A language family of the Caucasus Mountains, resembling nothing outside the Caucasus. It is probably distantly related to the North-West Caucasian family, and the two (NEC and NWC, as they're commonly written) can be grouped as North Caucasian. They mainly occur in the Russian republics of Daghestan, Chechnya, and Ingushetia, and in neighbouring Azerbaijan.

The main ones are Chechen (1 000 000 speakers), Avar (600 000), Lezgi (450 000), Dargwa (350 000), Ingush (250 000), Lak (100 000), and Tabassaran (100 000). The others are spoken by only a few thousand each and are generally unwritten: but I like the names of them so much that I'm going to list them all. The family is also known as Nakho-Daghestani. The Nakh languages are the North Central Caucasian group, consisting of Chechen (Nokhchiin), Ingush (Ghalghay), and Bats (Batsbi). The Daghestani (or East Caucasian) languages are Aghul, Akhvakh, Andi (Qwannab), Archi, Avar, Baghvalal, Bezhta, Botlikh, Budukh, Chamalal, Dargwa (Kubachi), Dido (Tsez), Ghodoberi, Hinukh, Hunzib (Gunzib), Karata, Khinalugh, Khvarshi, Kryts (Dzek), Lak, Lezgi (Lezgian), Rutul (Rutulian), Tabassaran, Tindi, Tsakhur, and Udi.

See Chechen for greater detail about one. I also have some Ingush material which I intend to work up into a node, but the others are much harder to find anything substantial on, though peculiar features mean they are often quoted in linguistics texts for specific points.

Phonetically they tend to have a rich consonant inventory, though not as full as in the NWC languages. Chechen and Ingush have pharyngeal consonants. Most have uvular stops and fricatives and pharyngeal fricatives; and in voiceless stops distinguish ejective from lax and tense aspirated, such as q' qh qqh. Avar, Akhvakh, and Karata have lateral affricate series, and Avar and Akhvakh also have uvular affricates.

They typically have ergative grammar, but may also show a typology known as active marking, in which the subjects of sentences are marked according to how deliberate or intentional the action was: "the child fell" could be marked to mean "fell accidentally" or "fell deliberately". The accidental marking is the same as that used for the direct object of transitive sentences.

Chechen has noun classes, like genders but more than three in number, with prefixes on adjectives and verbs to agree: again, I don't know how widespread this is in NEC.

One morphological feature they generally have in common is that number and case are marked by separate suffixes. Between those suffixes is another marker which is different for singular and plural. For example, in Akhvakh the plural marker is -di, so nido 'forehead, plural nidodi. That is in the absolutive case, with no explicit case ending. The genitive ending is -de. But before this or any other case ending comes singular -la-, plural -le-: so nidolade 'of the forehead', nidodilede 'of the foreheads'. The case endings themselves are simple, but the plural and oblique stem markers may be complex: in Lak there are around thirty possibilities for each one, with no real predictability of which one a particular noun uses.

Although the North Caucasian languages bear no resemblance to anything outside the region, a minority of linguists speculate about finding deeper links. They believe they can discern the so-called Dene-Caucasian phylum, grouping NWC and NEC with Basque, Sino-Tibetan, and Na-Dene (which includes Navaho and Apache).

Details of oblique stem marking from Greville Corbett, Number, CUP, 2000.