Eyak was the language of a small Alaska Native people who lived around the Copper River delta and the Gulf of Alaska coast, near present-day Cordova. It was never spoken by many — a few hundred people at most — and it sat, linguistically and geographically, between far larger neighbors: the Tlingit, the Alutiiq, and the Athabaskans. For most of the twentieth century its speakers slipped away one by one, until only a single fluent speaker remained.
That speaker was Marie Smith Jones. Born in 1918 in Cordova, she was the last full-blooded Eyak, the last fluent native speaker of the language, and the last traditional chief of her people. When she died on January 21, 2008, at the age of 89, Eyak became the first of Alaska’s twenty Native languages to lose its last fluent speaker in modern times — a milestone reported around the world.
Eyak’s importance to linguists was out of all proportion to the number of people who spoke it. It is a branch of the Na-Dené family, the closest relative of the large Athabaskan group, and its survival as a distinct language was a key piece of evidence for how those languages are related. Marie Smith Jones spent her last decades working with the linguist Michael Krauss to record, write down, and preserve as much of it as possible.
Eyak is not entirely silent. The dictionary and grammar she helped create have allowed a small revival effort to begin — including one young Frenchman who taught himself the language from the records and now teaches it. But the unbroken chain of native speakers, passed from parent to child for thousands of years, ended in a hospital in Anchorage in 2008.
Akkala Sámi — known in Russia as Babino Sámi — was an Eastern Sámi language of the Uralic family, spoken in the interior of the Kola Peninsula in northwest Russia. Its speakers lived in a handful of semi-nomadic reindeer-herding villages, principally Aʼkkel (Russian Babinsky, Finnish Akkala), Čuʼkksuâl (Ekostrovsky), and Sââʼrvesjäuʼrr (Girvasozero). Long mistaken for a dialect of neighboring Kildin Sámi, it is now recognized as a distinct language, structurally closest to Skolt Sámi, with which it shared a degree of mutual intelligibility.
Akkala Sámi was always small, spoken by at most a few hundred people, and it is one of the most poorly documented of all the Sámi languages. A handful of phonological descriptions, some archival recordings, and an 1897 Gospel translation make up much of the surviving record. There was never a stable written standard, and the community was too small and too pressured to develop one.
The language was extinguished by a century of upheaval. Tsarist Russification gave way to Soviet collectivization of reindeer herding in 1928-30, the consolidation and destruction of traditional Sámi villages, Stalinist repression, the disruptions of the Second World War, and a residential boarding-school system that removed children from their families and forbade their language. Settlement at places such as Yona concentrated survivors among Russian and Kildin Sámi speakers, and intergenerational transmission collapsed.
The last fluent native speaker, Maria (Marja) Sergina, died on December 29, 2003. The 2010 UNESCO atlas listed the language as extinct. A small number of elderly people retained partial knowledge afterward — two individuals around age seventy were noted in 2011 — and interest in Kola Sámi heritage persists, but Akkala Sámi has no remaining fluent speakers and no children learning it.
Nǁng — widely known by the name of its main surviving variety, Nǀuu — is a Tuu language of South Africa’s Northern Cape, the ancestral tongue of the ǂKhomani San. It is the last surviving language of the ǃUi branch of the Tuu family (formerly grouped under the misleading label “Southern Khoisan”), its nearest relatives, such as ǂUngkue, already extinct and its only living relative the distant Taa.
The language is famous among linguists for one of the largest sound inventories ever documented: by one count some 45 click sounds together with around 30 non-click consonants and 39 vowels, including rare click types found in almost no other language. This phonological richness makes Nǁng a treasure for the study of human speech — and made its near-silencing a loss felt well beyond the Kalahari.
Nǁng was driven to the brink by the dispossession of San peoples, the dominance of Afrikaans, apartheid-era stigma against San identity, and the scattering of the ǂKhomani community from the lands that are now Kgalagadi Transfrontier Park. Linguists had presumed the language extinct until the late 1990s, when around twenty elderly speakers were located among the ǂKhomani. By then no children were learning it, and the speakers were already old.
Nǁng must not be written off as a language with a single “last speaker” who has died. Its most prominent custodian, Katrina Esau — known as Ouma Geelmeid — has taught the language to children since 2002, helped produce a dictionary and a published reader, and was honored nationally for her work. The death of her brother Simon Sauls in June 2021 left her widely described as the last fully fluent speaker, but the language is being actively taught and reclaimed. It is best understood as critically endangered and dormant in everyday use, not gone.
Mbabaram (also spelled Barbaram) was a Pama-Nyungan language of the Atherton Tableland in the rainforest country southwest of Cairns, in north Queensland, Australia. The traditional lands of its speakers, the Mbabaram people, stretched roughly from west of Almaden across Lappa and east toward Atherton, near localities such as Irvinebank and Petford. Classified within the Paman subgroup, Mbabaram was the language of a comparatively small group; the linguist R. M. W. Dixon estimated the population at around 500 people before sustained European settlement reached the district in the late nineteenth century.
What made Mbabaram remarkable to linguists was its phonology. A long series of regular sound changes — including the loss of initial syllables — had pushed its word forms far from those of its neighbors, so that Mbabaram looked, on first inspection, almost like an isolate stranded among the rainforest languages around it. It was not mutually intelligible with adjacent tongues such as Yidiny, Dyirbal, Djangun, or Agwamin, and its speakers tended to learn their neighbors’ languages rather than the reverse. Only careful comparative reconstruction, much of it Dixon’s, revealed that beneath the divergence Mbabaram followed thoroughly ordinary Australian patterns.
The language is best known today for a single coincidence. When Dixon began eliciting vocabulary from the last fluent speaker, Albert Bennett, in the 1960s, the Mbabaram word for “dog” turned out to be dog (transcribed dúg) — a chance resemblance to English, with no shared history whatsoever, since the inherited regional root survives elsewhere as Yidiny gudaga and Dyirbal guda. Dixon called it a one-in-a-million accident of form and meaning between unrelated languages, and the example has been cited ever since as a caution against reading borrowing or kinship into surface similarity.
Mbabaram became extinct in 1972 with the death of Albert Bennett. By the time Dixon reached the community in the mid-1960s, mining, clearing of the rainforest, frontier dispossession, and the long machinery of “Protection”-era removal had reduced the fluent population to a handful of elderly people. Bennett and a few others preserved several hundred words and the outline of the grammar before the language fell silent.