1. Introduction & endangerment status
Ket is the last living member of the Yeniseian language family, spoken by the Ket people along the middle Yenisei River and its tributaries in the Turukhansky District of Krasnoyarsk Krai, central Siberia. Its sister languages — Yugh, Kott, Arin, Assan and Pumpokol — are all extinct, the last of them (Yugh) lingering into the late twentieth century. Ket therefore carries an entire branch of human language on its own.13
It is also disappearing fast. The number of ethnic Kets who speak the language fell from 1,225 in the 1926 Soviet census to 537 in 1989, and linguists now count fewer than thirty fluent speakers, nearly all elderly. Soviet-era sedentarisation and boarding schools, which moved nomadic taiga families into Russian-speaking settlements, broke the chain of transmission.1
Ket is, however, intensively documented by linguists precisely because it is so unusual. Heinrich Werner and Edward Vajda produced grammars and analyses, and Stefan Georg's A Descriptive Grammar of Ket (2007) is a standard reference. Ket is world-famous for one reason above all: Vajda's Dene-Yeniseian hypothesis argues that Yeniseian is related to the Na-Dene languages of North America — including Tlingit and Navajo — which would make Ket the Old World cousin of a whole American family.3 (See our Tlingit guide for the other end of that proposed link.)
A frank note for the learner: Ket is hard, and conversational resources are scholarly rather than app-friendly. This guide aims to make the language's astonishing structure intelligible and to point you to the real documentation. It is organised into the eight CEFR sub-levels, each with a visible heading.
Ket was an unwritten language until the twentieth century. A Cyrillic-based alphabet was later devised for it, adding special letters for sounds Russian lacks — for example ӄ for the uvular /q/, ӈ for /ŋ/, and marks for the back vowels. Scholars more often write Ket in a Latin/IPA transcription, which this guide uses alongside the Cyrillic.2
The greatest surprise for a newcomer is that Ket is tonal — unusual for Siberia. Descriptions identify several tonal/phonation contrasts (commonly counted as four), combining pitch with features such as glottalisation and length, so that the 'same' string of consonants and vowels can mean different things depending on its tone. Training the ear to tone is the first real task.
Ket also has a uvular series (/q/, /ɢ/, /χ/) made far back in the throat, and glottalised vowels written with a glottal-stop mark (ˀ), as in the numeral qoʔk ('one').
Most basic Ket nouns are short — typically one syllable — and belong to native Yeniseian stock. A striking number of core words have been compared to North American Na-Dene vocabulary in the Dene-Yeniseian proposal, which is part of why each Ket word repays attention.13
Because fluent everyday speech is now rare, learners begin with documented vocabulary lists and recorded texts rather than casual conversation. The aim at this level is simply to read and pronounce short, attested words correctly — respecting tone and the glottal catch.
Ket numerals one to ten, in the Cyrillic orthography with IPA, show a feature worth pausing on: 'eight' and 'nine' are built subtractively — literally 'two/one short of ten'.2
| # | Ket (Cyrillic) | IPA | Note |
|---|---|---|---|
| 1 | қо’к / қуусь | qɔʔk / quːsʲ | |
| 2 | ыын | ɨːn | |
| 3 | до’ң | dɔʔŋ | |
| 4 | сьиик | sʲiːk | |
| 5 | қаак | qaːk | |
| 6 | аа | aː | |
| 7 | о’н | ɔʔn | |
| 8 | ынам бәньсьаң қөө | ɨnam bənʲsʲaŋ qoː | ‘two short of ten’ |
| 9 | қус’ам бәньсьаң қөө | qusʲam bənʲsʲaŋ qoː | ‘one short of ten’ |
| 10 | қөө | qoː |
Ket nouns fall into grammatical classes — broadly animate masculine, animate feminine, and inanimate — and the class governs agreement elsewhere in the sentence. Learning a noun therefore means learning its class, much as German learners learn der/die/das.
Basic Ket clause order is generally subject–object–verb, with the verb carrying a great deal of grammatical information (see B1.2). Postpositions, not prepositions, mark spatial relations, and possession is shown by placing the possessor before the possessed. At this stage the goal is to recognise the skeleton of a simple sentence in a glossed text.
Because Ket packs so much into the verb, even 'simple' sentences can be short on the surface but dense in meaning — a single verb form may encode the subject, an object, tense and direction all at once.
At this level the three-way noun-class system and its agreement effects come into focus, along with plural formation and the set of postpositional/case markers that express location, direction and source. Ket's nominal morphology is modest compared with its verb, but the class system is pervasive and must be internalised to read accurately.
The Ket verb is the heart of the language and one of the most intricate verbal systems described anywhere. It is built from a stem plus numerous prefix and suffix positions that mark subject, object, tense, aspect and more — and the markers do not sit in a tidy row: different pieces of information are slotted into specific positions around the root, and some categories are expressed by combinations spread across the word.3
Linguists describe Ket verbs using 'position-class' templates with eight or more slots. The practical consequence for a learner is that you cannot build a verb by simply adding endings; you learn verb 'shapes' and the patterns that fill their slots. This is exactly the feature that makes Ket comparable, structurally, to the famously complex Na-Dene verb.
Advanced study of Ket is, realistically, philological: working through interlinear-glossed texts — folktales, taiga ecological knowledge, recorded narratives — with a grammar at hand. At this level you also engage the comparative literature: the Dene-Yeniseian debate, the reconstruction of Proto-Yeniseian from Ket plus its extinct sisters, and what Ket tells us about ancient connections across the Bering land bridge.
Ket's vocabulary encodes generations of knowledge about the Siberian taiga — its animals, rivers, weather and seasons — so reading Ket texts is also a way into a distinctive way of seeing the northern forest.
The capstone is the documentary record itself. Continuing study means Georg's descriptive grammar, Vajda's and Werner's work, archived recordings of the last fluent speakers, and the comparative Yeniseian and Dene-Yeniseian scholarship. Revitalisation efforts in Ket communities and schools, though small, mean some new material is being produced for younger learners.
To set Ket in its broadest context, read it next to our Tlingit guide: if the Dene-Yeniseian hypothesis holds, a learner who studies both is looking at two surviving ends of a single ancient family, separated by the width of a continent and an ocean.
Learning resources
- Ket language — phonology, classes & verb (Wikipedia)Reference — structure & status
- Ket numbers, with IPA (Omniglot)A2 — the numerals 1–10
- Yeniseian languages & the Dene-Yeniseian hypothesis (Wikipedia)Background — the family & the American link
- Ket Language — research overview (EBSCO Research Starters)Background — documentation & Vajda's fieldwork
Notes & Bibliography
- “Ket language,” Wikipedia, on Ket as the sole surviving Yeniseian language, the decline in native speakers from 1,225 (1926) to 537 (1989) to fewer than thirty today, and the monosyllabic native lexicon with comparisons to Na-Dene. [source] ↩
- Numerals one to ten (Cyrillic orthography with IPA), including the subtractive forms for eight and nine, from “Ket numbers,” Omniglot. [source] ↩
- “Yeniseian languages,” Wikipedia, on the family, Edward Vajda’s Dene-Yeniseian hypothesis linking Yeniseian to the Na-Dene languages of North America, and the position-class complexity of the Ket verb (cf. Stefan Georg, A Descriptive Grammar of Ket, 2007). [source] ↩