Linguistic, cultural markers could hold the key to caste enumeration method: Professor G.N. Devy

Linguistic and cultural markers could hold the key to solving the caste enumeration riddle, says scholar, linguist, author, and cultural activist Professor G.N. Devy. In an exclusive conversation with The Hindu, he explained that even if residents entered what they thought was their caste name, post-Census studies and a carefully layered scrutiny could analyse markers of language, ancestry, lifestyle, and kinships to arrive at a comprehensive list of castes that accounts for all groups, while being able to explain duplication, variations in names, and spellings.

“This model has been tried and tested for languages,” said Professor Devy, whose work leading the “Peoples’ Linguistic Survey of India” project resulted in the documentation of more than 780 languages in the country.

The Union government has planned to hold the next Census in 2026 and 2027. The first phase — houselisting — is set to be completed this year; the second phase — population enumeration (with caste) — is due in 2027. However, the methodology for caste enumeration has not yet been announced by the Office of the Registrar General and Census Commissioner of India.

There are discussions among activists, scholars, and community leaders about two possible methods: the first is to leave an open field in the Census form — what the 2011 Socio Economic and Caste Census (SECC) did; and the second is to compile a list of castes for people to select from — what Bihar’s Caste-based Survey did. The argument for the latter often draws its strength from the fact that the 2011 SECC ended up returning more than 46 lakh caste names.

Post-Census study crucial

Speaking to The Hindu, Professor Devy argued in favour of the first method, saying that the methodologies of surveying and enumerating languages could be used to condense not just the 2011 SECC data but also the data from the upcoming Census. However, he noted that this would require the government to keep the data open to scrutiny by scholars and involve institutions such as the Anthropological Survey of India (AnSI).

Explaining the approach, Professor Devy said the process may begin with collecting information on mother tongues. “The 2011 Census returned 19,000 mother tongues. But this was put through multiple layers of scrutiny that accounted for duplication, variation in spellings, errors, and yet another layer to filter for those with verified grammar. This narrowed the list down to 1,369 mother tongue languages,” he said.

Using the example of community classification, he continued, “Similarly, there is a community called Sansi in Punjab. The same community is called Kanjar in Rajasthan, Chhara in Gujarat, and Kanjar Bhat in Maharashtra. But it is one community because they have a shared language called Bhaktu. So, while the census will return four names, it can point to common language, ancestry, lifestyle, relatives, marriages, and kinship bonds. And the Anthropological Survey of India can certify this.”

Professor Devy added that works such as AnSI’s ‘People of India’ project could be a reference point.

DNT count needed

Professor Devy, a scholar who co-founded the Denotified, Nomadic, and Semi-nomadic Tribes – Rights Action Group (DNT-RAG), along with author Mahasweta Devi and has chaired the government’s 2006 Technical Advisory Group for DNTs, said that the Census office must declare its intention to explicitly count the DNT communities (formerly classified as “criminal” under the colonial-era Criminal Tribes Act, 1871).

He noted that if this opportunity is not taken, India risks alienating more than 10 crore people, presenting a “problem that can become much bigger than the problem of calculating, tabulating, and making a proper list”.