The endangered language crisis

by Ross Perlin

Language is a universal and democratic fact that cuts across all human societies: no human group is without it, and no language is superior to any other. More than race or religion, language is a window onto the deepest levels of human diversity. The familiar map of the world’s 200 or so nation-states is superficial compared with the little-known map of its 7,000 languages. Some languages may specialize in talking about melancholy, seaweed, or atomic structure; some grammars may glory in conjugating verbs while others bristle with syntactic invention. Languages represent thousands of natural experiments: ways of seeing, understanding, and living that should form part of any meaningful account of what it is to be human.

Users of Sanskrit, Latin, Greek, Hebrew, Arabic, Persian, Mandarin, English, and the like have continually proclaimed their languages holier, more perfect, or more adaptive than the unwritten, unstandardized “dialects” they look down on. But from a linguistic point of view, no language as used by a native speaker is in any way inferior, let alone broken. The vast majority have always been oral, with written language a derivative of comparatively recent vintage, confined to a tiny group of elites in a small number of highly centralized societies. Writing is palpably a trained technology of conscious coding, in comparison with the natural and universal human behaviors of speaking and signing.

Perceptions of linguistic superiority or inferiority are not based on anything about the languages themselves, but on the power, class, or status of their speakers. Every language signed or spoken natively is a fully equipped system for handling the core communicative demands of daily life, able to coin or borrow words as needed. “Languages differ essentially in what they must convey and not in what they can convey,” said the linguist and polyglot Roman Jakobson. In other words, it’s possible to say anything in any language, but each language’s grammar requires speakers to mark out certain parts of reality and not others, however unconsciously. This is the essence of what makes linguistics fascinating and revealing.

All languages may be equal in the abstract, but much harder to bridge are the social and historical disparities among their speakers. At present, about half of all languages are spoken by communities of 10,000 or fewer, and hundreds have just 10 speakers or fewer. On every continent, the median number of speakers for a language is less than 1,000, and in Australia this figure goes as low as 87.

Roughly 7,168 living languages in the world.

Today, these numbers reflect serious endangerment, and even languages with hundreds of thousands or a few million speakers can be considered vulnerable. In the past, however, small language communities could be quite stable, especially hunter–gatherer groups, which typically comprised fewer than 1,000 people. Likewise, most older sign languages, now critically endangered, evolved in so-called deaf villages, where the incidence of hereditary deafness in the population was significantly higher than elsewhere, though still rarely more than about two percent. Many hearing people in these villages could also sign, but the core group of signers was typically several hundred at most.

In general, sheer speaker or signer numbers have always mattered less than intergenerational transmission. A small language can apparently remain strong for centuries, as long as parents, grandparents, and other caregivers are using it with children. Take Gurr-Goni, an Aboriginal language from north central Arnhem Land in Australia, which has had just several dozen speakers as far back as anyone can remember. Far from being an isolated group, Gurr-Goni speakers maintained their language in a context of multilingual equilibrium, where each “father tongue” was integrally connected with certain ancestral lands and natural resources. It’s this kind of equilibrium that has been vanishing fast as colonial and national languages take over.

43% languages are classified as endangered.

But why does linguistic diversity matter in the first place? For a linguist, the answer is clear enough: little-documented, primarily oral languages are often the ones with the most to teach us about the nature and possibilities of human communication more generally. Without the Khoisan languages of southern Africa, we wouldn’t know how extensively and expressively clicks could be used. Without Warao, spoken in Guyana, Venezuela, and Suriname, we wouldn’t know that object-subject-verb could be the routine way of ordering a sentence. Without the Hmong-Mien languages of southeast Asia, we wouldn’t know that a language could have a dozen tones.

But it’s also what languages carry inside them: the poetry, literature, jokes, proverbs, and turns of phrase. The oral histories, the local and environmental knowledge, the wisdom, and the lifeways. Only a fraction of this can ever be translated into other languages.

If this still sounds theoretical, consider the even more immediate, practical consequences. A growing body of research shows that there is no substitute for mother-tongue education, and that language maintenance is an integral component of physical and mental well-being—perhaps especially so for long-marginalized Indigenous and minority peoples.

For this is the crux of it: languages are not “dying natural deaths,” but instead are being hounded out of existence.

Like biodiversity, linguistic diversity remains strongest today in remote and rugged regions traditionally beyond the reach of empires and nation–states: mountain ranges like the Himalayas and the Caucasus; archipelagoes like Indonesia and the Solomons; and what were once zones of refuge like the Amazon, southern Mexico, Papua New Guinea, and parts of West and Central Africa. But these too are now under tremendous pressure.

“Language has always been the companion of empire,” wrote Antonio de Nebrija in his 1492 Gramática Castellana, which aimed to raise vernacular Castilian Spanish to the level of Latin and other imperial languages, just in time for European conquests across the globe. Though languages have always changed and come and gone, the scope for linguistic imperialism has widened exponentially since Nebrija’s day. A comparatively small number of empires and nation–states, now bristling with 24/7 communication and education systems, cover every inch of the Earth. Worldwide, centuries of imperialism, capitalism, urbanization, environmental destruction, and nation-building are now coming to a head linguistically. With power behind them, a few hundred languages keep growing and getting all the resources, while the other 95 percent struggle.

Particularly dominant are just a few dozen languages of wider communication, less politely called “killer languages.” English, Spanish, and Chinese are on the march, but so are Nepali and Brazilian Portuguese. These languages are spreading through political, economic, and cultural conquest, and the consequences are seeping into everything. At the same time, only under extraordinary circumstances are a few new languages emerging, such as Light Warlpiri, which developed out of mixing English and the Aboriginal language Warlpiri in Australia’s Northern Territory.

In anglophone settler societies such as the US and Canada, genocide, expulsion, disease, and every form of prejudice and pressure exerted on Native peoples have profoundly altered the linguistic landscape. About half of the 300 distinct languages once spoken north of the Rio Grande have already been silenced, and most of those remaining are no longer actively used, with under ten native speakers. Only a few of the largest, including ᏣᎳᎩ (Cherokee), Diné Bizaad (Navajo) and Yup’ik can in any way be considered “safe,” though profoundly embattled, for the coming decades. Likewise, most of the hundreds of Aboriginal languages once spoken in Australia are either no longer spoken or else down to small clusters of elderly speakers, with just a few still heroically being transmitted.

Dominant-language speakers opine that everything would be easier and better (and peace on Earth!) if everyone would just speak their particular dominant language. But common languages don’t unify in and of themselves; look at many of the world’s civil wars, or the deep divisions in anglophone American society today. The imaginative challenge of big differences is quickly replaced with the narcissism of small ones: scrutinizing other people’s accents, sociolects, word choice, tone of voice.

The spheres of use for smaller languages and nonstandard varieties are continually shrinking: they often emerge only in private, yielding as soon as a speaker steps outside. Now the shift is happening inside homes as well. Families around the world are hitching their fate to English and other dominant languages—abandoning not just words but vast traditions of gesture, intonation, facial expression, conversational style, and perhaps even the culture and character behind all these. Only in the face of intense political, economic, religious, or social pressures do people stop passing on their mother tongues to children, but today these pressures are everywhere. The disruption of this basic natural process has come to feel almost normal.

Of course, English in particular, supercharged by business, pop culture, and the internet, after centuries of colonial expansion, is the real empire of our time—far more fluid and influential than any political entity. Many English speakers go their entire lives without encountering anything significant they can’t do or get in their language. Whatever the power dynamics of any given conversation, English is pure linguistic privilege, the reserve currency of communication. The push to learn it is an event of planetary significance, swelling a linguistic community of going on half a billion native English speakers worldwide, plus another one or two billion who know it as a second language. These numbers are growing every day.

Many people think the world, or at least their corner of it, is growing ever more diverse, but monolinguals are increasingly in charge. The monolingual mindset, bonedeep in almost every anglophone American, blocks any real urgency about other languages. A multilingual childhood, only now widely recognized as an inestimable cognitive advantage, can add a whole other dimension to one’s understanding of the world, with a sense of linguistic and cultural perspective. But to do it right, especially for monolingual parents, can require serious effort and resources.

What should a monolingual person do? Every time someone speaks, they embody an inherited chain of choices. It can be profoundly useful to be a native speaker of the dominant dialect of a dominant language. Representing the associated “mainstream” culture with every sound means being able to talk to many and sound good to most. Rarely does a dominant-language monolingual need to speak anyone else’s language, and it counts as a charming attempt if they do, a mark of open-mindedness and sophistication or an advanced party trick. Since reading and writing usually hew close to the dominant dialect, book learning is that much easier. A person “without an accent” is by default considered to be smarter or better educated as soon as they open their mouth.

88 million people speak languages at risk of extinction.

Yet people intensely aware of privilege based on gender, race, class, or sexuality seldom consider their linguistic privilege. English or Spanish or Mandarin or Urdu may simply seem like the air one breathes. The only cure for monolingualism is to learn other human languages, but it’s at least a start to learn about them from those who speak them. Maybe there should be a special kind of therapy for monolinguals, where you have to sit and listen to a language you can’t understand, without translation but with total patience.

For an academic linguist, this is an occupational hazard. On first meeting, a speaker knows you don’t know their language, but there is a useful ambiguity. If not learning languages, what exactly does a linguist do? Sound systems are entire ecosystems for the ear, but even on an initial listen you can try to make out the shapes of syllables, the qualities of vowels, the puff of aspiration, the bent tongue of a retroflex. There may be clues in the intonation patterns: variations in pitch, rhythm, loudness, voice quality, or the length of sounds, which convey not only vibes but essential information, like how rising pitch in English can signal a yes–no question. Under the rush of unfamiliar sound, flowing at hundreds of syllables a minute, you try to hold back from the scramble for meaning and suss out the structure.

From the glottis to the lips, the whole tract where spoken language happens is just five or six inches long. Across evolutionary eons, a space for eating and breathing gradually took on linguistic uses, not just anywhere but at certain places of articulation: the lips, the teeth, the alveolar ridge, the hard palate and the soft one behind it, the uvula that hangs like a little grape above the throat, the pharynx, and the larynx. The tongue—that near-universal symbol of language—darts and bends to make contact wherever it can. For signers, it happens in the hands. (I use terms such as speech and oral for the sake of simplicity, but virtually everything here also applies to sign languages.) There are also whistled languages, drum languages, and many other ways of emulating speech across space.

To document and describe languages while there is still time ought to be the first task for a linguist. Yet a linguist’s moment of discovery is also almost always the moment of grasping a disappearance. For any outsider claiming to “discover” any human society or culture or language—that is, announcing the existence of some smaller group to the ruthlessly joined-up juggernaut sometimes known as “us”—is also arriving at, and bound up in, the moment of its destruction. The same forces that bring an outside linguist in are bringing everything else as well.

The organized movement to preserve the world’s languages is recent. In 1992, the linguist Michael Krauss warned that linguistics could “go down in history as the only science that presided obliviously over the disappearance of 90 percent of the very field to which it is dedicated.” This helped light the spark. Inspired by the new push for biodiversity and the growing movement for Indigenous rights, a cohort of linguists and language activists vowed to use new technologies to record and preserve as much of the world’s vanishing linguistic heritage as possible. Ideally, speakers record and document their own languages, and this is now increasingly common.

Language documentation may sound like an obvious priority for linguistics, but it flies in the face of what most linguists have been focusing on for the past 70 years: language, not languages. Following Noam Chomsky, most have been chasing theoretical and computational questions, seeing themselves as Martians trying to document an essentially uniform language called Earthling. Their evidence has come mostly from the largest languages, which happen to be the dominant ones they’re familiar with. Few meaningful universals have emerged from all the armchair theorizing and laboratory testing. Theory has its place, of course, but it’s essential that languages be documented on their own terms. The real view from Mars, it turns out, is that linguistic diversity on Earth is far more profound and fundamental than previously imagined.

At the same time, there is an essential toolkit that every language should have: a substantial dictionary, a detailed grammatical description, and a representative corpus of recorded stories, oral histories, and other texts showing the language in action, and at least partially transcribed, translated, analyzed, and archived. To the extent that speakers are willing, these materials should be maximally accessible and archived for posterity. Speakers of larger languages take for granted effectively limitless resources in and about their languages. Forget Siri, speech recognition, automatic translation, spellcheck, and other nifty tools: imagine not having a dictionary, any established way of writing, or any authority on the language at all, aside from an elder you have to find and ask in person.

It’s one thing to help build arks, or at least archives, but linguists don’t and can’t “save” languages. By definition, every language is limitless as long as speakers are still speaking it or signers are signing it. No language ends on the last page of a dictionary. From a finite number of sounds, words, rules, and techniques, speakers form an infinite number of utterances. There is no single way that a community “really speaks,” nor any one authoritative type of data to preserve for all time. Language is too fluid. Unfortunately, many linguists also dwell on damaging, defeatist abstractions about language “death” and “extinction,” while Indigenous scholars state clearly that oppression is the threat, and that reclaiming Indigenous languages is about liberation and recovery from historical trauma. Linguistics, like anthropology, has skeletons in its disciplinary closet. Fighting for endangered languages can only mean fighting on the side of their speakers and signers, and ultimately it’s always up to communities whether and how to keep using their languages. Some have been struggling to do so for centuries; others are less concerned. Of course, there are not only pressures but also enticements to learn a dominant language, which may grant access, however limited, to the dominant culture’s resources.

For some members, the breakdown of a traditional community may feel like emancipation; for others, a disaster. Whether it’s a matter of survival or a question of common sense, the logic of abandoning a smaller mother tongue for work, education, migration, marriage, or any number of other reasons can seem unassailable. Once-valued or ingrained connections to ancestors, traditions, territories, and knowledge systems can easily seem irrelevant, obscure or simply impossible to access under contemporary conditions.

It’s a powerful-sounding truism, but not quite true, that language and culture are inextricably linked, since group identities in some cases persist after the loss of a language. Nor should anyone feel forced to stay within any particular culture. What matters is that individuals and communities have meaningful options for how they relate to their linguistic pasts and construct their linguistic futures. Given the normal and natural human capacity for multilingualism, maintaining a less widely spoken language need not preclude learning a more widely spoken one.

There are now hundreds of language revitalization movements around the world, most launched in just the past few decades, creating a wealth of experience for others to draw on. It can feel like nearly impossible work, where even a single new speaker of a highly endangered language counts as a serious triumph, requiring years of dedication. At the same time, scattered speakers are now finding one another in virtual spaces, where language-learning options are multiplying and reality-augmenting and artificial intelligence possibilities are on the horizon.

If one critical ingredient has been missing from language revitalization movements, it is real financial, political, and technical support from majority populations. Speakers of endangered languages almost never encounter outside interest in or knowledge about their languages, while persecution, mockery, and stigma are still common. To the extent that language policy or discussion is on the agenda at all, it relates to specific points of conflict in a few dominant languages, not the collapse of linguistic diversity itself.

For the revivers of endangered languages, a sense of radical futility may be waiting round the bend of every utterance. Where will I speak this? Who will understand me? Who can even tell me if I’m speaking correctly? Will I ever start thinking in the language? When almost no one else is doing it, matching a string of sounds to a meaning can seem downright arbitrary. And yet it is only after ingesting masses of often arbitrary-seeming words that people can process or produce them at speed, and only then that they can start feeling the indescribable sense of what it is to live in a particular language—just as an actor needs to get her lines down cold before even starting to get into character. To try to communicate with what is no longer a tool of communication—to resurrect a whole worldview that is almost over the horizon—is a wonderful madness.

Cover image from Language City: Ralph Fasanella, New York City (1957), Oil on canvas, 50 x 110. Collection of Nicholas and Shelley Schorsch.

This article appears in the 2025-26 Berlin Journal (39). It is an edited excerpt from Language City: The Fight to Preserve Endangered Mother Tongues, published by Grove Press UK.

Sources for statistics: Ethnologue, Statistica, unesco, Visual Capitalist