Tundra Nenets is a Uralic language spoken in European Russia and North-Western Siberia. It is the largest and best-preserved language in the Samoyedic group.[2]: 1–2
Tundra Nenets | |
---|---|
Nenec, Nenetsy, Nentse, Yurak, Yurak Samoyed | |
Native to | Northern Russia |
Ethnicity | Nenets |
Native speakers | 21,900 (2010)[1] |
Language family | |
Writing system | Cyrillic |
Language codes | |
ISO 639-3 | – |
Glottolog | tund1255 |
ELP | Tundra Nenets |
Tundra Nenets is classified as Definitely Endangered by the UNESCO Atlas of the World's Languages in Danger | |
This article contains IPA phonetic symbols. Without proper rendering support, you may see question marks, boxes, or other symbols instead of Unicode characters. For an introductory guide on IPA symbols, see Help:IPA. |
Tundra Nenets is closely related to the Nganasan and Enets languages, and more distantly to Selkup. Tundra Nenets and its sister language, Forest Nenets, are sometimes considered dialects of a single Nenets language, though there is low mutual intelligibility between the two. In spite of the large area in which Tundra Nenets is spoken, the language is very uniform with few dialectal differences.[3]: 13
Geographically, the Tundra Nenets territory spans the Nenets District of the Arkhangelsk Province, as well as parts of the Komi Republic, the Yamal-Nenets District in the Tyumen Province, and the Ust-Yeniseisk region of the Taimyr District in the Krasnoyarsk Region. This territory has been in constant growth over the past millennium, as Tundra Nenets settlers moved further east and engaged with other groups of Enets.[2]: 3
A 2010 census reported 44,640 Nenets, 49% of whom were speakers of the Nenets language. However, while the population of Nenets has been growing in the past few decades, the language itself has been in a decline, as many children are now educated in Russian-language schools and many other ethnic groups have begun settling in Tundra Nenets territories.[2]: 5–6 The language is classified as 6b (Threatened), indicating that it is still spoken by all age generations, but the number of speakers in decreasing.[4]
Tundra Nenets is spoken primarily within family circles and in traditional economic activities, such as hunting and herding reindeer. The language has no official status within the Russian Federation. In the mid 1930s, an orthography based on the Cyrillic script was developed, which is taught in local schools. However, many Tundra Nenets speakers are primarily literate in Russian. Nonetheless, there is a small amount of Tundra Nenets literature, as well as radio and television broadcasts.[2]: 7–8
The syllable structure of Tundra Nenets is generally CV(C), and syllables with initial, medial or final consonant clusters of more than two consonants are not allowed. Words normally do not begin with a vowel, except in western dialects of the language, mostly due to the loss of /ŋ/, so the standard Tundra Nenets word ŋarka ('big') is found as arka in western varieties.[5]
The number of vowel phonemes in Tundra Nenets is 10, which have 17 distinct allophones governed by palatality, which dominates whole sequences of vowels and consonants.[6] Vowel frontness is not segmentally contrastive.
Monophthong vowels are present in the chart below. Phonemes are marked in bold, with their palatal (on the left) and non-palatal (on the right) allophones marked underneath using the International Phonetic Alphabet.
Unrounded | Rounded | ||
---|---|---|---|
Close | Long | /í/ [(ʲ)iː], [ɨː] |
/ú/ [(ʲ)ʉː], [uː] |
Short | /i/ [(ʲ)i], [ɨ] |
/u/ [(ʲ)ʉ], [u] | |
Mid | Tense | /e/ [(ʲ)eː], [ɤː] |
/o/ [(ʲ)ɵː], [oː] |
Lax | /°/ [ə] |
||
Open | Tense | /a/ [(ʲ)aː], [ɑː] |
|
Lax | /ø/ [(ʲ)ɐ], [ʌ] |
There is also a vowel ⟨æ⟩, which is interchangeably realized as [æ͡e̘] or [æː]. This and the long close vowels only occur in word-initial syllables.
In much of the literature on Tundra Nenets and its sister dialect, Forest Nenets, a so-called reduced vowel is mentioned. This reduced vowel was thought to have two distinct qualities depending on whether it was found in a stressed or unstressed position. In stressed position it was transcribed as ⟨ø⟩ and represented a reduced variant of an underlying vowel, and in unstressed position it was transcribed as ⟨â⟩ and represented a reduced variant of /a/. Recently, however, it has become clear that the reduced vowels are in fact short vowels, counterparts to their respective long vowels. Today ⟨â⟩ should simply be replaced by ⟨a⟩, while ⟨ø⟩ simply represents a short vowel, although it is not specified which short vowel in this orthography.[7]
The number of consonant phonemes in Tundra Nenets is 27.[8] All labial and coronal consonants other than the semivowels /w/ and /j/ have plain and palatalized counterparts.
Bilabial | Coronal | Velar | Glottal | ||||
---|---|---|---|---|---|---|---|
Plain | Pala. | Plain | Pala. | ||||
Nasal | m | mʲ | n | nʲ | ŋ | ||
Stop | Voiceless | p | pʲ | t | tʲ | k | ʔ |
Voiced | b | bʲ | d | dʲ | |||
Affricate | Voiceless | ts | tsʲ | ||||
Fricative | Voiceless | s | sʲ | x | |||
Approximant | Semivowel | w[lower-alpha 1] | j | w[lower-alpha 1] | |||
Lateral | l | lʲ | |||||
Trill | r[lower-alpha 2] | rʲ |
All consonants can be found word-internally between vowels, but their occurrence in other positions is strongly limited.[10]
Tundra Nenets has a phonological process of sandhi: the simplification of consonant clusters, both within words (in e.g. inflection) and between words. This allows considering some of the consonant phonemes secondarily derived from underlying consonant clusters.[11]
As the citation form of a noun is the bare stem, a word ending in a glottal stop in isolation can thus underlyingly end either in a plain glottal stop, or in a nasal. The latter is sometimes called a "nasalizable glottal stop", and is in the orthography of the language written differently from the former.
Tundra Nenets has a (C)V(C) syllable structure, and the minimal word is CV. Thus, there are no word initial or word final consonant clusters, nor are there any three-consonant clusters. Moreover, syllables with zero onset typically cannot occur word-initially, but in Western dialects, the word-initial ŋ is lost, giving some vowel-initial words. For example, the Eastern dialect ŋəno 'boat' becomes əno in the Western dialect.[2]: 27 Word-internally, zero onset syllables only occur when ə or ° follow another vowel. For example, such vowel clusters can occur when forming the finite stem: me° 'he takes (3SG)' gives meə-s'° 'he took (3SG.PST).'[2]: 27–28
Tundra Nenets displays bisyllabic trochaic feet that are aligned to the left. Primary stress falls on the initial syllable. Secondary stress falls on subsequent odd syllables and on even-position syllables preceding a syllable with °, excluding the final syllable,[2]: 28 as illustrated in the following examples where ´ indicates primary stress on a vowel and ` indicates secondary stress on the preceding vowel:
mé-nake`-x°yu`-n'ə-s'°
take-PROB-DU.OBJ-1SG-PAST[2]: 28
mé-nake`-x°yu`-n'ə-s'°
take-PROB-DU.OBJ-1SG-PAST[2]: 28
Typical of the Uralic language family, Tundra Nenets has an agglutinating morphological structure with a wide variety of suffixes. There is no prefixation. The two primary word classes are nouns and verbs. Other word classes include adjectives, pronouns, numerals, adverbs, postpositions, conjunctions, particles, and interjections.[2]: 8–9
A noun can contain up to five morphemes, including the root, a derivational suffix, a possessive suffix, a number suffix, and a case suffix. A verb can contain up to six or seven morphemes, including the root, one or two derivational suffixes, a tense suffix, a mood suffix, a subject agreement suffix, and an object agreement suffix. Although the morphology is predominately agglutinating, there are some suffixes that express multiple meanings, as well as periphrastic clausal negation and some auxiliary verbs.[2]: 8–9
Tundra Nenets contains a few nominal derivational affixes that can be used to denote a cause, express an instrument, or refer to a location of action. For example, the noun xərwa-bco 'wish' can be derived from the verb xərwa- 'to want'.[2]: 31 There are also several mixed categories of nouns that have a syntactic distribution of a different word-class, yet share other properties with nouns. For example, the proprietive suffix -sawey° can be used to derive nouns with the meaning 'with X, having X', as in yī-sawey° 'intelligent' (from yī 'mind').[2]: 32
Tundra Nenets has two verbal aspectual classes, perfective and imperfective. There are several derivational aspectual suffixes which can change the aspectual class of a verb. For example, imperfectivizing suffixes can be used to express durative, frequentative, multiplicative, and iterative meanings, such as in tola-bə 'to keep counting' (from tola- 'to count').[2]: 45 There are also denominal verbs with the meaning 'to use as X, to have as X', which are formed from the accusative plural stem, such as in səb'i-q' 'to use as a hat' (from səwa 'hat').[2]: 46
Nouns are inflected for number (singular, dual, plural), case (nominative, accusative, genitive, dative, locative, ablative, prolative), and possessive, which can indicate the person and number of the possessor.[2]: 9 For example, the following noun is inflected for similative case and third person plural number.
Verbs are inflected for agreement, tense, and mood. Present tense is unmarked, but Tundra Nenets distinguishes inflectionally the past, future, habitual, and future-in-the-past tenses. There are sixteen moods, which include the imperative, hortative, optative, conjunctive, necessitative, interrogative, probabilitative, obligative, potential, and inferential.[2]: 9 For example, the verb below is inflected for subjunctive mood, first person singular agreement, and past tense.
ŋod'°q
DP
s'it°
you.ACC
məneq-yi-dəm-c'°
see-SUBJ-1SG-PAST
ŋod'°q s'it° məneq-yi-dəm-c'°
DP you.ACC see-SUBJ-1SG-PAST
'If only I could see you!'[2]: 89
Clitics undergo the same phonological processes and stress assignment as affixes. They can attach to an affirmative finite verb, a negative auxiliary, or a non-verbal final predicate, and follow any other inflection,[2]: 116 as shown with the following exclamative clitic:[2]: 117
pidər°
you
tola-mp'i-n°=n'uq!
read-DUR-2SG-EXCL
pidər° tola-mp'i-n°=n'uq!
you read-DUR-2SG-EXCL
'It turns out you are reading!'[2]: 117
Particles are primarily used for discourse. Common particles include yekar°q 'it is unknown', ŋod'°q 'hardly', tǣr'i 'just, very', and məs'iq 'maybe, perhaps.'[2]: 53 An example is given below:
yekar°q
DP
tūt-bə-ta=w°h
come.FUT-COND-3SG-DUB
yekar°q tūt-bə-ta=w°h
DP come.FUT-COND-3SG-DUB
'It's unlikely that he will come.'[2]: 53
There are some lexical noun-noun compounds in Tundra Nenets. As shown in the following example, the first element in the compound can always be modified and take a number.[2]: 167
A few irregular verbs show suppletion. The most frequent suppletive verbs are xǣ- ‘to go, to depart’, ŋǣ- ‘to be’, to- ‘to come’, ta- ‘to bring, to give’ and the negative auxiliary nʹi-. Some common suppletive forms for these verbs are given in the table below.[2]: 25
to go/depart | to be | to come | to bring, give | negative auxiliary | |
---|---|---|---|---|---|
3SG | xəya | ŋa | to° | ta° | nʹī |
CONNEG | xanʹ°q | ŋaq (~ŋǣq) | tuq | taq | - |
IMPF PART | xǣn(ʹ)a | ŋǣda | tona | tada~tana | nʹinʹa |
IMP 2SG | xanʹ°q | ŋaq | tuq | taq | nʹon° |
FUT 3SG | xan°tə° | ŋǣŋu | tūtə° | tətə° | - |
Tundra Nenets is predominantly a head-final SOV language.[2]: 9 Verb finality is the primary constraint on word order.[2]: 213 Below are examples of the basic word order for a transitive and intransitive sentence.
məy°mpə-da
cheerful-IMPF.PART
Wera
Wera
Maša-m
Masha-ACC
pad°ta°-da
draw-3SG > SG.OBJ
məy°mpə-da Wera Maša-m pad°ta°-da
cheerful-IMPF.PART Wera Masha-ACC {draw-3SG > SG.OBJ}
‘Cheerful Wera drew Masha.’[2]: 197
Wera-h
Wera-GEN
teda
reindeer.3SG
yuxu
get.lost
Wera-h teda yuxu
Wera-GEN reindeer.3SG get.lost
'Wera's reindeer got lost.'[2]: 223
However, although most simple sentences have SOV order, a more general trend is for the informationally new element to be immediately preverbal and to be preceded by the informationally old element. So, it is possible to have sentences where the direct object precedes the subject,[2]: 214 as illustrated below:
s’exari°-m
road-ACC
sira
snow
toxora°-da
cover-3SG > SG.OBJ
s’exari°-m sira toxora°-da
road-ACC snow {cover-3SG > SG.OBJ}
‘Snow covered the road.’[2]: 214
The possessor precedes the thing being possessed.[2]: 142
Comparative adjectives follow their standards, which take the ablative case.[2]: 174
t’uku°
This
pəni°
coat
taki°
that
pəne-xəd°
coat-ABL
səwa(-rka)
good-COMP
t’uku° pəni° taki° pəne-xəd° səwa(-rka)
This coat that coat-ABL good-COMP
‘This coat is better than that one.’[2]: 174
The determiner precedes the noun phrase.[2]: 141–142
tʹuku°
this
Wera-h
Wera-GEN
ti
reindeer
tʹuku° Wera-h ti
this Wera-GEN reindeer
‘this reindeer of Wera’s’[2]: 143
The alphabet of Tundra Nenets is based on Cyrillic, with the addition of three letters: Ӈ ӈ, ʼ, and ˮ.
The palatalized and plain vowel allophones are distinguished in the original orthography[3]: 36–37
phonemic transcription | a | e | o | i | u | æ | |
---|---|---|---|---|---|---|---|
Cyrillic | Plain | а | э | о | ы | у | э |
Palatalized | я | е | ё | и | ю |
The Cyrillic orthography does not distinguish the reduced vowel from a, nor the long ī and ū from their short counterparts i and u. ǣ is not found in a palatalized environment, and thus does not show up in the chart. The schwa, [ə], has no direct counterpart in the Cyrillic orthography and is in most cases not written. However, it may sometimes appear as ⟨а⟩, ⟨я⟩, ⟨ы⟩, ⟨ӗ⟩ or ⟨ŏ⟩. For example, xad°, ('snowstorm') is written as хад, and nix° ('power') is written as ныхы.[3]: 34–35
The consonants in the Cyrillic orthography can be seen in the chart below. Note that palatalized consonants are not included.[3]: 38
phonemic transcription | /m/ | /p/ | /b/ | /w/ | /n/ | /t/ | /d/ | /ts/ | /s/ | /j/ | /l/ | /r/ | /ŋ/ | /k/ | /x/ | /ʔ/ | /ʔ/ |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Cyrillic | м | п | б | в | н | т | д | ц | с | й | л | р | ӈ | к | х | ˮ | ʼ |
The letter ⟨ˮ⟩ marks a "plain" glottal stop, while ⟨ʼ⟩ marks a glottal stop derived from a word-final n. As in Russian, the consonants are palatalized using the soft sign, ⟨ь⟩. For example, the palatalized consonant m' is represented with ⟨мь⟩ in Cyrillic unless it is followed by a palatalizing vowel, such as ⟨ё⟩, so that m'o is written as ⟨мё⟩.[3]: 38
(Article 1 of the Universal Declaration of Human Rights)
Ет
Et
хибяри
xibjari
ненэць
nenėc’
соямарианта
sojamarianta
хуркари
xurkari
правада
pravada
тнява,
tnjava,
ӈобой
ṇoboj
ненээя
nenėėja
ниду
nidu
нись
nis’
токалба,
tokalba,
ӈыбтамба
ṇybtamba
илевату
ilevatu
тара.
tara.
Ет хибяри ненэць соямарианта хуркари правада тнява, ӈобой ненээя ниду нись токалба, ӈыбтамба илевату тара.
Et xibjari nenėc’ sojamarianta xurkari pravada tnjava, ṇoboj nenėėja nidu nis’ tokalba, ṇybtamba ilevatu tara.
All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood.[12]
| |||||
---|---|---|---|---|---|
List of Uralic languages | |||||
Finnic | |||||
Sámi |
| ||||
Mordvinic | |||||
Mari | |||||
Permic | |||||
Ugric | |||||
Samoyedic | |||||
Others |
| ||||
Reconstructed |
| ||||
|