The Ket ( ) language, or more specifically Imbak and formerly known as Yenisei Ostyak ( ), is the sole surviving language of the Yeniseian language family. It is spoken along the middle Yenisei basin by the Ket people.
The language is threatened with extinctionâÂÂthe number of ethnic Kets that are native speakers of the language dropped from 1,225 in 1926 to 537 in 1989. According to the latest reports from linguists, this number has since fallen to less than 30. A 2005 census reported 485 native speakers, but this number is suspected to be inflated. According to a local news source, the number of remaining Ket speakers is around 10 to 20. Another Yeniseian language, Yugh, became extinct in the 1970s.
The earliest observations about the language were published by Peter Simon Pallas in 1788 in a travel diary (, ). During the 19th century, the Ket were mistaken for a tribe of the Finno-Ugric Khanty. A. Karger in 1934 published the first grammar ( ), as well as a Ket primer ( ), and a new treatment appeared in 1968, written by A. Kreinovich.
Ket people were subjected to collectivization in the 1930s. In the 1950s and 1960s, according to the recollections of informants, they were sent to Russian-only boarding schools, which led to the ceasing of language transmission between generations. As of 2013, Ket is taught as a subject in some primary schools, but only older adults are fluent and few are raising their children with the language. Kellog, Russia, is the only place where Ket is still taught in schools. Special books are provided for grades second through fourth but after those grades there is only Russian literature to read that describes Ket culture. There are no known monolingual speakers as of 2006. A children's book, A Bit Lost by Chris Haughton, was translated into the language in 2013. Alexander Kotusov was a Ket folk singer and poet who died in 2019.
Only three localities, Kellog, Surgutikha and Maduika, retain a native Ket-speaking population in the present day. Other villages such as Serkovo and Pakulikha were destroyed in the second half of the 20th century, dispersing the local Ket population to nearby towns.
Ket has three dialects: Southern (Upper Imbat), Central and Northern (collectively Lower Imbat). All the dialects are very similar to each other and Kets from different groups are able to understand each other. The most common southern dialect was used for the standardized written Ket.
The three remaining Ket-majority localities natively speak different dialects. Southern Ket is spoken in Kellog, Central Ket in Surgutikha and Northern Ket in Maduika.
Georg classifies , , and as marginal phonemes.
Vajda analyses Ket as having only 12 consonant phonemes:
It is one of the few languages to lack both and , along with Arapaho, Una (Goliath), Obokuitai, Palauan, and Efik, as well as classical Arabic and some modern Arabic dialects.
There is much allophony, and the phonetic inventory of consonants is essentially as below. This is the level of description reflected by the Ket alphabet.
Furthermore, all nasal consonants in Ket have voiceless allophones at the end of a monosyllabic word with a glottalized or descending tone (i.e. turn into ), likewise, becomes in the same situation. Alveolars are often pronounced laminal and possibly palatalized, though not in the vicinity of a uvular consonant. is normally pronounced with affrication, as .
Descriptions of Ket vary widely in the number of contrastive tones they report: as many as eight and as few as zero have been counted. Given this wide disagreement, whether or not Ket is a tonal language is debatable, although recent works by Ket specialists Edward Vajda and Stefan Georg defend the existence of tone.
In tonal descriptions, Ket does not employ a tone on every syllable but instead uses one tone per word. Following Stefan Georg's model of Southern Ket tone, which is also adapted by the more recent works on Ket and Yeniseian, the following can be inferred:
In the 1930s a Latin-based alphabet was developed and used:
In the 1980s a new, Cyrillic-based, alphabet was created:
Ket is classified as a synthetic language. Verbs use prefixes, while suffixes are rare; incorporation is well-developed. The basic word order is subjectâÂÂobjectâÂÂverb (SOV).
Nouns have nominative basic case (subjects and direct objects) and a system of secondary cases for spatial relations. The three noun classes are: masculine, feminine and inanimate.
Unlike the neighbouring Siberian languages, Ket makes use of verbal prefixes. Ket has two verbal declensions, one prefixed with d- and one with b-. The second-person singular prefixes on intransitive verbs are .
Ket makes significant use of incorporation. Incorporation is not limited to nouns, and can also include verbs, adverbs, adjectives, and bound morphemes found only in the role of incorporated elements. Incorporation also occurs as both a lexicalized processâÂÂthe combination of verb and incorporate being treated as a distinct lexical element, with a meaning often based around the incorporated elementâÂÂand a paradigmatic one, wherein the incorporation is performed spontaneously for particular semantic and pragmatic effect.
Forms of incorporation include:
The division between morphemes is based on fusion. Sandhi are common as well. The name marking is of EzÃÂfe-type, the same as in predication.
Ket has two grammatical numbers, the singular and plural. This is usually expressed by the presence or absence of (individuated plural) or (collective plural), the plural suffixes. The old singulative suffix is present on certain singular forms, however, like the stem 'stone' > 'stones'. Some shape-classifying suffixes have developed and are mildly productive.
Ket has many loanwords from Russian, such as , 'sea'; there are also loanwords from other languages, such as Selkup: for example, the word , 'ox', comes from the Selkup word . Ket also has some Mongolian words, such as , 'tea', from Mongolian . There are also words from Evenki; for example, the word , 'tobacco', is probably borrowed from the Evenki word of the same meaning: .
Prefix positions in finite verbs are marked with superscript numerals for ease of inferring, where the superscript 0 marks the root morpheme and superscript 7 marks the verb incorporate position, as adapted from Vajda-Zinn (2004), Georg (2007) and Kotorova-Nefedov (2015). The following examples are adapted from Vajda-Zinn (2004):
The same verb root can be used in different configurations, using different verb incorporates for a variety of meanings: