The first decennial census of India was undertaken in 1871–2, which coincided with the foundation of the Anthropological Institute in London in 1871. This was not entirely accidental, since both events reflected the intellectual spirit of the age. The systematic, ‘scientific’ anthropology of India developed in conjunction with the censuses and reflected, among the British, the ‘relentless need to count and classify everything they encountered [that] defined much Victorian intellectual activity’. [1] By the early twentieth century, there was a huge official literature containing ethnographic data and anthropological theories about Indian society, especially about caste, tribe and race. Around 1871, however, anthropologists in Britain ceased to take much interest in India or, more exactly, in the villagers, mostly Hindus or Muslims, living on the plains. They retained an interest in archaic ‘survivals’ found among the lowest castes, as well as in the ‘primitive’ hill and forest tribes, whose members made up approximately 5 per cent of the subcontinent’s population. The vast majority of Indians, however, were too ‘civilised’ to matter much to metropolitan anthropologists, until the scope of the discipline changed after the Second World War and Indian Independence in 1947. This article discusses the colonial ethnology – social anthropology’s antecedent – of British India during the two decades before the changes inaugurated by the census in 1871. [2]
Colonial Knowledge of Indian Society
Orientalist scholarship on languages and classical literature was crucial for the development of Indian ethnology in the nineteenth century. The history is complicated and this brief outline of it mainly relies on the work of Thomas Trautmann and George Stocking. [3] For my discussion, the most important consideration is the ‘twinning of comparative philology and ethnology before mid-century and their separation in, roughly, the 1870s’. [4]
William Jones (1746–94) famously discovered the Aryan or Indo-European language family at the end of eighteenth century. At first, all Indian languages were thought to be derived from Sanskrit, but by the early nineteenth century it was known that the main south Indian languages belonged to a separate Dravidian family. In the 1840s and early 1850s, the ‘aboriginal’ languages – which included both the main south Indian languages and those of tribal peoples – were generally thought to belong to the one Dravidian family, but it then became clear that many tribal languages belonged to a third Austro-Asiatic, Munda or ‘Kolarian’ family. On the Himalayan and north-eastern frontier, Tibeto-Burman languages were spoken. During the nineteenth century, too, the ‘Aryan invasion’ thesis became conventional wisdom among writers on India. It postulated that in ancient times, light-skinned Aryans, who spoke Sanskrit or a cognate language and were relatively ‘civilised’, invaded the subcontinent from the north-west. They then spread east and south, and conquered or displaced the dark-skinned, more ‘primitive’ and predominantly Dravidian indigenous inhabitants, whom they subjugated. The origins of the caste system, it was further claimed, primarily lay in the racial inequality between the Aryans and the Dravidians, together with the Kolarians. Notwithstanding repeated admonitions that racial and linguistic divisions were not homologous, nineteenth-century ethnologists of India were preoccupied by the classification of castes and tribes as Aryan, Dravidian or Kolarian, each a supposedly distinct racial-cum-linguistic group with its own distinctive customs and institutions. By around 1870, however, among the leading metropolitan scholars, historicist ethnology allied to philology was being displaced by ‘classical evolutionism’, in Stocking’s phrase, to produce a new anthropology exploring the evolution of society and culture from ‘primitive barbarism’ to ‘advanced civilisation’. [5]
Another aspect of orientalism relevant to this article is that Jones and his colleagues translated The Laws of Manu and other classical Hindu texts describing the four ‘classes’ or varnas : Brahman priests and literati, Kshatriya kings and warriors, Vaishya farmers and traders, and Shudra servants. The relationship between the classical varnas and the countless castes or jatis now in existence was much debated by colonial authors. Mountstuart Elphinstone (1779–1859), in his History of India – published in 1841 and widely regarded as authoritative in the mid-nineteenth century – said Manu described a learned ideal, not the real state of affairs, but he also assumed that since Manu’s era the caste system had become much more differentiated, so that the number of castes had grown. [6] After Elphinstone, some ethnologists were convinced that the varna system was the true historical origin of the contemporary caste system, whereas others (like modern anthropologists) treated it as only a Brahmanical model.
A branch of colonial knowledge that also had some relevance to Indian ethnology was represented by the settlement reports compiled for every district of British India. Settlements determined who were the landholders responsible for paying revenue to the government and how much they should pay. Although settlement reports were mainly about the land, land tenure systems and other agrarian matters, some of them – especially on districts in the Punjab – included ethnographic material on local castes, tribes and religions as well. The official literature on Indian land tenure was important for Henry Maine’s scholarship, which I discuss below, but agrarian questions themselves fell outside the scope of ethnology and anthropology as normally understood in the colonial era.
George Campbell and the Ethnology of India, 1852–66
The great majority of ethnological and anthropological work carried out in British India was undertaken by administrative officials or other civilians employed by the government, or by army officers. In the literature, various terms have been used for these colonial anthropologists, including ‘official anthropologists’, which usefully indicates both their status as officials and the fact that their work – ‘official anthropology’ – was mostly undertaken on behalf of the government. In the late nineteenth century, the majority of the leading official anthropologists were members of the elite Indian Civil Service (ICS), but before then they occupied a wider range of positions within the Raj.
The most significant political events in India in the mid-nineteenth century were the great rebellion or mutiny of 1857, and the transfer of sovereign authority from the East India Company to the British Crown a year later. Partly because the British believed the uprising was provoked by their violation of caste and religious traditions, a better understanding of them was now seen as imperative, and that was one major reason for the growing commitment to systematic inquiry into Indian society and culture after 1858.
Before then, quite a lot of ethnographic data had been collected through land settlements, limited censuses, geographical surveys and other inquiries, as well as the work of several scholarly polymaths, for example, James Tod (1782–1835) and Brian Hodgson (1800–94), who were East India Company officers mainly serving in Rajasthan and Nepal, respectively. [7] Between 1850 and 1871, however, the most influential colonial, official ethnologist was George Campbell (1824–1892), who arrived in India in 1842 and began his career as a junior district officer in the North-Western Provinces (modern Uttar Pradesh). He was promoted through a series of posts in the same province, and later in the Punjab, Bengal and the Central Provinces. In the early 1860s, he was a high court judge in Calcutta and he ended his career as lieutenant-governor of Bengal from 1871 to 1874. [8]
Campbell’s voluminous history, Modern India, was published in 1852 and was heavily indebted to Elphinstone’s work. The book’s opening chapters described Indian society. According to Campbell, aborigines still lived in the mountains, but the Hindus had occupied the plains of India even ‘prior to history’ and, though their origins were unknown, they had belonged to the ancient eastern civilisation extending from India to Egypt. Like other ancient peoples, Hindus were divided into priestly, military, mercantile and labouring classes ; the Brahman priests and Kshatriya warriors became the ruling group. Manu, however, described a Brahmanical ideal of how things ought to be, not how they really had been. [9] In an early epoch, Hindus moved southwards and eastwards ; later, ‘republican tribes’ of Rajputs and Jats migrated into the Punjab from the west, subjugating those they conquered and establishing their own corporate villages and internally democratic institutions. The Rajputs were eventually conquered by the invading Muslims ; afterwards, other groups, notably Marathas and Sikhs, became regionally powerful. [10] Campbell drew attention to the need for a ‘combined philological and ethnological survey’, but also provided an outline that anticipated a longer study in 1866. [11] In his description of caste, Campbell insisted that other writers had relied too much on ancient texts. In contemporary India, he said, ‘the variety of castes is infinite’, and although some ranked higher than others, the system was ‘most decidedly a division of labour or of clans, and not a division of rank’, because caste ranking was ‘a mere matter of opinion, and exclusiveness is generally mutual’. When higher castes did enjoy superiority, ‘it is rather political than the result of caste’. The rules restricting eating and marrying were absolute, however, and they, ‘in fact, constitute caste’. [12]
Denzil Ibbetson (1847–1911) was the superintendent of the 1881 census of the Punjab and his report was the first important anthropological text to emerge from the decennial censuses. [13] Ibbetson did not cite Modern India, but it may have influenced him, since Campbell clearly anticipated Ibbetson’s occupational theory of caste as a special form of the division of labour, as well as his emphasis on the political basis of caste rank. Campbell notably did not explain caste as a hierarchical system with reference to race, as Herbert Risley (1851–1911), who criticised Ibbetson’s occupational theory, would do later. [14] Ibbetson and Risley, together with William Crooke (1848–1923), whose approach to caste was closer to Ibbetson’s than Risley’s, did most to develop the systematic anthropology of India in the late nineteenth and early twentieth centuries. [15]
In 1866, Campbell’s long article on Indian ethnology was published in the Journal of the Asiatic Society of Bengal in a ‘special number’ on ethnology, alongside an article on Kolarian tribes by E. T. Dalton (1815–80), an army officer and the commissioner of Chota Nagpur, a tribal region within Bengal (now in Jharkhand). The Asiatic Society of Bengal (ASB) was British India’s principal centre of scholarship in a variety of fields, including ethnology. [16] In 1866, as we shall see, the ASB was drawing up proposals for an ‘ethnological congress’ and urging the government to collect information comprehensively ; Campbell, who was one of its leading members, explained that his article set out an ‘ethnological skeleton’, which could assist inquirers to collect and classify data on the ‘races and classes’ of India in a uniform manner. [17] As was normal at the time, Campbell often used the term ‘race’ loosely to mean ‘people’, or caste or tribe, but sometimes more specifically to denote an ethnological division, such as Aryan or Dravidian ; similarly, ‘caste’ and ‘tribe’ were also used for different kinds of ‘class’ or social group in some contexts and interchangeably in others. Campbell’s article drew on personal observation, information from fellow officials and other people, and material he had read. He repeatedly emphasised the gaps in his account, but it was ‘the most authoritative synthesis of its day’, according to Trautmann. [18] His coverage of south India was scanty, however. Campbell identified five criteria for classifying the people : physical appearance (indicating racial division), language, religion, laws (relating to marriage, property, etc.), and manners and mental characteristics. [19]
Campbell divided the Indian population into three main categories. The first comprised the ‘black aboriginal tribes of the interior hills and jungles’, the remnants of a pre-Hindu race, who have ‘many ethnological features in common’. The second category, making up the vast majority, were ‘modern Indians’ belonging to various tribes and castes, both Hindu and Muslim. Moreover, Campbell argued, no true ethnological line existed between north and south, because although south Indians spoke Dravidian languages (like many aborigines), the Hindus were predominantly Aryan and south Indian society was Aryan in its structure, manners, laws and institutions. The third, small category was made up of ‘borderers’ on the subcontinent’s Himalayan and north-eastern frontiers, who belonged to tribal groups of mixed descent. [20] The principal division, however, was between the aboriginal tribes and modern Indians, and it was primarily racial, rather than linguistic, because people in the first category spoke Dravidian or Kolarian languages, and those in the second Dravidian or Aryan ones. [21]
The bulk of Campbell’s article was an ethnographic survey of various tribal groups, which was fairly short, and of different caste groups, which was much longer. [22] It described the Brahmans of different provinces ; Jats, Rajputs and other agricultural tribes ; mercantile groups ; scribes ; artisans ; ‘inferior and helot classes’, which included untouchable Dalits ; southern groups (covered very briefly) ; and finally, various miscellaneous groups. Campbell, as was common in his day, peppered his text with prejudicial comments about different groups, characterised as physically attractive or ugly, industrious or lazy, honest or dishonest, and so on and so forth. Nonetheless and despite his own caveats, too, Campbell’s survey was more comprehensive than anything previously published and it demonstrated his remarkably broad knowledge.
The Ethnological Congress and Early ‘Tribes and Castes’ Handbooks, 1865–72
In 1865, Joseph Fayrer (1824–1907), a professor in the Calcutta Medical College, wrote to the Asiatic Society of Bengal to suggest that it and the government should jointly organise an ethnological exhibition by assembling a large collection of people from different races, who could then be studied for the great benefit of ‘anthropological science’. Fayrer was primarily interested in the physical anthropology of race and thought that examining living people would be more fruitful than the usual method of collecting crania. In March 1866, at an ASB meeting, Campbell spoke about the initial plan for ‘a great ethnological congress of all the races in India in its widest sense’ – meaning the rest of southern Asia and the Pacific Archipelago as well – which would be held in Calcutta. The ASB supported the plan and wrote to the government of India, enclosing Fayrer’s letter, to explain that it would be immensely valuable to bring together typical living examples of each race, so that their physical characteristics could be observed, their spoken languages ascertained, and their prominent social customs investigated and described. The ‘Ethnological Congress’, as it was now called, would be a ‘fitting adjunct’ to the General Industrial Exhibition planned for 1869–70, because it could take advantage of ‘the specimens of many Indian and Asiatic tribes and races’, which would then throng Calcutta. The ASB was sure that the plan could be realised with government support and suggested it would also be useful to send all the provincial governments a circular asking for detailed statements of their inhabitants’ different races. The society was confident, too, that the government would give its support, because it had always done so for ‘any scheme tending to spread a knowledge of the benefits of civilization and to advance learning’. In May, the official circular was sent to provincial governments.
The ASB also approached the government of Bengal with a second proposal for a smaller congress of the tribes in Bengal, Nepal, British Burma, and the Andaman and Nicobar Islands, to be held in conjunction with a forthcoming agricultural exhibition. Campbell supported the proposal and thought it would be easy to organise in Calcutta, where there was already a huge variety of people in the bazaars and among the city’s labourers, who could be employed as coolies to build the exhibition and afterwards to sit as exhibits, ‘without in any way degrading men and brethren to the position of animals’. Campbell himself had a further suggestion for an equivalent congress in the Punjab where, unlike Bengal with its aboriginal tribes, the inhabitants included ‘the highest types of the human race’ – ‘the very purest Arians, fair, robust, high-featured, eminently handsome’ – who stood in contrast with many Europeans, whose ancestors, especially in the ‘labouring classes’, must have mingled with local aborigines. [23]
Yet another idea was put forward at the 1866 annual meeting of the British Association for the Advancement of Science (BAAS) by Walter Elliot, a retired Indian administrator, whose own ethnological work is discussed below. Although Elliot liked the ASB’s plan for the Ethnological Congress and the more limited one for Bengal, he more strongly favoured an intermediate ‘assemblage of all the races found in British India’, which he divided into three categories : those descended from ‘aboriginal races and the servile classes’ (mainly Dalits), from ‘Tamil or Dravidian races’, and from ‘Hindi immigrants whose language has been modified and perfected by Sanscrit’, by which he primarily meant ‘Aryans’. [24]
When speaking at ASB meetings, Campbell emphasised the ethnological significance of race more than in his published article. He also said it was important to collect data to resolve ‘great questions’ of the day, such as whether the Negro was a brother-man or fit only for slavery, as recently asserted at an Anthropological Society meeting in London. [25] Campbell’s own opinion on this question was unclear, but the overall tenor of discussion about the main ethnological congress and the supplementary Bengal one inconsistently mixed patronising concern to protect the people who would be exhibited, expectations that the government would make them appear, and racist generalisations about ‘wild creatures’ from aboriginal tribes and people from other subordinate groups. Campbell’s Punjab proposal and Elliot’s both envisaged the presence of caste Hindus. For the main congress and the Bengal one, however, only the exhibition of tribal people was discussed ; hence – to cite Gyan Prakash – an unavoidable ‘liminality’ would emerge, because when ‘the colonizer staged the colonized as man, he disavowed the racist polarity … that enabled his discourse’, so that the ‘subordinated “aborigine” emerged as ‘the kindred of the dominant European’ and was himself a ‘specimen’ who could represent ‘man’. [26]
Whether most ASB members, European or Indian, shared Campbell’s enthusiasm for the Ethnological Congress is not revealed by reports on its meetings. But in 1868, Fayrer – who was now the ASB’s president – acknowledged that financial difficulties were delaying the project, which had in any case been reduced in scale. At the same time, Fayrer highlighted the progress in ethnology being achieved with official assistance, despite the country’s economic problems. Fayrer also mentioned his recent letter to the government of India urging it to ensure that ethnological reports were submitted by all provincial governments ; he had drawn attention, too, to the BAAS’s recent letter to the British government, which emphasised ‘the great and urgent importance’ of obtaining reports on ‘the physical form, manners, and customs’ of the Indian population, especially the tribal groups. Replying to Fayrer, the government said reminders about the reports had been sent to provincial governments. But it also declared that it could not help the ASB to collect ‘specimens’ or offer further assistance, so that the society, together with other learned institutions and private individuals, would have to organise the congress itself. [27] Moreover, according to Dalton, several serious difficulties had emerged. For instance, the ‘wild tribes’ had to be fully represented at the congress, but sending ‘those strange shy creatures’ to Calcutta was impractical. The head of the Assam government said that twenty tribal ‘specimens’ could not be conveyed to and fro without risking serious casualties and he was concerned that if any of them did fall ill, it could cause ‘inconvenient political complications’. [28] Somewhat surprisingly, this potential problem never seems to have troubled Campbell, despite his many years of experience handling sensitive political issues in India.
In the end, neither the General Industrial Exhibition and the Ethnological Congress, nor any of the smaller ethnological congresses suggested by the ASB, Campbell and Elliot ever took place. The only one that did was a congress in Jubbelpore in the Central Provinces (Jabalpur in modern Madhya Pradesh), held alongside the city’s exhibition in 1866–7. Richard Temple (1826–1902), chief commissioner of the Central Provinces (and later lieutenant-governor of Bengal), was inspired to organise it by the ASB’s proposals. According to Campbell, Temple – in a particularly egregious example of the liminality identified by Prakash – said he would ensure, if necessary with ‘a little lucre’, that the ‘wild creatures’ came to be displayed ‘for the edification of their more civilized fellow-humans’. [29] Temple established an ethnological committee for the exhibition, chaired by Alfred Lyall (1835–1911), an intellectually distinguished official who wrote an early gazetteer containing ethnographic material, as well as highly influential essays on religion. [30] The ethnological committee’s report, which relied on reports by district officers of variable quality, amounted to an ethnographic survey of the region, although some basic measurements of bodily size for about thirty tribal men (and one woman) were tabulated in a final section. [31] Given the material available to him, Lyall produced a moderately interesting survey of the Central Provinces’ tribal groups, divided into Kolarian, Dravidian (chiefly Gond) and Helot (Dalit). But the absence of any reference to ethnographic information collected from individuals displayed at the exhibition suggests that actually there was none. Indeed, it is worth noting that when the various congresses were discussed at ASB meetings, their real utility for ethnographic inquiry never seems to have been examined.
Although Calcutta’s congresses never took place and Jubbelpore’s was of dubious value, the plans made for them were not entirely fruitless. At some stage, it was agreed that an exhibition catalogue for the Ethnological Congress, based on the provincial government’s reports, would be prepared by Dalton, who plainly had the necessary expertise. His article on the Kolarian tribes (especially Mundas and Santhals) in Chota Nagpur, much of it based on ethnographic material he had collected, appeared alongside Campbell’s in the ASB’s Journal in 1866 and was republished in England. [32] But when Dalton looked at the few provincial reports sent to Calcutta and forwarded to the ASB by late 1867, he found them largely useless. So he abandoned an exhibition catalogue and instead wrote Descriptive Ethnology of Bengal, which was published with the support of the society and the government of Bengal in 1872. Dalton’s lengthy book, which included much of his own data, collected in Assam as well as Chota Nagpur, was divided into nine sections, eight covering different groups of tribes and a final section on Aryans, primarily the Bengali castes. Tribes were grouped by their ‘most obvious affinities’ ; the principal division was between Dravidians and Kolarians, but there were also non-Aryan ‘Hinduised Aborigines’ and a variety of mixed peoples who were hard to classify. [33] Within sections, individual tribes and castes each had one entry of historical and ethnographic description, some more detailed than others. The unusually elegant volume also included lithographs showing ‘typical’ men and women from different tribal groups. Dalton’s work was a clear precedent for the official ‘tribes and castes’ handbooks for the provinces inaugurated twenty years later by Risley’s Tribes and Castes of Bengal, which actually incorporated a lot of Dalton’s material. [34]
Another noteworthy precedent was Henry Elliot’s ethnographic glossary for the North-Western Provinces. In the course of his official career, Elliot (1808–53) collected information about tribes and castes, rural life and other subjects, and in the 1840s he prepared a glossary. Although uncompleted, Elliot’s text was incorporated into a larger glossary of judicial and revenue terms published in 1855 by H. H. Wilson (1786–1860), the leading orientalist of his era. John Beames (1837–1902), an official who wrote extensively on languages and was also interested in ethnology, produced a revised edition of Elliot’s glossary. He separated the entries on different topics, so that the new work’s first part, published in 1869, consisted of an alphabetical series of ethnographic and historical descriptions of tribes and castes. [35] M. A. Sherring (1826–80), a Protestant missionary, also published a glossary of tribes and castes in Benares in 1872. He divided them into Brahmans ; Kshatriyas or Rajputs ; Vaishyas, Shudras and others ; and, finally, ‘aboriginal tribes and inferior castes’. He drew a distinction between Brahmans, Kshatriyas and some Vaishyas of ‘comparatively pure blood’, on the one hand, and the rest with ‘impure or mixed blood’, on the other. [36] The entries in Sherring’s glossary typically included information on legends and history, occupations and subdivisions, religious affiliations, and other miscellaneous matters. Elliot’s and Sherring’s volumes were both later cited as useful sources by some official anthropologists and they, too, were early prototypes for the official tribes and castes handbooks, most pertinently Crooke’s The Tribes and Castes of the North-Western Provinces and Oudh. [37]
The Ethnological Society of London and India in the 1860s
Some papers on India by official anthropologists were appearing in Britain by the 1860s, mostly published by the Ethnological Society of London (ESL) ; they had been few and far between earlier. The ESL’s Transactions reprinted Dalton’s article on the Kolarians (1868) and also published substantial ethnographic articles on tribes in the Darjeeling region by Archibald Campbell (1869a), the ESL’s vice president and a doctor who had been superintendent of Darjeeling before retirement, and on the Todas and allied tribes in the Nilgiri Hills in south India by John Shortt (1869), a member of the Indian Medical Service in Madras. [38] Another article on the Nilgiri tribes by W. Ross King (1870), an army officer, appeared in the Anthropological Society of London’s Journal of Anthropology. [39] Almost all the other papers on India, however, were brief, superficial sketches sent in by the prolific Shortt or other contributors. Because ‘the publication of information in Indian journals does not generally spread in England’, according to Campbell, ethnographic work published only in India, even by the ASB, was probably little read. [40]
In 1869, Thomas Huxley, the ESL’s president, chaired a meeting devoted to India, the first of a proposed series on the British Empire’s ethnology. Huxley told the audience it was ‘fitting that the greatest of the possessions of the Empire’ should come at the beginning and the papers would be by ‘Indian veterans, speaking from long practical experience of the country and its people’. He then outlined the geography and ethnology of the subcontinent, and stated plainly that the great majority of the population was divided into two groups : the Aryan people of Hindustan and the Dravidian people of the southern Deccan. [41] The meeting’s first paper was delivered by Walter Elliot and the second by George Campbell. In addition to theirs, the ESL’s Journal also published articles (which I shall not discuss) by Archibald Campbell on the Lepcha tribe of the Himalayan region, Philip Meadows Taylor, an army officer and former administrator, on prehistoric archaeology, and George Fosberry, also an army officer, on mountain tribes on the North-West Frontier. [42]
For ease of presentation, I begin with George Campbell’s paper, which was a shorter, revised version of his 1866 article. Campbell started with the aboriginal tribes, many existing ‘in the very lowest stage of barbarism’, although others ‘have become comparatively civilised’ and were engaged in agriculture and other occupations. [43] By language, the tribes were divisible into two or possibly three classes. The first comprised the Dravidians of south and central India, including the large Gond tribe, as well as some northern groups, such as the Oraons. The second class of Kolarians were found in the north and included, for example, the Santals. Elsewhere there were tribes, such as the Bhils in the west, who spoke modern north Indian dialects, and their classification as a distinct third class was problematic. In addition, there were also frontier tribes of non-Indian origin. The second part of Campbell’s paper examined the Aryans and covered similar ground to his earlier article, but contained more speculation about the Aryan race and the effect of successive waves of immigration. He provided ethnographic sketches of some major castes, but far less detailed than in his previous article. In some places, he modified his earlier comments. Thus, for example, discussing the ‘helots’, Campbell was unsure in 1866 whether Chamars in north India ‘are considered to be properly Hindus’ – which in some respects is still a significant sociological question – whereas in 1869, the issue of race was more in evidence when he observed that Chamars, like other similar groups in the north, were dark-skinned but lacked clear aboriginal features, despite being descendants of conquered peoples. Campbell’s paper had no clear conclusion except for his confident expression that India’s ‘races and castes’ made up ‘a great field for inquiry’. [44]
Walter Elliot (1803–87) served in India from 1821 until 1860, mostly in the Madras Presidency, where he started as a junior district officer, before being posted to the South Mahratta country (in modern Maharashtra). In 1837, he returned to Madras and ended his career as a member of the governor’s council. Elliot was a proficient linguist who was interested in Indian natural history, epigraphy, numismatics and many other subjects, as well as ethnology, and he amassed a huge quantity of material on castes and tribes through personal inquiry and from other sources. This material was not published, except for some short pieces in Robert Sewell’s memoir, but it is preserved in manuscript collections (compiled around 1845–55). Particularly interesting is a uniquely detailed account of a village goddess’s festival incorporating a buffalo sacrifice, which Elliot watched in 1829, most probably in the South Mahratta country. [45]
Elliot’s long paper for the ESL divided the Indian population into three classes : the aboriginal races ; the ‘civilised classes’, who were mixed descendants of the aborigines and later immigrants ; and the ‘latest intruders’. Indians divided themselves into northerners and southerners, but this did not produce an accurate ethnological classification, for which religion was important, owing to the intermingling of primitive beliefs and practices with later Brahmanical ones in contemporary Hinduism. Elliot applied his religious test most fully with reference to the Parias (Pariahs) – variously known as Parayas or Vettiyans in Tamil, Mahars or Mangs in Marathi, and so on – who were excluded from the village society of the higher castes. He described the village goddess’s festival mentioned above and observed that the officiating priests were Parias (actually Mahars and Mangs), who on this sole occasion could participate in village affairs ; he also explained that a very similar festival was held throughout southern India. On these two grounds in particular, Elliot deduced that the ancestors of the Parias or ‘servile classes’ were the subcontinent’s earliest inhabitants, whose other descendants were now found in a tribal state of freedom only in central India. Noting that Parias and tribal people were physically similar, he concluded that south India’s autochthones were an aboriginal race worshipping local deities with human and animal sacrifices, who were represented today by servile classes and free tribes. [46]
Elliot next discussed the south Indian Kurubars or Kurumbars and allied shepherd castes, who were distinguished by their steadfast truthfulness, like the tribal Santals of the north ; he argued that this virtue, alongside other features, pointed to a common origin, even though Kurubars spoke a Dravidian language and Santals a Kolarian one. Comparable evidence about moral character supposedly showed that all the fierce ‘predatory classes’ – such as Kallars in the south and Gujars in the north, and some tribal groups – were a distinct ‘race’ with a common origin. The southern ‘demon worshippers’, notably the Shanars, formed another group. Most south Indians, however, belonged to ‘an endless variety of castes’ of mixed Dravidian and Aryan origin, although the people of Malabar (Kerala) were separate. [47] Elliot admitted that his views were not widely shared and the overall picture remained confused, but he was sure, in contrast to Campbell, that Aryans had not deeply penetrated the south ; he also emphasised the need for knowledge about the ‘pre-Aryan population’ making up the ‘long oppressed races’, which the British should protect and lift up. [48] All in all, Elliot’s paper was rather a ragbag, which organised an eclectic mixture of ethnographic and other material into an intermittently opaque classificatory scheme. He probably impressed his London audience with his knowledge, but I suspect they found him a lot more confusing than Campbell.
After ethnology divided from philology and evolutionism emerged as the dominant paradigm around 1870, ‘race science’ developed strongly. Fair-skinned, civilised Europeans were presumed to be evolutionarily superior to dark-skinned ‘savages’, ‘barbarians’ and other ‘primitive’ people, which led in the Indian context to a ‘racial theory of Indian civilization’, whose fullest anthropological expression, according to Trautmann, was Risley’s racial theory of caste, first published in 1891 and modified in 1901. [49] Risley is not directly relevant to this discussion of Campbell and Elliot, because they were writing earlier. Nonetheless, the growing prominence of race as an explanatory principle in the colonial anthropology of India during the Victorian age clearly was a significant development. Trautmann detected signs of it in Campbell’s 1866 article, in which ‘strains were beginning to show between the classifications of languages and the classifications of races’, although it was actually more prominent in his comments on the Ethnological Congress, as I noted above. [50] The tension between linguistic and racial classifications was visible, too, in Campbell’s and Elliot’s 1869 papers, whereas for Huxley, race was plainly definitive and he was certain that an Aryan-Dravidian dichotomy explained Indian ethnological variation.
Trautmann is obviously right about the shift from language to race in British ethnology and anthropology, but I am unconvinced that it had as much effect on the official anthropology of India as he and others have suggested. It is true that Campbell and Elliot, who assumed that racial groups and language families were normally congruent, had difficulty making their classification schemes fit the facts about castes and tribes ; on the other hand, the two men often disagreed anyway, notably about the basis for unity of the aboriginal population, and how far Aryans and Dravidians had intermingled in the south. Significantly, though, they also employed additional criteria to bring order to their subject-matter, especially religion in Elliot’s case, and drew attention to cases where race or language were not definitive, such as Bhils speaking north Indian languages or Kurubas and Santals displaying the same moral worth. More striking than any reliance on either race or language in Campbell’s and Elliot’s classifications was, first, their recognition that neither was actually definitive, so that other factors mattered too, and, secondly, their repeated acknowledgement that they could not fully account for all the ethnological variation. The same provisionality was evident in the work of Dalton, Henry Elliot and Sherring. As Campbell particularly emphasised, much more systematic ethnographic data were needed to understand the structure of the Indian population and its division into castes and tribes.
Campbell and the other writers I have discussed were mainly interested in India, rather than general anthropology. As past or present members of the imperial ruling class, they were all sure of their superiority over their Indian subjects, but all of them had also spent so long in the subcontinent that they understood how much they did not know, how necessary it was to collect and collate more ethnographic evidence, and how premature was any theorisation about Indian society. Even by 1870, therefore, there was a divide between the official anthropologists of India and metropolitan ‘armchair theorists’, such as Huxley. Moreover, the division was not only about general theoretical as opposed to specifically Indian interests, because as it widened – irrespective of the impact of race theory – it also reflected anthropology’s growing preoccupation with what were believed to be the most rudimentary primitive societies. One significant factor in this change of direction had an Indian dimension that deserves some scrutiny.
Henry Maine and John McLennan
Henry Maine (1822–88) was the most esteemed writer on India of his age. Maine, of course, was primarily a legal scholar, not an anthropologist. Nevertheless, his ‘patriarchal theory’ – or more exactly criticisms of it – played an important part in the development of Victorian anthropology and India’s place within it.
Maine conventionally assumed that Aryans made up a racial group with its own distinctive civilisation, but he also believed it was divided into a ‘progressive’ European branch and a ‘stationary’ Indian one. Contemporary India could therefore shed light on ancient Aryan society and the British, for example, could learn about their own distant past ; ‘In India’, he wrote, ‘these dry bones live’. [51] Only occasionally did Maine refer to the non-Aryans in India. [52] Maine’s well-known thesis in Ancient Law, published in 1861, is that ancient society was made up of groups, not individuals, and the elementary unit was the patriarchal, agnatic family group. As societies evolved, kinship gave way to local contiguity as the basis for social and political solidarity, and the patriarch’s authority declined, so that personal rights and property increasingly rested on free agreement among individuals, and status gave way to contract.
In Ancient Law, the Indian village community played a small but significant role in the argument because, according to Maine, it was both ‘an organised patriarchal society and an assemblage of co-proprietors’ ; it was also ‘of immense antiquity’. [53] In north India, the village community was always ‘founded by a single assemblage of blood-relations’, but outsiders were periodically ‘admitted to the brotherhood’ ; in south India, there were communities originating in more than one family, but the tradition of ‘an original common parentage’ was preserved. Thus the Indian village community was always a kinship group either in reality or by fiction, like the Roman gens or house. [54]
From 1862 to 1869, Maine was in India as the Law member of the viceroy’s council. His second book, Village-Communities in the East and West, published in 1871, compared the Indian village community with its ancient Teutonic counterpart. Maine enjoyed immense authority as an expert on Indian affairs, and in Village-Communities he referred to his first-hand knowledge of official material, especially the settlement reports, which ‘constitute a whole literature of very great extent and variety, and of the utmost value and instructiveness’. [55] Maine’s book was based on a lecture series, so lots of footnotes might have been inappropriate. Nonetheless, hardly any sources were cited and E. B. Tylor, in his review, reasonably thought too many details had been suppressed, so that future editions might usefully contain an appendix with ‘positive particulars of village organizations’ in India and elsewhere. [56]
Others have criticised Maine’s use of evidence more directly. In 1896, B. H. Baden-Powell (1841–1901), British India’s leading expert on land tenure systems, strenuously avoided personal criticism and pointed to gaps in the settlement reports available to Maine when he was writing. [57] Even so, Baden-Powell made it plain that Maine had his facts wrong, especially because ‘joint villages’ were the norm only in north and north-west India, pre-eminently the land of the Aryans, and even there alternative forms of co-proprietorship existed. ‘Raiyatwari villages’, in which land is owned by separate holders, the raiyats, prevailed in the rest of the country, including the Dravidian south. The more primitive Dravidians, therefore, had individual property in land, whereas the more advanced Aryans had collective property. [58] Among modern scholars, Louis Dumont has highlighted several statements in Village-Communities that qualified or undermined the assertion in Ancient Law that the Indian village community was a patriarchal society and assembly of co-proprietors. [59] As Dumont commented, such statements suggest that Maine, in his second book, tacitly recognised the existence of different types of village community without being explicit about them. Yet even before Ancient Law was written, George Campbell, in Modern India in 1852, had drawn a distinction between two types of village communities, which Baden-Powell refined. Very briefly, in the more widespread ‘ancient type’ of the Hindus, property rights in land belonged to individual cultivators, whereas in the later, ‘democratic form’ commonest in the north where the Jat and Rajput tribes settled, rights in land – although individual cultivators enjoyed shares of it – belonged to a ‘proprietary community’ whose members claimed to be a ‘brotherhood or clan’. [60]
It is hard to believe Maine never read Campbell’s Modern India when writing Ancient Law. But the two men were acquainted in Calcutta, where Campbell saw a great deal of Maine during 1863 and came to admire a man with the ‘most acute intellect’ he had known. [61] Furthermore, Maine must have known about Modern India when writing Village-Communities, because he mentioned Campbell’s 1870 essay on land tenure, which summarised and cited Modern India’s discussion of villages. [62] Maine also thanked Campbell personally for confirming that the statements in his text ‘coincide in the main with the results of his [Campbell’s] own experience and observation’. [63] In any case, though, whatever he actually read and the two men talked about, Maine’s reiterated description of the Indian village community as ‘an assemblage of co-proprietors’ ignored Campbell’s discussion of individual ownership.
By 1871, however, John Lubbock (1834–1913) and J. F. McLennan (1827–81) had criticised Maine’s thesis that the earliest societies were founded on patriarchy, rather than primitive promiscuity, communal marriage or matriliny. [64] Indeed, in Primitive Marriage, published in 1865, McLennan ‘wrote more in refutation of Maine’s ideas than of anyone else’s’ according to Peter Rivière. [65] In his book, McLennan referred to material from all over the world, including India. One source was an ethnographic report from 1859 on the hill tribes of Manipur, north-east India, by W. McCulloch, an army officer and the political agent in Manipur. McLennan had discussed this report in an 1863 article, which had ‘evolutionary undertones’, as Rivière notes. [66] But the article was mainly about social, cultural and linguistic differences among the Manipur tribes, and McLennan’s cautious conclusion was that all customs, like languages, were liable to change and tended towards variety. [67] McCulloch’s report reappeared in Primitive Marriage, cited to support McLennan’s thesis about exogamy and marriage by capture. [68] Yet caution had now been thrown to the winds. Thus, for example, McLennan added one speculative hypothesis about ancient exogamy and marriage derived from McCulloch’s brief comment on marriage restrictions in Manipur to another one about Manu’s catalogue of Hindu marriage rules, even though there was never any connection between Manipur tribal customs and Brahmanical texts. [69]
Maine did not answer McLennan in any detail ; instead he tersely reiterated that ‘the Patriarchal Family’ is ‘a primary fact in the history of society’ supported by the bulk of the Indian evidence. [70] Maine, in fact, never truly rebutted his critics, especially his leading opponents, McLennan and L. H. Morgan (1818–81), who insisted that patriarchy was a very late development in human evolution. [71] Instead, Maine effectively narrowed his field of inquiry to early Indo-European society, which meant he could present himself as a sober historian, rather than a wild speculator like McLennan, although he was never in fact a systematic thinker. [72] No doubt, scholars like Maine and McLennan were and are important for their ideas, rather than their use of evidence. Even so, it is worth noting that Maine’s thesis about the patriarchal family and the co-proprietary kinship group neglected the known facts about Indian villages, while McLennan’s critique of Maine depended on conjectures about societies across the world that were no sounder than those about Manipur tribes.
Yet careless use of evidence has nothing to do with why Maine lost his argument with McLennan. Moreover, what matters is that the outcome of their dispute made one significant contribution to the rise of anthropology as an evolutionary science primarily concerned with ‘primitive’ society. Ancient Aryans, it now seemed clear, were not very ancient at all, especially when compared to Aboriginal Australians. Furthermore, although the small minority of Indians who belonged to hill and forest tribes were also ‘primitive’, albeit less so than the Australians, the great majority were not, so that unless they possessed some surviving customs from an earlier stage, the ‘civilised’ villagers settled on the plains – who were mostly Hindus or Muslims – ceased to interest leading metropolitan anthropologists. The latter mainly wanted to understand the evolutionary relationship between ‘primitive’ and ‘civilised’ society in general, whereas British India’s official anthropologists wanted to understand the classificatory relationships among castes, tribes and other social units within Indian society. Some topics, such as marriage customs and regulations, totemism and exogamy, or animism and its relationship with ‘higher’ religions, did interest both groups. Nonetheless, even when they shared common ground, there was a significant difference in outlook between metropolitan and official anthropologists, which meant that the study of Indian villagers, as well as their caste systems, became more and more peripheral to British social anthropology until its scope expanded to ‘complex’, peasant societies after the Second World War.
Conclusion
For some leading official anthropologists in the late nineteenth century, India’s growing peripherality was irksome. Thus Ibbetson emphasised that in the Punjab, despite the absence of aboriginal tribes, there was a large variety of social types raising ‘a series of problems sufficiently intricate to satisfy the most ardent ethnologist’. It was, thought Ibbetson, amazing and even shameful that Tylor, Lubbock, McLennan and others were gathering information about little known tribes, when ‘the everyday routine of every Panjab village would afford them infinitely better examples’. [73] In a paper first delivered at a BAAS meeting and later published in the Journal of the Anthropological Institute, Risley repeated complaints he had made earlier about ‘the comparatively scanty use that has been made of the great storehouse of ethnographical data which British rule has thrown open to European enquirers’. The available material on India was usually used unsatisfactorily, too, partly because ‘distinguished ethnologists’ such as Herbert Spencer and Lubbock were devoid of ‘some slight personal acquaintance with even a single tribe of savage men’, so that they could not evaluate their sources critically. [74] The solution, Risley believed, was systematic research undertaken by men on the spot, so that India’s vast ethnological diversity could be incorporated into scientific, anthropological inquiry. In their own eyes – and rightly so – India’s official anthropologists were never just fact-gatherers for theorists back home and they were criticising armchair anthropologists long before Malinowski and his colleagues did so.
Notwithstanding the criticisms voiced by Ibbetson and Risley, both men, together with Crooke and Lyall, were described by Tylor as administrators ‘entitled to speak with authority on various branches of anthropology’ in his presidential address to the RAI in 1892. [75] Risley’s eminence as an anthropologist was particularly recognised, not only in Britain, but in continental Europe, too, especially in France and Germany. A year before his death in 1911, Risley also became the RAI’s president, and Crooke in retirement was a member of its council. In England, therefore, British India’s official anthropologists were neither belittled nor excluded from the RAI’s inner circles. Nevertheless, as I have shown, Tylor and other influential, metropolitan anthropologists of his day were almost entirely unconcerned with the topics on which the official anthropologists had done much of their most important ethnographic and theoretical work, especially the Indian caste system.
When the ESL decided to hold a meeting on India in 1869, partly because, as Huxley said, it was the most important and most populous territory in the British Empire, this was actually the high-water mark of ethnological or anthropological engagement with the subcontinent in Britain during the colonial period. Thereafter interest ebbed among British and other metropolitan anthropologists, until India and its majority population started to command attention again after both the Second World War and the British Raj ended and the modern anthropology of village India began.
References
Books and Articles
Anderson, J. D. 1912. Obituary : Sir Herbert Risley. Man 12 : 1–4.
Baden-Powell, B. H. 1896. The Indian Village Community. London : Longmans.
Bose, P. N. 1885. Ethnology. In Centenary Review of the Asiatic Society of Bengal from 1784 to 1883, pt. 3, Natural Science, 90–100. Calcutta : Thacker, Spink.
Burrow, J. W. 1966. Evolution and Society : A Study in Victorian Social Theory. Cambridge : Cambridge University Press.
Campbell, A. 1869a. On the Tribes around Darjeeling. Transactions of the Ethnological Society of London 7 : 144–59.
Campbell, A. 1869b. On the Lepchas. Journal of the Ethnological Society of London 1 : 143–57.
Campbell, G. 1852. Modern India : A Sketch of the System of Civil Government. London : John Murray.
Campbell, G. 1866. The Ethnology of India. Journal of the Asiatic Society of Bengal 35, pt. 2 (Special number on ethnology) : 1–152.
Campbell, G. 1869. On the Races of India as Traced in Existing Tribes and Castes. Journal of the Ethnological Society of London 1 : 128–140.
Campbell, G. 1870. Tenure of Land in India. In J. W. Probyn, ed., Systems of Land Tenure in Various Countries (Cobden Club), 149–231. London : Macmillan.
Campbell, G. 1893. Memoirs of my Indian career, ed. C. E. Bernard. London : Macmillan.
Crooke, W. 1896. The Tribes and Castes of the North-Western Provinces and Oudh. Calcutta : Superintendent, Government Printing.
Dalton, E. T. 1866. The ‘Kols’ of Chota-Nagpore. Journal of the Asiatic Society of Bengal 35, pt. 2 (Special number on ethnology) : 153–98.
Dalton, E. T. 1868. The ‘Kols’ of Chota-Nagpore. Transactions of the Ethnological Society of London 6 : 1–41.
Dalton, E. T. 1872. Descriptive Ethnology of Bengal. Calcutta : Government of Bengal.
Dumont, L. 1970 [1966]. The ‘Village Community’ from Munro to Maine. In Religion, Politics and History in India : Collected Papers in Indian Sociology, chap. 6. Paris : Mouton.
Elliot, H. M. 1845. Supplement to the Glossary of Indian Terms, compiled by H. H. Wilson. Agrabe : Secundra Orphan Press.
Elliot, H. M. 1869. Memoirs, Folk-lore, and Distribution of the Races of the North Western Provinces of India : Being an Amplified Edition of the Original Supplemental Glossary of Indian Terms, ed. J. Beames. London : Trübner.
Elliot, W. 1869. On the Characteristics of the Population of Central and Southern India. Journal of the Ethnological Society of London 1 : 94–128.
Elphinstone, M. 1889 [1841]. The History of India : The Hindu and Mahometan Periods. London : John Murray.
Fosberry, G. 1869. On Some of the Mountain Tribes of the N. W. Frontier of India. Journal of the Ethnological Society of London 1 : 182–93.
Fuller, C. J. 2023a. Anthropologist and Imperialist : H. H. Risley and British India, 1873-1911. New Delhi : Social Science Press/London : Routledge.
Fuller, C. J. 2023b. Colonial Ethnography and Theories of Caste in Late-Nineteenth-Century India, in BEROSE International Encyclopaedia of the Histories of Anthropology.
Gottschalk, P. 2013. Religion, Science, and Empire : Classifying Hinduism and Islam in British India. New York : Oxford University Press.
Hiltebeitel, A. 1985. On the Handling of the Meat, and Related Matters, in Two South Indian Buffalo Sacrifices. L’Uomo 9 : 171–99.
Huxley, T. H. 1869. On the Ethnology and Archaeology of India : Opening Address of the President. Journal of the Ethnological Society of London 1 : 86–93.
Ibbetson, D. C. J. 1916. Panjab Castes. Lahore : Government of the Punjab. (Reprint of Report of the Census of Panjab 1881, 1, Text, chap. 6)
Kuper, A. 1991. The Rise and Fall of Maine’s Patriarchal Society. In A. Diamond, ed., The Victorian Achievement of Sir Henry Maine, 99–110. Cambridge : Cambridge University Press.
Kuper, A. 2005 [1988]. The Reinvention of Primitive Society : Transformations of a Myth. London : Routledge.
Lubbock, J. 1870. The Origin of Civilisation and the Primitive Condition of Man. London : Longmans.
Lyall, A. C. 1870. Gazetteer for the Haidarabad Assigned Districts, commonly called Berar. Bombay : Education Society’s Press.
Lyall, A. C. 1899. Asiatic Studies : Religious and Social. London : John Murray.
Maine, H. S. 1861. Ancient Law. London : John Murray.
Maine, H. S. 1883. Dissertations on Early Law and Custom. London : John Murray.
Maine, H. S. 1890 [1871]. Village-Communities in the East and West. London : John Murray.
McCulloch, W. 1859. Account of the Valley of Munnipore (Selections from the records of the Government of India, Foreign Department). Calcutta : Bengal Printing Co.
McLennan, J. F. 1863. Hill Tribes in India. North British Review 38 : 392–442.
McLennan, J. F. 1970 [1865]. Primitive marriage, ed., P. Rivière. Chicago : University of Chicago Press.
Meadows Taylor, P. 1869. On Prehistoric Archaeology of India. Journal of the Ethnological Society of London 1 : 157–81.
Metcalf, T. R. 1995. Ideologies of the Raj. Cambridge : Cambridge University Press.
Morgan, L. H. 1877. Ancient Society. London : Macmillan.
Pels, P. 2000. The Rise and Fall of the Indian Aborigines : Orientalism, Anglicism, and the Emergence of an Ethnology of India, 1833-1869. In P. Pels and O. Salemink, eds. Colonial Subjects : Essays on the Practical History of Anthropology, 82–116. Ann Arbor : University of Michigan Press.
Prakash, G. 1999. Another Reason : Science and the Imagination of Modern India. Princeton : Princeton University Press.
Risley, H. H. 1886. Primitive Marriage in Bengal. Asiatic Quarterly Review 2 (July–October) : 71–96.
Risley, H. H. 1890. The Study of Ethnology in India. Report of the British Association for the Advancement of Science, 59th meeting, 1889, 785.
Risley, H. H. 1891a. The Tribes and Castes of Bengal : Ethnographic Glossary. Calcutta : Bengal Secretariat Press.
Risley, H. H. 1891b. The Study of Ethnology in India. Journal of the Anthropological Institute 20 : 235–63.
Risley, H. H. 1915 [1908]. The People of India. Calcutta : Thacker, Spink and Co.
Ross King, W. 1870. The Aboriginal Tribes of the Nilgiri Hills. Journal of Anthropology 1 : 18–51.
Sewell, R. 1896. Sir Walter Elliot of Wolfelee : A Sketch of his Life, and a Few Extracts from his Note Books. Edinburgh : Private printing.
Sherring, M. A. 1872. Hindu Tribes and Castes as Represented in Benares, 1. London : Trübner.
Shortt, J. 1869. An Account of the Hill Tribes of the Neilgherries. Transactions of the Ethnological Society of London 7 : 230–90.
Stocking, G. W. 1987. Victorian Anthropology. New York : Free Press.
Trautmann, T. R. 1987. Lewis Henry Morgan and the Invention of Kinship. Berkeley : University of California Press.
Trautmann, T. R. 1997. Aryans and British India. Berkeley : University of California Press.
Tylor, E. B. 1871. Review of Maine’s Village-Communities. Quarterly Review 131 (July) : 176–89.
Tylor, E. B. 1892. Anniversary Address. Journal of the Royal Anthropological Institute 21 : 396–412.
Wilson, H. H. 1855. A Glossary of Judicial and Revenue Terms. London : W. H. Allen and Co.
Government and Learned Society Reports
Asiatic Society of Bengal, Proceedings, 1866. Calcutta, 1866.
Asiatic Society of Bengal, Proceedings 1868. Calcutta, 1868.
British Association for the Advancement of Science, Report of 36th meeting, 1866. London, 1867.
British Association for the Advancement of Science, Report of 59th meeting, 1889. London, 1890.
Census of India, 1901, vol. 1, India, pt. 1, Report, by H. H. Risley and E. A. Gait. Calcutta, 1903.
Report of the Census of Panjab 1881, vol. 1, Text, by D. C. J. Ibbetson. Lahore, 1883.
Report of the Ethnological Committee on Papers laid before them and upon Examination of Specimens of Aboriginal Tribes Brought to the Jubbulpore Exhibition of 1866-67. Nagpur, 1868.