Summary and Keywords
Languages from at least five genetically unrelated families are spoken in the Caucasus, but there are only three endemic linguistic families belonging to the region: Kartvelian, West Caucasian, and Northeast Caucasian. These families are rather heterogeneous in terms of the number of languages and the distribution of the speakers across them. The Caucasus represents a situation where languages with millions of speakers have coexisted with one-village languages for hundreds of years, and where multilingualism has always been the norm. The richness of Caucasian languages on every linguistic stratum is dazzling: here we find some of the largest consonant inventories, inflectional systems where the mere number of word forms strains credibility (one of the Caucasian languages, Archi, is claimed to have over a million and a half word forms), and challenging syntactic structures. The typological interest of the Caucasian languages and the challenges they present to linguistic theory lie in different areas. Thus, for Kartvelian languages, the number of factors at play in the verbal system make the task of the production of a correct verbal form far from trivial. West Caucasian languages represent an instance of polysynthetic polypersonal verb inflection, which is unusual not only for Caucasus but for Eurasia in general. East Caucasian languages have large systems of non-finite forms which, unusually, retain the ability to realize agreement in gender and number while their non-finite nature is determined by the inability to head an independent clause and to express certain morpho-syntactic categories such as illocutionary force and evidentiality. Finally, all Caucasian languages are ergative to some extent.
1 Indigenous Languages of the Caucasus: Overview
The term “Caucasian languages” usually applies only to languages that belong to one of the three linguistic families indigenous to the Caucasus: Kartvelian (also referred to as South Caucasian), West Caucasian (North West Caucasian, Abkhaz-Adyghe), and Northeast Caucasian (Nakh-Daghestanian). These languages are believed to have been spoken in the area for at least 4,000 years. Despite covering three unrelated linguistic families, the term “Caucasian languages” makes sense: the languages are spoken in a contiguous geographical area, which is characterized by cultural homogeneity and established trading contacts. The Kartvelian languages are spoken predominantly in Georgia, and West and East Caucasian languages are spoken mostly in the Russian Federation. South Caucasian languages are not genetically related to East and West Caucasian; despite that, the Caucasian languages have often been the subject of linguistic comparison.
Kartvelian is comprised of four languages: Georgian, Mingrelian, Laz, and Svan. Sociolinguistically these languages show considerable variation: Georgian has more than 4 million speakers, a very long literary tradition, and a status of a state language of Georgia. Mingrelian is spoken by 500,000 people in Georgia, Laz is spoken in Georgia and Turkey by 150,000 people, and Svan has between 15,000 and 40,000 speakers according to different sources (Tuite, 1997).
West Caucasian consists of five languages: Abkhaz, Abaza, Kabardian, Adyghe, and the recently extinct Ubykh. The largest language of the family is Kabardian, spoken by 516,826 people in the Russian Federation (Russian Census, 2010) and 1,628,500 people in total (Ethnologue, accessed on 18/07/2016). Abkhaz is spoken mainly in Georgia but also in Turkey and Russia. The total population of speakers is 152,740 (Ethnologue). Abaza is spoken by 37,831 people in the Russian Federation (Russian Census, 2010) and about 10,000 in Turkey (Ethnologue). Adyghe is spoken by about 130,000 people in the Russian Federation and 425,000 people in total (Koryakov, 2006, p. 22).
East Caucasian (Northeast Caucasian) is the largest linguistic family in the Caucasus in terms of number of languages: by the most conservative count, it is comprised of 26 languages which are divided into two main branches: Nakh (Chechen, Ingush, and Tsova-Tush) and Daghestanian languages. The number of Daghestanian languages varies in different counts rather dramatically. This is due to the fact that for political reasons some linguistic variations have been claimed to be dialects of the same language whereas linguistically they should be considered different languages. This situation is particularly striking in case of Dargwa, which has recently been claimed to be a group of 18 different languages (Koryakov, 2006), some of which are severely endangered.
Setting aside the complex situation with Dargwa, there are three larger Daghestanian languages: Avar (744,000 speakers), Lezgian (400,000 speakers), and Lak (157,000 speakers). These have long written traditions (in case of Lak starting in the 15th century), official status, and various media publishing and broadcasting in them. Smaller Daghestanian languages are all to some extent endangered. Many of them are spoken in just one village and are either unwritten or achieved literacy within the last 100 years. Northeast Caucasian languages are spoken in the Russian Federation with the exception of Budukh, Kryz, and Khinalugh, which are spoken in Azerbaijan, and Udi (Georgia). Table 1 shows the genealogical branching.
Table 1. East Caucasian (=Northeast Caucasian) Languages
All Caucasian languages are characterized by rich consonantism, rich inflections especially in verb systems, and ergative alignment at least in some parts of the system (Kartvelian languages are particularly complex in this respect). Of the three families, the Kartvelian languages are probably the best studied, though many of the materials are in Georgian.
Kartvelian languages are mostly famous for their complex verbal systems and split ergativity. The verbal system exhibits a two-way split: one concerning the lexicon and another concerning morphosyntactic features. First, one-argument verbs mostly have nominative subjects, but a sizable subclass has ergative subjects. Second, the argument structure depends on the verbal tense: the present series has nominative-accusative alignment, the aorist series has ergative-absolutive alignment, and the perfective verbs take dative-nominative alignment where the agent is in the dative case.
West Caucasian languages are typologically unusual in that they have very rich consonantal systems and minimal vowel inventory. Another typologically striking feature of West Caucasian is the polysynthetic structure of the verbal complex. Other Caucasian languages do not show this feature (Hewitt, 2005), and it is extremely rare in the languages of Eurasia in general; compare the list in Fortescue (2016), where West Caucasian languages are the only polysynthetic languages in Central Eurasia . Another feature that sets West Caucasian apart from other Caucasian languages is the morphological marking of definiteness in the form of articles on nouns.
Languages in the Northeast Caucasian family are famous for very large paradigms in nominal as well as verbal domains and a plethora of non-finite forms. One typologically interesting feature of NEC languages is the possibility for non-finite as well as non-verbal forms to agree: in some NEC languages adverbs, particles, infinitives, verbal nouns, and some case forms of personal pronouns agree in gender with the absolutive argument.
In this article I will outline the main characteristics of Caucasian languages in phonology, morphology, and syntax, emphasizing those that are typologically unusual and linguistically interesting.
Caucasian languages are famously rich in consonants and have relatively simple vowel systems. The richness of the consonantal systems is achieved by two means: first, the familiar place/mode of articulation grid is densely populated. In addition to this, several types of secondary articulation bring the number of consonants to the maximum of 80 (as is the case with Ubykh and Archi). Table 2 shows one of the richest systems among Northeast Caucasian languages, that of Archi (Lezgian group).
Table 2. Consonants in Archi
Using Archi as an example, we can see all the places and mode of articulation employed by the languages of Northeast Caucasian branch: from labial to laryngeal and from plosive to approximant. The lateral and palato-velar consonants are present in most, though not all, NEC languages, while other consonants are common for all language groups of the branch.
Several cells in Table 2 contain more than one consonant. These consonants are contrasted by several different features. First there is a familiar division into voiced and voiceless. This is relevant for almost all NEC languages. However, in Andic languages the voiced consonants are contrasted with aspirates. Next is the contrast of lenis and fortis consonants (notated Cː in the table). This contrast is relevant for almost all NEC languages and is realized as greater articulatory energy applied in the production of a segment. The next contrast, also present in most NEC languages, is the ejective–non-ejective consonants. Ejective consonants (notated C’ in the table) are pronounced with a glottalic egressive airstream. Only voiceless consonants can be ejective, and, with the exception of one language, only plosives and affricates can have this type of secondary articulation. Bagwalal is the only language which has this distinction in spirants: there is a contrast between non-ejective [s] and ejective [s’]. Finally, the contrast in labialized—non-labialized consonants (notated Cw) is characteristic of the Lezgian branch.
Table 2 does not show another secondary articulation present in many NEC language, that of pharyngealization (notated Cˤ). It can be present in consonants or vowels. During the articulation of the pharyngealized sound the epiglottis is constricted. Acoustically, it is characterized by a lower frequency of the third and a higher frequency of the first formant (Ladefoged & Maddieson, 1996, p. 307). In Archi it is phonologically distinctive on stressed vowels and uvular consonants but has an effect on all vowels in the word. Pharyngealization is present in other Lezgian languages (though not Lezgi itself), and also in Lak.
As for vowel systems, they tend to be relatively simple. Below the simplest and the most complex vowel systems in NEC languages are shown—the former is represented by Avar and the latter by Hunzib:
The richness of the Hunzib vowel inventory is achieved by having contrast in length and nasalization throughout the vowel system. The phonological contrast of long and short vowels is also found in Lezgian languages, and nasal vowels are found in Andic languages; to complete the list of Northeast Caucasian vowels we only need to mention a front low vowel [ä] in Dargwa.
As mentioned before, the vowel inventory can be augmented by the presence of pharyngealized vowels—they are found in Archi, Lak, Tabassaran, Tsakhur, and Tsez.
The West Caucasian family is also famous for abundant consonantal systems; Table 3 shows the consonantal system of Adyghe as an example.
Table 3. Consonants in Adyghe
Like Northeast Caucasian languages, West Caucasian languages have lateral, uvular, laryngeal, fortis, ejective, and labialized consonants. However, the richness of these systems is largely due to the variety of affricates that spread over three places of articulation and exhibit two types of additional articulation: ejectiveness and labialization. In addition to what is shown for Adyghe, Kabardian has the labiodental ejective fricative [s’]. It is, however, the West Caucasian vowels that are more important for phonological theory. For example, the vowel system of Kabardian took linguists decades of debate to finally establish that West Caucasian vowel inventories consist of a minimum of two phonemes (see Colarusso, 1992).
Compared to NEC and West Caucasian languages, Kartvelian languages do not have such rich consonantal inventories. In Georgian, there are bilabial, labio-dental, pre-alveolar, alveolar, velar, and uvular voiced and voiceless stops and spirants, one pharyngeal voiceless stop, and one laryngeal voiceless spirant. Voiceless stops also have ejective counterparts.
Morphologically, Caucasian languages show substantial differences between families but languages within a family have similar features. Generally, it is possible to say that Kartvelian and West Caucasian languages have relatively simple nominal systems and very complex and inflectionally rich verbal systems, whereas Northeast Caucasian languages are inflectionally rich and morphologically complex both in the nominal and verbal domains.
In each of the two following sections on nominal and verbal systems, I will start with morphological characteristics of the Northeast Caucasian languages and then turn to Kartvelian and West Caucasian languages.
3.1 Nominal Systems
Nouns in NEC languages have lexical gender, inflect for case and number, and also have complex systems of locative forms. There are three NEC languages and one dialect—all in the Lezgian branch—which have lost the gender feature: Lezgian, Aghul, Udi, and the Khiv dialect of Tabassaran; all other NEC languages have genders. The average NEC language has three (Avar and most Andic languages) or four (Lezgian branch) genders, plus a handful of words with some deviations in terms of gender agreement. Usually it means that a word uses one gender value in the singular and a different value in the plural: thus, a word for ‘child’ in Khvarshi takes gender 3 in the singular and gender I/II in the plural (based on Corbett, 1991, p. 170). It has been claimed that Batsbi (Tsova-Tush) has 8 genders, but see Corbett (1991, pp. 170–172) for discussion. In all gender systems masculine and feminine genders include only nouns referring to human males and human females, respectively, whereas the distribution of other nouns into different remaining genders is never fully semantic. The gender of the noun can only be seen in agreement. One recurrent feature of Northeast Caucasian languages is that in the plural there is one agreement form for the first and second genders (i.e., for the nouns denoting humans) and another form for the rest; hence the term ‘human plural from’ employed in the works on NEC languages. Example (1) shows gender forms in Tsakhur (Lezgian, 4 genders), example (2) shows those in Bagwalal (Andic, 3 genders):
The number feature is relatively simple—it can be singular or plural. One interesting characteristic of Daghestanian languages is the abundant usage of the standard (additive) plural in the function of associative plural (Daniel & Moravcsik, 2013). For example, in Archi, the standard plural suffix –tːu/–du can attach to first names and means ‘X and his family’: Pat’ima-tːu ‘Patimat and her children,’ Juq’ub-du ‘Jaqub, his wife and children.’ With common nouns this suffix denotes additive plural, thus parčaħ-tːu (king-pl) can only mean ‘kings’ but not *‘king and his family.’ Nakh, West Caucasian, and Kartvelian languages realize the associative plural by special morphological means distinct from additive plural realization. Thus, the Chechen standard plural suffix –š/-aš cannot attach to proper names, but there a special associative plural suffix -ɣar: Zaːra-ɣar ‘Zara and her relatives,’ but the form with standard plural suffix is ungrammatical: *Zaːra-š (Zarina Molochieva, personal communication).
The system of grammatical cases (all cases excluding the locative forms) in Northeast Caucasian can be as large as 17. The core cases that are typically found in a NEC language are absolutive, ergative, genitive, dative, and comitative. Some languages (such as Andic) have a special case to code the experiencer role, termed the affective case. Lak is the only Northeast Caucasian language where the roles normally coded by the ergative and the genitive are coded by one and the same case.
There is a common morphological pattern for case realization: the case forms are produced by adding the endings to the noun stems. A noun in Northeast Caucasian languages has a number of stems, typically four. Table 4 shows nominal stems in Archi which instantiate a typical Daghestanian paradigm.
Table 4 Noun Stems in Archi
In Archi, the direct stems of a noun always coincide with their absolutive case form, and for nearly all nouns there is a one-to-one correspondence between its oblique stems and the forms used for the ergative case. The oblique stem is also used as the base to which other case markings are added, including the spatial cases. As the Archi example shows, the stems show a high degree of morphological irregularity, whereas the case endings are invariant and do not change in form according to the number value or morphological regularity of the stem. Table 5 shows Archi non-spatial case forms.
Table 5. Archi Non-Spatial Cases
In terms of morphosyntactic functions, the core grammatical cases are used to code clause arguments, while others (such as comparative, substitutive, etc.) are used to code adjuncts. All NEC languages manifest (morphological) ergativity, i.e., the agent argument of the transitive verb takes the ergative case, and the patient/object of a transitive and the only argument of an intransitive verb take the absolutive case. The verb normally agrees with the absolutive argument in gender and number, but see below on languages with person agreement. The ergative can also be used to code the instrument, so it is possible to have two ergative cases in a sentence; compare ʕalimu ‘Ali.erg’ and nak’ǝnni ‘key.erg’ in the following sentence:
Verbs denoting cognitive functions and emotions code the experiencer in the dative (unless there is an affective case in the system) and the stimulus in the absolutive. The verb then agrees with the absolutive; compare the two sentences below:
Most Northeast Caucasian languages have a genitive case which is used to code the adnominal modifier or the possessor. Deviations from this are represented by Tsakhur (Lezgian), where the genitive is absent from the case system (or at least not easily defined), and Lak, where the same case is used for adnominal modification, possession, and the agent of the transitive verbs, so some descriptions use the term Ergative-Genitive. Bagwalal (Andic) has an interesting genitive realization which depends on the gender of the noun. If the noun in the genitive belongs to the second or third gender, the genitive ending is -ɬ as in sajuz-i-ɬ ‘of the Union’ in example (5). If the noun belongs to the first gender, the genitive ending is the gender-number marker agreeing with the head noun. Table 6 summarizes this, adding the plural values.
Table 6. Genitive Endings in Bagwalal
Example (6a–c) illustrates genitive formation for a singular first-gender noun; (6d) shows a noun in the plural:
In addition to the nominal cases discussed above, Northeast Caucasian languages express many spatial meanings by means of bound morphemes. These forms are organized into subparadigms, which are clearly different from the grammatical case paradigms. A spatial form normally includes two separately coded categories, localization and orientation, where the former defines space in relation to the landmark and the latter denotes the movement of the object. The spatial forms are morphologically built on the oblique stems, just like the grammatical cases. The examples below are from Archi:
Archi represents an ‘average’ Northeast Caucasian system with five localizations—IN, ‘in a hollow space’; INTER, ‘in a filled space’; SUPER, ‘on’; sub, ‘under’; and CONT, ‘in close contact with, next to’—and six orientation cases—essive, ‘be at/in the LOC’; elative, ‘from the LOC’; lative, ‘to the LOC’; allative, ‘toward the LOC’; terminative, ‘to the LOC and no further’; and translative, ‘through the LOC.’ Note that the essive case (7b) is realized by zero marking. This is again typical for NEC languages, but Dargwa represents an interesting deviation from this: not only is the essive case in Dargwa more marked than the lative, the marking itself is unusual in that the essive is realized by gender-number marking:
There are languages that have more localizations in their systems: thus Aghul (Lezgian) has 8 localizations, while Bezhta and Tsez (Tsezic) have 9 (Daniel & Ganenkov, 2011). Tsez adds a distal distinction to the picture, having a 30-cell spatial paradigm as a result (Polinsky, personal communication).
Spatial forms in NEC languages make a paradigm that is morphologically distinct from the paradigm of grammatical cases, but their morpho-syntactic usage overlaps that of non-spatial cases. Thus, although spatial cases never control agreement, they can code core arguments: for example, cont-allative is a typical case for the addressee of a speech, sub-elative often codes the source (or cause) argument of the verb ‘to be afraid,’ and so on.
In contrast to these rich systems, the West Caucasian and Kartvelian languages do not have the category of noun gender, and the case feature values vary from four in Kabardian and Adyghe (ergative, absolutive, instrumental, and adverbial) to seven in Georgian: nominative, vocative, dative, ergative, genitive, instrumental, and adverbial.
3.2 Verbal Systems
The verbal systems in the three Caucasian linguistic families are inflectionally rich and complex. Compared to nominal systems, there is greater variation between verbal systems in different families. Generalizing roughly, the Northeast Caucasian verbal systems are mostly interesting for their aspectual meanings and non-finite forms, while the Kartvelian system displays very interesting personal agreement and a rich array of temporal and evidential categories. Kartvelian verbs are also well known for the tense-based split ergativity. West Caucasian verbal systems are famous for the amount of morpho-syntactic information packed in a verb form. Below are some highlights from each Caucasian family.
3.2.1 Verbal Systems in Northeast Caucasian
A typical Northeast Caucasian verb has at least two aspectual stems (perfective and imperfective), agrees with the absolutive of the clause, and has four types of non-finite forms: converbs, participles, verbal nouns (traditionally called masdars), and infinitives. The category of tense is very often expressed by periphrastic forms (consisting of a converb and an auxiliary, contrary to the familiar Indo-European model with participles). Evidentiality is rarely expressed by a bond morpheme; typologically familiar evidential usage of a (periphrastic) perfect is more common.
The Lezgian language Archi represents a typical NEC verbal system. An Archi verb has three aspectual stems and a form of the imperative as its principal parts: each of these four forms has to be learned, but the rest of the verb paradigm can be deduced from them. Agreement is realized in the stem, too, as shown in example (9). Only forms for gender III and gender IV in the singular are shown (gender IV is realized by zero) (see Chumakina & Corbett, 2015, for a detailed discussion of the Archi verb inflection).
These forms have two functions: they can be used independently, as separate words (main or dependent predicates), and they can serve as a morphological base for other verb forms such as converbs or participles. The infinitive illustrates one of the important features of NEC verbs: together with other non-finite forms, the infinitive retains the possibility to agree, so the non-finite forms are defined as such only on the base of the fact that they cannot head an independent clause.
Infinitives normally head purpose clauses and complements of matrix verbs such as ‘want,’ ‘begin,’ and ‘finish.’ Participles head relative clauses, masdars head complements of the matrix verbs such as ‘know’ and ‘say,’ and converbs head all kinds of clausal adjuncts. Simple converbs express secondary action, whereby the perfective converb denotes an action completed before the action of the main clause and the imperfective converb denotes an action simultaneous with the main clause action. Converbs with bound morphemes attached to them (complex converbs) express various temporal, conditional, concessive, and causal meanings. Examples from Archi below illustrate each type of non-finite forms.
In most NEC languages verbs agree in gender and number only. However, in Dargwa, Lak, and Tabassaran an innovative system of personal agreement has emerged. Example (15) is from DargwaDargwa, where the verb agrees in person by prefix d-:
3.2.2 Verbal Systems in West Caucasian
The verb in West Caucasian is polypersonal and in fact shows a polysynthetic type of inflection and can have more than 20 morphological slots. Below is the verb template for Adyghe:
The first position is reserved for the absolutive agreement marker. The next slot, dir, stands for directional prefix, which indicates movement toward the speaker. pv+obj denotes a complex of morphemes consisting an indirect object prefix and a preverb indicating its role (beneficiary, comitative, etc). There can be more than one such complex in a word. The indirect object (oblique) prefix following this complex reflects the lexically specified valence of the root. A prefix agreeing in person with the agent (the ergative) follows it. Next come prefixes of the optative mood, negation, and causative. The first suffixal position (incorp) contains incorporated elements, mostly of an adverbial nature. The suffixes following it are called ‘propositional elements’ (Arkadiev, Lander, Letuchiy, Sumbatova, & Testelets, 2009, p. 46) and include mostly lexical aspect (aktionsart) suffixes such as habilitive, reversive, and simulative, and suffixes of hypothetical mood. The plural suffix following it agrees with the absolutive. The dyn suffix is used with present tense dynamic verbs in dependent clauses. The final position in reserved for various illocutory and pragmatic markers such as markers of negation, question, and emphasis. Below are some examples of Adyghe verbs:
3.2.3 Verbal Systems in Kartvelian
I will exemplify the structure of Kartvelian verb system using Georgian. Georgian has tense/aspect-based split ergativity with three alignment-determining sets of tenses: present, aorist, and prefect series. Each series is associated with a specific set of affixes that apply to the verb and with a specific case encoding of the verbal arguments (Hewitt, 1995, p. 218).
In the structure of a verb, 11 morphological slots are recognized, as in (18). Examples of a Georgian verb form are given in (19) and (20):
The coding of verbal arguments in Georgian is determined by tense. Example (21) illustrates three coding possibilities available: tenses from the present series require nominative-accusative alignment (21a), the aorist series require ergative-nominative alignment (21b), and the perfect series require dative-nominative alignment. Notation SUBJ and OBJ in the glossing indicates the agreement controller.
Note that regardless of tense, verb agreement distinguishes subject and object by using two sets traditionally referred to as Set A and Set B, the hyphen showing their position relative to the stem (Table 7).
Georgian verb inflection combines morphological, syntactic, and semantic information in a non-trivial way: to result in a correct inflected form, one needs to know tense, aspect and mood values, and argument structure and then apply the endings from the right set. Table 7 shows some examples.
Table 7. Some Forms of the Georgian Verbs ašenebs ‘build’ and c’ers ‘write’
In this section I can only highlight some characteristic and typologically interesting syntactic properties of the Caucasian languages. The basic word order in all three Caucasian families is SOV, although variation is possible for certain pragmatic purposes or a specific information structure. In the noun phrase, the modifier typically precedes the head noun, although very heavy relative clauses can occasionally be postponed. Typical for SOV is the use of postpositions rather than prepositions. Probably the most salient syntactic properties of Caucasian languages are ergativity and (pervasive) agreement.
All Caucasian languages display ergativity to some extent. NEC languages are most prominent in this respect: all NEC languages have ergative alignment of the arguments where the subject of the intransitive verb and the object of the transitive verb take the absolutive case and the subject of the transitive verb takes the ergative (except for erbs of emotion and cognition, mentioned earlier, which take DAT-ABS alignment). Most NEC languages have the ergative agreement, where the agreement in the clause is controlled by the absolutive case. Only the languages that have agreement in person (Dargwa, Lak, and Tabassaran) represent in some parts of the system the situation where the subject of the transitive verb (the ergative) can control the agreement.
However, the ergativity found in the NEC languages is only surface (morphological) ergativity; on a deeper syntactic level, the ergative argument displays some subject properties. For example, the ergative argument c-commands the absolutive, which can be seen in the behavior of the reflexive. The following examples are from Archi:
NEC languages are known for agreement, which has been termed in the literature as pervasive, exuberant, and extreme. This means that the agreement is expressed not just by familiar targets, such as modifier and verb, but by all parts of speech, and agreement markers can appear more than once in the same word. The following sentence from Archi illustrates this:
The attributive modifier doːˁzub agrees in gender and number with the noun it modifies, χˁon ‘cow.’ In the clausal domain, the lexical verb abu ‘make’ agrees in gender and number with the head of the absolutive object of the clause: doːˁzub χˁon ‘big cow.’ The ergative subject nenabu ‘we’ (incl), dative indirect object belabu ‘us’ (incl), and the adverb ditːabu ‘quickly’ also agree in gender and number with the absolutive object.
Agreeing pronouns make Archi the most extreme case of pervasive agreement; however, agreeing adverbs and postpositions are found in other NEC languages as well:
In Dargwa, the postposition sala-w governs the genitive but agrees with the absolutive:
Boeder, W. (2005). The South Caucasian languages. Lingua,115, 5–89.Find this resource:
Chirikba, V. A. (1996). Common West Caucasian: The reconstruction of its phonological system and parts of its lexicon and morphology. Leiden, The Netherlands: Research School CNWS.Find this resource:
Colarusso, J. (1988). The Northwest Caucasian languages: A phonological survey. New York: Garland.Find this resource:
Comrie, B. (2008). Linguistic diversity in the Caucasus. Annual Review of Anthropology, 37, 131–143.Find this resource:
Forker, D. (2012). The bi-absolutive construction in Nakh-Daghestanian. Folia Linguistica, 46, 75–108.Find this resource:
Harris, A. C. (1985). Diachronic syntax: The Kartvelian case. Orlando, FL: Academic Press.Find this resource:
Harris, A. C. (1990). Georgian: A language with active case marking. Lingua,80, 47–65.Find this resource:
Helmbrecht, J. (1996). The syntax of personal agreement in East Caucasian languages. Sprachtypologie und Universalienforschung,49(2), 127–148.Find this resource:
Hewitt, G. (1995). Georgian. A structural reference grammar. London: John Benjamins.Find this resource:
Kalinina, E., & Sumbatova, N. (2007). Clause structure and verbal forms in Nakh-Daghestanian languages. In I. Nikolaeva (Ed.), Finiteness (pp. 183–249). Oxford: Oxford University Press.Find this resource:
Kibrik, A. (1995). Direct-oblique agreement of attributes in Daghestanian. In F. Plank (Ed.), Double case: Agreement by Suffixaufnahme (pp. 216–229). New York: Oxford University Press.Find this resource:
Lacroix, R. (2009). Description du dialecte laze d’Arhavi (caucasique du sud, Turquie): Grammaire et textes (Unpublished PhD diss.). Universite Lumiere, Lyon .Find this resource:
Lander, Y., & Letuchiy, A. (2010). Kinds of recursion in Adyghe morphology. In H. van der Hulst (Ed.), Recursion and human language (pp. 263–284). Berlin: Mouton de Gruyter.Find this resource:
Lyutikova, E. A. (2000). Reflexives and emphasis in Tsaxur (Nakh-Dagestanian). In Z. Frajzyngier & T. S. Curl (Eds.), Reflexives: Forms and functions (pp. 227–255). Amsterdam: John Benjamins.Find this resource:
Smeets, R. (Ed.). (1994). The indigenous languages of the Caucasus. Vol. 1: The Kartvelian languages. Vol. 2: The North West Caucasian languages. Vols. 3 and 4: The North East Caucasian languages. Delmar, NY: Caravan Books.Find this resource:
Testelets, Y. (1998). Word order in Daghestanian languages. In A. Siewerska (Ed.), Constituent order in the languages of Europe (pp. 257–280). Berlin: Mouton de Gruyter.Find this resource:
Arkadiev, P., Lander, Y., Letuchiy, A., Sumbatova, N., & Testelets, Y. (2009). Vvedenie. Osnovnye svedenija ob adygejskom jazyke. In Y. Teslelets (Ed.), Aspekty polisintetizma: očerki po grammatike adygejsko jazyka. Moscow: Russian State University for the Humanities.Find this resource:
Chumakina, M., & Corbett, G. G. (2015). Gender-number marking in Archi: Small is complex. In M. Baerman, G. G. Corbett, & D. Brown (Eds.), Understanding and measuring morphological complexity (pp. 93–116). Oxford: Oxford University Press.Find this resource:
Colarusso, J. (1992). A grammar of the Kabardian language. Calgary, Canada: University of Calgary Press.Find this resource:
Corbett, G. G. (1991. Gender. Cambridge, U.K.: Cambridge University Press.Find this resource:
Daniel, M., & Ganenkov, D. (2011). Case marking in Daghestanian: Limits of elaboration. In A. Malchukov & A. Spencer (Eds.), The Oxford handbook of case (pp. 668–685). Oxford: Oxford University Press.Find this resource:
Daniel, M., & Moravcsik, E. (2013). The associative plural. In M. S. Dryer & M. Haspelmath (Eds.), The world atlas of language structures online. Leipzig: Max Planck Institute for Evolutionary Anthropology.Find this resource:
Fortescue, M. (Ed.). (2016). The Oxford handbook of polysynthesis. Oxford: Oxford University Press, 2016.Find this resource:
Harris, A. C. (1981). Georgian syntax. A study in relational grammar. Cambridge, U.K.: Cambridge University Press.Find this resource:
Hewitt, G. (2005). North West Caucasian. Lingua,115, 91–145.Find this resource:
Kibrik, A. E., Kodzasov, S. V., Olovjannikova, I. P., & Samedov, D. S. (1977a). Opyt strukturnogo opisanija arčinskogo jazyka, I: Leksika. Fonetika. (Publikacii otdelenija strukturnoj i prikladnoj lingvistiki, 11). Moscow: Izdatel’stvo Moskovskogo universiteta.Find this resource:
Kibrik, A. E. (1977b). Opyt strukturnogo opisanija arčinskogo jazyka, II: Taksonomičeskaja grammatika. (Publikacii otdelenija strukturnoj i prikladnoj lingvistiki, 12). Moscow: Izdatel’stvo Moskovskogo universiteta.Find this resource:
Kibrik, A. E. (1977c). Opyt strukturnogo opisanija arčinskogo jazyka, III: Dinamičeskaja grammatika. (Publikacii otdelenija strukturnoj i prikladnoj lingvistiki, 13). Moscow: Izdatel’stvo Moskovskogo universiteta.Find this resource:
Koryakov, Y. (2006). Atlas kavkazskix jazykov. Moscow: Pilgrim.Find this resource:
Ladefoged, P., & Maddieson, I. (1996). The sounds of the world’s languages. Oxford: Blackwell.Find this resource:
Russian Census 2010, official site. (2010). Retrieved from http://www.gks.ru/free_doc/new_site/perepis2010/croc/perepis_itogi1612.htm.
Samsel, J. (1992). A role and reference grammar analysis of Georgian morphosyntax (Unpublished master’s thesis). University of California Davis.Find this resource:
Sosenskaja, T. (1999). Poslelogi. In A. E. Kibrik (Ed.), Èlementy caxurskogo jazyka v tipologičeskom osveščenii. Moscow: Nasledie, 1999.Find this resource:
Sumbatova, N. (2009). Constituent questions and argument-focus constructions: Some data from the North-Caucasian languages. In J. Helmbrecht, Y. Nishina, Y.-M. Shin, S. Skopeteas, & E. Verhoeven (Eds.), Form and function in language research: Papers in honour of Christian Lehmann (pp. 313–328). Berlin, New York: Mouton de Gruyter.Find this resource:
Tuite, Kevin. (1997). Svan. Munich: Lincom.Find this resource: