The term Caucasian languages is loosely used to refer to a large and extremely varied array of languages spoken by more than 7 million people in the Caucasus region of West Asia, between the Black Sea and the Caspian Sea. Linguistic studies allow those languages to be classified into several language families, with little or no discernible affinity to each other. Some of those language families have no known members outside the Caucasus area.
The four widely accepted language families that are presently spoken only in the Caucasus region are:
- Northeast Caucasian or Dagestan family. It includes Aghul , Akhvakh , Andi , Archi , Avar, Bagvalal , Bezhta , Botlikh , Budukh , Chamalal , Dargwa , Gigatl , Ghodoberi , Hinukh (Ginukh), Hunzib , Kajtak , Karata , Khinalugh , Kryts , Kubachi , Khvarshi , Lak , Lezgi (Kuri), Rutul , Tabassaran , Tindi , Tsakhur , Tsez (Dido), and Udi . S. A. Starostin and the late I. M. Diakonoff have claimed connections between this family and the extinct languages Hurrian and Urartian, though this claim is not widely accepted by Caucasologists.
Other languages of the Caucasus area can be placed into families with a much wider geographical distribution:
For a more detailed classification of these languages, see the articles on the corresponding families.
Affinities with other languages
Since the birth of comparative linguistics in the 19th century, the riddle of the apparently isolated Caucasian language families has attracted the attention of many scholars who have strenuously tried to relate them to other languages outside the Caucasus region. While no links with living languages have been found, there seem to be clear connections between the Northwest and Northeast Caucasian families and some extinct languages, formerly spoken in Anatolia and northern Mesopotamia.
Some scholars have seen affinities between the Northwest Caucasian (Circassian) family and the extinct Hattic language. Hattic was spoken in Anatolia (Turkey), in the area around ancient Hattusa (modern Boğazk÷y), until about 1800 BC, when it was replaced by the Indo-European Hittite language.
Hurrian and Urartian
Some scholars — primarily but not exclusively the Russian historical linguists I. M. Diakonoff and Sergei Starostin — also see similarities between the Northeast Caucasian languages 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 centered in the area of Lake Van in Turkey, that existed from 1000 BC or earlier to 585 BC.
Hurrian is primarily attested from Hittite and Mesopotamian cuneiform tablets of the second millenium, while Urartian is mostly known from monumental inscriptions of a much later date.
The two extinct languages have been grouped into the Hurro-Urartian family, and the name Alarodian has been proposed for the union of the Northeast Caucasian and Hurro-Urartian families. Caucasologists in general have been quite skeptical, as many (e.g., Johanna Nichols) point out that too little is understood about the historical phonology and morphology of Northeast Caucasian languages to take historical comparison one step further into history. Also, many of Diakonoff's and Starostin's proposed cognate sets for Proto-Northeast-Caucasian (NEC) and Hurro-Urartian (HU) are semantically weak and often beg the question (e.g., the PNEC for 'reed' and the HU word for 'canal'), and the vast disparity in the overall number of phonemes between NEC (around 80) and HU (around 30) all suggest that any potential relationship would have involved huge changes and great time spans. However, the similarities between the complicated verbal and nominal templatic morphologies of NEC and HU are so striking that the proposal cannot be discounted on its face. One prominent Caucasologist, Wolfgang Schulze, strongly opposes this theory, but his own reconstructed proto-homeland of the PNEC family lies somewhere to the south of the Caucasus, which would put PNEC geographically closer to the historically attested locations of HU communities, thus making the connection slightly more plausible. It is probably too soon to make any categorical claims about this relationship.
Many of the Caucasian languages have case systems (noun inflection rules) of a particular kind, known as ergative, which sets them apart from most European languages. The fact that Basque, an isolated language spoken in the Pyrenees, also has an ergative case system has led many scholars to propose it as a displaced member of some Caucasian family. However, the resemblances between the case systems of Basque and of the Caucasian languages have been found to be rather superficial; most linguists, and almost all Kartvelologists, agree that the underlying structure of Georgian and some unrelated languages such as Batsbi is not ergative, but Split-S or Fluid-S. What is perhaps more damaging to this claim, ergativity arises not infrequently among the world's languages (some estimates are given at around a quarter of all languages, depending on the definition): Iranian languages are known to have descended from a language with a primarily nominative-accusative morphosyntactic alignment, and yet they developed a kind of ergative construction in the past tenses. Thus, the fact that a language has some ergative properties is not sufficient evidence to warrant the claim of some historical linguistic proto-unity.
Proposed higher-level classifications
Another topic that has attracted much research since the 19th century is the classification of the four major Caucasian families into larger groups, and their possible connections with other major linguistic families. Unfortunately this field is quite sensitive given the complex ethnic and political situation of the region, both before and after the extinction of the Soviet Union. As in many other regions of the globe, linguistic arguments are often used to back up or dismiss territorial disputes and separatist movements. Given the general paucity of linguistic and historical evidence for inter-family relationships, those political implications often dominate the debate.
The name Hetto-Iberian (or Proto-Iberian) was proposed by Georgian historian Simon Janashia for a superfamily comprising the South Caucasian languages, other Caucasian language groups, Hattic and other languages of ancient Anatolia. (The Iberian in the name refers to Caucasian Iberia, a kingdom centered in Eastern Georgia which lasted from the 4th century BC to the 5th century AD; it is not related to the Iberian Peninsula.)
Alarodian and North Caucasian
Some linguists see the three northern families as having a common origin, about five thousand years ago, and have proposed a family of North Caucasian languages with the following genetic classification:
The reality of the North Caucasian family is not yet widely accepted; many of the cognates that have been claimed between its branches may actually be loanwords.
Since the early 1980's, some linguists — for instance, Sergei Starostin, Merritt Ruhlen, and John Bengtson — have argued for the existence of a Dene-Caucasian superfamily including the North Caucasian, Sino-Tibetan, Na-Dene languages, and occasionally Basque.
The South Caucasian and North Caucasian families are unrelated phyla even in Greenberg's deep classification of the world's languages. Nevertheless, some scholars — notably Georgian linguist Arnold Chikobava — have proposed the single name Ibero-Caucasian languages (or Iberian-Caucasian languages) for all four Caucasus-specific language families, i.e., the union of the North and South Caucasian families. However, given the lack of significant common traits between these two families, this grouping is still only a convenient geographical term, not yet a linguistic concept.