Understanding Language Endangerment

Policy implications from Northeast India

Ritwiz Sarma

Madras School of Economics

December 28, 2023

The central problem

“Over 40 percent of the world’s approximately 7,000 languages are at risk of disappearing.”

Source: Endangered Languages Project

Language endangerment is about more than a speaker count. It is also about which language people can use at home, in school, and in public life.

Three working definitions

  • Endangered language: a language at risk of dying out because it lacks speakers.
  • Language shift: speakers move from their mother tongue toward a dominant language.
  • Majority tongue: the area’s default language for education and public affairs.

Endangerment is concentrated in the Northeast

The Northeast accounts for 43 of 117 languages on the SPPEL endangered-languages list used in the source presentation.

Zone Languages
North East 43
Northern 25
Southern 20
East Central 15
Andaman and Nicobar 9
West Central 5

Source: SPPEL Endangered Languages list.

How many languages are there in India?

Source and year Reported count
Linguistic Survey of India, 1927 179 languages and 544 dialects
Census, 1961 1,652 languages, later revised to 1,100
Census, 1971 109 languages with more than 10,000 speakers
Census, 2011 121 languages
People’s Linguistic Survey of India, 2013 780–880 languages

Scheduled and non-scheduled languages

Scheduled and non-scheduled language speakers across Northeast India.

The source map distinguishes scheduled and non-scheduled language speakers in rural and urban areas.

Legend for the rural and urban shares in the map.

Source: Language Atlas 2011, Census of India. Pie sizes indicate population; slices indicate language category and residence.

Speaker shares have shifted

26.7%

Growth in scheduled-language speaker share in Arunachal Pradesh during the 1990s

223%

Growth in scheduled-language speaker share in Mizoram during the 1990s

6.53%

Growth in scheduled-language speaker share in Assam during the 2000s

20.5%

Growth in scheduled-language speaker share in Nagaland during the 2000s

Percentage changes refer to shares, not percentage-point changes.

Why shift away from a mother tongue?

A language can be numerically large yet functionally marginal—or locally dominant while marginal in wider institutions.

“I want her to go to law school. I do not want her to be a maid when she grows up. She should know English and Marathi.”

Tulu-speaking mother quoted in Padharipande (1992), as reproduced in the source presentation. Numeric/functional distinction attributed there to Srivastava (1984).

The Tai Khamti case

Namsai marked in red.
  • Resident in Namsai District
  • Two-level social stratification
  • A majority or minority language at different subnational levels
  • Spoken across economic strata

Speaker-count estimates

Source Speakers
Census 2011 12,890
Gaur et al. (2022) 5,000
Morey (2005) 7,500

Methodology

Interviews

  • Ten semi-structured interviews
  • Government workers, teachers, independent artistes, and clergymen
  • Respondents aged 25–56; all literate and educated up to graduate level

Analytical framework

Fishman’s (1991) reversing language shift framework guides the case study. Interviews examine:

  • Language identification and maintenance
  • Indigenous knowledge and documentation
  • Linguistic culture, including honorifics
  • Narrative forms

Findings: identification and knowledge

Community identification

  • Hindi versus Assamese in everyday speech
  • Language maintenance from elders to younger speakers
  • “We learn English only in school” — Namchoom
  • Tai Khamti Heritage and Literary Society
  • “Two or three generations from now, Khamti will still be spoken.”

Indigenous knowledge

  • No dictionary; script digitalisation in progress — Mannoi
  • No textbooks; teachers use Myanmarese material
  • Indigenous context and its uniqueness (Abbi, 2021)
  • Some words cannot be translated and may die with the language

Findings: honorifics and narratives

Honorifics

  • Honorifics carry over into majority-tongue speech without translation.
  • Translation becomes onerous when requested.
  • They exemplify “linguistic culture” (Schiffman, 1996).

Narrative forms

  • Almost all have heard folktales; few can remember them.
  • Stories are unique to the culture, not translations.
  • Songs are performed at festivals and distributed physically and online.

What community-led work needs

Cultural programmes

  • Ceremonial use reaffirms identity.
  • Programmes provide a forum for the community.
  • Festivals need resources.

Local capacity

  • Include community stakeholders in research and decisions.
  • Provide documentation training and tools.

Data collection

  • Repositories require technical infrastructure.
  • Gaur et al. (2022) report outdated primary data.

Reference materials

  • Their absence limits language teaching and maintenance.
  • Community initiative needs support for resource-intensive work.

Fund, follow, facilitate

  • Fund documentation, archiving, local libraries, native scholars, and heritage festivals.
  • Follow speaker numbers for all languages and priorities set by indigenous leaders and scholars.
  • Facilitate local organisations, training, and language teaching.

Closing thought

“Only a community can keep its language alive. No governmental or non-governmental initiative is sufficient in this regard.”

Gaur, C. R., Dey, S., & Mamta, K. (2022). Tribal and Indigenous Languages of India. UNESCO.