent Assembly. Questioning H. V. Kamath's proposition that every citizen should have the right to bear arms, Ambedkar decried, ‘it would be open for thousands and thousands of citizens who are today described as criminal tribes to bear arms … [I]t is not possible to allow this indiscriminate right’.Footnote ¹⁴⁹ A similar discrepancy arose with regard to the right to move freely throughout the territory of India. H. J. Khandekar argued that through the use of sub-clauses in the Constitution, ‘what had been granted by the right hand has been taken away by the left’. He noted that it would be ‘extremely unjust’ if the principle was not intended to apply to the criminal tribes ‘who are also citizens of India’. But as Deshbandhu Gupta replied, Khandekar did not have in mind ‘the right type of freedom’: ‘[W]hy should not restrictions be imposed on the movement of the criminal-tribe people, when they are a source of danger to other law-abiding citizens?’ he asked. Freedom of movement without qualifications for those like criminal tribes would be a ‘freedom of the jungle’. As Algu Rai Shashtri put it, the government now had ‘certain obligations and responsibilities’ to protect society at large.