16 | Risley, Grierson and the 1901 Census: How Empire Made Linguistic Categories in Colonial India

By Nishita Mandava & Shivakumar Jolad

Published on: 26 June 2026

‘India is a land of contrasts, and nowhere are these more evident than when we approach the consideration of its vernaculars. There are languages with words of a few hundred roots, and there are others with opulent vocabularies rivalling English in copiousness and accuracy of idea-connotation.’

— G.A. Grierson, 1901

In 1901, the British Census Commissioner Herbert Risley wrote to the linguist George Abraham Grierson with an update on the Brahuis, a community spread across Baluchistan and Sindh. They were now being anthropometrically measured, Risley informed him, and these measurements would place their ‘physical type beyond dispute.’ He then asked Grierson a related question: were there linguistic connections between the Brahuis and the Todas of southern India? (Majeed, 2019).

Both men, officers of the empire, were involved in surveying and ethnographic investigations across the Indian subcontinent. In the 1890s, Risley began undertaking anthropometric measurements, a pseudo-science approach which he insisted revealed enduring racial distinctions between supposedly ‘advanced’ Aryans and ‘primitive’ Dravidians.

In short, castes were really races for Risley (Banerjee-Dube, 2011). He believed one’s caste could be scientifically identified through bodily markers such as one’s nose size (nasal index). Meanwhile, Grierson was orchestrating the Linguistic Survey of India (LSI), an ambitious effort to catalogue and classify the subcontinent’s various languages since 1894 (Kidwai, 2019).

What Risley sought from Grierson, then, was evidence that the Brahuis language supported the racial distinctions he believed anthropometric measurements had already uncovered. If linguistic similarities linked the Brahuis to Dravidian-speaking groups such as the Todas, this would, in Risley’s view, reinforce the idea that racial identity, physical characteristics, and language all corresponded to one another. Grierson’s response confirmed this assumption.

Recent genetic studies (Pagani et. al. (2017) and Singh et. al. (2025), however suggest that the Brahui population’s connection with South Indian Dravidian-speaking peoples is primarily linguistic rather than demonstrably genetic. Brahui speakers genetically resemble neighbouring populations of Balochistan—such as the Baloch, Sindhi and Pathan—more closely than Dravidian-speaking groups from India. Even the 2025 study, which included the Oraon/Kurukh population—the closest linguistic relatives of Brahui—found no distinctive shared recent ancestry. The findings therefore support the interpretation that Brahui represents an ancient Dravidian language retained despite extensive population mixing or replacement.

Linguistic classification and racial anthropology in the 19^th century came to complement each other and the correspondence discussed above is a case in point. These men were working in a context wherein anthropometry’s scientific status was assured and this reflected in the correspondence between them. Their engagement became more substantive during the 1901 census. Acknowledging the advances made by the LSI since the previous census, Risley asked Grierson to contribute to the census language chapter and assist in resolving disputed linguistic classifications.

In this essay, we explore how the languages of the subcontinent were categorized, mapped and enumerated in the 1901 census, and the debates and tensions that such an exercise generated.

It might seem puzzling today that language has so often become the basis of intense political flashpoints across India. If we understand language simply as a tool of communication, such struggles can appear outlandish. But it has historically been far more than a means of expression, serving as a marker of community, culture, territory, and belonging. ‘There are few issues, aside from religion, that can mobilise and sustain such passion as the status of language because it is central to collective identity.’ (Esman, 1992). 

Colonial exercises such as ethnographies, surveys and censuses were sites of language study. Colonial officers such as William Jones and Henry Thomas Colebrooke devoted extended periods of their careers to studying the subcontinent’s language families. Missionaries and colonial officials produced dictionaries on Indian languages, which would be utilized as standard grammars by colonial officials.

Expanding these efforts, the census helped transform languages into identities that could be categorized and converted standard languages and linguistic groupings. In the process, they laid the groundwork for many of the linguistic debates and political contests that would shape modern South Asia. In the process, they laid the groundwork for many of the linguistic debates and political contests that would shape modern South Asia.

In this respect, the census taken at the turn of the 20^th century provides fascinating insights into some of the debates and complexities that arose around the British Raj’s efforts to enumerate the many languages of India.

LSI and the Census

In October 1900, Risley wrote to Grierson explaining that the LSI had greatly expanded official knowledge about the subcontinent’s languages since the previous census. To avoid confusion, Risley instructed census officials to refer any ‘doubtful’ language or dialect names to Grierson, who would determine how they were supposed to be classified and where they were geographically spoken (Majeed, 2019). Risley insisted that the census had to follow Grierson’s linguistic framework.

Language as a category had interested colonial officials from the first census report (1872) itself. Initially, it was treated mainly as an indicator of ‘nationality’ and population movement. The 1881 census continued this approach, using birthplace, caste and language to identify broad social groupings, but there was still no standard framework for classifying the colony’s many languages and dialects (Majeed, 2019).

Risley complained that the 1891 census had produced ‘wild’ and inconsistent language classifications. To bring greater uniformity to the process, Grierson was tasked with preparing notes for provincial census superintendents, explaining the linguistic history and characteristics of different regions (Majeed, 2019). These would form the basis of provincial census reports and would guide the local observations from officials in the field.

Grierson went onto author the chapter on language in the 1901 census, which was an early draft of the first chapter of the Linguistic Survey itself. More importantly, the census became the first in colonial India where a trained linguist worked closely with census authorities to establish standardized categories for recording languages. Grierson continued to be associated with the census efforts well until 1931, cementing a close connection between the census and the LSI (Majeed, 2019).

The LSI and the census were quite different in scope. The LSI was an exhaustive multi-volume survey on the various languages and dialects, while the census did not engage with such granular details, especially dialects, as it had to render the data accessible for administration. Accounting for this, Grierson modified his ‘Index of Languages and Dialects’ to be utilized by census officials. Several dialects were omitted ‘on the principle that only main dialects were to be included in the classification’ (Majeed, 2019). These excluded dialects, which were considered to be a mix of multiple languages, were considered to be ‘too’ complex for the census to tabulate.

The enumeration exercise during the 1901 census made it clear that genuine confusion existed regarding the labels of languages and dialects, both for the colonial officers and the speakers.

They contradicted the assumption that a language belonged exclusively to a single region or people, instead stretching across geographies from China to Europe as part of a larger language family.

Multi-Lingual Realities

Drawing on examples from his fieldwork, Grierson observed that linguistic identities were often fluid and could not always be treated as reliable markers of descent. The Brahuis of Baluchistan, for instance, spoke a Dravidian language despite being surrounded by Balochi speakers, while the Kharias were found speaking Munda, Dravidian, or Bengali depending on locality (Grierson, 1903).

He further suggests that linguistic frontiers in India were rarely fixed, with communities frequently abandoning ancestral languages in favour of those spoken by neighbouring groups. In regions such as Malda in Bengal, multilingual villages had even evolved a shared lingua franca for inter-community communication while retaining separate languages for internal use (Its akin to Nagamese, an Assamese-lexified creole language that serves as the primary lingua franca of Nagaland, India. It was developed as a marketplace and trade language between various Naga tribes and the neighboring plains of Assam).

Such examples led Grierson to problematise simplistic attempts to infer racial origins from linguistic evidence alone, emphasising instead the historical processes of migration and adaptation that shaped India's linguistic landscape.

Grierson extended this critique to the census category of ‘tribal dialects,’ arguing that many such classifications rested on misleading assumptions. He takes up the example of ‘Jatki,’ which census officials had described as the language of the Jat tribe. According to Grierson, this designation obscured the fact that Jatki was spoken widely across western Punjab and was not exclusive to the Jats.

Similar problems appeared in the classification of ‘Chibhali’ in the Murree hills (part of Western Himalayas, present-day Pakistan), where a diverse set of local dialects were artificially grouped together and identified as the language of the Chibh tribe, disregarding the fact that the Chibhs themselves conversed in multiple dialects (Grierson, 1903).

Despite these criticisms of linguistic classification, Grierson did not entirely abandon the assumption that language could reveal insights into racial histories. While rejecting what he termed the ‘unholy alliance’ between race and language in contemporary ethnology, he still maintained that languages could sometimes provide clues to racial ancestry (Grierson, 1903).

To demonstrate this, we are offered an example of the Malto language. Spoken in the Rajmahal hills (Jharkhand) and classified as Dravidian, Malto was surrounded by Indo-Aryan languages and appeared to be declining under their influence. Grierson interpreted this situation as evidence that its speakers were originally Dravidian: ‘we are fairly entitled to assume that the dying language is the older tribal one, and that it gives a clue to the latter’s racial affinities.’ (Grierson, 1903).

He argues that the persistence of a language could illuminate the racial history of a population even where other forms of evidence were inconclusive. Despite acknowledging the multi-lingual realities, he remained committed to the Empire’s broader ethnological project of classifying India’s population into coherent racial and social groups.

Language and Language Families

The 1901 census recorded 147 distinct languages as vernacular within the Indian subcontinent. These forms of speech were classified into four primary families: Malayo-Polynesian, Indo-Chinese, Dravido-Munda, and Indo-European. The Indo-European family, particularly the Aryan sub-family, comprised the largest group with over 221 million speakers. The following infographics displays the language families , with the major languages under them (note 'dialects' have been excluded) are represented. Note that the Linguistic Survey was not yet complete, and some of the languages (such as Munda sub-family) were reclassified in later Census. 

Movement and Migration

Just as geologists reconstructed extinct worlds from fossils, philologists believed they could reconstruct extinct peoples and ancient migrations from surviving languages. Before the late eighteenth century, European scholars usually explained languages through biblical narratives such as the Tower of Babel. Language was understood primarily within a theological framework, and Hebrew occupied a privileged position as the language believed to stand closest to humanity’s divine origins (Olender, 2002). 

The emergence of comparative philology transformed this understanding. Scholars such as William Jones noted systematic correspondences between Sanskrit, Greek, Latin, and other languages. This meant that 

language needed to be studied as a historical object that changed over time according to identifiable patterns (Olender, 2002). This dramatically grew the ambitions of philologists who did not simply classify languages but sought to reconstruct histories from how languages evolved. By tracing similarities and divergences between languages, scholars believed they could recover evidence of prehistoric migrations and cultural encounters. 

Operating within this philological tradition and intellectual climate, Grierson classified languages according to comparative philology, and because comparative philology understood language as evidence of historical movement, the resulting classifications could then be used to narrate migration beyond the borders of British India. 

He sought to organize the immense diversity of speech in British India into a coherent classificatory system in the following manner: 

However scientific these classifications appear, languages and dialects were continually evolving through migration, multilingualism, literary standardisation and political mobilisation. The categories recorded in the census would themselves become objects of contestation as communities increasingly organised around linguistic identities during the twentieth century.

Particularly interesting among these language families is Grierson’s description of the Indo-Chinese language family which was a vast linguistic domain extending from Central Asia to the Malay Peninsula.

Grierson drew upon contemporary philological scholarship to propose a common homeland for the peoples who spoke Indo-Chinese languages, locating it in north-western China around the upper reaches of the Yangtze and Yellow Rivers. From this region, he shows that successive waves of migration had spread southwards and westwards over centuries. Some populations entered Burma, others moved into Assam and the Brahmaputra valley (Grierson, 1903).

These migrations formed the backbone of Grierson’s narrative of the Mon-Khmer, Tibeto-Burman, and Siamese-Chinese branches of the Indo-Chinese family. The Mon-Khmers were presented as some of the earliest inhabitants of India.

Subsequent migrations of Tibeto-Burman-speaking peoples gradually pushed them towards coastal regions and isolated upland enclaves. Following the courses of the Brahmaputra, Chindwin, and Irrawaddy rivers, Tibeto-Burman groups spread across Assam, the Naga Hills, Manipur, and Burma, establishing new centres of settlement while leaving behind a mosaic of related languages. Later came the Tai-speaking peoples contributing to the rise of the Shan states and the Ahom kingdom in Assam (Grierson, 1903).

The landscapes of eastern India, according to his account, were a product of repeated movements and encounters that now persist as linguistic traces long after the migrations themselves had faded away from inter-generational memory.

What made him certain about these movements was linguistic evidence. He emphasised that shared grammar rather than words offered more convincing evidence. While words can be easily borrowed between languages owing to their long-standing interactions, grammar structures, according to him, endured. It was these underlying grammatical correspondences that allowed philologists to distinguish between superficial similarity and historical descent. 

Hence, he suggested that languages need to be grouped based on shared grammar structures and not words. This comparative method gave Grierson confidence that the linguistic families presented in the 1901 census reflected genuine historical relationships rather than merely contemporary patterns of speech. In this sense, the 1901 census produced a historical narrative of migration, settlement and linguistic descent. 

The Indo-Aryan Debate

A constant obsession of the colonial state with the Indian subcontinent was the question of Aryan origins, which has continuously captivated the imagination of generations of historians, linguists and archaeologists. British officials and indologists prominently Sir William Jones ( ~1780-1800), Max Müller, and Risley have sought to explain India’s social and cultural diversity through the framework of the Aryan migration theory.

According to this view, a distinct Aryan race of people had entered the subcontinent from outside, bringing with them Indo-European languages and a superior civilisation that subsequently shaped Indian society.

Language was central to these arguments. The discovery of similarities between Sanskrit and European languages such as Greek, Latin, and English fired their imaginations of a shared Aryan ancestry, and linguistic relationships were frequently interpreted as indicators of racial relationships (Olender, 2002).

Languages as diverse as Hindi, Persian, and English which though sound and look very different, were found to share a common linguistic ancestry.

It was within this intellectual climate that Grierson was writing on the Indo-European languages in the 1901 census. Quoting Max Müller in the language chapter, he writes:

‘his oft-repeated warning that the existence of a family of Indo-European languages does not necessarily postulate the existence of one Indo European race, has too often been ignored by writers who should have known better.’ (Grierson, 1903). 

Grierson here attempts to distance himself from those conflated linguistic and racial identities. But he had limited success in doing so and rather slipped into a more ambiguous position on the relationship between language and race. He oscillates between speaking of Aryans as speakers of related languages and as a racial category.

He surveyed competing theories that had placed the Aryan homeland in the Caucasus, the Hindu Kush, North-Western Europe, Armenia, or the Oxus region. He ultimately favours the then-recent view of Otto Schrader, a philologist who located the original home of the Indo-Europeans in the steppe lands of Southern Russia, on the borderlands between Asia and Europe.

The Indo-Aryans entered India through the north-western passes and eventually spread across the Punjab and northern India. The Iranian and Indo-Aryan languages developed from a common ancestral speech before diverging into distinct linguistic families, much of this has also been accepted by contemporary historical scholarship. Modern genetic studies have given strong evidence of steppe ancestry, migration and native intermixing of Indo-European language speakers in India. 

Grierson repeatedly argues that these migrations brought Aryan-speaking peoples into contact with earlier populations. Indigenous groups were often absorbed into Aryan-speaking societies and adopted Aryan languages. This explains why linguistic and physical characteristics did not necessarily correspond (Grierson, 1903).

Even though he explicitly rejected treating Aryan languages as synonymous with an Aryan race in the census, later in his career he suggested that LSI’s linguistic findings could be useful to ethnological inquiry. Grierson oscillated between insisting on a distinction between philology and ethnology and allowing linguistic evidence to inform ethnological theories of race.

As Javed Majeed (2019) rightly observes, the discourse of race in Grierson’s writing is ‘strained.’ While he sought to distance linguistics from race science, he never fully severed the connection owing to the institutional dominance of racial ethnology within the colonial state. 

Grierson made several influential observations on the Indo-Aryan languages. He recognised that certain languages spoken in the north-western mountains, including Shina, Khowar and the languages of Kafiristan, did not neatly fit into a Sanskrit-centred account of Indian linguistic history. To account for this, he distinguished between Sanskritic and non-Sanskritic Indo-Aryan languages and held that the latter preserved older layers of Indo-Aryan speech that had developed independently of Sanskritic influence. 

He further divided the Sanskritic languages into ‘Inner’ and ‘Outer’ families, interpreting their geographical distribution as evidence of successive phases of Aryan migration across the subcontinent (Grierson, 1903). Although Grierson’s Inner-Outer migration model has largely been abandoned by historical linguists, linguistic evidence continues to play a central role in reconstructing the early history of Indo-Aryan languages (Cardona & Jain, 2003). 

It is in such discussions that we see how the 1901 census represented a meeting point between administrative enumeration and comparative philology. For the very first time India’s linguistic diversity was being organised according to historical theories about language families, migration, and linguistic descent. 

Persianisation and Sanskritisation of ‘Hindostani’

The 19^th and 20^th centuries witnessed raging debates over Hindi and Urdu between colonial administrators, Indian nationalists and missionaries. In the mid-19^th century interest in Persian as a language dwindled and was replaced by vernacular languages in many provinces as the official language (Banerjee-Dube, 2011).

It was by the late 18^th and early 19^th centuries when Hindi and Urdu began to emerge as separate literary traditions. Urdu was distinguished by its Persian script and Persian and Arabic vocabulary. Hindi, on the other hand, was written in the Devanagari script and drew more from Sanskrit words.

While those who articulated strict boundaries between these literary traditions often had ideological alignments, both languages remained mutually intelligible in spoken form, reflecting their shared origins. This common linguistic tradition was known by many names, but more popularly as Hindustani.

Hindi and Urdu increasingly came to be associated with Hindu and Muslim communities, respectively, contributing to the emergence of new forms of linguistic identification. Missionaries and colonial administrators also contributed to such discourse. Such developments were not inherently communal as they would become later (Banerjee-Dube, 2011). Instead, they reveal the depths of British criticism of Indian society to legitimise their civilising mission and the bearings this had on the Indian psyche.

Further, he writes that what people called ‘High Hindi’ was not an ancient literary language but a relatively recent creation. Before the nineteenth century, most Hindus wrote either in Urdu or regional dialects like Awadhi (Grierson, 1903). The emergence of modern Hindi, for him, was closely associated with Fort William College in the early nineteenth century, wherein languages were being standardised for the purpose of administrative ease (Banerjee-Dube, 2011).

By replacing Persian vocabulary with Sanskrit-derived words while retaining the grammar of Hindustani, a new literary standard was created. In other words, Grierson describes Hindi as a process of linguistic standardisation. He criticises attempts to eliminate Persian words from Hindi, arguing that many such words had long become part of everyday speech. At the same time, he is equally critical of highly Persianised Urdu.

In the subsidiary table on languages spoken in India, Grierson does not list Hindi or Urdu as a separate entry. They were instead recorded within the broader category of Western Hindi. We need to remind ourselves that he was a philologist who grouped languages based on grammar, phonology and historical development. Here, Western Hindi was a linguistic group within the Indo-Aryan family, not a single standard language. It referred to a cluster of closely related dialects spoken primarily in the Delhi, Agra, Meerut and Kanpur region and adjacent areas of the United Provinces (Uttar Pradesh) and Punjab.

Hence, Urdu and Hindi were grouped within the census table reflecting a philological classification. In contrast, the Hindi-Urdu controversy revolved around concerns over script, literary tradition, education, and communal identity. Grierson repeatedly observed that speakers often identified with local dialects rather than with broader linguistic categories. Categories such as ‘Western Hindi’ were not intended to represent a language that people themselves claimed to speak but instead functioned as analytical constructs designed to capture shared origins between related languages.

The same logic informed his later classification of Maithili, Magahi, and Bhojpuri as belonging to a ‘Bihari’ language group in the LSI. In both cases, Grierson was less concerned with contemporary politics and identities than with reconstructing linguistic genealogies and tracing the historical connections between languages (Brass, 1974). Of course, these views have not gone uncontested.

Prior to the 1901 census, a philological framework was not employed. For instance, in the case of Bihar from 1881 to 1891, ideological aspects of the Hindi-Urdu controversy came to inform language classifications and not any philological underpinnings (Brass, 1974). It was only under Grierson that census officials came to demonstrate philological knowledge and were increasingly put to use in the twentieth century. In 1911, the census would go on to record Hindi and Urdu separately.

The Afterlives  

The earlier censuses repeatedly lamented the difficulty of recording India’s diverse social identities. Grierson did not eliminate these hurdles, but the 1901 census marked an important shift in how they were addressed. Rather than relying solely on administrative judgement, it drew upon the classificatory frameworks of comparative philology and, to a lesser extent, anthropology and ethnology to record and rationalise languages spoken across the Indian subcontinent. 

Later, in 1909, in the Imperial Gazetteer, several linguistic maps of India were prepared under the direction of Grierson, and the data utilised for this exercise came from the 1901 census. These maps utilised the census data to visualise linguistic territories, making language appear as a spatial and territorial phenomenon.

One such map was titled the ‘Prevailing Languages,’ which displayed the distribution of Aryan languages across the Indian subcontinent:

Grierson’s reconstruction of Indo-Aryan linguistic evolution divided northern India into a series of linguistic tracts on the basis of criteria such as grammar, phonology, and patterns of linguistic change. 

This does not mean that this work was politically neutral. Grierson was working within a philological tradition that correctly identified historical relationships between languages and treated linguistic evidence as a valuable source for reconstructing the past. However, like many scholars of his generation, he often extended linguistic findings into broader claims about migration, ethnicity, and race that modern scholarship threads with greater caution.

Along with providing data for such maps, the 1901 census also opened the twentieth century, which witnessed some of the fiercest language battles the subcontinent was going to witness. This included the 1905 Partition of Bengal, which Lord Curzon cited was for administrative ease, but the Indian National Congress contested that linguistic states were a better alternative to reshape India’s map (Cohen, 2014). From here on, Congress continued to demand linguistic re-organisation.

In 1920, as M.K. Gandhi rose to prominence, he re-organised the local congresses along linguistic lines and he urged for the usage of Hindustani both in the form of Devanagari and Urdu script as a way to fan the communal tensions. In 1928 came the Nehru Report that put linguistic states on Congress’s party agenda. By the time the Constituent Assembly met in the late 1940s, language had become one of the most contentious questions facing independent India. Members debated the status of English, the place of Hindi as a national language, and the recognition of regional languages (Cohen, 2014).

These discussions eventually culminated in the linguistic reorganisation of states after independence through the States Reorganisation Act of 1956. After 1971, censuses stopped listing languages spoken by fewer than 10,000 people and began to lump them together under the ‘other languages’ category.

Such rationalisation of language data and categorisation as ‘other’ has led to ‘invisibilisation of languages spoken by “minority” people.’ The issue here is that often the speakers of such minoritised languages often exceed that of many small countries (Jolad and Agarwal, 2022). This minimises the linguistic diversity of India rather than capturing a faithful picture of it. 

Across these discourses, census reports, linguistic maps, and the findings of LSI were frequently invoked as evidence for where one linguistic region ended and another began. Languages and the way they were enumerated in this sense had enormous bearings on how community and religion were understood and political futures were imagined.

(Authors: Nishitha Mandava is an independent researcher and research Consultant for

Center for Legislative Education and Research, FLAME University, Pune;

Dr. Shivakumar Jolad is the Chair of Center for Legislative Education and Research and Director of India State Stories, FLAME University, Pune

Nishitha contributed to research and primary writing;  Shivakumar contributed to conceptualization, research and editing)

Bibliography

Banerjee-Dube, I. (2014). A History of Modern India: (1st edn). Cambridge University Press. 

Brass, P. R. (1974). Language, religion and politics in North India. Vikas Publishing House. 

Cardona, G., & Jain, D. (Eds.). (2003). The Indo-Aryan languages. Routledge. 

Cohen, B. (2014). Negotiating differences: India’s language policy. In S. H. Williams (Ed.), Social difference and constitutionalism in Pan-Asia (pp. 27–52). Cambridge University Press. 

Esman, M. J. (1992). The state and language policy. International Political Science Review, 13(4), 381–396. 

Jolad, S., & Agarwal, A. (2022, March 8). India’s linguistic diversity: How the census obscures it. The India Forum. 

Kidwai, A. (2019). The People’s Linguistic Survey of India volumes: Neither linguistics, nor a successor to Grierson's LSI, but still a point of reference. Social Change, 49(1), 154–159. 

Majeed, J. (2019). Colonialism and knowledge in Grierson’s linguistic survey of India. Routledge. 

Olender, M. (2002). The languages of paradise: Aryans and Semites, a match made in heaven (A. Goldhammer, Trans.). Other Press. 

Risley, H. H., & Gait, E. A. (1903). Census of India, 1901: Vol. I. India. Part I—Report. Office of the Superintendent of Government Printing, India. 

Appendix

Maps from Imperial Gazetteer of India-1909, vol 26, Atlas (based on 1901 Census)