Northeast Caucasian languages

Historically, Northeast Caucasian phonemic inventories were thought to be smaller than those of the neighboring Northwest Caucasian family. However, more recent research has revealed that many Northeast Caucasian languages are much more phoneme-rich than previously believed, with some languages containing as many as 70 consonants.[1]

In addition to numerous front obstruents, many Northeast Caucasian languages also possess a number of back consonants, including uvulars, pharyngeals, and glottal stops and fricatives. Northeast Caucasian phonology is also notable for its use of numerous secondary articulations as contrastive features. Whereas English consonant classes are divided into voiced and voiceless phonemes, Northeast Caucasian languages are known to contrast voiced, voiceless, ejective and tense phones, which contributes to their large phonemic inventories. Some languages also include palatalization and labialization as contrastive features.[2] Most languages in this family contrast tense and weak consonants. Tense consonants are characterized by the intensiveness of articulation, which naturally leads to a lengthening of these consonants.

In contrast to the generally large consonant inventories of Northeast Caucasian languages, most languages in the family have relatively few vowels, although more on average than the Northwest Caucasian languages.[3] However, there are some exceptions to this trend, such as Chechen, which has at least twenty-eight vowels, diphthongs and triphthongs.[4]

These languages can be characterized by strong suffixal agglutination. Weak tendencies towards inflection may be noted as well. Nouns display covert nominal classification, but partially overt cases of secondary origin can be observed too. The number of noun classes in individual languages range from two to eight. Regarding grammatical number, there may be a distinction between singular and plural, plurality itself may impact the class to which a noun belongs.[5] In some cases, a grammatical collective is seen. Many languages distinguish local versus functional cases,[6] and to some degree also casus rectus versus casus obliquus.

The inflectional paradigms are often based on partially classifying productive stem extensions (absolutive and oblique, ergative and genitive inflection. Localization is mostly conveyed by postpositions, but it can be also partly based on preverbs. Noun phrases exhibit incomplete class agreement, group inflection (on the noun) with partial attributive oblique marking, which may, in turn, carry a partially determining function.

Verbs do not agree with person, with a few exceptions like Lak, in which first and second persons are marked with the same suffix and verbs agree with the P argument, and Hunzib in which verbs agree with A argument. Evidentiality is prominent, with reported, sensory and epistemic moods all appearing as a way of conveying the evidence. Epistemic modality is often tied to the tense.

Most Northeast Caucasian languages exhibit an ergative–absolutive morphology.[7] This means that objects of transitive sentences and subjects of intransitive sentences both fall into a single grammatical case known as the absolutive. Subjects of transitive sentences, however, carry a different marking to indicate that they belong to a separate case, known as the ergative.[8] This distinction can be seen in the following two Archi sentences. Note that objects and subjects of intransitive sentences carry no suffix, which is represented by the null suffix, -∅. Meanwhile, agents of transitive sentences take the ergative suffix, -mu.

[9]

Northeast Caucasian languages have between two and eight noun classes.[3] In these languages, nouns are grouped into grammatical categories depending on certain semantic qualities, such as animacy and gender. Each noun class has a corresponding agreement prefix, which can attach to verbs or adjectives of that noun. Prefixes may also have plural forms, used in agreement with a plural noun.[10] The following table shows the noun–adjective agreement paradigm in the Tsez language.

[10]

Unusual agreement targets

In many Northeast Caucasian languages, as well as appearing on adjectives and verbs, agreement can also be found on parts of speech which are not usually able to agree in other language families – for example on adverbs, postpositions, particles, and even case-marked nouns and pronouns.[11][12] In the example from Archi below, doːʕzub ‘big’ and abu ‘made’, but also the adverb ditːabu ‘quickly’ and the personal pronouns nenabu ‘we’ and belabu ‘to us’, all agree in number and gender with the argument in the absolutive case, χʕon ‘cow’.

nena<b>u doːʕzu-b χʕon b-ela<b>u ditːa<b>u χir a<b>u

1PL.INCL.ERG<III.SG> be.big.ATTR-III.SG cow(III)[SG.ABS] III.SG-1PL.INCL.DAT<III.SG> quickly<III.SG> behind <III.SG>make.PFV

‘We quickly drove the big cow to us (home).’ [13]

This kind of clausal agreement has been labelled ‘external agreement’.[14] The same term is also used for the (cross-linguistically even rarer) phenomenon where a converb agrees with an argument which lies outside the converb's own clause. This is seen in the following example from Northern Akhvakh, where mīʟō ‘not having gone’ has a masculine adverbial suffix (-ō), agreeing with hugu ek’wa ‘the man’.

[ĩk’a ri-da-la m-īʟ-ō] hu-gu ek’wa-la w-uʟ’-u-wudi.

long time-INT-ADD N-go.NEG-M[ADV] DIST-LL man-ADD M-die-M-PF3

‘Shortly after that (lit. ‘long time not having gone’), the man died.’ [15]

Spoken in the Northwest Dagestan highlands and western Dagestan. Avar is the lingua franca for these and the Tsezic languages and is the only literary language. Schulze (2009) gives the following family tree for the Avar–Andic languages:

• Avar (761,960)
• Andic languages
  • Andi (Qwannab) (5,800)
  • Akhvakh–Tindi
    • Akhvakh (210 as of 2010)
    • Karata–Tindi
      • Karata (Kirdi) (260 as of 2010)
      • Botlikh–Tindi
        • Botlikh (210 as of 2010)
        • Godoberi (130 as of 2010)
        • Chamalal (500 as of 2010)
        • Bagvalal–Tindi
          • Bagvalal (1,450)
          • Tindi (2,150)

Figures retrieved from Ethnologue.[19] These languages are spoken in the following rayons of Dagestan: Axvax, Botlikh, Buynaksk (Shura), Čarodinsky (Tsurib), Gergebil, Gumbetovsky (Baklul), Gunib, Karabudaxkent, Kazbekovsky (Dylym), Lavaša, Tsumada (Agvali), Untsukul, Xebda, Xunzaq and Zaqatala rayon in Azerbaijan.

Spoken by 492,490 in Dagestan, as well as Azerbaijan, Central Asia and Ukraine.[20] Dargwa proper is a literary language.

• Dargwa (Dargva)
• Kajtak
• Kubachi
• Itsari
• Chirag

Dargwa is spoken in the following rayons of Dagestan: Aquša, Kaitak, Kayakent, Kubači, Sergokala.

Spoken in Quba rayon of Azerbaijan.

• Khinalug (Xinalug) (1,000 speakers)[21]

Spoken in the Central Dagestan highlands. Lak is a literary language.

• Lak (152,000 speakers)[22]

Lak is spoken in two rayons of Dagestan: Kumux and Kuli (Vači).

Spoken in the Southeast Dagestan highlands and in Northern Azerbaijan. The Lezgian language or, as the Lezgian people themselves call it, Лезги чlал (lezgi ch'al), is the biggest in terms of the number of native speakers of all the languages of the Lezgic group (other languages from this group include Tabasaran, Udi, Tsakhur and Rutul). They are spoken in the following rayons of Dagestan: Agul, Akhty, Derbent (Kvevar), Kasumxur, Kurakh, Magaramkent, Rutul, Tabasaran, Usukhchay, Khiv and Quba and Zaqatala in Azerbaijan.

Tabasaran was once thought to be the language with the largest number of grammatical cases at 54, which could, depending on the analysis, instead be the Tsez language with 64.

Lezgian and Tabasaran are literary languages.

• Peripheral: Archi (970 speakers)
• Samur (or Nuclear Lezgian)
  • Eastern Samur
    • Tabasaran (128,900)
    • Lezgian (655,000)
    • Aghul (29,300)
    • Udi (6,590)
  • Southern Samur
    • Kryts (5,000)
    • Budukh (1,000)
  • Western Samur
    • Rutul (47,400)
    • Tsakhur (23,673)

All figures retrieved from Ethnologue.[23]

Spoken in Chechnya, Ingushetia and Georgia. Chechen and Ingush are official languages of their respective republics.

• Bats (3,420 speakers in Georgia in 2000)[24]
• Vainakh languages
  • Chechen (1,350,000)[25]
  • Ingush (322,900)[26]

Spoken mostly in Southwest Dagestan. None are literary languages. Formerly classified geographically as East Tsezic (Hinukh, Bezta) and West Tsezic (Tsez, Khwarshi, Hunzib), these languages may actually form different subgroupings according to the latest research by Schulze (2009):

• Tsez–Hinukh
  • Tsez (Dido) (12,500)
  • Hinukh (Hinux, Ginukh) (5 as of 2010)
• Bezhta–Hunzib–Khwarshi
  • Bezhta (Kapucha) (6,800)
  • Hunzib (Gunzib) (1,420)
  • Khwarshi (Khvarshi) (8,500)[27]

All figures except for Khwarshi were retrieved from Ethnologue.[28] These languages are spoken in the Tsunta and Bezhta areas of Dagestan.

Some linguists such as Sergei Starostin think that the Northeast and Northwest Caucasian languages are part of a wider North Caucasian family,[29] citing shared vocabulary and typological features as evidence.[30] This proposed family does not usually include the neighboring Kartvelian languages.[30] This hypothesis is not well demonstrated.[31]

Some linguists—notably Igor M. Diakonoff and Starostin—see evidence of a genealogical connection between the Northeast Caucasian family and the extinct languages Hurrian and Urartian. Hurrian was spoken in various parts of the Fertile Crescent in the 3rd and 2nd millennia BC. Urartian was the language of Urartu, a powerful state that existed between 1000 BC or earlier and 585 BC in the area centered on Lake Van in current Turkey. The two languages are classified together as the Hurro-Urartian family. Diakonoff proposed the name Alarodian for the union of Hurro-Urartian and Northeast Caucasian.

Some scholars, however, doubt that the language families are related[32] or believe that, while a connection is possible, the evidence is far from conclusive.[33][34]

The Proto-Northeast Caucasian language had many terms for agriculture and Johanna Nichols has suggested that its speakers may have been involved in the development of agriculture in the Fertile Crescent and only later moved north to the Caucasus.[36] Proto-NEC is reconstructed with words for concepts such as yoke (*...ƛ / *...ƛƛ’), as well as fruit trees such as apple (*hʕam(V)c / *hʕam(V)č) and pear (*qur / *qar; *qʕur ?),[35] that suggest agriculture was well developed before the proto-language broke up.