Classifying and counting the Dalits in the late colonial period: The conundrum of the Jatavs

Abstract

The Jatavs of the United Provinces were not legally recognised as a separate caste until 1942; and then only as a consequence of an exceptional revision of legislation. Yet, for some considerable time before, the provincial authorities had routinely treated Jatavs as if they had already been granted the status of an officially recognised distinct caste grouping. This was despite a ruling in 1933, endorsed jointly by the India Office, the Government of India, and the provincial government, that the Jatavs were not a separate caste. The case of the Jatavs is examined here in the context of the contradictions and confusion in the policies of the colonial authorities, first towards the Depressed Classes, and later in the construction of the category that eventually became the Scheduled Castes. In addition, it is argued that those contradictions also created interstices of ambiguity that many Dalit representatives explored, interrogated, and exploited as they generated the space in which to assert their agency. The history of the Jatavs is an important instance of subaltern politics participating in the procedures of the colonial regime rather than operating in some separate autonomous domain of activity. By engaging with the existing power structures and processes, Jatav leaders created the opportunity to expose and take advantage of the contradictions generated by the confusion in the exercises of classifying and counting conducted by the colonial state.

It has been uncertain for some time whether a particular caste known as the Jatavahs should be included among the scheduled castes or whether they were a branch of another caste which is included or of a third caste which is not scheduled. After careful inquiry the Government of the United Provinces in which most of the Jatavahs live have come to the conclusion that they are a substantive scheduled caste. ¹

 This sort of change (adding to approved lists of Scheduled Castes) is not by any means a minor matter or one infertile of unfortunate possible consequences; however, it may appear in London. ²

Introduction

Like all colonial powers, the British in India sought to develop the means by which they were able ‘to name, to identify, to categorise, to state what is what and who is who’. ³ This article is an account of the attempts to arrive at definitive classifications and estimates of the numbers of those groups which became known as the Scheduled Castes. ⁴ The focus of the article is on the United Provinces (Uttar Pradesh, henceforth UP) and it specifically examines the example of the Jatav caste. ⁵ It is a study that seeks to analyse the fragility and contradictions in much of what has collectively come to be known as ‘colonial knowledge practices’. Additionally, it will be suggested that the friability in the foundations of the knowledge/power nexus—the cynosure seen by some as the font of imperial dominance—should lead us to question approaches that emphasise the pervasive and hegemonic character of the colonial state. The Jatavs of western UP provided a confusing and perplexing challenge to the colonial regime as its various branches struggled to arrive at an authoritative classification of their caste status. At the same time Jatav activists exploited every ambiguity they could to take advantage of the muddled colonial practices of classifying and counting.

Classifying and Counting in the Colonial Mind

The last major episode in the colonial project of attempting to construct comprehensive social taxonomies of the population of the Indian subcontinent was mounted with the object of delineating, classifying, and counting the ‘untouchables’. The official aspiration was to define a detailed and precisely enumerated administrative category, which could be utilised, among other things, in the calculations of the constitutional reform programme of the 1930s. That exercise was pursued in the context of a long tradition of attempts by the colonial power to identify, if possible count, and then determine the place of the discrete and compartmentalised groups that it imagined made up the population of the subcontinent.

The efforts to draw up lists, establish the boundaries, and count the numbers of what were thought to be the separate component segments of Indian society have long been identified by scholars as some of the main preoccupations of the colonial regime. For many British officials, India was best understood as ‘a vast collection of numbers’ providing ‘a particular form of certainty to be held on to in a strange world’. ⁶ The British concern with classifying the Indian population by attempting to specify its supposedly constituent elements and conducting systematic body-counts of its various components has been called a ‘virtual obsession’. ⁷ The creation of categories, the listing of groups and the counting of numbers have all been cited as important practices by which India was ‘ordered’ for administrative purposes, discipline imposed, and the ‘masses managed’. ⁸ Certainly, what are sometimes referred to as the ‘surveillance’ and ‘disciplinary’ functions of the colonial state, and their deployment in administration and governance, were often informed by the collection of this sort of information. As one prominent imperial practitioner of classifying and counting warned, ‘British ignorance … of the people among whom we dwell … involves a distinct loss of administrative power to ourselves’. ⁹

However, what has been frequently occluded in the extensive studies of this subject over time, or even totally ignored, is the extent to which the individuals, procedures, and institutions involved in the collection and processing of this information were equally responsible for the generation of substantial ambiguity, inconsistency, and even contradiction within colonial officialdom. Colonial knowledge was often far from being either consistent or coherent. There was not some form of reliable conveyor belt by which the accumulation of information was simply passed on and ordered into the ‘knowledge systems’ of the colonial regime. Neither were these bodies of knowledge effortlessly integrated into the exercise of discipline and power, nor were they absorbed within the practices of governance in the colonial order. The ‘capillaries’—imagined to ‘swarm’ over the social formation, and freely to collect, diffuse and process both power and knowledge, while simultaneously reinforcing each other in an endless symbiotic circularity—were often sclerotic or blocked. ¹⁰

A number of obstacles stood in the way of the orderly transmission and deployment of knowledge as a component of power and governance. The most important was that the personnel and processes of the colonial state were quite simply not always up to the task, even at the most basic of levels, of organising, or deciding how to utilise, or agreeing on the importance or relevance of the material they were collecting. They were frequently perplexed by the ‘huge and unmanageable flow of information’ that came their way. ¹¹ The gathering of information and its collation as ‘knowledge’, more often than not, resulted in both administrative and cognitive overload rather than providing any sort of instrumentality that could inform or assist governance—the ‘information order’ was endlessly subject to ‘information panics’. ¹² To put it in the colloquial vernacular of the era, the servants of Empire were not always quite ‘up to snuff’ when it came to collecting, agreeing on, or making good use of the information they had made such efforts to assemble.

In addition, the conclusions of ethnography and enumeration in their various guises and empirical detail often painted a picture of substantial—even contradictory— variation and difference. Consequently, exercises in colonial enquiry seldom produced a stream of uniform and manageable information that could reinforce ‘power- knowledge’ in any sort of straightforward manner. Imperial enquiries did not always support the hegemony of colonial knowledge and the integrity of the discursive practices which were thought to underpin ‘cultural technologies of rule’. ¹³ The findings and subsequent consideration and elaboration of these enquiries were just as likely to foster an agenda of ambiguity and disagreement rather than inscribe any map of certainty that could bolster the notion of a confident and coherent colonial gaze. Increasingly it also came to be realised that many of the categories of social taxonomy employed by the colonial regime as it embarked on its investigations provided little utility or guidance for public policy or the practice of governance. ¹⁴ As Fuller has argued, the main foci of official enquiry were providing fewer answers for the authorities and ‘from around 1909 to 1939 ethnography and anthropology became progressively less salient for high policy in the imperial government’. ¹⁵

While the prominence of ethnography and enumeration in the discourse of governance may have waned it did not disappear, and it was still the case that the different layers and functions of the colonial state deployed the use of classification and numbers as

[T]he fuel for a series of nested struggles between Indian officials at the lowest levels of the bureaucracy, up the system to governor-general of India, through a series of cross-cutting committees, boards, and individual officeholders, who conducted a constant internal debate about the plausibility and relevance of various classifications and the numbers attached to them. ¹⁶

These struggles about social taxonomy, enumeration, and increasingly the consideration of how the outcome of these practices might inform decisions about constitutional reform and political representation, acquired even more of an edge as clashes between the different, and often entrenched, regional traditions of colonial ethnography and administration took hold. In such encounters, the perspectives which had evolved and been endorsed institutionally at the provincial level stood their ground and defended their corner. These were views of how to interpret the world that were enthusiastically advocated and jealously guarded by the local official practitioners of classifying, listing, and counting. These provincial ‘schools’ of social taxonomy were just as developed as their counterparts that adhered to particular approaches to questions such as land settlement systems or methods of revenue collection, and just as likely to engage in vigorous debates and confrontations with advocates of different persuasions in other regions. The possibility of the existence of a coherent and pervasive subcontinental imperial gaze underwriting the exercise of colonial power seems unlikely. Any such notion certainly needs to be examined carefully and critically in the light of the refractions dispersed by the prism of these very different traditions and practices in the individual provinces.

Disputes about counting and classifying were also widely diffused as, more frequently than not, unresolved differences or confirmation of decisions were referred back to the imperial capital with the routine involvement of the bureaucrats of the India Office, the Secretary of State, or even the Cabinet and Prime Minister. The construction of the category that became known as the Scheduled Castes was an important example of these sorts of tensions and struggles. As one significant contribution to this area of research concluded, the history of deciding the position of the Dalits in the new constitutional order in late colonial India was a ‘convoluted process of institutional reform’ negotiated between ‘multiple arenas, split between Whitehall, Westminster, New Delhi and the provincial capitals of India’. ¹⁷

The imperial project to identify, describe, and enumerate discrete local components of the Indian population was also frequently confronted—and sometimes confounded—by the organised collective interventions of the very people that the authorities sought to classify, count, and slot securely into their allotted spaces. ¹⁸ Classifying and counting the population had acquired an additional and more immediate political significance in the context of the impending constitutional reforms of the 1930s, and as a consequence challenges to official taxonomy now became urgent and better organised. The spaces defined by the often-artificial labels and classifications invented as supposedly objective categories, were now occupied by an increasingly informed and politicised population. ¹⁹ Rather than accepting the results of colonial enquiry and acquiesce in their allotted spaces, many members of previously arbitrarily segmented populations ‘began to wonder what they would look like, and what would happen, if they combined’. ²⁰ Other groups sought to mark out their future by further segmentation and separation from earlier collective identities defined by colonial taxonomy. In a regime of growing systematic classification and enumeration, which was ever more likely to be linked directly in the near future to claims for a political and electoral presence, many Indians began to challenge with a new energy the descriptive map produced by official investigations like the census and other number-based enquiries. The results of these explorations were consequently seen less as an authoritative and given representation of putative collectivities, but instead as just one view of the world that could be challenged.

As the debates about constitutional reform intensified after the 1920s, Dalit representatives sought to register their presence and stamp their influence on the course of events. Our understanding of this period has been transformed in the last couple of decades by the explosion of interest in the Dalit movements of northern India in the late colonial period. When Gail Omvedt, who in many ways initiated the interest in studies of Dalit activism in the colonial period, published her early works she felt the need to ‘apologise’ for the ‘bias’ she had shown towards western India. ²¹ She hardly needed to make excuses for any omission—the evidence was indeed sparse and any solid documentation apparently missing from both the official and the informal archive. However later studies, particularly of the Punjab, UP, and Bengal, have enriched our knowledge of the Dalit movement in colonial northern India, when it was once thought to be confined largely to the western and southern regions. No longer can anyone make comments like ‘there were no socio-cultural movements among Dalits in UP in the colonial period. They also experienced delayed political mobilisation’. ²² These misapprehensions have been put to bed by a new generation of research.

We now know that Dalit leaders picked and probed at every stage of the reform process in an iterative engagement seeking to gain whatever advantage they could as they articulated their case and explored their options. They repeatedly strove to adapt, re-interpret, and even appropriate elements of the reforms for their own purposes. Of importance here were the efforts to expose the uncertainties, ambiguities, and inconsistencies in colonial epistemology and especially in its knowledge practices relating to caste. The involvement of Dalit representatives in the debates about how to define, label, classify, count, and represent the Depressed Classes and the Scheduled Castes in the evolving political settlement was a decisive intervention. Contrary to the claim that colonised Indians did not have the opportunity—or themselves refused—to create their identities within the domain of ‘bourgeois civil-social institutions’, ²³ it is suggested here that Dalits leaders actively engaged and even collaborated with the colonial state and campaigned within—rather than outside—its processes and structures. They were engaged in a struggle that challenges the idea that ‘subalternity emerged only on the underside of a rigid theoretical barrier between “elite” and “subaltern”, which resembles a concrete slab separating upper and lower space in a two-story building’. ²⁴ An important aspect of the emergence of the new wave of Dalit studies has been the emphasis it has placed on subaltern participation in so-called elite institutional politics. As Dwaipayan Sen explored in the context of Bengal, here was ‘subaltern politics expressed in an elite domain’, a politics that decisively straddled the essential dichotomy central to the historiographical imagination that influences much of the analysis in the tradition of Subaltern Studies. ²⁵ Similarly Anupama Rao—while acknowledging the influence of Subaltern Studies on her work—stressed the need to recognise those caste subalterns ‘who transformed conceptions of nation, citizenship and political rights by working within, rather than outside, state institutions’. ²⁶

It was by participating in the institutions and processes of the colonial state that those being classified and counted began to challenge the way they were being slotted into what they perceived to be the jumbled boxes and confused categories of colonial convenience and practice. Particularly significant was the success with which Dalit leaders engaged with the institutions of the state by articulating an account of the oppression of untouchability in the demotic idiom of everyday caste discrimination. At the same time this quotidian narrative was translated into demands for social justice and political representation that insisted on being heard in the formal processes of constitutional reform and in the language of liberal democracy.

The energy and persistence of the Dalits was fuelled in no small part by an overriding concern that any extension of democracy—and eventually independence—could be hijacked by the socially dominant, and manipulated to maintain the old hierarchical order. ²⁷ British delegates attending the second Round Table Conference had been advised in a secret briefing that ‘electorally the caste Hindus will inevitably find a way of dominating the Depressed Classes’. ²⁸ The British government minister leading the enquiry into the franchise similarly identified the issue when he predicted that the political position of the Depressed Classes would become ‘far worse under responsible government’, which he thought they feared would result in the return ‘of a majority pledged to protect religious and social orthodoxy’. ²⁹ Dalits engaging with the colonial state, on their own behalf and in their own interests, also strengthened their stance as they secured the means by which they could ‘step outside the suffocating embrace of a cultural nationalism dominated by upper-caste elites’. ³⁰

The persistence and intensity with which Dalit representatives participated in the institutions and processes of the colonial regime stands as a caution to those approaches that have seen subaltern politics confined to some separate, autonomous domain independent of the formal structures of politics. The late colonial period was marked by the constant involvement of subaltern representatives in the official structures and processes of colonial state formation. By engaging in this manner, they repeatedly sought to exploit the incoherence frequently exhibited by the colonial power. It was not an engagement characterised merely by a sort of narrowly focussed behaviouristic response, nor was it a simple passive reflex reaction to the limited opportunities being offered by the colonial state. The Dalit intervention was a purposeful and creative endeavour—carefully calibrated from a position of weakness—to take maximum advantage of the gaps and contradictions in colonial thinking and practice.

Classifying and Representing the Dalits: The Example of the United Provinces

Throughout the 1920s, the colonial policy towards the Depressed Classes in UP was marked by a number of inconsistencies. Initially there was a stubborn reluctance to recognise the Depressed Classes in any way at all. The UP government had gone so far as to dismiss the Depressed Classes as ‘mostly criminal tribes of people’. ³¹ The national Census Commissioner noted a similar predisposition in 1921 when he complained that while every other province had responded to a request to provide estimates of Depressed Class numbers, in UP ‘extreme delicacy of official sentiment shrank from facing the task of attempting even a rough estimate’. ³² The UP government accommodated its need to give at least some minimal recognition to the Depressed Classes through its patronage and promotion of Jatav leader Khem Chand of Agra. He had been nominated as the sole representative of the Depressed Classes in the legislature, and the authorities further legitimised his authority by nominating him to a range of other official positions and calling on his services at official hearings. For his part Khem Chand, like many other Dalit leaders at this time, channelled his involvement with the regime through often excessively obeisant statements of loyalty to British rule. ³³

The provincial authorities resisted repeated pressures to acknowledge any sort of Depressed Class collective interest or to supplement Khem Chand as the single representative in the legislature. The UP Governor claimed in 1925 that their interests had not suffered through a lack of ‘direct representation’ and declared that ‘politically the depressed classes are in no sense a community in themselves’. ³⁴ After the Muddiman review of the reforms, many provinces increased the number of Depressed Class nominees in the legislature, but UP did not. ³⁵ There was no shortage of evidence suggesting widespread scepticism among officials in the province about the usefulness of nominating Dalit members to representative bodies. Hailey’s most trusted advisor in UP, for example, thought that far more good would come from measures to improve employment opportunities than by increasing political representation. ³⁶

Yet, nominated Depressed Class appointments were being made to district and municipal local bodies of various kinds across the province. The 1916 UP Municipalities Act had made no provision for the representation of the Depressed Classes on urban bodies. However the UP Government later pledged that where possible it would nominate representatives to the urban bodies ‘provided persons fitted by education and otherwise, to further and safeguard their interests are available’, but then only as vacancies became available. ³⁷ The 1922 UP District Boards Act however, partly as a consequence of pressure from Khem Chand, did provide for one of the two nominated members to be a representative of the Depressed Classes. ³⁸ By the end of 1924 a nominee had been appointed to all but one of the 48 District Boards and by 1928 there were also representatives on 26 Municipal Boards. ³⁹

However, the manner in which these nominations were decided tended to meet with Dalit disapproval. The crucial issue was that the individuals selected were frequently not drawn from the ranks of the Depressed Classes themselves but rather ‘an Indian Christian or sometimes an Indian of superior caste who takes a special interest in depressed class work’. ⁴⁰ Khem Chand complained that the benefits of nominating Depressed Class representatives to local boards had been ‘neutralised by the specious claims of other communities to be included among them’. ⁴¹ One Chamar organisation in Kanpur reported to the Simon Commission that ‘Even this one seat has been usurped in many cities by high-caste Hindus posing themselves as our sympathisers’. ⁴² The same accusation was made at the provincial conference of the Adi-Hindus meeting in Lucknow in July 1930 that protested against the ‘gross injustice … of the method of nomination in this province’ which had been subverted by people who had ‘posed as our sympathisers’. ⁴³ When the matter was debated in the UP Legislative Council at the end of 1931, one member remarked that ‘in most districts I have seen that one of the two nominated members both in the Municipal Board and the District Board is a missionary’. However, the issue had now been confronted, and it was at this same session of the legislature that it was finally decided that:

Nominations can only be made of men who actually belong to the depressed classes and not of those who claim or somehow or other get the support of the depressed classes to represent them. ⁴⁴

Even in the face of increased nominations there was still ample evidence of the indifferent attitude towards Depressed Class representation exhibited by influential officials. For example when asked to explain the meaning of the term ‘Depressed Classes’ during the Simon hearings, the relevant Government Secretary in UP is reported to have readily confessed to his complete ignorance of the subject. ⁴⁵ Even more remarkably, in view of his responsibility for the portfolio, when asked in open session whether one nominee on District Boards was enough to represent the interests of the Depressed Classes, he admitted he really did not have an opinion, as he had never bothered to investigate the issue. He gave the following astonishing reply:

I could not answer that. I have never seen the members at work. I have often wondered what the single member of the depressed classes actually does on the Boards, but it is impossible for me to guess what influence he at present exercises. Consequently I am unable to say whether there should be two or more members. ⁴⁶

The slow modification in the position of the UP government eventually gave way to a complete abandonment of its previous stance. It was precipitated in part by the influence of the investigations of the Lothian Committee into the franchise that had made clear to the authorities in UP that they were simply not taking the issue of the Depressed Classes at all seriously. ⁴⁷ In a matter of months in 1932 the position of UP shifted from barely acknowledging the existence of the Depressed Classes to an acceptance that the province probably had the largest Depressed Class population in British India; consequently the question of appropriate representation now had to be addressed. But the reversal of policy was adopted extremely reluctantly, with Governor Hailey’s chief advisor complaining that he did not accept the revised thinking, but grudgingly conceding that ‘so long as the Franchise Committee insist on the present definition [of untouchability], we must comply’. ⁴⁸

Hailey later tried to explain the change in policy to the Viceroy, pointing out that previously the issue of representing the Depressed Classes had been approached as:

[M]erely a question of nominating individuals to the Council or to local bodies in order to voice the feelings of those who suffered in a general sense from social disabilities and when it mattered little to what particular castes such individuals belonged, provided that those castes were definitely low. ⁴⁹

However, he admitted to Willingdon that the earlier policy towards the Depressed Classes in the province had not been at all—as he now put it—‘consistent’. But the Governor conceded that circumstances ‘have now compelled us to study the problem more carefully’. ⁵⁰ A lengthy and detailed statement further justifying the abrupt change of policy followed soon after Hailey’s initial volte-face, wherein the government adhered to its earlier position that the ‘Depressed Classes have no common bond save varying degrees of social disability’. ⁵¹ The confused and contradictory nature of the policies towards the Depressed Classes in UP and to the representation of their interests was beginning to become obvious.

Counting and the Ambiguity of Numbers

While UP was beginning to come to terms with an official recognition of the general position of the Depressed Classes, it was no closer to deciding precisely who they were or especially what their numbers might be. The determination of the aggregate total to be assigned to the ranks of what would soon be called the Scheduled Castes was still a remote goal. It got off to a bad start as the first attempts to resolve the composition of the Depressed Classes in UP were made and revealed widely differing estimations. Ambedkar castigated the findings:

To the question what is the population of the Untouchables the replies received were enough to stagger anybody. Witness after witness came forward to say that the Untouchables in his Province were infinitesimally small.… In the United Provinces the Census Commissioner in 1931 had put the population of the Untouchables at 12.6 millions, the Provincial Government at 6.8 millions but the Provincial Franchise Committee at 0.6 millions only! ⁵²

Even after Lothian’s Franchise Committee had completed its work, Ambedkar continued to criticise the dependability of the investigations in UP and make personal complaints to Lothian. ⁵³ For his part, Lothian openly conceded the difficulty he had experienced settling on the accuracy of numbers. He admitted that during the conduct of his enquiry, questions concerning the Depressed Classes had been the ‘the most acute controversies which we encountered’, primarily because of disagreements about definitions and the ‘great controversy about their number’. ⁵⁴ And as had happened with the Simon Commission before, Lothian marked up his special concern when it came to dealing with UP, where he reported ‘the greatest divergence of opinion’ had been revealed. ⁵⁵ Ambedkar also publicly accused the UP government of covering up the extent of its official enquiries into the size of the Depressed Class population and consequently depriving some of the larger estimates of their reliability and authority. ⁵⁶ Ambedkar’s mocking verdict at the time was that he found all the UP estimates of numbers ‘more amusing than convincing’. ⁵⁷ And when he came to sum up the efforts of the authorities in the province, his assessment was withering. He warned of the possibility of a ‘disaster’ if any provincial government engaged in ‘such strange oscillations’ about the numbers of the Depressed Class population. ⁵⁸ But the problems of definition were not just confined to UP, nor were they short-lived. The hapless Parliamentary Counsel working at a very late stage on the draft of the White Paper was driven to request the Treasury Solicitor (Maurice Gwyer) to provide him with a definition of a ‘depressed class’. ⁵⁹ While the Communal Award and the revisions added by the Poona Pact were supposed to provide the framework for a settlement, the ambiguities persisted. The same uncertainty plagued discussions about how the number and location of constituencies might be decided. From the beginning of the process, when Hoare put up his estimates for the consideration of the UK cabinet, he made it clear in his introductory remarks that the number of special Depressed Class seats ‘has inevitably been assigned somewhat arbitrarily’. The substantive cabinet paper explained that ‘The precise definition in each Province of those who (if electorally qualified) will be entitled to vote in the special Depressed Class constituencies has not yet been finally determined’. ⁶⁰ The extremely vague provisions of the Poona Pact exacerbated the confusion. As late as 1936, while engaged in the task of settling specific constituency boundaries, the Delimitation Committee remarked of the Poona Pact that ‘this apparently simple document has given rise to so much discussion’ and expanded with more detail:

The agreement was, it appears, concluded under conditions which precluded any close examination of all its implications, nor does there appear at the time to have been any detailed discussion as to how its aims could best be realized. ⁶¹

The first attempts to decide which castes should be defined as ‘depressed’ or ‘scheduled’ had been made by the Franchise Committee in 1932, the first provisional comprehensive list was contained in the White Paper of 1933, but the final definitive list to take effect in the 1935 Act was only finalised by the Orders in Council in 1937. ⁶²

The general uncertainty exhibited in both London and New Delhi provided the background to the oscillation and inconsistency that continued to characterise the policy and behaviour of the authorities in UP as they wrestled to settle on the names and numbers of those to be included in the Scheduled Castes. The most far-reaching proposition put forward was that all urban-dwelling Chamars should be excluded from the Scheduled Castes on the grounds that they had achieved a more prosperous way of life and freedom from the social oppression of untouchability prevalent in the countryside. This proposal would have disproportionately affected the Jatavs of the towns and cities of the upper doab, particularly Agra, and removed a total of some two million persons from the scheduled lists. Again, Ambedkar stepped in to condemn the suggestion, pointing out that such a division, removing the more prosperous educated urban population ‘is bound to leave the depressed class rank and file without any leaders to lead them’. ⁶³

The process of finally deciding the caste groups to be designated as ‘scheduled’ was accompanied by a variety of explanations and justifications about who and how many should be included on the list. On the one hand, additional inclusions were sometimes justified solely on the grounds of pragmatism. Caste groups were suggested on the grounds that it was from the ranks of these particular caste groups that some of the most suitable political representatives might emerge. In April 1934, the UP government informed the central Reforms Department that it was now inclined to think ‘that some of the best leaders of the Depressed Classes may well be found in classes [previously] excluded from the list’. ⁶⁴ At other times revisions were explained as a routine administrative adjustment made in the light of a better understanding of the circumstances of individual caste groups and a greater clarity about the criteria for selection.

However, the long-standing confusion about how to define, delineate, or enumerate the Depressed Classes with any sort of precision often re-emerged as the authorities reiterated and retreated to its conviction that it had ‘throughout deprecated any attempt to arrive at a definition by any precise formula’. ⁶⁵ On some occasions, the UP government completely reversed its position in a relatively short space of time. For example, in early 1933 it announced that the listing of the Scheduled Castes had been ‘as full and complete … as it is possible to devise under the circumstances’. ⁶⁶ However, a little more than a year later it was admitting that the lists that had been produced ‘were not exhaustive but illustrative’. Remarkably, it also confessed that it was relatively unconcerned about the integrity of many of the decisions it was making:

In any case the Government will always have considerable doubt as to the exact position of certain castes … and there would be no grievous breach of the principles on which the lists were prepared if they are put on one side of the line rather than the other. ⁶⁷

This sort of ambiguity and inconsistency simply fuelled the intensity with which various caste groups challenged the reliability of the official classification and the consequent total enumeration of the Scheduled Castes. It was no surprise that as the provisional list of Scheduled Castes was published and its contents then clarified, it was scrutinised with increasing severity. Some caste groups now pleaded for inclusion while others petitioned to be left out, and the question of the eventual aggregate total was left as unresolved as ever. ⁶⁸

The Chamar population was not bifurcated, as had been suggested, and was included as the largest Scheduled Caste grouping of more than six million persons. But there was no mention of the Jatavs. The Jatav efforts to be recognised as a distinct caste group from the Chamars had faltered, and their efforts to be classified separately now intensified. By the end of 1933, their claims had been considered and completely rejected in a declaration jointly agreed by the India Office, the Government of India, and the Government of UP. ⁶⁹ Jatavs were to remain included, and their numbers counted within the more general caste-category of Chamar. However, notwithstanding the ruling, the UP government continued to acknowledge the Jatavs as a caste and in some instances to extend further recognition to them. The inconsistencies and confusion persisted.

Commissioner Yeatts and the ‘Authority’ of the Census

As the issue of the uncertain status of the Jatavs continued to be debated and batted back and forth, increasing recourse was made to the office of the Census Commissioner as the possible source of a definitive ruling. Eventually, after many years of exchanges between the various branches of the colonial state, the India Office finally declared that the unresolved issues were not at all acceptable and a speedy resolution was needed. The Reforms Commissioner in London expressed his dissatisfaction to his counterpart in New Delhi and informed him that the Secretary of State had decided that any further delay and indecision were unacceptable. For Sir Vernon Dawson the route to a solution was evident and he was in no doubt that the Census Commissioner ‘is the authority in a position to determine’ the position of the Jatavs and should do so as soon as he was available. ⁷⁰

The new Census Commissioner, responsible for the coordination and management of the operation across the whole of British India was M. W. M. Yeatts. He had been the provincial superintendent in Madras for the 1931 census and had subsequently worked closely with Lewis in the Reforms Office in Delhi. Yeatts was appointed in March 1939 but routine leave, a period of sabbatical study, and bouts of illness meant he did not take up his duties until nearly ten months later. ⁷¹ In December he delivered his first opinion on what he called ‘an obscure and ambiguous origin for these Jatavs’. Yeatts was sceptical about the usefulness of any further enquiries and given what he claimed to be the ‘rather excessively formal definition of the issues’ adopted by the Secretary of State, thought the matter should be referred back to the authorities in UP. He was of the view that as it was the province making the enquiries and wanting answers, they should conduct any investigations as well. ⁷² In the interim, in a slew of correspondence—within the secretariat and between New Delhi, London, and Lucknow—Yeatts took up a number of other issues.

First, he pointed out that existing austerity measures, now exacerbated by the outbreak of war, meant the impending census operation was likely to be much more limited than those of the past. He thought there was little likelihood of the upcoming census encompassing the sort of investigations into the Jatavs that the India Office was suggesting. Second, he reminded everyone that the issue of caste was in any event almost certainly to be relegated following the suggestions made by the previous two Commissioners. ⁷³ It seems that using the census to determine the status of a caste—and particularly to decide whether a caste was ‘untouchable’—was fast going out of fashion. The mood had been voiced in 1931 by Hutton, who had made the point that the officers of the census ‘cursed the day’ when it had been first decided to ‘draw up a list of castes according to their rank in society’. ⁷⁴

Yeatts reserved his severest criticism for what he saw as the excessively loose criteria being applied for inclusion within the Scheduled Castes. He had a long record of expressing his disapproval of the numbers being included within the Scheduled category. During the earlier debate about the composition of the Scheduled Castes in Bengal he had always been sceptical about the suggested numbers that had resulted in, as he put it, ‘almost half of the Hindu population’ consisting ‘of what is generally known as depressed classes’.

I have suggested more than once that the proportion is excessive and that in this over-generous classification lies the real root of the Bengal troubles.… There seems to be considerable reason to look with some suspicion on the Bengal Government’s classification of these communities. ⁷⁵

He now included the methods of UP in his criticism. He accused the province of adopting a ‘highly subjective approach’ to the definition of untouchability that had resulted in ‘the phenomenon of nearly a third of the Hindu population of this province being “scheduled”’. Yeatts contrasted the numbers in UP with the much lower proportion in Madras which he cited as being ‘hitherto considered the chief illustration of the depressed caste as an element in Hindu life’. He was withering about the methods that had been adopted in UP for classifying the scheduled castes and warned:

But the UP, to adopt a popular locution, have ‘bought’ the present issue.… Prima facie it would seem a decision of the gravest moment to add to the already remarkable dimensions of the allegedly depressed proportion of the Hindu community. ⁷⁶

Yeatts was also keen to impress upon his colleagues that the inclusion of any one caste group in a particular province could well have implications elsewhere, and drew attention to what he saw as the implications of including Jatavs:

The effects of this contemplated decision may go further than is realised and it is conceivable that other provinces in which these so-called Jatavs are represented may have something to say about this addition … we may be quite certain that the Jatavs elsewhere will make some play with it. ⁷⁷

In addition to the detailed objections, Yeatts had a more fundamental disquiet about the function of the census. It is quite clear that he was starting to have difficulties defending the ability of the census to produce supposedly objective or scientific information on the question of caste. The Reforms Commissioner in London may have seen the Census Commissioner as ‘the authority’ to determine the status of the Jatavs, but Yeatts was far less certain. He warned his colleagues thus: ‘Actually in the present difficult and shifting field of Indian caste nomenclature I doubt if any one could speak with authority on such a subject’. ⁷⁸ Later he disparaged ‘the false notion that the decision [on the Jatavs] could be reposed on something of the nature of an authoritative pronouncement on a caste essence and position’. ⁷⁹ Yeatts was now reflecting a more general concern about the overall process of classification and enumeration, an unease that the colonial project of social taxonomy was faltering in the face of ‘the present difficult and shifting field of Indian caste nomenclature’. In addition he had worries about the ways in which the matter of the Jatavs had been passed around, batted backwards and forwards between the various branches of the state machine and delayed without there being any resolution of the matter. Having studied the whole history of the matter, he complained to his colleagues in New Delhi about the actions of London:

The more one studies this file the more one sees that the India Office letter of 25th July made it last nine months longer than it need have. In effect, the Secretary of State tried to pass this baby to us. We quite correctly declined. ⁸⁰

Whatever doubts Yeatts may have had about the process only reinforced his conviction that the Jatavs should not be granted the official status of a separate caste. He based this position on two broad grounds. First, he argued that the colonial record contained ‘little positive information’ about the ‘obscure and ambiguous origin’ of the Jatavs. Second was his long-held belief that inclusion in the Scheduled Caste category was being granted too readily and acceptance of the Jatavs would further swell their number and precipitate additional demands from other provinces. Both of these assertions by Yeatts, the use he made of them, and their effect as they entered the general discourse of decision-making need to be examined closely.

Yeatts’s claim of lack of information is curious and difficult to comprehend except as a remarkable example of the absence of any degree of fungibility in the information utilised by the various branches of colonial power. The Census of 1891 had explicitly identified the Jatavs—albeit labelling them as Jatia—as one of the main, and the most numerous, sub-caste of the Chamars. Across the province Jatavs were thought to constitute more than 20% of the total Chamar population. That census had also provided a detailed tabulation of the geographical distribution of the Jatavs, even to the level of the district, showing their population overwhelmingly concentrated in the western regions of the province. ⁸¹ When Briggs published his authoritative account of the Chamars in 1920 he also drew attention to the importance of the Jatavs as the grouping that could reasonably claim ‘to be the highest of all the sub-castes of the Chamars’ and he also drew attention to their geographical spread. ⁸² Perhaps even more startling was the fact that the great populariser of north Indian ethnography and enumeration, William Crooke, had also pointed clearly to the importance of the Jatavs who he also recorded were primarily to be found in Meerut, Agra, and the Rohilkhand. ⁸³ Furthermore, Crooke consolidated the complicated and elaborate tabulation of the 1891 Census into a conveniently condensed table showing the distribution of the Jatav population across the districts of the province. ⁸⁴

The Jatavs had been clearly identified by the usual machinery that the colonial regime used to collect information. Those procedures had mapped their population, determined their geographical distribution, and enumerated the Jatavs in great detail; there could be no question that the population the colonial authorities had identified as the Jatavs had been clearly classified and carefully counted. However, Yeatts reiterated his position of years earlier in almost identical language. In 1933, he had stated that the position of the Jatavs was ‘ambiguous’ and there was ‘little positive information’ about them to be gleaned from the census, concluding that the claim of the Jatavs to be included among the Scheduled Castes should be rejected. ⁸⁵ He now asserted additionally that there was ‘nothing in any census table, central or provincial’ from which he could begin to estimate their numbers nor any basis on which he could begin to investigate ‘its accuracy or even its components’. ⁸⁶ Finally the Reforms Office in Delhi confirmed officially to the UP government that Yeatts could see no advantage in further investigations by the census staff because he was unable to declare on the basis of all the available evidence if ‘the so-called Jatavs’ were in fact a separate caste. ⁸⁷

Yeatts’ conclusion illustrated the tensions and ambiguities in the collection and utilisation of colonial knowledge. His opinion flew in the face of all the evidence accumulated over the years, but he nevertheless persisted in his contention that he could discover no information about the Jatavs and that there was no basis on which he could make a judgement about their status or position. The contradictions in the official view of the Jatavs were now glaring. Colonial knowledge concerning the Jatavs was revealed to be anything but coherent and authoritative, but rather was the source of a series of deeply contested disputes.

Similarly, Yeatts’ assertion that the inclusion of the Jatavs would add millions to what he considered the already inflated ranks of the Scheduled Castes was a contentious claim but one that he repeatedly made. He particularly emphasised the electoral implications:

It would seem to mean an addition of four million to the scheduled caste population of British India. This total, particularly as accompanied by certain electoral privileges, has received very considerable criticism and to add four million to it is in itself a serious step. ⁸⁸

This assertion was completely unfounded, and Yeatts—despite his years of experience both on the census operation and in the Reforms Office—was misrepresenting the situation and contributing to the uncertainty. The Jatavs as part of the Chamar caste group were already enumerated within the Scheduled Castes and the inclusion of them as a separate caste would make no difference at all to the aggregate number. In fact, the Reforms Commissioner openly contradicted Yeatts. In a report which he prepared to summarise the position for the Viceroy, he pointed out that the Jatav’s claims should be regarded ‘less as an addition to the ranks of the Scheduled Castes than a change urged by the memorialists to get rid of the stigma of being Chamars’. ⁸⁹

The hesitancy, confusion, and disagreements of the various agencies and departments of the colonial authorities, the differences of opinion between the Government of India and the provincial administrations, and the disputes within and between the provincial governments themselves, all provided additional opportunities for the Jatavs to step in and make their claims.

The Jatav Response

The two decades after 1920 witnessed an incremental increase in the official recognition of the Jatavs in UP despite the clear rejection of their formal representations to be considered as a separate caste. Notwithstanding the decision made at the end of 1933 and agreed upon by all parties, in effect the authorities in Lucknow increasingly behaved, in complete contradiction of official policy, as if the Jatavs were a separate caste.

Recognition of the Jatavs came in a number of contexts. At the most informal and personal level, it was revealed in the routine acknowledgment of Khem Chand as a Jatav in official settings such as the hearings of the Simon Commission in Lucknow. ⁹⁰ At the same hearings, the Jatav identity received a rather more official imprimatur in a statement submitted by the Local Self-Government Secretary, which detailed the Depressed Class membership of District Boards and specifically not only named Jatavs but also specifically distinguished between Jatavs and Chamars. ⁹¹ The Jatavs were named, though not separately enumerated, in the census of 1931which recorded claims to include the name Jatav from seven different districts, including two independent ones from Bulandshahr. ⁹² By 1932 District Magistrates in UP had been instructed to allow individuals to be entered as ‘Jatav’ in revenue records if they so wished. ⁹³ The UP Education Department also sanctioned the use of the name in an official order of October 1935 in which it declared that it had ‘no objection’ to the use of the term ‘Jatav’ in the records of educational institutions. ⁹⁴

In all of these decisions, the authorities in UP were recognising the Jatavs as a separate caste while the governments in Delhi and London had done no such thing. Indeed, not only had they refused to acknowledge the Jatav claims to be a separate caste but had explicitly rejected them. Unsurprisingly, the Jatavs now embarked on a series of representations that placed the contradictions in government policy at the heart of its campaign. As they put it in one submission to the Viceroy:

It is the irony of fate that in spite of our repeated requests and memorials our Jatav community has been meted out a step-motherly treatment at the hands of the Government in the matter of the classification of the Scheduled Castes … it is an anomaly of a very peculiar type that the Government recognise the Jatav caste as a depressed class on the one hand, and the Government have not included the Jatav caste on the list of the scheduled castes on the other hand without any rhyme or reason. ⁹⁵

A later petition complained in almost identical terms that the UP Government appeared to be ‘hesitative’ to include the Jatavs ‘without any rhyme or reason’. ⁹⁶ When no action was taken a further letter was dispatched:

It is evident that the Jatavas are regarded as a ‘depressed class’ by the Government for all practical purposes. But it is passing strange that that in spite of the de facto recognition of the Jatavas by the Government they are de jure excluded from the list of the Scheduled Castes. ⁹⁷

Alongside petitioning the Viceroy at this time, the Jatavs also made direct representations to London, dispatching a petition to Secretary of State Zetland pointing out once again the contradictions in government policy. ⁹⁸ The Jatav leadership was now making a concerted effort to expose the gaps between the actual practice of the UP government and the stated policy of the Government of India and in addition were seeking to recruit the support of the India Office to come to their assistance and intervene in the matter. They had now engaged with all three branches of the imperial state in an effort to achieve their goals. It would be difficult to find a clearer example of the sort of activism and engagement that Sen refers to as ‘subaltern politics expressed in an elite domain’. ⁹⁹

The vigour of the Jatav representations on the contradictions in policy was given a boost by the provincial elections of 1936–37 and by Congress eventually assuming office in July. Not only had both Chamars and Jatavs taken a prominent role in the election campaign, but some of their most significant leaders now entered the legislature. ¹⁰⁰ Jatavs were particularly well represented among the new legislators and arguably had been favoured both by the rules of the franchise and the distribution of seats. While Jatavs made up only around 10% of the Scheduled Caste population they won in 20% of the reserved seats. ¹⁰¹

Pant, the newly elected Chief Minister, spent the first few months in office testing the boundaries of the threats of intervention from the Governor as he explored the limits of the ‘special powers’ contained in the new constitution. ¹⁰² In April 1938, the new Premier addressed a large demonstration of Dalit protestors who marched to the Assembly Chamber in Lucknow. He assured the protestors that the Congress ministry was adopting many policies ‘to benefit the depressed classes’. ¹⁰³ Among the measures that had been introduced was an instruction to the Education Department about the classification of castes. In January 1938, P. M. Kharegat (Chief Secretary, UP Education Department) moved to harmonise the various departmental statements made previously and to enable Jatavs ‘to call themselves Jatavas and yet be eligible for the stipends and scholarships sanctioned for the children of the depressed classes’. This he did by announcing that in future the official stipulation was that ‘Chamars include Jatavas’. ¹⁰⁴ This formulation conformed to the terms of the Scheduled Caste Order of 1936 which specified that it included ‘the castes, races or tribes, or parts of or groups within castes, races or tribes which are to be treated as the scheduled castes’. ¹⁰⁵ While Kharegat’s solution may have been designed to suit the administrative convenience of the UP Government, may have been consistent with the law, and almost certainly was intended to assist the Jatavs, it ran directly counter to the aspirations and objectives of the Jatav campaigners. The whole purpose of their campaign was precisely that their numbers not be defined and included as a part of or a group within the Chamar caste. The reaction was immediate and led by those who had struggled to end the inclusion of the Jatavs within the Chamar caste group.

Ram Prasad Soni, on behalf of the All India Shri Jatav Mahasabha and the UP Depressed Classes Federation pointed out that being included as a component of the Chamars meant the word Jatav would ‘continue to carry the same stigma’. Soni added ‘We desire that “JATAV” be included in the list of “SHEDULED (sic) CASTES” as a separate CASTE’. ¹⁰⁶ The Working Committee of the All India Jatav Youth League, in a plea to the Viceroy, pointed out that ‘the Jatavas have no caste connections’, including dining and marriage, with the caste known as Chamar. In addition to ‘strongly protesting’ against the wording of the statement, it re-asserted the demand that Jatavs should be included ‘as a separate and distinct caste in the list of the Scheduled Castes’. ¹⁰⁷

Pant eventually took up the issue, but pointed out there was nothing he could do about the legislation governing the listing of Scheduled Castes. He promised to forward the concerns of the Jatavs to London but in the prevailing circumstances did not think the British government would be ‘prepared to take up this small matter’. However, he did offer some amelioration by instructing the government to see if there was any objection to using just the term Jatav across all departments and in government papers and documents. ¹⁰⁸ Eventually, in June 1939 the UP government, while pleading that the ambiguity of the Jatav position ‘still exists’, informed New Delhi that it had ‘issued instructions to the effect that the Jatavas should be treated as a scheduled caste’. ¹⁰⁹ In October it was announced that the ‘Government have decided in future the caste of the community known as Jatav should be recorded as “Jatav” only and not as “Jatav (Chamar)” in all Government records’. ¹¹⁰

By the time of the resignation of the Congress ministry in UP in November 1939, it had recognised the Jatavs as a separate caste and enacted measures to implement the policy. It had also forwarded to London its case for doing so. The Government of India however was far from convinced and some of its prominent representatives, including Census Commissioner Yeatts, Reforms Commissioner Lewis, and his deputy V. P. Menon, were deeply sceptical about the merits of the case. Menon, who at this stage came to play an increasingly prominent role in the deliberations about the Jatavs, was convinced that all the accumulated evidence suggested that the Jatavs were not a separate caste and added that his investigations yielded ‘purely negative results’. Menon circulated a draft policy in June 1939 that stated he was in no doubt that the best policy was ‘to defer consideration of this question until the next census’. ¹¹¹ This opinion, sanctioned by Reforms Commissioner Lewis, was forwarded to London. The matter lingered for the next six months until the impatient Reforms Commissioner in London reminded the New Delhi secretariat of the matter and asked to be brought up to date. Lewis excused the delay by claiming that he had been ‘on the point of replying’, but he felt that the issues had been carefully considered and concluded that no evidence supported the Jatav claim. He recommended that the Secretary of State refuse the Jatav request. The Secretary of State decided he was not minded to follow the advice and intended to grant them separate status. In addition the UP government, after some discussion, also expressed its support for an amendment to the Scheduled Caste Order of 1936. ¹¹²

Menon now floated the idea around the secretariat that, following the resignation of the Congress ministry in UP and the imposition of Governor’s rule, the provincial government ‘may take a more realistic view of the position’. ¹¹³ With this in mind Lewis now played his final cards with London. He wrote again to Dawson pointing out that ‘account may require to be taken of other considerations before the provisional conclusion of the Secretary of State … were implemented’, although in reality there appeared to have been very little that had been ‘provisional’ in the position announced earlier. In particular, Lewis pointed out the range of reservations and implications that had been raised by Yeatts about Jatavs in other provinces, and the potentially burgeoning numbers of the Scheduled Caste population. Lewis emphasised that the proposed action could end up ‘giving rise to embarrassments’ and such a move ‘might seem to require very special justification’. ¹¹⁴

Dawson’s response was brusque and made it clear that London had resolved to proceed without delay and saw ‘no reason … why embarrassment should result’. ¹¹⁵ The Secretary of State had intervened decisively in an attempt to settle the issue, repudiate the position the Government of India had been taking, and give official sanction to the policy that the UP government had been following informally for almost a decade. The India Office had now moved in an attempt to dispel the confusion surrounding the issue and to try to resolve the contradictions that had blighted policy over the previous decade.

The Order in Council that G. B. Pant had thought the British government would not be prepared to change was amended on the 9 December 1941, forwarded to India, and eventually promulgated by the Reforms Office in New Delhi in March 1942. ¹¹⁶ It could be said that the Jatavs at last, despite all the contradictions and confusions in colonial policy, had eventually been recognised as a distinct, separate, and officially sanctioned caste group. On the other hand it might equally be argued that the recognition of the Jatavs had come to pass, in large part, precisely because of the confusions and contradictions in policy, and the inventive exploitation of them by Jatav activists. ¹¹⁷

Conclusion

The history of the circumstances surrounding the consideration and resolution of the case of the Jatavs directs our attention to three conclusions.

First, in general terms we need to question the notion of the coherent functioning and pervasive control of the state in the late colonial period. An examination of the case of the Jatavs indicates very strongly an imperative to look closely at the tensions and contradictions within the structures and processes of the state apparatus that could lead us to doubt its hegemony and coherence. We need to exercise caution and examine in detail the shortcomings of the functioning of the state lest we are led down the route that ends in seeing colonial dominance as something ‘homogenized, abstracted from internal tension, and presented as all-pervasive, virtually irresistible within its own domain’. ¹¹⁸ The official treatment of the Jatavs certainly suggests that the colonial state formation in India was neither ubiquitous nor integrated and the functioning of its various elements was neither coherent nor consistent.

Second, more specifically we need to be aware of the contested nature, the fragility and friability of much of the colonial knowledge that is seen by many as an essential component of the power-knowledge nexus and the hegemony of the state. Classifying and counting may well have been the practices at the foundation of the ‘ethnographic state’, and certainly contributed to elements of its power. The accumulation of colonial knowledge however did not always lead automatically to the accretion of colonial power. The close consideration of the example of the Jatavs demonstrates the incoherence and contested utility of the outcomes of many ‘colonial knowledge practices’. Recent scholarship has pointed generally to the problematic utility of much colonial knowledge and to the ways in which many elements of it were simply inapplicable to the practicalities of governance. This sort of knowledge, spurious as much of it may always have been, was also increasingly losing its usefulness and applicability in the late colonial period. Most importantly, in the new circumstances, where it was confronted by more organised and better-informed challenges to its legitimacy, it was being deprived of its confidence and authority. The notion—in large part derived from Edward Said—of oriental subjects that never answered back, never questioned the definitions imposed on them by the imperial power, and never articulated an alternative version of social reality, is undermined here. The idea that Indians were merely mute ‘passive onlookers to the colonial encounter’ in which ‘European colonizers alone were responsible for introducing new knowledges and practices’ ¹¹⁹ is confronted by the evidence of a robust subaltern challenge to the colonial knowledge practices of classifying and counting. For surely it is the case that we must take much more into consideration the ways in which colonial knowledge, and indeed the whole orientalist perspective—as has been suggested by Aijaz Ahmad—‘might have been received, accepted, modified, challenged, overthrown or reproduced by the intelligentsias of the colonized countries’. ¹²⁰

Finally, the case of the Jatavs reveals the importance of recognising the significance of the engagement of subaltern Dalits with the colonial state. It was as a consequence of their involvement with the formal processes of politics that Dalits were empowered to exploit the ambiguities and contradictions of the practices of the state and engage with the opportunities offered by inconsistencies of colonial practice. It was by exploring, exploiting, and contesting the spaces created by the colonial power that Jatavs, like other Dalits, generated the opportunity to assert their agency and intervene in the reform process. They directly involved themselves with the institutions and the everyday processes of the Raj. Their engagement and intervention were creative endeavours, not some reflexive simple stimulus-response mechanism. It was a calculated, purposeful, and carefully directed intervention. From the earliest involvement of the likes of Khem Chand of Agra, often characterised by a painfully obsequious attitude to the Raj, to their later efforts to secure the best advantage in the outcome of the reforms, Dalits were constantly engaged in asserting their agency through an iterative engagement with the state. Their presence was expressed by an involvement with the structures and processes of the imperial power and not by acting in some unconnected separate domain. And that engagement was in no small part assisted by the contradictions in colonial knowledge practices, by the interstices of uncertainty opened up by the colonial authorities as they falteringly had gone about their flawed business of classifying and counting.