Excerpted from A Part Apart: The Life and Thought of B.R. Ambedkar by Ashok Gopal, published by Navayana on the occasion of Babasaheb’s 132nd birth anniversary.
BEFORE the inevitable rift between Ambedkar and Gandhi, and before the negotiations leading to the Poona Pact of 1932, a basic question regarding the Depressed Classes (DCs) came to the fore: How were they to be enumerated for the purpose of political representation? Ambedkar had definite views on the subject, which were influential, and their consequences are felt even today.
The issue of enumerating the DCs properly had bothered colonial officials for several decades, and the matter gained urgency after the first two Round Table Conferences, as the British government had shown readiness to give these groups political representation in some form or another. As was evident from his first meeting with Ambedkar in Mumbai, even Gandhi had partially accepted this inevitability. The issues that remained were the quantum of representation and the manner in which it was to be provided. Irrespective of the way in which these matters were going to be decided, the government had to arrive at an exact count of the DC population. An exercise to arrive at province-wise numbers began even before the Poona Pact was signed. Underlying the effort were some basic questions. On what basis was a group to be adjudged as being in a ‘Depressed’ state? If only Untouchable groups were to be considered ‘Depressed’, by what parameters were they to be identified? What constituted untouchability exactly?
As the foremost leader of the DCs and as a person whose opinion commanded weight, Ambedkar had a significant role in determining the answers to these questions. The whole process was also framed by earlier efforts undertaken by the British government to define Untouchables, and the basic challenge that was faced, that is, untouchability manifested in varying ways and intensities across jatis and locations.
For Indian political players who identified themselves primarily as Hindus or Muslims, the definitional challenge was framed by another reality— the quantum of political representation that would be given to Muslims vis-à-vis Caste Hindus. Population numbers had become a political weapon, and the official number of the DCs could tilt the Hindu–Muslim power balance.
The first challenge faced by the British administrators was the absence of reliable listings of Untouchable groups. There wasn’t even a commonly used term for all these groups, and their historical origins were unclear.
As Ambedkar noted in The Untouchables: Who Were They and Why They Became Untouchables? (1948), on the basis of P.V. Kane’s monumental History of Dharmasastra, ‘Ashprashya’, literally ‘untouchable’, was used in some ancient Hindu texts to refer to some social groups, but other texts used terms such as ‘Antya’, ‘Bahya’, ‘Antyavasin’ and ‘Antyaja’ for groups that were apparently treated as untouchable. The listing of social groups under these categories was not consistent, and descriptions and rules applied to them showed that their common attribute was ‘impurity’ rather than untouchability. Impurity was either circumstantial or recognised only by Brahmins, whereas untouchability implied a permanent, hereditarily transmitted status recognised by all groups in society. Hence, Ambedkar argued, the aforementioned terms did not refer to Untouchables. These groups arose only after the Buddha posed a challenge to Brahmin claims to authority. Groups that refused to join the Brahmin counter-reaction were deemed Untouchable (BAWS 7: 359–69).
In a more recent analysis, the Savarna leftist historian Vivekananda Jha found that in ancient Dharmasutra texts only one group, Chandala, was linked to “permanent and hereditary pollution” (1997: 23). [Jha dismissed Ambedkar’s theory on the origin of untouchability as the handiwork of a partisan political leader who had lacked the discipline of a historian (1975: 35). There was more than that consideration in his response to Ambedkar. In a paper on caste and untouchability, Jha said the Bhagavad Gita was “in favour of eradication of untouchability”, and Gandhi’s “arduous campaign to eradicate untouchability”, which was derived from the Gita, “could not be ignored”. As for Ambedkar, Jha thought it fit to mention him only in an endnote: “B.R. Ambedkar’s singular contribution in this regard (‘self-assertion by the Dalit themselves’) is a matter of record and duly acknowledged” (1997: 28, 30).] Shvapaka and Antyavasayin may also have had the same status. Subsequently, according to Jha, a few more groups like Pulkasa and Meda were added to the category, along with Mritapa and Matanga, which may have been subgroups of Chandalas (24). Chandala, Nesada, Pukkusa, Vena and Rathakara were described as despised castes in the Buddhist Pali canon, and Chandala, also known as Matanga, and Sovaga were “equally despised” in Jain texts (25). An eleventh-century Persian traveller, Alberuni, identified twelve Untouchable castes in north India, and by the end of the early medieval period, the number had risen to around twenty. Subsequently, the number increased steadily according to Jha (26).
Across India, no collective name was in use to identify Untouchable groups. That itself showed they had no place in Hindu society. Challenging this condition, Jotirao Phule coined ‘Atishudra’, to identify all groups suffering from “the prohibition of untouchability” (2012: 95). As both Shudras and Atishudras were regarded as “lowly and contemptible” by “Arya Brahmins” (201), Phule hoped to form a united Shudra–Atishudra front against the Brahmins. But this hope was not fulfilled, and the term he had coined for the Untouchables did not gain currency.
On what basis was a group to be adjudged as being in a ‘Depressed’ state? If only Untouchable groups were to be considered ‘Depressed’, by what parameters were they to be identified? What constituted untouchability exactly? As the foremost leader of the DCs and as a person whose opinion commanded weight, Ambedkar had a significant role in determining the answers to these questions.
In the Madras province, ‘Pariah’ was used by colonial officials for several Untouchable jatis. According to one of the several explanations given for the origin of the term, it was derived from ‘para’ or ‘parai’, the name for a drum. Hence it was inferred that the term originally referred to a low-ranked jati of temple drummers. However, M.C. Rajah pointed out that the so-called Pariahs did not beat drums, and the people who did beat drums had another name, Ambattans. Hence, he argued, ‘Pariah’, was a corruption of ‘pariyan’, and probably referred to the original “lords of the soil”, as ‘par’ meant land and ‘iyan’ meant lord in Tamil (Rajah 1925/2012: 42–43). Rajah’s interpretation reflected versions of the Aryan invasion theory that had gained ground in the province, positing that Dravidians were the original inhabitants of the land. An early DC thinker, Iyothee Thass, had argued that the Untouchables were the original Tamils (Poorva Thamizhan), and from that idea, a new term emerged around the turn of the nineteenth century, Adi Dravida, or original Dravidians. But it could obviously not become an acceptable term across India.
In his pioneering study of the ‘bahishkrut’ population of India done in 1908, V.R. Shinde observed that ‘antyaja’ was used often to refer to jatis that were said to be neech, or low and ignoble. However, the term, which literally meant ‘last’, did not connote untouchability. ‘Asprushya’ did, but was not commonly used, and was not easy to pronounce, Shinde argued (1908/1999: 186).
By the 1880s, some British officials had started using a term that became part of the language of government and Indian politics: Depressed Classes. The use of the term, whose origin is not known, was ambiguous. [In England, the term ‘submerged tenth’ was used to describe people who were, in the opinion of the Brahminical theosophist Annie Besant, “ignorant, degraded, unclean in language and habits; people who perform many tasks which are necessary for society but who are despised and neglected by the very society whose needs they minister” (in Natesan 1912/1977: 43). The term had become popular after it was used by William Booth, the founder of the Salvation Army, in his book In Darkest England (1890).] Neither the causes nor symptoms of the ‘depressed’ status were identified, and the term was not used exclusively for groups that suffered untouchability. What was meant by ‘class’ was not defined, and caste was ducked.
The evasion of caste was the outcome of a practical problem faced by the census officials. They had decided that the caste system was the backbone of Indian society, but they could not identify, categorise or describe castes easily. This was clear in the first large-scale census conducted in the subcontinent, in 1871–72. Across provinces, different terms and categories were used by census officials to list castes. An extreme case was Bengal and Assam. Only sixty-nine castes were identified, but these were put under as many as thirteen categories that reflected either status ranking, such as ‘superior’ and ‘intermediate’, or occupations like trading, farming and weaving (Waterfield 1875: 21). A number of ‘outcaste’ groups were listed in a consolidated statement. However, in the provincial reports that formed the basis of the statement, untouchability had not been explicitly and uniformly used as a criterion. In many cases, ‘lower caste’ or ‘semi-Hindooised aborigines’ had been treated as ‘outcaste’. The notion of hereditary pollution was suggested only with respect to a few jatis by the use of epithets such as ‘impure’ and ‘unclean’ by the enumerators (Waterfield 1875: 22–23).
W.C. Plowden, census commissioner for North-Western Provinces for the 1871–72 census and later census commissioner for India, observed that the “whole question of caste” was so “confused” that he hoped an effort to obtain information on the subject would not be made again (Dirks 2002/2010: 202). Subsequently, in the 1881 census, the issue of caste categories and caste names was treated superficially. Only three categories— Brahmins, Rajputs and ‘other castes’— were used, and surveyors were told to club different caste names under ‘some generally known’ names (208).
The superficial treatment of caste is clearly seen in the Gazetteer of the Bombay Presidency issued in over twenty volumes from 1873 onwards. The volume on Pune district, published in 1885, listed social groups as ‘communities’, of which only some such as ‘Kunbis’ and ‘Brahmans’ were listed as castes (Govt of Bombay 1885/1989a: Ch. III). Most of the other identified groups were also castes with distinctive names, but these were classified as occupational groups under categories such as traders, craftsmen, servants, labourers and beggars. Non-occupational categories were also used: Depressed Classes, ‘unsettled tribes’, and religious categories such as ‘Musalmaans’. Under Depressed Classes, the gazetteer listed only four castes (431–44): Dhor; Halalkhor, or Lal Begi, a term used loosely to include Bhangi, Dhedi (Dhed) and Mhetar (Mehtar), Mhar (Mahar), and Mang. How these groups were ‘depressed’ in comparison to several other groups of low socio-economic condition was not explained, and untouchability was not used as the main criterion, as several jatis that suffered untouchability, such as Chambhar and Holar, were not included in the category.
Across India, no collective name was in use to identify Untouchable groups. That itself showed they had no place in Hindu society..
The merging of occupation and caste status was repeated in the 1891 census, but the practice was criticised by H.H. Risley, the commissioner of the 1901 census, who believed that castes could be identified by establishing a relation between status positions and the ‘physical type’ of different social groups. To get evidence on the second variable, which was based on a racial theory of caste, Risley embarked on an anthropometric exercise, to find patterns in the dimensions of the noses and skulls of people of different groups. And to establish the relative social status of different groups, he told the superintendents of the 1901 census that they had to classify groups by “social precedence as recognised by native public opinion” (Risley 1915: 111).
The anthropometric exercise could not be carried out throughout India for all castes, but Risley was determined to identify caste hierarchies, at least separately for each province. Accordingly, he gave provincial census superintendents some guidelines. Displaying remarkable knowledge of caste practices, he told the superintendents to take cognisance of the fact
that Brahmans will take water from certain castes; that Brahmans of high standing will serve particular castes; that certain castes though not served by the best Brahmans, have nevertheless got Brahmans of their own, whose rank varies according to circumstances; that certain castes are not served by Brahmans at all, but have priests of their own; that the status of certain castes has been raised by their taking to infant-marriage or abandoning the remarriage of widows; that the status of some castes has been lowered by their living in a particular locality; that the status of others has been modified by their pursuing some occupation in a special or peculiar way …. (Risley 1915: 111–12)
Mapping those indicators loosely to the chaturvarna model, Risley suggested a five-level grouping of castes: (i) twice-born castes, namely those linked to Brahmin, Kshatriya and Vaishya varnas; (ii) Satsudra, or ‘high-level’ Shudras such as Kayasths; (iii) Jalacharania Shudra, or people belonging to Shudra jatis from whom Brahmins and members of the higher castes could accept water; (iv) Jalabyabahariya Shudra, or people belonging to Shudra jatis from whose hands a Brahmin was not supposed to accept water, and (v) Asprushya Shudra, or “castes whose touch is so impure as to pollute even Ganges water” (Enthoven 1902: 186).
Risley’s effort to establish caste ranking provoked a vigorous reaction from many Indians. Numerous bodies claiming to represent particular jatis sprang up and submitted petitions demanding a high varna ranking in relation to the claims of certain other jatis. That reaction added to the more serious, long-term effect of the government’s effort to identify castes, which was pointed out two decades later by one of the census superintendents of Punjab:
We deplore the caste system and its effect on social and economic problems, but we are largely responsible for the system which we deplore. Left to themselves such castes as Sonar, Tarkhan and Lohar would rapidly disappear and no one would suffer… Government’s passion for labels and pigeon-holes has led to a crystallisation of the caste system, which, except amongst the aristocratic castes, was really very fluid under indigenous rule. (Middleton and Jacob 1923: 343)
The point about fluidity of boundaries among low-ranked castes is questionable, but the enumeration of specific castes definitely led to the crystallisation of caste identities. Risley’s move accelerated the process. But he did not see it that way. Instead, he saw the flood of petitions from jati associations as proof of his hunch that caste was an institution of “remarkable vitality” and of importance to even those educated people who were “sometimes alleged to be anxious to free themselves from the trammels of the caste system” (Risley 1915: 112).
The evasion of caste was the outcome of a practical problem faced by the census officials. They had decided that the caste system was the backbone of Indian society, but they could not identify, categorise or describe castes easily.
Risley’s five-level classification could not be applied across India. For example, it was of little use in the Bombay province, as Brahmins there generally accepted water only from other Brahmins, and that too from members of their own sub-caste. R.E. Enthoven, the superintendent of the 1901 census operations in Bombay province, also observed that identification of ‘twice-born’ castes was likely to “give more trouble than is worth”, as several jatis “freely” claimed Brahmin, Kshatriya or Vaishya origin (1902: 186). The most significant of these claims was of Marathas, who claimed Kshatriya status.
Bombarded by claims and counter-claims on caste status, the 1901 census could not conclusively establish the social gradation of castes, and no further attempt to fix caste hierarchies was made in subsequent censuses. However, the 1901 census had two lasting effects. First, the effort of various castes to officially upgrade their status continued, through various caste associations (Blunt 1931: Ch.XI; Rudolph and Rudolph 1960: 5–22). Second, the introduction of a new category, Asprushya Shudra, led to the first official use of the term ‘Untouchables’ (Charsley 1996: 2). Among the various census reports of 1901, the term was used in only the report for Rajputana, which listed ‘Untouchable castes’, as the last of seven groups among Hindu and Jain castes. The Untouchable group was further divided into two sections, one comprising castes engaged in leather work, and the other comprising ‘the lowest castes’ (Bannerman 1902: 133). In other provincial census reports, the term used was ‘Unclean Castes’— the Bombay report listed around thirty jatis under this category (Enthoven 1902: 192). Nevertheless, ‘Untouchables’ had slipped into the language of the government, and untouchability came to be officially recognised as a phenomenon determining social boundaries and living conditions.
Two people who did not belong to the category of Untouchables played an important role in bringing the term to the mainstream: Sayajirao Gaekwad, the ruler of the princely state of Baroda, and G.A. Natesan, a nationalist Tamil journalist and publisher who had started a monthly, The Indian Review, in 1900 (Charsley 1996: 6–7). Between 1909 and 1911, Natesan published a series of articles on the Depressed Classes in his monthly, and subsequently reprinted most of the articles in a book, The Depressed Classes (1912). Sayajirao Gaekwad was the main contributor, and he highlighted the harm caused by the “theory of untouchableness”. He argued that untouchability added to common causes of low living status such as poverty and illiteracy, and so ‘Depressed Classes’ was too ‘elastic’ and ought to be replaced by ‘Untouchables’ (Charsley 1996: 6–7). Sayajirao had already been using the term, as in an address to the Depressed Classes Mission of Bombay in 1909 (Speeches 1928: 244–45). S.V. Ketkar, whose studies in the US were funded by Sayajirao, also used it as an adjective in Chapter II of his The History of Caste in India (1909). [The first use of ‘Untouchable’ as a noun in an academic paper was most probably in Stanton 1920.]
The British government took decisive action in tune with Sayajirao’s suggestion only some two decades later. Until then, ‘Untouchables’ was the most common term of use in the public domain, along with ‘Depressed Classes’. Other popular terms included: Panchama, which referred to a supposed fifth varna; Suppressed Classes, a term Gandhi borrowed from Vivekananda, and used for some time after his return to India from South Africa (CWMG 18: 375); Adi Dravida, which was officially adopted by the Government of Madras in 1922 (Geetha and Rajadurai 1998: 176); Adi Hindu, a term used in a movement led by urban, literate Untouchables in the United Provinces, to convey Bhakti egalitarianism as well as ancient racial origin (Singh 2009: 574–85); and ‘so-called Untouchables’, a term used by nationalist newspapers.
Bombarded by claims and counter-claims on caste status, the 1901 census could not conclusively establish the social gradation of castes, and no further attempt to fix caste hierarchies was made in subsequent censuses.
The effort to define and enumerate the Untouchables had meanwhile taken a political turn. In 1906, a group of elite Muslims had succeeded in securing an assurance from Viceroy Minto that the interests of Muslims, as a category of people different from Hindus, would be considered while granting increased political representation to Indians in legislatures. In his address to Minto, the head of the Muslim delegation, Mahomed Shah Aga Khan, had made a contentious claim, which is italicised in the following extract from a newspaper report cited by Ambedkar in The Untouchables: Who Were They and Why They Became Untouchables?:
The Mahomedans of India number, according to the census taken in the year 1901, over sixty-two millions or between one-fifth and one-fourth of the total population of His Majesty’s Indian dominions, and if a reduction be made for the uncivilised portions of the community enumerated under the heads of animist and other minor religions, as well as for those classes who are ordinarily classified as Hindus but properly speaking are not Hindus at all, the proportion of Mahomedans to the Hindu majority becomes much larger. (BAWS 7: 312)
The implied claim that Untouchables were not ‘properly speaking’ Hindus came into the forefront when the census of 1911 was conducted. Untouchables were going to be counted as Untouchables. At the same time, the demand for assured representation for Muslims was expected to be met by the British government. Hence, the enumeration of Untouchables raised the fear that it was a ploy to weaken the case of Hindus against the demand of the Muslims. When the censuses of 1881, 1891 and 1901 were conducted, Savarna officials resisted the British attempts to classify Untouchables as Hindus (Mendelsohn and Vicziany 2000: 27–28), but now the situation had reversed. In “From millions to fractions”, an incomplete essay that would have been written sometime in the mid-1930s, Ambedkar reported:
It was said that this attempt of the Census Commissioner was the result of a conspiracy between the Musalmaans and the British government to divide and weaken the Hindu community. It was alleged that what was behind this move was not a genuine desire to know the population of the Untouchables, but the desire to break up the solidarity of the Hindu community by separating the Untouchables from the Touchables. Many protest meetings were held all over the country by the Hindus [who] condemned in the strongest terms this plan of the Census Commissioner. (BAWS 5: 232)
The census commissioner, E.A. Gait, went ahead with the enumeration of the Untouchables, and plunged into another contentious question: How do you define a Hindu? The task was more difficult than identifying followers of monotheistic faiths such as Islam and Christianity.
Despite protests, the census commissioner, E.A. Gait, went ahead with the enumeration of the Untouchables, and plunged into another contentious question: How do you define a Hindu? The task was more difficult than identifying followers of monotheistic faiths such as Islam and Christianity. In earlier censuses, Hindus were identified in remarkably practical ways, as Denzil Ibbetson, the author of the 1881 Punjab census report explained: “Every native who was unable to define his creed or describe it by any other name than that of some recognised religion, or a sect of such religion, was held to be classed as a Hindu” (Kaul 1912: 109).
Gait was not satisfied with this approach, and devised a set of ten tests to help census surveyors enumerate Hindus by a process of exclusion. The surveyors had to identify “genuine Hindus”— by which he meant Caste Hindus— as those who did not meet any one of the following criteria: (i) They denied the supremacy of Brahmins. To be included in this category were (a) sectarian groups that owed their origin to a revolt against Brahminical supremacy, (b) aboriginal tribes, (c) ‘low castes’, who being denied the ministration of Brahmins, had “retaliated” by “professing to reject” the Brahmins. (ii) They did not receive a mantra from a Brahmin. (iii) They denied the authority of the Vedas. (iv) They did not worship ‘great’ Hindu gods. (v) They did not have any “good” Brahmins as family priests. (vi) They did not have any Brahmin priests at all. (vii) They could not enter “ordinary” Hindu temples. (viii) They were believed to cause “pollution” by touch, or within a certain distance. (ix) They buried their dead. (x) They ate beef, and did not revere the cow (Gait 1913: 117).
Gait’s elaborate list did not make the job of classification easy for census officials. If anything, they found a number of variations they could not deal with according to the list. For example, the Arya Samajists in Punjab denied the supremacy of the Brahmins, but they did not regard themselves as non-Hindus. Then, there were many ‘so-called Hindus’ whose religion had a ‘strong Muhammedan flavour’, and there were many Hindus across India who made pilgrimages to Muslim shrines. Contrariwise, there were many converts to Islam who were not ‘genuine’ Muslims. In Gujarat, for instance, a group called Matia Kunbis were followers of a Sufi pir, Imam Shah, and his successors, but whose religious ceremonies were administered by Brahmin priests. There was another group known as Sheikhadas, who called on both Muslim and Hindu priests for their wedding ceremonies. To Gait’s dissatisfaction, the superintendent of the census in the Bombay province chose to use the term “Hindu–Muhammedans” to enumerate such groups, rather than relegate “the persons concerned to the one religion or the other as best as he could”. But Gait was being overzealous, for as he himself admitted, the “boundary line” between Hindus, Sikhs and Jains was “even more indeterminate” (1913: 117–18).
With regard to tests that could be applied to enumerate Untouchables, the picture was similarly confusing (Gait 1913: 117). In the Central Provinces and Berar, more than half the population was reported to be not receiving a mantra from a “recognised Hindu guru”, but only a quarter did not worship the “great Hindu gods”, and were not served by “good” Brahmin priests. A third were denied access to temples, but the number of those who were reported to “cause pollution by touch” was considerably lower. Of the thirteen castes reported to be causing pollution by touch, nine did not eat beef, and of the eight castes that ate beef, four were not regarded as polluting, and two were allowed access to temples. On the other hand, in Bengal, Bihar and Odisha, there were fourteen castes that ate beef, and none were allowed to enter temples.
Gait had not recognised that the Untouchables shunned the Brahmins as much as the Brahmins shunned them, and he did not inquire into these phenomena
Gait’s conclusion was that the “extent” to which his tests were satisfied varied in different parts of the country (1913: 117). But this guarded admission could not hide the fact that he had failed to achieve what he had set out to do, that is, arrive at a clear, countrywide enumeration of categories such as Hindus and Untouchables. The superintendent of the Punjab census, a Kashmiri Brahmin, Hari Kishan Kaul, was more candid in admitting failure. In a “modern state of society”, he said, it was “impossible” to draw boundaries using tools like Gait’s tests. Kaul’s report for the Punjab only listed the groups that fulfilled each of the ten tests, and the total number of people matching each test. For the rest, he left it to “critics” to “arrive at such conclusions as they may” (1912: 109).
Years later, Ambedkar noted in The Untouchables: Who Were They? that Gait’s tests of enumerating Untouchables had been “one-sided” and missed the main point (BAWS 7: 313–14): Gait had not recognised that the Untouchables shunned the Brahmins as much as the Brahmins shunned them, and he did not inquire into these phenomena. “It is the ‘why’ of these facts which is more important than the existence of these facts,” Ambedkar said.
Photographs from Vijay Surwade’s archival collection, courtesy of Navayana. These pictures are for one-time use in The Leaflet, and cannot be reproduced or shared anywhere without the permission of Vijay Surwade, the archivist of the Dalit movement from Kalyan, Maharashtra.