On a morning in February 2010, a 96-year-old woman named Boa Senior died in the Andaman Islands, a remote archipelago in the Bay of Bengal administered by India. She had been, for the last several years, the last fluent speaker of Bo, one of the ten Great Andamanese languages. With her death, the language went silent. No recording captures her full lexical knowledge; no written grammar preserves the complete structure of the language she had spoken since childhood. The National Geographic Society sent a linguist to document what could be learned from Boa in her final years, and some recordings exist -- but the living, full, transmitted language is gone.

The Great Andamanese languages are believed to represent some of the oldest surviving linguistic lineages on Earth, descended from the first waves of anatomically modern humans who migrated out of Africa along the southern coast of Asia perhaps 65,000 years ago. The people of the Andaman Islands arrived there early and remained in relative isolation for millennia. Their languages therefore preserve features of deep human prehistory -- sounds, structures, and organizational principles that provide evidence about the range of possibilities human language has explored across its entire history. When Boa Senior died, a lineage of linguistic knowledge tens of thousands of years in the making ended with a single woman's last breath.

This is what language death looks like at its quietest.

Key Definitions

Language death -- the process by which a language ceases to have any native speakers; distinguished from language extinction (no speakers at all) by the possibility that some semi-speakers or rememberers survive; almost always results from language shift rather than the physical death of all speakers.

Language shift -- the process by which a community progressively abandons its heritage language in favor of a dominant language, typically over two to three generations; the primary mechanism of language death.

Language attrition -- the loss of proficiency in a language by individual speakers as another language becomes dominant in their lives; speakers gradually lose vocabulary, grammatical complexity, and fluency in a language they once spoke fully.

The tip point -- the demographic threshold beyond which language death becomes nearly inevitable; generally reached when the last cohort of monolingual speakers or fluent elderly speakers is no longer producing new fluent speakers through normal child-rearing; at this point, the language can only be maintained through deliberate intervention.

Endangered language -- a language whose community of speakers is declining in size or age profile, such that the language faces significant risk of extinction within a few generations; UNESCO categorizes languages on a scale from vulnerable to critically endangered.

Language revitalization -- deliberate efforts to reverse language shift and restore active use of a threatened language, typically involving educational programs, media creation, policy change, and community mobilization.

Intergenerational transmission -- the passing of a language from parents to children during the critical period of language acquisition; its cessation is the defining event of terminal language decline.

Biocultural diversity -- the concept, developed by Luisa Maffi at Terralingua, that linguistic, cultural, and biological diversity are correlated and interdependent; regions with the highest language diversity tend to have the highest biodiversity, and the forces destroying one tend to destroy the other.

Language nest (kohanga reo) -- an immersion childcare setting in which a minority language is the sole medium of interaction; pioneered by the Maori in New Zealand from 1982 as a model for early-childhood language transmission.

GIDS (Graded Intergenerational Disruption Scale) -- Joshua Fishman's 1991 eight-stage model describing the progressive disruption of language transmission, from fully functioning community language (Stage 1) to language shift effectively complete (Stage 8); the most widely used framework for assessing language endangerment and planning intervention.

The Scale of the Crisis

The starting figure is 7,000. According to Ethnologue, the authoritative catalog of the world's languages published by SIL International, approximately 7,168 languages are currently spoken worldwide. This number is approximate -- the boundary between a language and a dialect is famously contested -- but it provides a baseline.

Against this, the projections are stark. A 2003 UNESCO report, Language Vitality and Endangerment, estimated that approximately 50% of the world's languages are endangered or could cease to be actively transmitted within the 21st century. The linguist David Crystal, in Language Death (2000, Cambridge University Press) -- one of the foundational popular accounts of the crisis -- estimated that one language goes silent approximately every two weeks. Michael Krauss, a linguist at the University of Alaska Fairbanks who was among the first to document the scale of the crisis systematically, estimated in a 1992 paper in Language (Vol. 68, No. 1, pp. 4-10) that 90% of existing languages might not survive the next century if current trends continued.

The demographic distribution underpinning this crisis is extreme. Approximately 96% of the world's population speaks one of approximately 300 languages. The remaining 4% of the world's population speak the other approximately 6,800 languages -- the vast majority of human linguistic diversity. This means that Mandarin, Spanish, English, Hindi, Arabic, Portuguese, Bengali, Russian, Japanese, and Punjabi together account for a population larger than all other languages combined. The death of any of these would be a massive cultural disruption. The death of all the other 6,800 languages would pass largely unnoticed by most of the world's people.

The geographic concentration of language diversity is equally striking. Papua New Guinea, a country of approximately 10 million people and 462,840 square kilometers, contains approximately 840 languages -- roughly 12% of all the world's languages. Nigeria has approximately 520. Indonesia has approximately 700. The Vanuatu archipelago in the Pacific, with a population of roughly 320,000, has approximately 110 languages -- giving it the highest language density per capita on Earth. These tropical and subtropical regions are, not coincidentally, also among the most biologically diverse on Earth.

How a Language Dies

Languages do not typically die catastrophically. They die through a gradual process that unfolds across generations, and understanding the stages is essential to understanding both why it happens and where intervention is possible.

Stage 1: Bilingualism and Prestige Shift

The first stage is the emergence of widespread bilingualism in a community. A minority language community comes into sustained contact with a dominant language through colonialism, national political integration, economic migration, or media penetration. Initially, both languages coexist -- older community members may be monolingual in the minority language, younger members become bilingual.

At this stage, a prestige asymmetry typically develops. The dominant language is associated with economic opportunity, education, government, media, and social mobility. The minority language is associated with home, tradition, rural identity, and sometimes poverty or marginalization. This asymmetry is not merely economic -- it is deeply social. Speakers of minority languages may experience shame or stigma associated with their heritage language, particularly when that language has been actively suppressed or denigrated by authorities.

Stage 2: Language Shift in Child-Rearing

The critical transition occurs when parents begin raising children primarily or exclusively in the dominant language. This is not usually a deliberate decision to destroy a language -- it is almost always a decision made out of love and practical concern for a child's future. If speaking the majority language is the pathway to education, employment, and social acceptance, parents rationally choose to give their children the language of opportunity.

Once intergenerational transmission stops -- once children are no longer acquiring the minority language as a native language in the home -- the language's long-term survival becomes extremely difficult to ensure even with intervention.

Stage 3: Attrition and the Last Speakers

As transmission ceases, the community's fluent speakers age. The youngest fully fluent speakers become older with each passing decade. Younger community members may have receptive knowledge -- they understand their grandparents -- but cannot speak the language with full grammatical complexity. These are called "semi-speakers" in the linguistics literature.

At the tip point, the community has entered a phase where natural recovery without deliberate and substantial intervention is essentially impossible. The last monolingual elders die. The last fully fluent speakers age. The language survives only in recordings, written grammars if any exist, and the partial competence of semi-speakers.

Stage 4: Dormancy or Extinction

After the last fluent speakers die, a language may be dormant -- preserved in recordings and documentation, with community members who have some familiarity with it -- or extinct, with no surviving speakers of any competence. Dormant languages can theoretically be revived through intensive documentation-based reconstruction, though this has rarely been accomplished successfully and the result is always somewhat artificial.

Causes of Language Death

Colonialism: The Primary Historical Driver

The most powerful driver of language death in the modern world has been colonialism. European colonial expansion from the 15th through 20th centuries imposed European languages on vast populations through conquest, administration, religious conversion, and -- most devastatingly -- through systematic educational suppression of indigenous languages.

In North America, residential school systems operated from the late 19th century through the mid-20th century (in Canada, the last federally funded residential school closed in 1997). Indigenous children were forcibly removed from their families, prohibited from speaking their native languages under threat of physical punishment, and immersed in English-language instruction designed explicitly to eliminate indigenous identity. "Kill the Indian, save the man" was the explicit philosophy articulated by Richard Henry Pratt, founder of the Carlisle Indian Industrial School in Pennsylvania in 1879. The consequence was the near-total destruction of dozens of languages and severe damage to hundreds more.

Similar systems operated in Australia, New Zealand, Latin America, Africa, and Asia. The Australian "Stolen Generations" -- Aboriginal children removed from their families and placed in institutions or with white families -- documented in the 1997 Bringing Them Home report, resulted in the severance of language transmission across entire communities. In many cases, the languages these children never learned are now spoken by only a handful of elderly community members or are extinct.

Economic Pressure

Outside the context of direct suppression, economic pressure operates as a slower but pervasive force for language shift. In a global economy, proficiency in major languages -- English, Mandarin, Spanish, Arabic -- is a direct economic asset. Proficiency in a minority language spoken by a few thousand people in a remote region provides no comparable economic leverage. Rational actors therefore invest their time and their children's time in acquiring economically valuable linguistic capital.

This force operates independently of any malicious intent. No one needs to tell a mother in rural Papua New Guinea that her child should learn Tok Pisin and English rather than only the village language of 500 speakers -- she can observe this reality herself. The economic logic of language shift is compelling precisely when the speaker has genuine freedom of choice. This is why some linguists resist the framing of language death as pure loss: in many cases, speakers are making rational choices about their lives.

Urbanization

Languages are embedded in communities. They carry specific vocabulary for local geography, traditional occupations, kinship systems, ceremonies, and ecological knowledge. When speakers leave their home communities for cities -- as billions have done during the 20th and 21st centuries -- they enter environments where their heritage language has no practical function. In a city of five million, a language spoken by 500 people in a distant valley is not economically useful, not socially recognized, and not transmittable to children who have no contact with the community in which it is embedded.

Urbanization has been the single largest demographic transformation in human history. In 1900, approximately 13% of the world's population lived in urban areas. By 2007, urban population exceeded rural for the first time. By 2050, approximately 68% of the world's population is projected to be urban. The social conditions that historically maintained minority language communities -- geographic isolation, economic self-sufficiency, dense intergenerational community contact -- are exactly what urbanization dissolves.

The Mass Media Effect

Radio, television, and the internet have operated as powerful homogenizing forces in the global language landscape. When radio broadcasting became widespread in the 1920s and 1930s, it was conducted almost exclusively in major national languages. Television from the 1950s onward was similarly dominated by major national and international languages. Entertainment, news, education, and commerce were all conducted in languages other than the world's thousands of minority languages.

The internet has been, in its early decades, similarly unequal. English dominated internet content through the 1990s and much of the 2000s. As the internet has expanded globally, Mandarin, Spanish, Arabic, and other major languages have grown substantially. But the roughly 7,000 languages with fewer speakers than the population of a mid-sized city remain effectively invisible in digital space -- no search engine indexes them meaningfully, no major platform supports them as interface languages, and almost no entertainment content is produced in them.

The one significant reversal of this trend involves deliberate policy intervention -- Welsh on S4C and BBC Cymru Wales; Irish on TG4; Maori on Maori Television -- and these cases are notable precisely because they represent exceptions to the overwhelming pattern.

What Is Lost

The question of what is actually lost when a language dies has two distinct dimensions: the scientific and the cultural.

Scientific Loss

Every language is, in the technical sense, evidence. The 7,000 languages currently spoken represent the accumulated linguistic experimentation of the last 100,000-plus years of human communication. Each language is a data point in the scientific project of understanding what human language is, what structures it can take, what categories it obligatorily marks, and what variation exists across the full range of human cognitive and social life.

The value of this data becomes apparent in specific cases. Stephen Levinson at the Max Planck Institute for Psycholinguistics conducted decades of research on spatial language and cognition, summarized in Space in Language and Cognition (2003, Cambridge University Press). His work on Guugu Yimithirr, an Australian Aboriginal language spoken on the Cape York Peninsula of Queensland, revealed that the language encodes spatial reference exclusively in absolute terms: speakers use only cardinal directions (north, south, east, west) to describe spatial relationships. They do not use relative terms like "to my left" or "in front of me."

This is not a trivial linguistic difference. Levinson found that Guugu Yimithirr speakers maintain a constant implicit awareness of absolute orientation -- they always know which direction is north -- and perform differently from speakers of European languages on spatial cognition tasks. The linguistic difference corresponds to a cognitive difference: the categories a language requires speakers to track and mark influence what speakers habitually attend to and remember.

When a language with such distinctive structural properties disappears, the cognitive data it represents disappears with it. The full range of human spatial conceptualization becomes less legible. Languages differ in tense systems (some have no grammatical tense), evidentiality (many languages obligatorily mark whether you know something directly, by report, or by inference), number systems, and color terminology (the number of basic color terms varies from 2 to 12 across languages). Each of these structural properties is evidence for the science of language.

Daniel Nettle and Suzanne Romaine, in Vanishing Voices: The Extinction of the World's Languages (2000, Oxford University Press), argued for an additional dimension of scientific loss: the ecological knowledge encoded in minority languages. Minority languages, particularly those spoken by communities living in complex natural environments, encode precise vocabulary for local ecological features -- species names, microclimate phenomena, traditional ecological practices, phenological knowledge about seasonal patterns -- that exists nowhere else. When a language dies, this environmental knowledge typically dies with it.

Cultural Loss

The cultural dimension of language loss is harder to quantify but perhaps more immediate for the communities involved. A language is not merely a communication tool -- it is the medium in which a culture's poetry, humor, oratory, ceremonial life, kinship terminology, and cosmological concepts are expressed. Many of these do not translate fully.

The Sapir-Whorf hypothesis -- the claim, associated with linguists Edward Sapir and Benjamin Lee Whorf in the early 20th century, that language shapes thought -- has been controversial in its strong form (that language determines what you can think). But the weak version -- that language influences habitual cognition and structures available conceptual categories -- is now broadly supported by the cognitive science literature on linguistic relativity. What Guugu Yimithirr requires its speakers to track (absolute direction), what Mandarin requires (whether an action is ongoing or completed), what certain languages do with obligatory evidentiality marking -- these represent different ways of organizing experience that are genuinely embedded in the language and its associated cultural practices.

When a community loses its language, it loses more than communication; it loses the medium in which its cultural memory, ceremonial practice, and intergenerational identity are stored. The grief documented by indigenous language communities confronting language loss is not mere sentimentality -- it reflects the recognition that what is being lost is a dimension of collective identity that cannot be reconstructed in a different medium.

The Geographic Pattern: Biocultural Diversity

The correlation between biological diversity and linguistic diversity is one of the most striking regularities in comparative geography. The same regions that contain the highest density of endemic plant and animal species -- tropical rainforests, coral reef systems, montane ecosystems -- contain the highest density of languages.

Papua New Guinea exemplifies this pattern. With approximately 840 languages (estimates range from 800 to 900 depending on the dialect/language boundary criterion), Papua New Guinea holds approximately 12% of the world's languages in roughly 0.1% of its land area. The island is also one of the most biologically diverse places on Earth -- among the highest levels of bird and plant species richness anywhere. The Sahul plate's long isolation from Southeast Asia allowed separate evolutionary trajectories for both biological and human diversity.

Terralingua, an organization founded by linguist Luisa Maffi, has quantified the biocultural diversity correlation globally. Their publications show that the ten most linguistically diverse countries -- Papua New Guinea, Indonesia, Nigeria, India, Mexico, Cameroon, Australia, Brazil, Democratic Republic of Congo, Philippines -- are also among the most biologically diverse. The correlation is not merely coincidental: the same conditions that produce high biodiversity (stable climate over long periods, topographic complexity, ecological productivity) also produce the geographic isolation and subsistence diversity that generates linguistic diversification.

The implication is that language loss and biodiversity loss are driven by the same forces -- economic globalization, land-use change, urbanization -- and that they reinforce each other. When a community loses its language, it loses the ecological knowledge embedded in that language. When it loses its land base, it loses the ecological context in which both biological knowledge and the language itself were embedded.

Cases of Revitalization

Hebrew: The Unique Full Revival

Hebrew is universally recognized as the only case in which a language passed through a phase of no native speakers -- functioning only as a liturgical and scholarly language -- and then was successfully revived to become the daily spoken language of a large community.

By the late 18th century, Hebrew was not spoken as a vernacular anywhere. It was used for religious study, prayer, and scholarly correspondence by Jewish communities worldwide, but no community used it as a language of daily life. The vernacular languages of Jewish communities were Yiddish (Eastern Europe), Ladino (Sephardic communities), Arabic (Middle East and North Africa), and the national languages of their respective countries.

Eliezer Ben-Yehuda (born Eliezer Yitzhak Perlman in 1858 in what is now Lithuania) immigrated to Palestine in 1881 committed to a mission: making Hebrew the spoken language of a reconstituted Jewish homeland. He established Hebrew as the language of his household from the moment of arrival. His son, Ben-Zion Ben-Yehuda, born in 1882, became what Ben-Yehuda himself called the first child in centuries to learn Hebrew as a mother tongue.

Ben-Yehuda's project involved not only household transmission but the systematic creation of modern vocabulary for concepts that biblical or rabbinic Hebrew lacked -- words for newspaper, dictionary, towel, and hundreds of others were coined or adapted. He published Hebrew-language newspapers and dictionaries and campaigned for Hebrew-medium education.

The revival succeeded because of a combination of ideological commitment (Zionist settlers shared a motivation to create a new national identity distinct from the diasporic past), practical necessity (immigrants from dozens of countries needed a common language), institutional support through Hebrew-medium schools (the first Hebrew-medium kindergarten was established in 1898), and the eventual political institutionalization of Hebrew as an official language.

By the founding of Israel in 1948, Hebrew was an established vernacular spoken by several hundred thousand people. Today it is the first language of over five million Israelis and widely spoken as a second language by millions more.

No other language has undergone anything comparable. The sociolinguist Ghil'ad Zuckermann has argued in Revivalistics (2020, Oxford University Press) that modern Israeli Hebrew is better understood as a semi-engineered Semito-European hybrid language, restructured through contact with European languages, particularly Yiddish, rather than a pure restoration of Classical Hebrew. The debate is linguistically interesting but does not diminish the achievement: a language that had no native speakers now has millions.

Welsh: Policy-Driven Partial Revival

Welsh provides the best-documented case of a language in a democratic polity being partially revitalized through deliberate policy intervention.

In 1960, the Welsh language was in sharp decline. The census of 1961 found that approximately 26% of the Welsh population spoke Welsh -- down from approximately 50% in 1901. The trajectory appeared to point toward minority status and eventual extinction. Several factors drove this: the industrialization of South Wales in the 19th century, which brought large numbers of English-speaking workers; compulsory English-medium education; and the general association of English with economic advancement.

The revitalization began with political mobilization. Cymdeithas yr Iaith Gymraeg (the Welsh Language Society), founded in 1962, conducted civil disobedience campaigns including the defacement of English-only road signs and demonstrations at government buildings. The Welsh Language Act of 1967 gave Welsh equal legal validity with English in legal proceedings in Wales. Welsh-medium schools expanded significantly from the 1970s and 1980s. The Welsh Language Act of 1993 established the Welsh Language Board, which required public bodies to develop Welsh language schemes. S4C (Sianel Pedwar Cymru), a Welsh-language television channel, was established in 1982 after sustained political pressure, providing for the first time significant Welsh-language broadcast media.

The results have been positive but incomplete. The 2021 census found that approximately 17.8% of the Welsh population reported being able to speak Welsh -- approximately 538,000 people. While this is lower as a percentage than the 1961 figure, the absolute number of speakers has increased, and -- more importantly -- the age profile has improved. The proportion of young people who speak Welsh is substantially higher than among older age groups, suggesting that intergenerational transmission has been partially restored.

Census Year Welsh Speakers (%)
1901 50.1%
1931 37.1%
1961 26.0%
1971 20.8%
1981 18.9%
1991 18.6%
2001 20.8%
2011 19.0%
2021 17.8%

The data since 2001 reflects both methodological changes in the census question and the growth of the non-Welsh-speaking population through immigration, complicating interpretation of the percentage figures. The absolute number of Welsh speakers is higher in 2021 than it was in 1961.

Maori: Language Nests and Institutional Support

Te Reo Maori, the language of the indigenous Polynesian people of New Zealand (Aotearoa), had approximately 100,000 fluent speakers in 1970 but was experiencing rapid decline as urban migration and English-medium education severed intergenerational transmission.

In 1982, Maori leaders and educators established the kohanga reo (language nest) program -- immersion childcare centers operated by Maori elders where only Maori was spoken. The program drew on the principle that early childhood immersion is the most effective mechanism for first-language acquisition. By the late 1980s, hundreds of kohanga reo were operating nationwide.

The kohanga reo model was extended into primary (kura kaupapa Maori) and secondary (wharekura) schools. In 1987, the Maori Language Act gave Te Reo Maori official language status in New Zealand alongside English. Maori Television, established in 2004, provided broadcast media in the language.

The results have been partial. The 2018 New Zealand Census found that approximately 185,000 people -- approximately 4% of the New Zealand population -- could hold a conversation in Maori. This represents a significant improvement from the nadir of the early 1980s, though still a fraction of the Maori population. The kohanga reo generation demonstrates that immersion education can produce some fluency but that broader community and media environments remain essential for full language vitality.

Hawaiian: Immersion Education as Recovery

Hawaiian faced severe decline. By the early 1980s, the language had approximately 2,000 native speakers, almost all of them elderly and residing on the island of Niihau. The language had been effectively banned from Hawaii's public schools from 1896 to 1986.

The Hawaiian language revitalization movement, Punana Leo (language nests), began in 1984 and established the first Hawaiian-language immersion preschool, modeled directly on the Maori kohanga reo. The movement eventually succeeded in establishing a complete K-12 Hawaiian immersion school system and, in 1997, a Hawaiian-language track at the University of Hawaii at Hilo.

By 2010, the number of Hawaiian speakers had grown substantially, with several thousand speakers at various proficiency levels and a small but growing number of children acquiring Hawaiian as a primary language through immersion education. The revival is ongoing but fragile -- dependent on continued institutional support and community engagement.

What Works in Revitalization

The comparative evidence from successful and unsuccessful revitalization efforts points to a set of conditions that are necessary, though not sufficient, for reversing language shift. No single intervention succeeds in isolation.

Factor Effect on Revitalization Example
Immersion early childhood education High; creates new native-speaker cohort Kohanga Reo (Maori), Punana Leo (Hawaiian), Mudiad Meithrin (Welsh)
Official language status and legal protection Moderate; provides institutional framework Welsh Language Acts 1967/1993, Maori Language Act 1987
Television and radio in minority language Moderate; normalizes language in daily environment S4C Wales (1982), TG4 Ireland (1996), Maori TV (2004)
Economic incentive to know the language High; without this, shift continues Welsh civil service and education employment
Critical mass of fluent young speakers Essential; without this, transmission cannot occur Welsh strongholds (Y Fro Gymraeg)
Community will and identity investment Essential; without this, no intervention succeeds All successful cases

Joshua Fishman's GIDS model identifies Stage 6 -- the restoration of intergenerational informal oracy, that is, the transmission of the language in the home from grandparents and parents to children -- as the most critical stage for any revitalization effort. Fishman's analysis in Reversing Language Shift (1991, Multilingual Matters) argued that all the institutional interventions in the world are insufficient if families are not using the language in the home. School programs can produce literate users of a language; they cannot by themselves produce the next generation of native speakers. That requires the home.

The Ethics of Language Preservation

The strongest argument for language preservation is the scientific and cultural value of linguistic diversity: each language represents irreplaceable evidence about the range of human cognition and cultural organization, and communities have a right to maintain their linguistic and cultural identity. David Crystal articulates this position compellingly in Language Death, arguing that the uniformity of a monolingual world would represent a profound impoverishment of the human heritage.

The most serious challenge to this position comes not from indifference but from respect for speaker agency. John McWhorter, a linguist at Columbia University, has argued in various publications that language death, while genuinely producing losses, is also often the result of people making reasonable choices about their lives -- choices that linguists and activists should not second-guess. If a community chooses English over its heritage language because English provides economic opportunity and social mobility, is it paternalistic to try to reverse that choice?

The most defensible position steers between these poles. Michael Krauss argued for an "endangered language documentation imperative": regardless of whether revival is possible or desired by the community, every language should be documented as fully as possible before it falls silent. This serves scientific purposes -- preserving the linguistic data -- without requiring that communities maintain living languages against their interests or preferences.

Beyond documentation, the conditions for language survival can be created without mandating survival. What Welsh, Maori, and Hawaiian have shown is that the primary requirements for successful revitalization are: sufficient community will; institutional support through media, education, and legal status; economic contexts in which the language has some practical value; and enough fluent speakers -- particularly young fluent speakers -- to make normal use of the language in daily life.

"Languages are the oldest human institutions -- far older than writing, agriculture, or states. Each one that falls silent takes with it something that cannot be recovered: a way of knowing the world that belongs to no other people, encoded in no other form." -- David Crystal, Language Death (2000)

Cross-References

References

  • Krauss, M. (1992). The world's languages in crisis. Language, 68(1), 4-10. https://doi.org/10.1353/lan.1992.0075
  • Crystal, D. (2000). Language Death. Cambridge University Press. https://doi.org/10.1017/CBO9781139106856
  • Nettle, D., & Romaine, S. (2000). Vanishing Voices: The Extinction of the World's Languages. Oxford University Press.
  • UNESCO. (2003). Language Vitality and Endangerment. United Nations Educational, Scientific and Cultural Organization. Document CLT/CEI/DCE/ELP/PI/2003/1.
  • Levinson, S. C. (2003). Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge University Press. https://doi.org/10.1017/CBO9780511613609
  • Fishman, J. A. (1991). Reversing Language Shift: Theoretical and Empirical Foundations of Assistance to Threatened Languages. Multilingual Matters.
  • Maffi, L. (Ed.). (2001). On Biocultural Diversity: Linking Language, Knowledge, and the Environment. Smithsonian Institution Press.
  • Grenoble, L. A., & Whaley, L. J. (2006). Saving Languages: An Introduction to Language Revitalization. Cambridge University Press. https://doi.org/10.1017/CBO9780511615931
  • Zuckermann, G. (2020). Revivalistics: From the Genesis of Israeli to Language Reclamation in Australia and Beyond. Oxford University Press. https://doi.org/10.1093/oso/9780199812790.001.0001
  • Simons, G. F., & Fennig, C. D. (Eds.). (2023). Ethnologue: Languages of the World (26th ed.). SIL International.
  • Bringing Them Home: Report of the National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children from Their Families. (1997). Human Rights and Equal Opportunity Commission, Australia.

Tags

language linguistics endangered languages language death language revitalization cultural preservation colonialism

Frequently Asked Questions

How many languages are there and how many are endangered?

Approximately 7,000 languages are spoken worldwide today, though the count varies by source depending on where linguists draw the boundary between a language and a dialect. The SIL International Ethnologue database, the most comprehensive catalogue, lists 7,168 languages in its 2023 edition. Of these, UNESCO's Atlas of the World's Languages in Danger classifies hundreds as vulnerable, endangered, severely endangered, critically endangered, or extinct in the past generation. Researchers David Crystal, Daniel Nettle, and Suzanne Romaine estimated in the early 2000s that approximately 50% of living languages could fall silent by 2100 if current trends continue. The linguist Michael Krauss, in a 1992 paper in the journal Language that is widely credited with drawing academic and public attention to the crisis, estimated that 90% of languages alive today would be gone by 2100. The distribution of languages is strikingly uneven: approximately 96% of the world's population speaks roughly 4% of its languages. This means that several hundred languages — English, Mandarin, Spanish, Hindi, Arabic, and a few dozen others — serve the vast majority of the world's communicative needs, while the remaining several thousand languages are each spoken by tens of thousands or fewer people, and hundreds by fewer than a thousand.

What causes a language to die — is it inevitable?

Language death is almost never caused by all speakers dying — that outcome (called language extinction or physical extinction) is comparatively rare. The typical process is language shift: a community progressively abandons its heritage language in favor of a dominant one over one to three generations, usually because speakers perceive the dominant language as offering superior economic opportunities, social prestige, or practical utility. Joshua Fishman's 1991 Graded Intergenerational Disruption Scale (GIDS) models this as a series of stages from fully functioning community language (GIDS 1) to language shift complete (GIDS 8), with the critical threshold being the cessation of intergenerational transmission: parents stop speaking the language to children. Once children grow up without the language, the community faces a narrow window to reverse the shift before the last fluent speakers age and die. Language death is not inevitable — it results from specific social, economic, and political conditions, and in principle those conditions can change. But reversing language shift once it is well advanced is extremely difficult, and reversing extinction (reviving a language with no living fluent speakers) has succeeded only in the unique and unrepeatable case of Hebrew.

Why did colonialism kill so many languages?

European colonial expansion from the 16th century onward killed languages through several mechanisms operating at different timescales. Physical violence and epidemic disease reduced indigenous populations catastrophically in the Americas, Australia, and the Pacific, eliminating many speech communities entirely. Colonial administration and economic systems required facility in European languages for participation in commerce, law, and government, creating powerful incentives for indigenous speakers to shift. Most directly and lastingly: colonial education policies systematically suppressed indigenous language use. In the United States, Canada, and Australia, residential and boarding school systems — compulsory, forcibly removing children from their families and communities — explicitly prohibited indigenous language use among children, punishing violations physically. Canada's residential school system, which ran from the 1870s to 1996, removed generations of children from language communities during the critical period for language acquisition. Similar systems operated in the United States under the motto attributed to Richard Henry Pratt: 'Kill the Indian, save the man.' The psychological consequences — shame attached to speaking indigenous languages, trauma associated with indigenous identity — continued to suppress language transmission for generations after the explicit policies ended. In Australia, similar policies operated under the broader 'Stolen Generations' framework. The result is that many indigenous language communities experienced a near-total disruption of intergenerational transmission across two or three generations during the 20th century, leaving many languages with small elderly speaker populations and few or no younger fluent speakers.

Why does Papua New Guinea have 840 languages while most countries have far fewer?

Papua New Guinea has approximately 840 living languages — roughly 12% of the world's total in a country occupying about 0.3% of the world's land area, making it the most linguistically diverse country on Earth. Several factors explain this. Papua New Guinea has been continuously inhabited by modern humans for approximately 40,000 to 50,000 years — among the longest uninterrupted occupations anywhere outside Africa. The extreme topography of the country — steep highland valleys separated by difficult terrain, river systems, and the island geography of the coastal and island regions — isolated small populations from each other over very long periods, allowing languages to diverge independently. Population sizes in traditional Papuan societies were small and territorialized, and ceremonial and cultural practices were often highly localized. Crucially, Papua New Guinea was never unified under a single empire or administrative state before colonialism, so no single language ever acquired the political dominance that Latin, Mandarin, or Arabic achieved in Europe, China, or the Arab world, displacing regional languages. Papua New Guinea's linguistic diversity also reflects a broader global pattern: tropical and subtropical regions that have been continuously inhabited the longest and were not subject to large-scale prehistoric state consolidation tend to have the highest language density. Linguist and biologist Mira Bhatt and others working on biocultural diversity have documented the correlation between linguistic diversity and biodiversity: the 25 countries with the highest number of endemic languages overlap significantly with the 25 countries with the highest number of endemic species.

Can a dead language be revived? What does Hebrew show us?

Hebrew is the only documented case of a language being successfully revived from a primarily liturgical state to become the daily mother tongue of a large living community. In the late 19th century, Hebrew was used in Jewish liturgy, religious scholarship, and some pan-Jewish written communication, but had no living community of native speakers. Eliezer Ben-Yehuda, born Eliezer Perlman in Lithuania in 1858, immigrated to Palestine in 1881 with the explicit mission of reviving Hebrew as a spoken vernacular as part of the Zionist national project. He instituted the rule of Hebrew-only speech in his home. His son Ben-Zion Ben-Yehuda, born in 1882, is considered the first native Hebrew speaker of the modern revival. Ben-Yehuda created thousands of new Hebrew words for modern concepts absent from biblical and Talmudic texts — dictionary, newspaper, dictionary, towel — using Hebrew roots and the language's morphological system. The revival succeeded through a combination of ideological commitment, the establishment of Hebrew-medium schools, newspaper publication in Hebrew, and the practical need for a common language among Jewish immigrants arriving from dozens of countries speaking Yiddish, Ladino, Russian, Arabic, and other languages. By 1948, hundreds of thousands of people spoke Hebrew as a mother tongue. The Hebrew case is unique and probably unrepeatable: it required an extraordinary ideological motivation (nation-building), a mass immigration of people who needed a common language, political control of an educational system, and the absence of an existing dominant vernacular. No other revived language has achieved native-speaker transmission at this scale, though Welsh, Irish, Maori, and others have achieved partial stabilization.

What is being done to document and preserve dying languages?

Two distinct activities are often conflated: documentation and revitalization. Documentation — recording and analyzing endangered languages before they fall silent — is the primary goal of academic linguistics programs worldwide. The Endangered Languages Documentation Programme (ELDP), based at SOAS University of London and funded by the Arcadia charitable fund, has supported hundreds of field documentation projects. The Endangered Language Fund and the Hans Rausing Endangered Languages Project fund grammar writing, dictionary compilation, and audio-visual recording of remaining fluent speakers. The Endangered Languages Project, a Google-supported resource, maintains a public database. Documentation creates a permanent record but does not preserve a living language community. Revitalization programs aim to create new speakers, and the evidence suggests they can succeed under specific conditions. The Maori-medium early childhood program in New Zealand, Kohanga Reo (language nests), established in 1982, provided full immersion in Maori from infancy and became a model for indigenous language programs worldwide including Hawaiian Punana Leo schools. The Mudiad Meithrin nursery program in Wales operates on similar principles. Research suggests that immersive early childhood education, a sufficient mass of speakers, community engagement, and institutional support (media, education, official status) are all necessary conditions for revitalization to produce intergenerational transmission. Documentation alone, without community engagement and political will, typically produces an archive rather than a living language.

What is lost when a language dies?

The question of what is lost when a language dies has practical, scientific, and philosophical dimensions. Practically, languages encode ecological and technical knowledge that may exist nowhere else: indigenous pharmacological knowledge of plant species, navigation systems, agricultural practices, and ecological management strategies encoded in vocabulary and discourse patterns that do not translate straightforwardly. Biopiracy researchers and ethnobotanists have documented cases where indigenous language communities possess pharmacologically active knowledge that has yielded or could yield medical applications. Scientifically, each language is a data point for the theory of language: it demonstrates what grammatical structures, phonological systems, and semantic categories human language can instantiate. When a language disappears, this data is permanently lost, making it harder for linguists to understand the full range of human linguistic capacity. The Guugu Yimithirr language of Queensland, Australia, which uses absolute spatial reference (cardinal directions) rather than relative reference (left, right) for all spatial description, provides one documented example of a language encoding a distinct spatial cognition system — potentially lost if the language community vanishes. Philosophically and culturally, the death of a language marks the end of a tradition of oral literature, naming practices, ceremonial language, kinship terminology, and ways of marking social relationship that may have no analogue in the dominant language. The argument from human rights is also made: UNESCO's 2003 Convention for the Safeguarding of Intangible Cultural Heritage frames language as intangible cultural heritage, and the suppression of a language community's heritage as a violation of cultural rights.

Share this article