The Gradual Disappearance of a Eurasian Language Family – the

Case of Yenisseyan
Stefan Georg, Leiden

The Yenisseyan languages are generally referred to as one of the so-called

Palaeoasiatic (also known as Palaeosiberian) language families. This designation
was introduced in the 18th century as a catch-all term for those indigenous languages
of Asiatic Russia/Siberia, which do not belong to one of the greater language families
which dominate this part of the world, and which extend far beyond the borders of
this territory. Palaeoasiatic is thus a purely negatively defined term, referring to
languages, which are spoken exclusively in Siberia, between the Ural mountains in
the West, the Ocean in the East and North, and the borders of Kazakhstan, Mongolia
and China in the South. This criterion, together with the second one of not having
demonstrable genetic relatives outside of this territory, defines the following
language isolates and small families as belonging to this group1: _ukotko-Kam_atkan,
Nivkh, Yukagir, and Yenisseyan.
The logical antonym of Palaeosiberian would be, of course, Neosiberian, and, though
rarely used these days, this term has been coined by CZAPLICKA (1914) to unite all
indigenous languages of Siberia belonging to larger, well-established language
families, which extend beyond the confines of Asiatic Russia, viz. Uralic, Turkic,
Tungusic, and Mongolic languages.
The criteria which led to this dichotomy are only partly valid today2, yet all
"Palaeoasiatic" languages do have in common that they are
- at best distantly related to their immediate neighbours3
- found on the very fringes of the vast Siberian territory (Nivkh and _ukotko-
Kam_atkan languages on the Eastern shores of the continent, Yukagir in the Far

1
More precisely, this would define the group of Palaeosiberian languages; some linguists, as e.g. the
editors of the Palaeoasiatic volume of the Russian series Jazyki Mira ("Languages of the World",
VOLODIN et. al. 1996) extend the latter term to include other language isolates of Asia as well, such as
Burushaski (in Pakistan), or Ainu. Conventionally, Soviet and Russian linguists often add Eskimo-
Aleut languages to their number of Palaeoasiatic families.
2
Thus, the theory that Yukaghir may not be a language isolate after all, but form a part of Uralic (or,
then, Uralo-Yukaghir) is gaining ground in specialist circles (cf. N IKOLAEVA 1988); this family may
very well form part of a still larger grouping comprising Eskimo-Aleut and Chukotko-Kamchatkan, cf.
FORTESCUE 1998 and, for a very thorough reasoning in favour of such a grouping on the basis of non-
trivial correspondences in morphology, SEEFLOTH to appear 2001.
North, Ket just below the Polar Circle), which lends support to the hypothesis that
they
- represent the remnants of formerly more widespread populations, which gradually
lost ground to (resp. were pushed into residual areas by) speakers of Neosiberian
languages, as the latter advanced into Siberia from more southerly areas (particularly
Turkic and Tungusic). I will try to show below that various observations strengthen
the assumption for such a scenario especially for the Yenisseyan languages.
Few indigenous languages of Siberia may be viewed as not endangered at the
beginning of the 21st century. In fact all of them constantly lose ground to Russian, the
one language which offers prospects for higher education, economic success in urban
life etc. to native Siberians. The number of persons who claimed one of the
palaeoasiatic languages of Siberia as their native language in 1989, the date of the last
general census held in the USSR, is given in the following table4:

population speakers %
Chukchi 15.106 11.163 73.9
Koryak 8.942 5.168 57.8
Eskimo 1.703 933 54.8
Ket 1.084 589 54.3
Yukagir 1.112 398 35.8
Aleut 644 184 31.4
Nivkh 4.631 1.199 25.9
Itelmen 2.428 563 23.2

In this list, the only Yenisseyan language mentioned is Ket (from k e/t pl. d e/N
"person, human being"), and, indeed, it is the only language of the family still spoken.
However, at least six different, and clearly distinct, languages make up the whole
family: apart from Ket, Yugh, Arin, Assan, Pumpokol, and Kott have been, at least
partially, recorded and identified as forming the close-knit language-family, which

3
Nivkh is certainly a true isolate, theories arguing for external relations of Yenisseyan are extremely
peoblematic, cf. GEORG 2000
came to be known as Yenisseyan, since all varieties belonging to it are or were spoken
on the banks of the river Yenissey or within the area of its large tributary system.
Their genetic relationship has always been clear and uncontroversial, so we need not
dwell on this here5 (the question whether the genetic unit Yenisseyan may, after all,
form part of a larger genetic grouping together with other languages and families of
the Old, or even the New, world, is, however, very controversial and continues to be
debated vigorously).
In the following, I'll try to summarize the history of the Yenisseyan languages in a
rough chronological order, beginning from reconstructed Proto-Yenisseyan and its
discernable interactions with other language families of Inner and Northern Asia
down to the present day. While the history of every language which finds itself in the
situation of serious endangerment in our times will naturally be a history of decline
and retraction, the case of Yenisseyan is somewhat special, among the languages of
Northern Asia at least, since it can be shown that this language family once not only
occupied a considerably greater territory than its remnants do now (or did as late as at
the time of the Russian conquest), but that its speakers may once have played a quite
important role in one of the more influential theatres of world history, namely the
South Siberian/Inner Asian steppe zone. Arguments in favour of this may be drawn
from two types of sources, toponymy and etymology.
Ever since it was first developed in the context of Indo-European linguistics, the study
of toponymy has been regarded as a potentially powerful device for the determination
of prehistoric ethnolinguistic territories and boundaries; toponyms, especially
hydronyms, are generally known as diachronically highly stable, often surviving
repeated changes of populations, which continue to use the names for rivers and other
natural objects in spite of the fact that they may be semantically meaningless for
them. The wealth of place-names in the Americas, which have been accepted and
taken over by European settlers is a well-known testimony to this fact. Outside of
Europe and the Americas, the historical study of placenames as an aid to determine
the ethnolinguistic makeup of a region for the time before written records appear has
not been employed widely, but a remarkable exception is Western and Central

4
Adapted from JANHUNEN 1991; the first column lists the number of persons who claimed membership
in the respective ethnic group (or "nationality"), followed by the number of native speakers of the
language and the percentage of speakers among the ethnic group.
5
For the reconstruction of the parent language ("Proto-Yenisseyan") cf. STAROSTIN 1982, VERNER
1990
Siberia. Due to the efforts of A.P. DUL'ZON, one of the pioneers of Ket/Yenisseyan
studies in Russia after World War II and the founder of the Ketological school at the
Pedagogical University of Tomsk, thousands of place-names of the region have been
collected and linguistically analyzed. One of the most remarkable results of this large-
scale project was, that a large territory in Central Siberia is characterized by place-
and river-names, which are not analyzable from the languages spoken there in
historical times, but rather from Yenisseyan. This territory extends well over the great
rivers Ob' and Irtysh in the west, reaches the Irkutskaja Oblast' in the east, and finds
its southern boundary only in the Sayan/Altay mountain range in Southern
Siberia/Western Mongolia.
The toponomastic "leading fossils", which reveal the Yenisseyan origin of these
names are frequently found suffixal elements, which are matched by Yenisseyan
words for "water" and "river", such as –ses (cf. Ket s'es' "river"), -_es (Yugh), -tet
(Pumpokol), -set (Arin), -_et (Kott), or –ul (cf. Ket ul' "water"). This vast area is now
mostly inhabitated by Uralic-speaking peoples, Ob’-Ugrians and Samoyeds (now
extinct Samoyedic languages, Kamass, Koibal, Tawgy, Mator, were also spoken here,
Kamass well into the 20th century); the other ethnolinguistic group which is now
dominating this territory is Turkic, with Tuvan and Tofa, Khakass, Shor, Ojrot/Gorno-
Altaj, and the _ulym-Turkic language being present in the territory where Yenisseyan
place-names dominate. One of the earliest groups of Turkic written documents comes
from exactly this region - the Yenissey-Runic inscriptions.
Yenisseyan loanwords in Southern Samoyed and Turkic languages like Khakass and
Shor, which cannot have been in contact with Yenisseyan for at least half a
millennium, further add to this picture6
However, both the reconstructable Yenisseyan proto-language and the names of rivers
and villages do not provide any clue to answering the question of dating, i.e. when did
Yenisseyan speakers come to be present in this region, when did they begin to exert
linguistic influences on other languages of the region and when did they cease to do
so.
In the absence of direct historical sources, or the clear and unambiguous identifiability
of an archaeologically observable culture with early Yenisseyan speakers this
question may well remain unanswerable forever, but one indirect hint that we may be

6
Cf. _ISPIJAKOV 1976, _ISPIJAKOVA 1992
dealing with a considerable time-depth indeed comes from areal-typological
observations.
It is well known that Turkic languages do not tolerate initial liquids, like /l-/ or /r-/,
which is in fact a phenomenon very widespread in Eurasia, and it should not surprise
us that Yenisseyan, to a degree, shows almost the same restriction; while it does
tolerate initial /l-/, initial /r-/ is never found in native words. Another phonological
restriction of Turkic is, however, not so widespread, neither in Eurasia, nor on a
global scale, and this is the Turkic horror nasalium, neither /n-/ and no /m-/ may
occur initially in native Turkic words, unless another nasal consonant further in word
is present, which may assimilate an initial stop regressively. Actually, the same rule
holds for Yenisseyan, initial nasals are restricted to loan words and words with a
second nasal in inlaut position. This pattern seems too specific, and typologically too
unusual to think of anything else here, than of the direct consequence of early
language contact, already at the level of the respective proto-languages. While it
remains unclear, which language family was the donor and which the recipient of this
highly marked feature, one further observation, from the domain of lexical etymology,
may shed some light on early Yenisseyan-Turkic contacts, strengthening the
assumption that an early variety of Yenisseyan – or undifferentiated Proto-Yenisseyan
– was not only present in the region, when the earliest historically datable nomadic
confederations entered documented history, but that its speakers may at times have
played a rather important political and cultural role in some of them.
As argued in detail in GEORG (2001), one of the most widespread and culturally
significant terms of early Turkic (and Mongolian), which for internal reasons has
always been regarded as a foreign loan in Turkic (and a Turkic loan in Mongolian)
finds a satisfying etymological explanation in Yenisseyan. The Turko-Mongolic word
for the “sky”, the “sky-god”, or “God”, /tängri/, attested from the earliest Turkic
sources (8th century CE) onward, is doubtlessly a loan from Proto-Yenisseyan *tıNgır
"high"7.
The earliest attestation of this term in any source, however, antedates the era of
documented Turkic considerably. The word occurs in Chinese sources of Han times,
and is there explicitly labelled as belonging to the language of the Xiongnu, a nomadic

7
The vowel of the second syllable is somewhat unclear, the difference in form can be accounted for in
terms of well-documented processes internal to Turkic, the shift of meaning "high" > "sky" has many
parallels in Inner and Northern Asian languages.
confederation of as yet undetermined, but doubtlessly heterogeneous, ethnic elements,
which threatened Northern China from the 4th c. BCE to the 2nd c. CE. The language
of the Xiongnu8 (if there ever was a single, or a single dominant Xiongnu language)
remained unwritten, but the possibility that some of the isolated words and phrases9 of
this mysterious language might be Yenisseyan in origin has been discussed since the
1950s10. These, now together with /tängri/, provide a strong indicator of Yenisseyan
presence – and importance – among the ethnic groups which formed one of the most
powerful nomadic states of early Inner Asia.
After that the Yenisseyans and their language(s) disappear from documented history
until the advent of Russian fur-traders (and later settlers), who first appear east of the
Ural mountains with Ermak's famous quest in 1581. The river Yenissey was reached
by Russians in 1610 (the important fortresses and later cities of Yenisseysk and
Krasnoyarsk were founded in 1619 and 1628 respectively, right within the territory of
Yenisseyan-speaking tribes). At this time, the name of "Yenissey Ostyaks" was
coined for the Kets, which lingered on in the literature until the 20th century. Russian
colonists tranferred the name of the Ostyaks of the Ob'11 to peoples further eastward,
which happened to resemble West-Siberian Ob'-Ugrians in terms of outward
appearance, ways of life etc. Language did not play a role in that12.
Few accounts from these earliest days of Russian-Yenisseyan contact are known
which could clarify the ethnolinguistic situation, but Jasak (= tax) lists for the various
regions which came under Russian control are available from the middle of the 17th
century. Using these sources, B. O. DOLGICH13 could ascertain that speakers of

8
The Xiongnu of Chinese sources are often equated with the Huns of European history, but cf.
DOERFER 1973 for the penumbra of difficulties connected with such a rash identification.
9
The only "text" in "Xiongnu" is a two-line poem found in a Chinese chronicle from the 1st c. CE,
which now is read and analyzed as Yenisseyan by A. VOVIN 2000. This identification is not without
problems, but likely to be basically on the right track.
10
Cf. LIGETI 1950, PULLEYBLANK 1962
11
Nowadays known as the Khanty nationality, living, together with the Mansi (formerly known as
Vogul) in the Chantimansijskij Nacional'nyj Okrug (cap. Chantimansijsk); both Khanty and Mansi
speak Ob'-Ugrian languages, which form, together with Hungarian, an independent branch of the
Finno-Ugric language family.
12
Another ethnic group which recieved the name of "Ostyaks" this way were the Ostyak-Samoyeds,
now known as Sel'kups. In the 30s of the 20th century, the official Soviet terminology for the various
ethnic groups of the Union was changed, introducing self-designations in place of traditional names,
some of which were felt as derogatory terms. Thus, e.g., Zyryenes became Komi, Lamuts became
Ewens, Lapps became Saami and so on. As a curiosity, it should be mentioned that, of all groups which
went by some form of the "Ostyak" name in the 19th century, the Kets were the only ones who
embraced it themselves. Thus, alongside the official self-designation ke/t (often also Keto , which is
originally the vocative form), Ket speakers even today refer to their language as ostÈgan' .
13
DOLGICH 1960
Yenisseyan languages were living in an area much smaller than that defined by the
extension of Yenisseyan-type toponyms. Though in the South East, the confines of the
Irkutskaja Oblast' were still reached, and in the South the city of Krasnoyarsk, the
western boundaries of the language family had been withdrawn largely to the
watershed, which divides the tributary systems of the Ob' and the Yenissey.
Only in the Far North, where Yenisseyan toponyms are rare and recent, the language
family was still winning ground. While the northern boundary of Ket roughly
coincided with the river Yeloguy and the settlement of Verchneimbatsk in the 17th
century (where the Ket dialect spoken today is called Southern Ket), Ket speakers
advanced as far as the river Kurejka (a right tributary of the Yenissey, north of
Turuchansk), some 300 km north of Verchneimbatsk and well beyond the polar circle,
during the following centuries.
Apart from language-shift to the idioms of neighbouring, Samoyed, Turkic, and
Tungus groups, a further factor which led to the reduction of speakers of Yenisseyan
and other native Siberian languages during the 17th century - a close analogy to the
situation in post-columbian America - was the spread of imported diseases.
Epidemies of smallpox are recorded for the Yenisseyan territory in 1630/1, and the
1660s. No exact figures of victims are known, but the impact on Kets, Sel'kups,
Khanty, Mansi, and Ewenks are reported as devastating.
The first reports about the actual languages spoken by these people reached the
scholarly world only considerably later. The names of D.G. MESSERSCHMIDT (1723),
J.E. FISCHER (1730-47), G.F. MÜLLER and his co-traveller J.G. GMELIN (1733-43),
who travelled through remote Siberia, often with more botanical than linguistic
interests in their minds, but nevertheless keeping their eyes open for Siberia's
linguistic diversity, must be mentioned. From their notes and travel diaries, sometimes
published during their lifetimes, but often only considerably later unearthed in
Russian archives and analyzed by Yenisseyanists of the 20th century, we have
linguistic information on the six Yenisseyan languages mentioned above14. For Arin
(on the left bank of the Yenissey, south of Yenisseysk), Assan (on the right bank of
the river, south of the Angara) and Pumpokol (NW of Arin), everything we possess
are lexical lists. By the end of the 18th century, all these three languages had

14
Or five, if we stick to the traditional terminology, which viewed Yugh as a mere dialect of Ket (=
Sym-Ket). The current view, that Ket and Yugh have to be regarded and described as seperate
completely disappeared15, leaving only Ket, Yugh and Kott as the sole surviving
Yenisseyan idioms16.
For Kott, we are more fortunate. The Finnish pioneer-linguist, and true founder of
Yenisseyan linguistics, Matthias Alexander CASTRÉN , brought an exhaustive
grammatical description of Ket, Yugh, and Kott home from his Siberian journey
(1845-48). The Kott language was, already then, on the brink of extinction. Castrén
found only five persons, who still spoke this southernmost Yenisseyan language,
which was earlier found on the shores of the Kan' and Birjusa rivers. In his university
lectures (CASTRÉN 1857, 88) he writes:

Diese fünf Personen waren übereingekommen ein kleines Dorf am Agul anzulegen, wo sie
ihre Nationalität aufrecht erhalten wollten, theils aus Liebe zu derselben, theils auch aus der
Ursache, weil Sibiriens Eingeborne der russischen Regierung geringere Abgaben als die
Russen zahlen. An diese Colonisten haben sich später einige von den Kotten herstammende
Familien angeschlossen, welche bereits ihre Muttersprache vergessen haben und Russen
geworden sind. Indessen liegt es auch diesen gegenwärtig sehr am Herzen, sowohl sich selbst
als ihren Kindern die kottische Sprache beizubringen und es ist möglich, dass die kleine
Colonie noch lange ihre Nationalität, welche bereits als erloschen angesehen wurde,
beibehalten werde17.

Nothing is known about the further fate of this village, but by the end of the 19th
century the Kott language had finally ceased to exist.
No continuous texts in the Kott language have been recorded, but, with the glossary
and the numerous morphological paradigms Castrén managed to save, Kott is

languages was only developed from the 60s of the 20th century onwards. The language now generally
referred to as "Ket" is the Imbatsk-dialect of the earlier literature.
15
Most of the surviving lexical material have been collected and edited by DUL'ZON 1961 (appx. 550
lexical entries in total)
16
But the process of the gradual disappearance of Yenisseyan languages has started earlier. Of course it
was well underway already before contact-time, as evidenced by the discrepancy between the territory
of Yenisseyan toponymy and the ethnolinguistic picture drawn by 17th century sources. Additionally,
several Yenisseyan tribes and their languages disappeared between the first Jasak-records and the
advent of non-fur-trading, scientifically minded Europeans. Thus, we have indirect indications that the
tribes of the Baikot, Yarin, and Yastin spoke Yenisseyan languages or dialects, too, albeit no data at all
of these hypothetical languages have survived. One of the first dialects of Ket ever recorded, however
superficially, that of the Eed-shesh river – the exact location of which is unknown – vanished shortly
after MESSERSCHMIDT'S journey, cf. WERNER 1997a, 4.
17
Tr.: "These five persons had agreed to found a small village on the river Agul, where they intended to
keep their nationality alive, partly out of love for it, partly because the indigenous peoples of Siberia
pay less taxes than Russians. These settlers were later joined by some families of Kott origin, who have
already forgotten their native language and have become Russians. However, they, too, have the strong
wish to learn the Kott language and to pass it on to their children, so that it seems possible that this
small settlement might preserve their nationality, which had already been regarded as extinct."
considerably better known than the abovementioned Southern Yenisseyan languages.
These data suffice to allow for the hypothesis that the main factor which led to the
disappearance of Kott was language shift to some Southern Siberian variety of Turkic.
For, Kott, especially when compared to Ket and Yugh, shows strong traits of areal
influence from Turkic. Not only do Turkic loanwords abound in Kott18, but several of
its structural-typological characteristics suggest that the language had been already
under prolonged and intensive Turkic areal pressure at the time of its recording.
Thus, instead of the complicated vowel system of Ket with three central vowels,
differentiated by height (/È/, /´/, /ø/), Kott shows only one central vowel, namely /È/,
just like most Turkic languages.
More striking is the transformation of the verbal system. In both Ket and Kott, the
predominant morphological technique is agglutination, but, whereas in Ket and Yugh
affixes of person agreement can, as a rule, only be prefixes19, in Kott these affixes are
exclusively suffixes. Especially in the singular, the typological parallels between Kott
and, say, Khakas (South-Siberian Turkic), are quite striking, when compared to a
(rather untypically simple) prefixing Ket paradigm:

Kott Khakas Ket

“to wash” “to throw” "to go"

1.P.Sg. urka:k-N at_a-m bOgOtn'

2.P.Sg. urka:k-u at_a-zıN kugOtn'
3.P.Sg. urka:k-ø at_a-ø OgOtn' (m.), u- gOtn' (f.)
1.P.Pl. urka:gan-toN at_a-bıs døNOtn'
2.P.Pl. urka:gan-oN at_a-zar køNOtn'
3.P.Pl. urka:gan-ø at_a-lar ONOtn'

Few scholars, even fewer linguists, visited the Yenissey region after CASTRÉN, and we
have to wait well into the 20th century to learn more about the state and fate of the

18
These were studied chiefly by TIMONINA 1978, 1985. Cf. also WERNER 1997b, a work which
assembles all available knowledge on Kott. The recorded Turkic loans of Kott are often also found in
extinct Southern Samoyed languages, which themselves have been absorbed by Turkic languages
during the last two centuries, possibly by closely related Turkic-speaking groups (cf. JOKI 1952,
GEORG 1999).
19
Better: they can only occur before the root, since an intricate system of preverbs and other affixes
may leave the actual personal affixes somewhere in the middle of the affix-chain, rather than always at
the leftmost position.
Yenisseyan languages. In the 1920s and 30s H. FINDEISEN20 and K. DO N N E R21
collected ethnographical data, as well as a few texts, but the first grammatical
description of Ket after
was published only in 1934 by KARGER22. KARGER was also responsible for the first
attempt to introduce a Latin-script writing system for the Ket language, based on the
Central Ket dialect. At least one school primer of Ket got published23, before the
project of alphabetizing the Ket language was discontinued24.
From ca. 1950 onwards, Tomsk and its Pedagogical Institute (later The Pedagogical
University) became the centre for Yenisseyan studies (as well as for studies on most
indigenous languages of the Tomsk and Krasnoyarsk oblasti, such as Khanty, Sel'kup,
_ulym Turkic, Enets, Nenets and others). The energetic personality of A. P. DUL'ZON
initiated large-scale linguistic field-work activities (as well as archaeological
investigations and the thorough study of Siberian toponymy, s.a.) on the vanishing
languages of the region, and put Yenisseyan studies on a firm methodological and
factual footing. All aspects of Ket and Yugh grammar were studied extensively by
D UL 'ZON and his pupils; the impressive scholarly output of the Tomsk school is
epitomized by DUL 'ZON 's own Ket grammar25 and the encyclopedic series of
monographs published by Heinrich WERNER (G.K. VERNER) in the 1990s26. Under the
supervision of the latter, a second attempt to create a written language for Ket was
launched in the late 80s of the last century. This time, the dialectal basis is the
Southern Ket dialect (as spoken in Kellog, which is informally known as the Ket
"capital") and a modified cyrillic alphabet is used. So far, three school primers and
some readers have been published. Instruction in Ket has begun in a few elementary
schools, but the prospects of success may be regarded as shaky, since the Ket
language now seems to be rather irreversably on the decline, as a closer look at some
statistical data will show.

20
E.g. FINDEISEN 1929
21
DONNER 1933, and especially 1955; another valuable ethnographic description from these times is
the rare booklet by DOLGICH 1934.
22
KARGER 1934a
23
KARGER 1934b
24
KARGER himself did not survive the Soviet purges of the late 30s; he got executed under the charge
of having propagated "Ket nationalism".
25
DUL'ZON 1968
26
see the list of references; apart from Tomsk-based scholars, E.A. KREJNOVI_ must be mentioned as
one of the most prolific and influential Ketological scholars in the second half of the 20th century, cf.
i.a. KREJNOVI_ 1968.
For the 19th century, the development of the Ket nationality presents itself as
follows27:

1836/39 1859 1897

"Yenissey-Ostyaks" of Upper and 740 911 888
Lower Imbatsk and the
Podkamennaya Tunguska
"YO" of the Symsko-Kasov-Tribe 164 161 94
total 904 1072 982

The "Yenissey Ostyaks" of the Symsko-Kasov-Tribe were speakers of the Yugh

language (Sym-Ket); later Soviet censuses did not differentiate these from Ket
speakers. The decline of Yugh is obvious, and by the time of the 1980s the last
speakers of this language had died in Vorogovo and Jarcevo.
During the Soviet era, the following census data became available:

1926 1959 1970 1979 1989

Kets (general) 1428 1017 1182 1122 1084
number of Ket native speakers ? ? ? 885 589
language retention ratio 78,9 % 54,3 %

On the whole, the number of individuals claiming Ket ethnicity for themselves seems
to be remarkably stable, but the language retention rate is decreasing rapidly. This
observation is considerably strengthened by the wealth of data we find in the work of
V.P. KRIVONOGOV28, who visited every single settlement where Kets still live
between 1991 and 1995 and managed to assemble a great amount of fine-tuned
sociolinguistic data. According to KRIVONOGOV , Ket was spoken in the following
villages and settlements (from N to S)29:

population number of Kets % of Kets

27
These data are adapted from PATKANOW 1912;
28
KRIVONOGOV 1998
29
The difference between the number of Kets given in the most recent census reports and the overall
number of Kets found in the region by KRIVONOGOV is, of course, accounted for by the fact that a
certain number of Kets now live in Siberian cities like Krasnoyarsk, or in places as far afield as
Moscow or St. Petersburg.
Kurejka 600 8 1.3
Madujka 86 53 61.6
Svetlogorsk 2.000 16 0.8
Goro_icha 212 58 27.4
Turuchansk 8.400 23 0.3
Star. Turuchansk 330 3 0.9
Kostino 85 10 11.8
Baklanicha 84 23 27.4
Vere__agino 216 25 11.6
Surguticha 299 91 30.4
Kangatovo 58 21 36.2
Verchneimbatsk 820 25 3.0
Kellog 405 247 61.0
Bachta 283 44 15.5
Sumarokovo 93 16 17.2
Sulomaj 243 154 63.4
Bor 4.500 15 0.3
Vorogovo 1.300 17 1.3
Zotino 800 5 0.6
Sym 135 10 7.4

Regarding the internal dialectal division of Ket, all settlements south of (and
including) Kangatovo are home to Southern Ket, in Baklanicha, Vere__agino and
Surguticha Central Ket is spoken, and north of this is the territory of Northern Ket30.
Only in three of these settlements Ket speakers form the majority, which in no case
exceeds 2/3 of the population. Madujka, with its very few inhabitants, does not really
count, while the two villages with the greatest percentage of Kets, Kellog and
Sulomaj, are situated in quite inaccessible regions on tributaries of the Yenissey. It
seems highly likely that this remoteness itself is a decisive factor for the ethnic
composition of these settlements (Madujka, too, is on a right tributary of the
Yenissey); the villages right on the bank of the Yenissey, which is a major water-road
between the densely populated south around Krasnoyarsk and the Far North with the
industrial city of Noril'sk, always attracted more Russian and other non-local settlers
than places, which are rarely served by regular transportation means and offer little
economic prospects.
All in all, only 454 five ethnic Kets live in settlements, where this nationality forms
the majority. In addition to the decreasing language retention rate (48,3 % in 1989),
the future prospects of Ket as a living language looks even less promising when we

30
The dialectal differences between these variants are quite sharp and may at times stand in the way of
mutual comprehension. Northern Ket is the least studied Ket variety, most extant data pertain to
Southern Ket.
take into account Krivonogov's data on language proficiency. Of all ethnic Kets in the
Yenissey regions, the following ratios (in %) are found regarding their ability to use
Ket and other languages in all situations of daily life:

language fluently with difficulties with great difficulties passively not at all
spoken

Ket 21.9 7.6 12.5 23.1 34.9

Russian 95.8 2.8 1.2 0.1 0.1
Sel'kup 2.0 1.3 1.1 3.3 92.3
Ewenki 0.8 0.4 0.1 2.0 96.7

The low figure of 21,9 % full speakers (not more than 190 persons) is already
alarmingly low, but, as might be expected, if we have a closer look at the percentage
of speakers in various age groups, it becomes clear that Ket may be fairly alive among
persons over the age of forty, while for individuals under thirty years of age the Ket
language is certainly no longer a means of daily communication.
Language abilities by age roups (Ket only):

age group full Ket speakers with difficulties not at all

70+ 92.9 7.1 0

60-69 88.1 11.9 0
50-59 58.2 27.3 14.5
40-49 55.3 36.2 8.5
30-39 36.4 54.5 9.1
20-29 14.0 59.8 26.2
10-19 1.3 54.7 44
0-9 1.3 29.6 69.1

One word of caution might be added, lest the percentage of persons who speak Ket
"with difficulties" is viewed with too much optimism. All too often, at least in the
places which the present writer was able to visit in 1999 and 2000 (Baklanicha and
Vere__agino), this label has to be tacitly translated into "with considerable
difficulties". It is very hard to find fluent speakers, and it is even harder to overhear a
spontaneous Ket conversation between people under forty. Even some elderly
persons, which classify themselves as native speakers, who have not begun to use
Russian before entering primary school (some of which have served as language
consultants in their youth and keep lively memories of Ketologists like A. P. DUL'ZON
working with them in the 50s) may now at best be classified as semi-speakers, whose
Ket speech is patched with long stretches of Russian. KRIVONOGOV offers some
further interesting statistics, which show that it is mostly the older generation which
still uses Ket spontaneously, or which can expect being addressed in Ket these days.
Asked, which language they use most in their daily lives, Ket speakers gave the
following answers (in %):

Ket Russian both

70+ 50 28.6 21.4

60-69 45.2 38.1 16.7
50-59 5.5 70.9 23.6
40-49 6.4 78.7 14.9
30-39 2.3 83.3 14.4
20-29 1.1 86.6 12.3
10-19 0 97.3 2.7
0-9 0 97.8 2.2

The most dramatic caesura is obviously that between the age cohorts of the 50-59
years old (born between 1930 and 1940) and older persons. While some of the latter
have been exposed to elementary schooling in Ket in their youth, the next cohort had
to grow up in the turmoil of World War II, when fathers fought on the front and a
great number of deported persons (mainly Volga Germans and Balts) were resettled in
villages of the North (where they and their descendents still live) with whom Russian
was the only possible means of communication.
And on the question with whom and when Ket is spoken, and when other languages
are used, Krivonogov obtained the following figures (in % of those who speak Ket):

language spoken Ket Russian (and other lg.) both

with parents 30.3 23.1 23.1

with spouse 14.8 14.5 14.5
with siblings 10.4 12.5 12.5
with children 9 16.5 16.5
when working 3.2 19.9 19.9
Ket, thus, keeps losing its social functions. It is mainly used by the elderly, or by
(adult) children communicating with their parents. From KRIVONOGOV'S data we also
learn that no single monolingual Ket speaker exists31 (and that a minuscule ratio of 2.8
and 1.2 % admit to speak Russian only "with (great) difficulties"); the prospects of
keeping Ket alive for more than the lifespan of the present middle generation thus
seem to be very low.
The most threatened of all Ket varieties still spoken is certainly the Central Ket
dialect, especially its subvariety which is known in the literature as the dialect of
Pakulicha. While it was chosen as the basis for the short-lived written language of
the 1930s, it is now spoken only by very few individuals, and used on a daily basis by
almost none of them. Though this dialect is now spoken in Baklanicha and
Vere__agino, on the banks of the Yenissey, it was formerly the main language of the
settlement of Pakulicha, which was situated on the left tributary of the Yenissey of
the same name, until it was officially dissolved by Soviet authorities in the early
1960s as one of the numerous Siberian villages "without perspective". The transfer of
Pakulicha's population to their present homes is felt, by those who still remember, as
the major caesura of their lives. Shortly after their resettlement, in the early 1970s,
they were forced to abandon their traditional activity of reindeer breeding. After the
end of the Soviet Union, most adults lost any opportunity to work and stayed
unemployed ever since. Pensions are payed erratically only. Fishing, hunting and
gathering wild berries and mushrooms are thus the only means left to ensure survival.
Needless to say, these circumstances do not encourage younger people to stay in their
villages, and most of them plan to move to one of the Siberian cities, or even further
away. Among those who stay, the omnipresent abuse of alcohol takes its toll, too.
Languages are, first and foremost, problem-solving devices. For most Ket speakers,
situations where only the use of Ket will allow them to solve real-life communicative
problems will be extremely rare in their daily lives. However intensively linguists and
local teachers strive to save the Ket language from oblivion, the one indispensable
prerequisite for maintaining a native language and passing it on to the next generation
is the need to use it. When in each communicative situation a potential, but restricted,
Ket speaker will be confronted with, adifferent language, Russian, will be available, a

31
The 0.1 %, who do not speak Russian have Sel'kup as a second language.
language which is known by all members of the community, while only the elderly
remain full speakers of Ket, little can be done to prevent the ultimate end of, first the
use of, finally the knowledge of Ket, and, with it, the Yenisseyan language family.

References

CASTRÉN, M.A.
(1857) Ethnologische Vorlesungen über die altaischen Völker nebst
samojedischen Märchen und tatarischen Heldensagen, St.
Petersburg: Akademie
(1858) Versuch einer jenissej-ostjakischen und kottischen Sprachlehre
nebst Wörterverzeichnissen aus den genannten Sprachen, St.
Petersburg: Akademie
_ISPIJAKOV, _.F.
(1976) _orsko-ketskie paralleli v leksike, Jazyki i Toponimija 3,
Tomsk, 73-76
_ISPIJAKOVA, F.G.
(1992) Istori_eskie plasty v leksike kondomskogo dialekta, in:
An_iganova, O.P. et al. (red.): Chakasskaja dialektologija,
Abakan, 144-154
CZAPLICKA, M.A.
(1914) Aboriginal Siberia. A study in social anthropology, Oxford:
Clarendon Press (repr. ibid. 1969)
DOERFER, G.
(1973) Zur Sprache der Hunnen, CAJ 17, 1-50
DOLGICH, B.O.
(1934) Kety, Moskva/Irkutsk: Ogiz
(1960) Rodovoj i plemennoj sostav narodov Sibiri v XVII v., Moskva:
Nauka
DONNER, K.
(1933) Ethnological Notes about the Yenissey-Ostyak (in the
Turukhansk Region), Helsinki: Suomalais-Ugrilainen Seura
(1955) Ketica. Materialien aus dem Ketischen oder Jenissei-
Ostjakischen, Helsinki: Suomalais-Ugrilainen Seura
DUL'ZON, A.P.
(1961) Slovarnye materialy XVIII v. po ketskim nare_ijam, UZTGPU
XIX/2, 152-189
(1968) Ketskij jazyk, Tomsk
FINDEISEN, H.
(1929) Reisen und Forschungen in Nordsibirien. Skizzen aus dem
Lande der Jenissejostjaken, Berlin
FORTESCUE, M.
(1998) Language Relations Across Bering Strait. Reappraising the
Archaeological and Linguistic Evidence, London/New York:
Cassell
GEORG, St.
(1999) Rev. of. Werner 1997b, UAJb N.F. 16, 304-309
(2000) Methodologische Bemerkungen zum Problem der äußeren
genetischen Beziehungen der jenissejischen Sprachen, in: O.A.
Osipova (red.): Sravnitel'no-istori_eskoe i tipologi_eskoe
izu_enie jazykov i kul'tur, Tomsk: TGPU, 128-139
(2001) Türkisch-mongolisch tengri "Himmel, Gott" und seine
Herkunft, SEC 6
JANHUNEN, J.
(1991) Ethnic death and survival in the Soviet North, JSFOu 83, 111-
122
Joki, A.
(1952) Die Lehnwörter das Sajansamojedischen, Helsinki: Suomalais-
Ugrilainen Seura
Karger, N.K.
(1934a) Ketskij jazyk, in: Jazyki i pis'mennost' narodov Severa III.,
Leningrad: Nauka, 223-238
(1934b) Bukvar (na ketskom jazyke), Leningrad
KREJNOVI_, E.A.
(1968) Glagol ketskogo jazyka, Leningrad: Nauka
KRIVONOGOV, V.P. Kety na poroge III tysja_eletija, Krasnojarsk: KPI
(1998)
LIGETI, L.
1950 Mots de civilisation de Haute Asie en transcription chinoise,
AOH 1, 141-188

NIKOLAEVA, I.A.
(1988) Problema uralo-jukagirskich svjazej, Diss. Kand. fil. nauk,
Moskva
PATKANOW, S.
(1912) Über die Zunahme der Urbevölkerung Sibiriens, Keleti Szemle
9, 54-94
PULLEYBLANK, E
1962 The consonantal system of Old Chinese, AM N.F. 9, 58-144,
204-265
SEEFLOTH, U.
(to appear, 2001) Die Entstehung polypersonaler Paradigmen im Uralo-
Sibirischen, ZAS 30

STAROSTIN, S.A.
(1982) Praenisejskaja rekonstrukcija i vne_nye svjazi enisejskich
jazykov, in: Alekseenko et al. (edd.): Ketskij sbornik.
Antropologija, etnografija, mifologija, lingvistika, Leningrad:
Nauka
TIMONINA, L.G.
(1978) Tjurkskie zaimstvovanija v kottskom jazyke, SovTjurk 1978/3,
7-13
(1985) Tjurkskie zaimstvovanija v enisejskich jazykach v sravnitel'no-
istori_eskom osve__enii, Avtoref. Kand. fil. nauk, Leningrad
VERNER, G.K.
(1990) Sravnitel'naja fonetika enisejskich jazykov, Taganrog: TPI
VOLODIN, A.P./VAKHTIN, N.B. KIBRIK, A.A. (edd.):
(1996) Jazyki Mira. Paleoaziatskie jazyki, Moskva: Indrik
VOVIN, A.
(2000) Did the Xiong-nu Speak a Yenisseyan Language ?, CAJ 44/1,
87-104
WERNER, H.
(1995) Zur Typologie der Jenissej-Sprachen, Wiesbaden: Harrassowitz
(1997a) Die ketische Sprache, Wiesbaden: Harrassowitz
(1997b) Abriß der kottischen Grammatik, Wiesbaden: Harrassowitz
(1997c) Das Jugische (Sym-Ketische), Wiesbaden: Harrassowitz
(1998) Probleme der Wortbildung in den Jenissej-Sprachen, München:
Lincom-Europa