7Jezikoslovni zapiski 28  2022  1
1
peter arkadiev
moRphology oF the CauCasian languages:  
a typologiCal oveRview
Cobiss: 1.01
Morfologija kavkaških jezikov: tipološki pregled
Jeziki avtohtonih jezikovnih družin na Kavkazu (naško-dagestanske, abhaško-adigejske 
in kartvelske družine) se odlikujejo po raznolikosti svojih oblikoslovnih sistemov. Prav na 
oblikoslovni ravnini so si ti jeziki tudi med seboj različni – medsebojno se razlikujejo tako 
jezikovne družine kot predstavniki posameznih družin. Prispevek predstavlja teoretični 
in tipološki pregled osnovnih posebnostih oblikoslovnih sistemov kavkaških jezikov, pri 
čemer izpostavlja netipične in s tipološkega stališča redke pojave v njihovi morfologi-
ji. Osredotoča se na posebnosti izražanja skladenjskih razmerij med jedrom in odvisnim 
delom ter na različne tipe razmerij med morfološkimi pomeni in sredstvi njihovega izra-
žanja; prav tako obravnava primere netipičnih razmerij med obliko in pomenom ter tipe 
polisintetizma, ki so značilni za kavkaške jezike.
Ključne besede: kavkaški jeziki, tipologija, oblikoslovje, označevanje morfoloških po-
menov, razmerja med obliko in pomenom, polisintetizem 
The indigenous languages of the Caucasus (Nakh-Daghestanian, Abkhazo-Agyghean and 
Kartvelian) present a remarkable degree of diversity in their morphological systems, both 
between and within larger genealogical units and even closely-related varieties. This ar-
ticle is an attempt to present a theoretically and typologically informed overview of the 
major parameters of morphological variation of the Caucasian languages and to highlight 
a number of non-trivial and rare phenomena found in their morphology. The phenomena 
discussed include locus of marking, types of affixal and non-affixal exponence, non-triv-
ial form-meaning relations, and manifestations of polysynthesis.
Keywords: Caucasian languages, typology, morphology, exponence, form-meaning re-
lations, polysynthesis
intRoduCtion
The Caucasus1 is commonly recognized as the most linguistically diverse region of 
Western Eurasia, with about fifty languages belonging to several language families 
1 This article is based on lectures given at the Workshop on Theoretical Approaches to the Lan-
guages of the Caucasus at the CreteLing Summer School (Rethymno, July 2019), at the Linguis-
tics Colloquium of Johannes-Gutenberg-Universität Mainz (November 2019) and at Linguistic 
Circle of the Faculty of Arts of the University of Ljubljana (May 2021). I thank the organisers 
and audiences of all these events, in particular, Maria Polinsky, Walter Bisang and Mladen 
Uhlik, for their support, as well as Nino Amiridze, Gilles Authier, B. Blasebalg, Arnaud Four-
net, Marc Gandarillas, Martin Haspelmath, Yury Lander, Timur Maisak, Juho Pystynen, Ilja 
Seržant, Nina Sumbatova, Yakov Testelets, Dmitry Zelensky and two anonymous reviewers for 
their useful feedback. I also thank Yakov Testelets and Nina Sumbatova for introducing me to 
the languages of the Caucasus, and the native speakers of the Abkhaz-Adyghean languages for 
their generous help. All faults and shortcomings are mine.
https://doi.org/10.3986/Jz.28.1.01
8 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
9Jezikoslovni zapiski 28  2022  1
10 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
11Jezikoslovni zapiski 28  2022  1
12 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
populating a relatively compact territory (for a recent comprehensive overview of the 
languages of the Caucasus and their structural features, see Polinsky 2020). Three of 
the language families spoken in the Caucasus are autochthonous to the area; these 
are the Northeast Caucasian or Nakh-Daghestanian (NEC), consisting of the Nakh, 
Avar-Andic, Tsezic, Lezgic and Dargic branches and isolates Lak and Khinalug; the 
Northwest Caucasian or Abkhaz-Adyghean (NWC), consisting of the Abkhaz-Aba-
za and Circassian branches and the extinct isolate Ubykh; and the South Caucasian 
or Kartvelian (SC), comprising Georgian, Svan, Mingrelian and Laz. Other language 
families of the area are Indo-European (Armenian, Ossetic, and Tat) and Turkic 
(Azeri, Karachay-Balkar, Nogay and Kumyk). The genealogical relationships be-
tween the three indigenous families have been a matter of debate. The so-called 
Ibero-Caucasian theory advanced primarily by the Georgian linguists of the Soviet 
era and assuming relatedness between South and North Caucasian languages has not 
been substantiated by historical-comparative evidence and did not gain support out-
side of Georgia (see Tuite 2008). By contrast, historical-comparative reconstruction 
linking NWC and NEC has been proposed (see Trubetzkoy 1926; Nikolaev – Sta-
rostin 1994; Chirikba 1999; 2016) and is upheld by many specialists. On the rather 
controversial hypotheses concerning potential distant relationships of the indigenous 
families of the Caucasus, see Comrie 2008: 134–135.
In areal terms, the Caucasus is a paradigm example of a “residual” or “accretion” 
zone (Nichols 1992: 13–16; 1997; Comrie 2008), where diversity increases over time 
due to both internal diversification of the existing linguistic varieties and the arrival of 
new languages from outside. Another feature of accretion zones relevant for the Cau-
casus is a lack of a single dominant language or language family and of a single lingua 
franca (this was true at least up to the arrival of Russian in the 19th and 20th centuries), 
and instead the existence of a complex network of local patterns of multilingualism 
and language contact (Chirikba 2008: 30–32; Dobrushina et al. 2020: 47–65). As a 
result, the question whether the Caucasus can be regarded as a Sprachbund in the 
classic sense of the word remains unsettled, since there are not many typologically 
non-trivial linguistic features common to all languages of the area and demonstrably 
originating through language contact (see e.g. Tuite 1999 vs. Chirikba 2008).
The languages of the Caucasus exhibit a remarkable degree of structural diver-
sity and feature a number of rare phenomena, as seen both against the background 
of the European languages as well as the languages of Western Eurasia and the Near 
East, and also from a world-wide typological perspective (cf. Comrie 2008). The 
domain where internal diversity and typological uniqueness of the languages of the 
Caucasus manifest themselves most clearly is the morphology. This article attempts 
to present a concise overview of some of the interesting morphological phenomena 
found in the languages of the Caucasus from a typological standpoint. By necessity, 
I limit myself to the three indigenous families, only cursorily mentioning the In-
do-European and Turkic languages of the region. The selection of phenomena to be 
discussed is inevitably rather subjective. The structure of the remainder of the paper 
13Jezikoslovni zapiski 28  2022  1
is as follows. In section 2 I briefly present the extent of variation of the Caucasian 
morphologies across a number of parameters current in typological literature. In 
section 3 I discuss non-trivial morphological exponents found in the Caucasian lan-
guages, while section 4 is devoted to non-trivial form-meaning relations. In section 
5 I focus on the Northwest Caucasian family with its polysynthetic morphology, for 
which I have firsthand fieldwork experience. Section 6 concludes.
majoR paRameteRs oF moRphologiCal vaRiation
All languages of the Caucasus possess rich and complex morphological systems, 
which show considerable variation both between and within language families. 
Below I give an overview of some of the more general typological parameters of 
this variation (i.e. dominance of suffixation vs. prefixation, head vs. dependent 
marking, agglutination vs. fusion), before turning to more intricate phenomena in 
the next sections.
In terms of suffixation vs. prefixation (cf. Dryer 2013), one finds predom-
inantly suffixing languages (e.g. Avar, Lezgian, Ossetic, Turkic) as well as lan-
guages with moderate (Agul, Tsez) and highly developed (Kartvelian, Northwest 
Caucasian) prefixation. Consider a Lezgian verbal form with a chain of suffixes in 
1, a Kabardian verbal form with a chain of prefixes in 2,2 and a Laz verbal form 
with both prefixes and suffixes in 3.3
Lezgian (NEC > Lezgic, Russia, Azerbaijan; Haspelmath 1993: 397)
[1] aku-nwa-č-ir-t’a=ni
see-prf-neg-pst-cond=even
‘although he hadn’t seen (him) yet’
Besleney Kabardian (NWC > Circassian, Russia; own fieldwork, textual example4)
[2] šjə-z-e-r-e-ʁa-ŝ’e
pvb-rfl.io-dat-3sg.erg-dyn-caus-know
‘he learns (lit. causes himself to know) it there’
Laz (SC > Zan, Turkey; Lacroix 2009: 192; transcription and glosses adapted)
[3] ko-mo-b-o-č’apx-i-dort’un
afrm-pvb-1sg.sbj-val-apply-aor-plsq
‘I had applied it there’
In terms of dependent vs. head marking typology (Nichols 1986; Nichols – Bickel 
2013; Lander – Nichols 2020), the languages of the Caucasus also differ widely. 
There are both almost exclusively dependent-marking languages (most of NEC, 
2 Of course, this example does not imply that Kabardian lacks suffixes altogether – only that their 
presence is not obligatory.
3 Unless stated explicitly, all examples are given in the transcription/transliteration of the source. 
I have unified the marking of ejective consonants, which is always indicated by an apostrophe.
4 Texts in Besleney Kabardian include oral narratives and dialogues collected in the village Ulyap 
(Republic of Adygheya) in 2011–2012.
2
14 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
Ossetic, Armenian and Turkic) and almost exclusively head-marking languages 
(Abkhaz and Abaza), as well as double-marking languages (Kartvelian, Circas-
sian and some of the East Caucasian, e.g. Tabassaran). Consider these marking 
strategies in noun phrases and clauses for Ingush (dependent-marking) in 4, Abaza 
(head-marking) in 5 and West Circassian (double-marking) in 6.
Ingush (NEC > Nakh, Russia)
[4] a.  desh-a  mawan
 word-gen  meaning
 ‘the word’s meaning’ (Nichols 2011: 417)
b. Muusaa-z	 	 zhwalie-na			 ghadzh	 tiex-ar.
 Musa-erg  dog-dat  stick(abs) strike-witn.pst
 ‘Musa hit the dog with a stick.’ (Nichols 2011: 467)
Abaza (NWC > Abkhaz-Abaza, Russia; own fieldwork, elicited5)
[5] a. s-án    l-ašjá
  1sg.pr-mother  3sg.f.pr-brother
  ‘my mother’s brother’
 b. a-phwə́spa	 	 ĉ’a	 	 j-l-s-t-t’
  def-girl  apple  3sg.n.abs-3sg.f.io-1sg.erg-give(aor)-dcl
  ‘I gave an apple to the girl.’
West Circassian (NWC > Circassian, Russia; own fieldwork, elicited6)
[6] a. pŝaŝe-m ǝ-ʔe
  girl-obl 3sg.pr-hand
  ‘the girl’s hand’
 b. čj’ale-xe-m				pŝaŝe-xe-m	 qeʁaʁe-xe-r   a-r-a-tə-ʁe-x
  boy-pl-obl   girl-pl-obl flower-pl-abs   3pl.io-dat-3pl.erg-give-pst-pl.abs
  ‘The boys gave flowers to the girls.’
Variation in terms of dependent vs. head vs. double marking in the languages of 
the Caucasus extends beyond the noun phrase and clausal core into the domain of 
spatial meanings, whose grammatical expression is one of the prominent features 
of the region. Thus, while most NEC languages have rich multidimensional sys-
tems of nominal locative marking (see e.g. Kibrik 2003; Ganenkov 2010), see ex-
ample 7 from Hinuq, NWC languages, by contrast, feature comparably elaborate 
systems of verbal spatial marking, see example 8 from Kabardian; finally, some 
NEC languages as well as SC and Ossetic possess both verbal and nominal spatial 
expressions, cf. example 9 from Agul.
Hinuq (NEC > Tsezic, Russia; Forker 2013: 90)
[7] b-iƛ’i-yo	 	 hibayɬu=tow	 maydan-i-ƛ’o-r
 hpl-go-prs that.obl=emph square-obl-super-lat
 ‘They are going to that square.’
5 Data from Abaza has been collected in the villages Inzhich-Chukun, Krasnyj Vostok and Psyzh 
(Karachay-Cherkess Republic) in 2017–2021.
6 Data on West Circassian has been collected in the villages Hakurinohabl and Pshicho in 2004–
2005, 2008 and 2010.
15Jezikoslovni zapiski 28  2022  1
Besleney Kabardian (Lomize 2011: 11)
[8] baʒe-r	 karobke-m	 	 q’ə-de-pšə-čj’-a
 fly-abs box-obl  csl-loc:inside-crawl-elat-pst
 ‘The fly went out of the box (towards the observer).’
Agul (NEC > Lezgic, Russia; Maisak 2014: 10; glossing adapted)
[9] ruš.a		 gardan.i-q  šarf  q-ix.i-ne
 girl.erg neck-post  scarf(abs) post-put.pfv-aor
 ‘The girl put a scarf on her neck.’
Finally, in terms of the classic albeit problematic distinction between “aggluti-
native” and “fusional” morphotactic techniques (see e.g. Plungian 2001 and 
Haspelmath 2008 for critical assessments), Caucasian languages also show both 
cross-linguistic and, notably, language-internal variation. While it is probably safe 
to characterise NWC as predominantly agglutinating with almost no non-phono-
logically conditioned allomorphy or fusion and very little cumulation of several 
meanings within one morpheme, other indigenous languages of the Caucasus can-
not be assigned to a single morphotactic type. Thus, while nominal inflection in 
Kartvelian is mostly agglutinating, verbs fall into lexically determined inflection 
classes; fusion and complex morphophonology is rare in Georgian, but abundant 
in Svan (Tuite 1997: 9–11). For NEC it is common to possess both more fusional 
and more agglutinative subsystems in both nominals and verbs, apparently reflect-
ing different historical layers. Thus, verbs in the Dargic and Lezgic languages dis-
tinguish between the perfective and imperfective stems whose formal relations are 
often opaque and unpredictable, while most other forms are built from those stems 
in an agglutinative fashion often betraying a transparent analytic origin (Arka-
diev – Maisak 2018: 134–137). Table 1 shows some verbal forms of Mehweb 
Dargwa (aspectual stems are bolded; hyphen indicates the gender prefix). Note 
that while the suffixes of the Preterite and the Imperative show some allomorphy, 
the exponents of the Future, Optative and Conditional are completely uniform.
Table 1: Some verbal forms of Mehweb Dargwa (Daniel 2019: 88)
‘come’ ‘put on’ ‘fly’ ‘read’
pfv ipfv pfv ipfv pfv ipfv pfv ipfv
Preterite -ak’ib -ik’ib ik’ub irk’wib arcur urcib -elč’un luč’ib
Imperative -ak’e(na) -ik’e(na) ik’wa(na) irk’we(na) arce(na) urce(na) -elč’a(na) luč’e(na)
Future -ak’iša -ik’iša ik’wiša irk’wiša arciša urciša -elč’iša luč’iša
Optative -ak’ab -ik’ab ik’wab irk’wab arcab urcab -elč’ab luč’ab
Conditional -ak’ak’a -ik’ak’a ik’wak’a irk’wak’a arcak’a urcak’a -elč’ak’a luč’ak’a
For the nominal morphology of NEC see section 4.
16 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
3 non-tRivial moRphologiCal exponents
The morphological exponent most widely attested in the world’s languages is a ca-
nonical affix, i.e. a phonologically bound continuous string of phonemes either fol-
lowing (suffix) or preceding (prefix) another string of phonemes (stem). Less trivial 
kinds of exponents include non-affixal operations such as vowel and consonant al-
ternations or shifts of stress or tone, and non-canonical affixes. The latter fall into 
several subtypes. First, an affix may not have a fixed linear position, in some cases 
following the stem and preceding it in others; such affixes are called ambifixes (see 
Arkadiev – Lander 2020 for a preliminary typology and references). Second, an affix 
may consist of two parts, one prefixal and one suffixal, which do not occur alone 
(circumfixes). Third, an affix may occur inside the stem rather than linearly attach to 
it; such infixes are quite widespread cross-linguistically (cf. Yu 2007). Finally, there 
exist transfixes, known primarily from Semitic languages, which are discontinuous 
strings of phonemes interlaced with similarly discontinuous stems.
While the majority of morphological exponents in the Caucasian languages 
are canonical affixes, one finds in these languages specimens of all non-canonical 
types as well. Thus, negation is expressed by means of ambifixes in Abkhaz, Aba-
za and Ubykh verbs, see 10 (the affix is bolded and the stem underlined).
Abkhaz (NWC > Abkhaz-Abaza; Chirikba 2003: 44; glosses added)
[10] a. dǝ-r-ga-wá-m
  3sg.h.abs-3pl.erg-carry-ipf-neg
  ‘They do not carry him/her.’
 b. d-rǝ-m-gá-jt’
  3sg.h.abs-3pl.erg-neg-carry-dcl
  ‘They did not carry him/her.’
Circumfixes constitute a characteristic morphological trait of the Kartvelian lan-
guages (Harris 2002a; Testelets 2020: 526), where they are found in both inflec-
tion and derivation, see Georgian examples in 11.
Georgian (SC)
[11] a. caritive ‘without’ u-√-o: pul-i ‘money’ ~ u-pul-o ‘pennyless’ (Vogt 1971: 234)
 b. intensive u-√-es: met’-i ‘more’ ~ u-met’-es-i ‘still more’ (Vogt 1971: 236)
 c. nomen abstractum si-√-e: maγal-i ‘high’ ~ si-maγl-e ‘height’ (Vogt 1971: 239)
 d.  nomen loci sa-√-e: pul-i ‘money’ ~ sa-pul-e ‘wallet’ (Vogt 1971: 240)
 e. nomen agentis me-√-e: baγ-i ‘garden’ ~ me-baγ-e ‘gardener’ (Vogt 1971: 242)
 f.  ‘teen’ t-√-met’-: ekvs-i ‘six’ ~ t-ekvs-met’-i ‘sixteen’ (Vogt 1971: 56)
 g. ordinal me-√-e: or-i ‘two’ ~ me-or-e ‘second’ (Vogt 1971: 58)
 h. participle m-√-ar: mo-k’vd-a ‘s/he died’ ~ mo-m-k’vd-ar-i ‘dead’ (Vogt 1971: 249)
Harris (2002a: 308–312) shows that the circumfix t-√-met’ forming teen numerals 
originates from the Old Georgian compounds of the type at-otx-met’-	“ten-eight-
-more”. Note that in modern Georgian such numerals can attach the ordinal cir-
cumfix me-√-e thus forming words with two layers of circumfixation: me-[t-[ert]-
-met’]-e ‘eleventh’ (Harris 2002a: 311).
17Jezikoslovni zapiski 28  2022  1
NWC languages also feature circumfixes, which originate from idiomatic 
combinations of locative or applicative preverbs with verbal roots, into which oth-
er lexical roots may be incorporated (Кумахов 1964: 143–146; Arkadiev – Maisak 
2018: 125–126). The second root of such compounds grammaticalises into a direc-
tional suffix always co-occurring with a particular preverb, cf. 12.
West Circassian (НКъ 2007: 57)
[12] šjebzašje-r-jə	 waŝʷe-m de-bəbə-je
 arrow-abs-add sky-obl up-fly-up
 ‘The arrow flies up into the sky.’ (cf. de-je-n ‘to go up’)
Infixes are attested in various branches of NEC in a number of functions, usually 
alongside other morphological processes, both affixal and non-affixal. For exam-
ple, infixation as well as vocalic ablaut work in the formation of aspectual stems 
in Dargic languages, cf. Table 2. Note that in some verbs the infix occurs in the 
imperfective stem, and in others in the perfective stem.
Table 2: Aspectual stems of Sanzhi Dargwa verbs (Forker 2020: 207–211)
Gloss Perfective Imperfective
‘throw’ ixw- i〈r〉xw-
‘stick, attach’ kat’- ka〈l〉t’-
‘milk’ -i〈r〉c:- ic:-
‘turn, grind’ -e〈l〉q’- -uq’-
‘understand’ arʁ- irʁ-
‘saw’ erč- urč-
Similarly, verbal markers of gender agreement can be both prefixal and infixal in 
some languages, cf. the data from Archi in Table 3, showing that the inanimate 
gender marker b- occurs prefixally if the infixal position is occupied by the im-
perfective marker -r- (in fact, the situation is more complex, and the prefixal vs. 
infixal position of the gender marker is not always determined by the presence or 
absence of an imperfective infix, see Chumakina – Corbett 2015 for details).
Table 3: Gender markers in Archi verbs (Lezgic, Russia; Кибрик 1977: 80)
Gloss Durative Terminative Finalis
‘drive’ b-a〈r〉k’ur a〈b〉k’u a〈b〉k’as
‘let’ b-a〈r〉tir a〈b〉t:i a〈b〉tis
‘measure’ b-a〈r〉sin a〈b〉sni a〈b〉smus
A cross-linguistically rare phenomenon akin to infixation is endoclitics, i.e. 
clitics able to occur inside their hosts. The most famous example of endoclisis 
comes from the Lezgic language Udi (Harris 2002b), where markers of pronom-
inal agreement and negation are clitics which attach to the focused constituent 
(13a), including verbal forms, and under some conditions occur between the 
18 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
components of complex verb stems7 (13b) and even inside simplex verbal roots 
(13c). 
Udi (NEC > Lezgic, Azerbaijan, Georgia; glossing modified)
[13] a. äyel-en	 p’ạ	 	 ẹš=ne	 	 aq’-e
  child-erg two  apple=3sg  take-aor
  ‘The child took two apples.’ (Harris 2002b: 119)
 b. nana-n	 	 bụγa=ne=b-e	 	 p’ạ	 					ačik’alšey
  mother-erg find=3sg=do-aor two      toy
  ‘Mother found two toys.’ (Harris 2002b: 122)
 c. pasčaγ-un	 	 γar-en	 gölö	 	 bẹ=ne=γ-sa	 	 			met’a-laxo
  king-gen  boy-erg much  look〈=3sg〉-prs     this.gen-on
  ‘The prince looks at this for a long time.’ (Harris 2002b: 125)
Vocalic, and more rarely, consonantal alternations are attested as morphological 
exponents primarily in the NEС languages. We have already seen ablaut as one of 
the means of formation of aspectual stems in Dargic languages in Tables 1 and 2 
above. Another example is constituted by vowel quality and quantity alternations 
and consonant gemination serving as the means of formation of oblique nominal 
stems in Bezhta (Tsezic), see Table 4.
Table 4: Nominal stems in Bezhta (Комри et al. 2015: 241–244; transcription 
adapted)
Gloss Nominative Oblique (= Ergative)
‘roof’ ƛ’amo ƛ’ama
‘bull calf’ biše biši
‘neck’ boɬo biɬa
‘moon’ boco bico
‘year’ ƛi ƛi:
‘woman’ aqo aq:a
In Circassian languages, alternation of the final vowel of verbal stems marks direc-
tion (lative vs. elative) and valency (bivalent vs. antipassive), see Кумахов 1974 
and Arkadiev – Letuchiy 2021; consider examples in 14 and 15.
West Circassian (Кумахов 1974: 82; transcription adapted, glossing added)
[14] a. jǝ-šje-n     b. jǝ-šjǝ-n
  loc-lead.lat-msd   loc-lead.elat-msd
  ‘to lead inside’    ‘to lead outside’
[15] a. s-e-txǝ     b. s-e-txe
  1sg.erg-dyn-write.tr   1sg.abs-dyn-write.antip
  ‘I am writing it.’    ‘I am writing.’
7 For detailed argumentation regarding the status of complex verbs in Udi as single phonological 
and morphosyntactic words, see Harris 2002b: 76–87. For a recent discussion of endoclitics in 
Andi, another NEC language, see Maisak 2021.
19Jezikoslovni zapiski 28  2022  1
4
Stress shift can also serve as morphological exponent, cf. the cross-linguistically 
very peculiar marking of causativisation in Tsugni Dargwa in Table 5 or the use 
of stress and intonation for finiteness distinctions in Budugh (Lezgic, Azerbaijan; 
Authier 2010).
Table 5: Causativisation by stress shift in Tsugni Dargwa (Сулайбанов – 
Сумбатова 2022: 121; infinitive forms, non-human singular absolutive)
Gloss Base verb Causative
‘break’ b-áˁč-i b-aˁč-í
‘dry’ b-érʁw-i b-erʁw-í
‘get lost / lose’ b-it-áq-i b-it-aq-í
Vocalic alternations, when they become highly regular, may be reanalysed as 
transfixes. This has apparently happened in Budugh, consider partial verbal para-
digms in Table 6, where vocalic transfixes inserted into consonantal stems express 
gender and valency.
Table 6: Transfixes in Budugh verbs (Authier 2009)
Gloss Valency Gender Perfective Imperfective
‘sleep’ base masculine eχir arχar
animal öχür orχor
causative masculine eχir erχi
animal öχür örχü
‘sit’ base masculine aq’ul alq’al
animal oq’ul olq’ol
causative masculine eq’il elq’i
animal öq’ül ölq’ü
According to Authier (2009; Ms.), Budugh transfixation is a result of phonologi-
cal change that has affected stem vowels in combination with consonantal gender 
infixes and the causative suffix going back to the verb i- ‘do’, cf. 16:
Budugh (Authier 2009)
[16] a. öχür	< *e-w-χ-ir	‘it (animal) slept’, cf. feminine e-r-χ-ir
 b. orχor	<	*a-w-r-χ-ar	‘it (animal) sleeps’
 c. elq’i	<	*a-lq’-	+	*i-	‘makes him sit’
 d. ölq’ü < *a-w-lq’- + *i- ‘makes it (animal) sit’
non-tRivial FoRm-meaning Relations
Much of the variation and complexity in morphological systems of the world’s 
languages is grounded in form-meaning relations on syntagmatic and paradig-
matic axes (see, e.g., Anderson 2015; Baerman et al. 2017). Canonically, one 
form expresses one meaning, and vice versa (cf. Carstairs 1987: 12–13); how-
20 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
ever, this ideal is only rarely found in real languages, and various deviations 
from it have been subject to theoretical and typological investigations during 
the last several decades (see e.g. Stump 2016 and Baerman et al. 2017 for over-
views). 
Consider the West Circassian case-number paradigm in Table 7 (cf. 
Аркадьев 2014). While the Absolutive shows a neat one-to-one mapping be-
tween meanings and exponents, all other forms exhibit complications. In the 
Oblique Plural, alongside the combination of the regular Plural suffix -xe with 
the regular Oblique suffix -m, we see cumulative exponence of case and number 
in the marker -me, as well as multiple exponence of Plural by the combination 
of the Plural suffix with the Oblique Plural suffix; besides that, the coexistence 
of several competing expressions for the same paradigmatic cell is a case of 
overabundance. Further, comparing nouns with pronouns, we encounter lex-
ically-conditioned allomorphy of the Oblique case suffix (-m vs. -šj). Final-
ly, taking into account the Instrumental, we see that its marker attaches to the 
Oblique case form rather than to the bare stem, as other case suffixes (the situa-
tion is in fact more complicated, see Serdobolskaya 2011). 
Table 7: West Circassian case-number paradigm (Рогава – Керашева 1966: 
62, 85)
‘boy’ ‘this’
Singular Plural Singular Plural
Absolutive čj’ale-r čj’ale-xe-r mǝ-r mǝ-xe-r
Oblique čj’ale-m čj’ale-xe-m,  
čj’ale-xe-me,
čj’ale-me
mǝ-šj mǝ-xe-m,
mǝ-xe-me
Instrumental čj’ale-m-čj’e čj’ale-xe-m-čj’e mǝ-šj-čj’e mǝ-xe-m-čj’e
This example shows that even a small and apparently simple inflectional subsys-
tem can exhibit a large number of deviations from the canonical ideal. One type of 
such deviation that has proven particularly challenging for morphological theories 
and that is well-attested in the languages of the Caucasus is multiple exponence, 
i.e. co-occurrence of several (identical or distinct) exponents of the same meaning 
in one word. An insightful discussion of multiple exponence, its typology and his-
torical origins has been recently provided by Harris (2017). Harris distinguishes 
four types of multiple exponence, three of which are amply attested in the Cauca-
sian languages.
Periodic multiple exponence “occurs when a bound morpheme [= a carrier 
morpheme] must be accompanied by an exponent of feature F, while the stem 
must also be accompanied by an exponent of F” (Harris 2017: 55). This type of 
multiple exponence is attested in many NEC languages, especially in gender-num-
ber agreement, cf. the Batsbi example in 17 with three instances of gender marker 
d- agreeing with the noun ‘house’.
21Jezikoslovni zapiski 28  2022  1
Batsbi (a.k.a. Tsova-Tush; NEC > Nakh, Georgia; Harris 2009: 268)
[17] tišin	 c’a	 	 	 	 daħ	 	 d-ex-d-o-d-an-iš
 old house(nom)  pvb  gm-destroy-gm-prs-gm-evid-2pl.erg
 ‘Y’all are evidently destroying the old house.’
Multiple exponence of the kind found in Batsbi historically arises via univerbation 
of constructions with auxiliaries each carrying gender agreement (Harris 2017: 
115–130).
Reinforcement multiple exponence “characteristically involves exponents 
that are identical in feature representation but not identical in form” (Harris 2017: 
61). A good example comes from Khinalug (NEC, Azerbaijan), where many nouns 
feature two or even three plural suffixes (whose vowels alternate according to the 
rules of harmony), each of which can occur on its own, see Table 8.
Table 8: Khinalug plural suffixes (Khvtisiashvili 2013: 96–99)
Gloss Singular Plural
‘grandchild’ xıdıl xıdıl-ır
‘goat’ taka taka-d
‘drop’ kixir kixir-d-ir
‘corner’ kunǯ kunǯ-ur-d-ur
Another example of this type is provided by Abaza, where negation in finite ver-
bal forms is expressed twice – by the already familiar common West Caucasian 
ambifixal marker -m- and by the innovative prefix gj- stemming from an emphatic 
particle (Пазов 2019), see 18.
Abaza (fieldwork data, textual example)
[18] jə-gj-sə-m-də́r-t’
 3sg.n.abs-neg-1sg.erg-neg-know(aor)-dcl
 ‘I did not know that.’
The Abaza double negation is a cross-linguistically fairly common instance of 
grammaticalisation of the originally pragmatic reinforcement of negative markers 
(Jespersen 1917; van Gelderen 2008).
What Harris (2017: 64) calls accidental multiple exponence “involves expo-
nents in a subset or overlapping relationship”, i.e. if one or each of the exponents 
in addition to the multiply realised feature also express some other meanings (as 
e. g. in the case of West Circassian Oblique Plural forms in -xe-me pl-pl.obl). A 
remarkably complex and systematic pattern of multiple exponence of this kind is 
found in Ubykh, especially in the speech of its last fluent speaker Tevfik Esenç 
(1904–1992) (see Dumézil – Esenç 1975: 161–162; Smeets 1997; and Fenwick 
2011: 135–136). Here the number of the absolutive (S/P) argument of the verb 
is expressed by such elements as (i) the person-number prefixes, (ii) the plural 
suffixes -a and -n(e), (iii) the Retrospective suffixes -jt’ sg ~ -jλ(e) pl, (iv) the cau-
sative prefixes də- sg ~ ʁe- pl, and (v) root suppletion with a considerable number 
22 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
of verbs. Combinations of these different markers can yield verbal forms with up 
to four exponents of number, cf. 19.
Ubykh (NWC > Ubykh, extinct; transcription adapted, glosses added)
[19] a. šjǝ-w-ʁe-qʷ’e.χe-q’e-n
  1pl.abs-2sg.erg-caus.pl-stop.pl-pst-pl
  ‘You (sg) made us stop.’ (Dumézil – Esenç 1975: 173)
 b. šj-kj’-a-ne-jλe-me
  1pl.abs-go-pl-dyn-retro.pl-neg
  ‘We were not going.’ (Dumézil – Esenç 1975: 165)
The indigenous languages of the Caucasus constitute one of the “hotbeds” of mul-
tiple exponence in the languages of the world, and many of the cases of multi-
ple exponence in these languages are highly systematic and cannot be “explained 
away” as accidental quirks or historical residues.
Some of the complex form-meaning relations become apparent only when 
paradigmatic structures are taken into account. One of the famous cases of this sort 
is constituted by NEC noun inflection analysed in Kibrik 1991. Consider a partial 
paradigm of an Archi noun in Table 9.
Table 9: Partial paradigm of Archi noun ‘cup’ (Kibrik 2003: 60)
Singular Plural
Nominative gel gel-um
Ergative gel-li gel-um-čaj
Genitive gel-li-n gel-um-če-n
Dative gel-li-s gel-um-če-s
In Table 9 alongside the unequivocal suffixes of Plural number (-um) and Genitive 
(-n) and Dative (-s) cases we also see two markers whose status is not immediately 
obvious, -li and -čaj/-če. On the one hand, they seem to be cumulative exponents 
of the Ergative case and Singular or Plural number; on the other, they serve as the 
stems to which other oblique case suffixes attach, resembling the West Circassian 
Instrumental in Table 7. In fact, as Kibrik (1991: 257) argues, the appropriate 
analysis is to treat both Nominative and Ergative as expressed by zero markers 
attached to distinct stems: the nominative stem and the oblique stem. Prima facie 
evidence for this comes from other languages where the Ergative case has overt 
exponents attached to the Oblique stem, cf. the Tsakhur (Lezgic, Russia) partial 
paradigm in Table 10.
Table 10: Partial paradigm of the Tsakhur noun ‘road’ (Лютикова 2017: 669)
Singular Plural
Nominative jaˀq jaˀq-bɨ
Ergative jaˀq-ɨ-n jaˀq-b-iš-e
Dative jaˀq-ɨ-s jaˀq-b-iši-s
23Jezikoslovni zapiski 28  2022  1
The most general structure of the NEC noun paradigm is schematised in 20 from 
Kibrik 2003: 61. 
[20] nom.sg       =   ROOT → pl       = nom.pl
         ↓    ↓
oblique ←         obl.sg         obl.pl   →      oblique
cases sg                    cases pl
This schema accounts for both Archi and Tsakhur examples, however, there are 
many deviations from it giving rise to considerable inter- and intralinguistic varia-
tion in paradigmatic structure (see Kibrik 1991; 2003: 61–67). Thus, for instance, 
in Rutul (Lezgic, Russia) many inanimate nouns form their oblique plural stem on 
the basis of the oblique singular stem, cf. Table 11.
Table 11: Partial paradigm of the Rutul noun ‘moon’ (Махмудова 2001: 34; 
transcription and segmentation adapted)
Singular Plural
Nominative waz waz-bɨr
Ergative waz-ɨr-ɨra waz-ɨr-mɨ-ra
Dative waz-ɨr-ɨs waz-ɨr-mɨ-s
Oblique stems in NEC languages are formed in a variety of ways, including suffix-
es, infixes, vowel alternations (cf. Bezhta in Table 4), stress shift (e.g. in Khwar-
shi), and combinations thereof, as well as suppletion. Importantly, each language 
possesses a whole set of oblique stem formations, whose distribution is partly 
predictable from the noun’s phonological shape or semantics and partly lexically 
determined (Kibrik 2003: 69–72), cf. the example of Lezgian in Table 12.
Table 12: Oblique stem formations in Lezgian (Haspelmath 1993: 74–77)
Exponent Condition Example
-di default buba ‘father’: buba-di
-a consonant-final personal names + some 
common nouns
Farid: Farid-a 
apaj ‘father in law’: apaj-a
-i abstract nouns with the suffix -wal 
verbal nouns with the suffix -(u)n 
all plural suffixes but -bur
jaru-wal ‘redness’: jaru-wil-i 
k’el-un ‘learning’: k’el-un-i 
buba-jar ‘fathers’: buba-jr-i
-u plurals in -bur jaru-bur ‘red ones’: jaru-bur-u
-Adi monosyllabic nouns that denote a 
non-discrete mass
čig ‘dew’: čig-edi
-rA monosyllabic nouns that denote animals lam ‘donkey’: lam-ra
-Uni various monosyllabic nouns kam ‘trap’: kam-uni
-U lexical (monosyllables only) q’ünt ‘elbow’: q’ünt-ü
-Ci lexical (monosyllables only) čar ‘paper’: čar-či 
žin ‘ghost’: žin-ži
24 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
One may wonder whether the nominative vs. oblique distinction in the nominal 
inflection of the NEC languages is a purely formal complication or has any ex-
tramorphological function. Remarkably, the latter turns out to be true, since the 
nominative vs. oblique distinction has obvious morphosyntactic repercussions 
in at least some of the languages of the family (cf. Тестелец 2019). The first 
piece of evidence for the syntactic relevance of the “obliqueness” feature comes 
from noun-phrase-internal concord. Thus, in the Nakh languages attributive ad-
jectives distinguish between nominative and oblique forms, cf. the Chechen ex-
ample in Table 13.
Table 13: Attributive adjective inflection in Chechen (Nichols 1994: 29)
‘high fence’
NomSg leqan kyert
NomPl leqan kyertaš
GenSg leqaču kyertan
DatSg leqaču kyertana
GenPl leqaču kyerti:n
Some languages have two genitive forms for nouns, the one used with the head 
noun in the nominative case and another occurring when the head is in one of the 
oblique cases (Kibrik 1995), cf. examples from Bezhta in 21.
Bezhta (Kibrik 1995: 220)
[21] a. abo-s   	 is
  father-gen.dir  brother.nom.sg
  ‘father’s brother’
 b. abo-la   	 is-t’i-l
  father-gen.obl  brother-obl-dat
  ‘to father’s brother’
In some languages, the nominative vs. oblique distinction becomes relevant for 
noun phrase syntax. Thus, in Bagwalal (Andic, Russia; Кибрик 2001: 691–693) 
only modifiers of nominative nouns can be focused, either by means of a fo-
cus particle (22a) or by moving the attribute before the verb (23a), while noun 
phrases in oblique cases are opaque both for focus particles (22b) and focus 
movement (23b).
Bagwalal (Кибрик 2001: 691, 693; transcription and glossing adapted)
[22] a. [ʕisa-w-ʁ-ō   waša]NOM 	 w-ā. 
  Isa-gen-foc-m  son.nom  m-come
  ‘isa’s son came.’
 b. *[ʕisa-w-ʁ-ō	  waša-š:u-r]OBL awal	 	 					ǯē-rā-χ. 
  Isa-gen-foc-m  son-obl-erg house.nom      build-ipf-cvb
  expected: ‘Isa’s son is building a house.’ 
25Jezikoslovni zapiski 28  2022  1
[23] a. q’alam-dari	 	 	 di-ha		 	 š:iš:u-r  r-ah-a!
  pencil-pl.nom  1sg.obl-dat red-npl  npl-buy-imp
  ‘Buy red pencils for me!’
 b. *q’alam-li-r	 	 	 š:iš:u=b qwa-ra!
  pencil-obl-erg  red-n  write-imp
  expected: ‘Write with a red pencil!’
All this implies that oblique stems in at least some NEC languages express a sui 
generis morphosyntactic feature (“obliqueness”) distinct from case proper, and 
that the phenomenon cannot be simply an instance of multiple exponence. This 
kind of “layered” nominal inflection, which is also attested in some Indo-Irani-
an languages, such as Romani (Elšík 2000) and Ossetic (Беляев 2014),8 presents 
clear challenges for the theories of morphology and the morphology-syntax inter-
face and opens potential windows into the history of nominal inflection.
A particularly complex example of a lack of one-to-one mapping between 
forms and meanings is presented by Kartvelian verbal morphology, which is 
largely organised according to the principle of distributed exponence, defined 
by Caballero and Harris (2008: 170) as situations where “no single morphological 
marker can truly be said to realize a feature or category; the feature is, rather, re-
alized by a combination of morphemes”. As a relatively simple example, consider 
selected tense-aspect-mood forms of the Georgian regular verb ‘hide’ in Table 14.
Table 14: Partial paradigm of the Georgian verb ‘hide’ (3rd person singular 
subject and object; personal knowledge)
Active Passive
Present mal-av-s i-mal-eb-a
Imperfective past mal-av-d-a i-mal-eb-od-a
Future da-mal-av-s da-i-mal-eb-a
Conditional da-mal-av-d-a da-i-mal-eb-od-a
Aorist (perfective past) da-mal-a da-i-mal-a
Optative (subjunctive) da-mal-o-s da-i-mal-o-s
While some of the affixes in Table 14 seem to have a clear meaning (e.g. the prefix 
i- expressing Passive; this association, however, is only valid for a particular verb 
class to which ‘hide’ belongs, see e.g. Boeder 1967; Gurevich 2006), the distribution 
of most of them is not linked to any particular feature value. Thus, the prefix da- (one 
of the spatial-aspectual preverbs) occurs both in the unequivocally perfective Aorist 
and in the forms whose aspectual interpretation is not so clear-cut (Future, Condi-
tional and Optative); likewise, the so-called “thematic elements” -av and -eb occur 
in the forms which do not seem to have any common morphosemantic feature. The 
same can be said about the -d and -od suffixes (Imperfect and Conditional) as well 
as about the suffixes -s and -a, which express 3rd person singular subject but are 
8 See, however, Erschler 2018 for an alternative interpretation.
26 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
distributed across the different tense-aspect-mood-voice subparadigms in a non-uni-
form way. Thus, while each tense-aspect-mood value is uniquely expressed by a 
particular combination of affixes, none of the latter represents a dedicated exponence 
of any of the former. Affixes cannot be assigned any meanings on their own and only 
acquire a meaning as parts of words (cf. Gurevich 2003). 
The propensity of Kartvelian languages to show distributed exponence is also 
manifested in the already mentioned frequent use of circumfixes and prefix-suf-
fix combinations in general (see Harris 2002: 315–320). Distributed exponence is 
challenging both for morpheme-based morphological theories, since it does not 
involve “morphemes” as Saussurean signs where form and meaning are coupled 
together, and for grammaticalisation theory, since the origin of such systems lies 
in processes of functional redistribution and adjustment (“featurisation”; Dahl 
2004: Ch. 9), rather than transition from lexemes to affixes.
polysynthesis in abkhaz-adyghean languages
NWC languages stand out among the languages of the Caucasus and Western 
Eurasia as the only truly polysynthetic languages of the region (Lander – Testelets 
2017; Arkadiev – Lander 2020). Polysynthesis is commonly understood as ex-
treme syntagmatic complexity of morphology (e.g. Mithun 1988: 442). In NWC 
this complexity is manifested both in verbs and nominals, as the following Kab-
ardian examples illustrate.
Besleney Kabardian (fieldwork data)
[24] a. z-a-q’ə-šjə-r-a-ʁe-pλə-hə-ne
  rfl.abs-3pl.io-csl-pvb-dat-3pl.erg-caus-look-circum-fut
  ‘they will let them look around there’ (textual example)
 b. d-jə-ʁʷəneʁʷ–bzəλxʷəʁe–daxe–dede-m 
  1pl.pr-poss-neighbour–woman–beautiful–very-obl
  ‘our very beautiful lady-neighbour’ (elicited, Yury Lander, p.c.)
According to a recent definition by Fortescue (2017: 122), polysynthetic lan-
guages display holophrasis (i.e. are able to represent a whole clause, including 
information about all core arguments in a single verb) and integrate more than 
one “semantically heavy morpheme”, either affixal or lexical, in the verb. Lan-
guages corresponding to this broad characterisation vary along several parame-
ters (cf. Mattissen 2004; 2017), such as the presence of productive compound-
ing (incorporation), available types of so-called “lexical affixation” (Mithun 
1997; Mattissen 2006: 297–333) and morphotactic organisation (rigid and often 
opaque templatic ordering vs. semantically-driven scopal ordering of affixes).
Manifestations of polysynthesis found in NWC include (i) polypersonalism 
facilitated by a rich system of semantically specialized applicatives introducing 
peripheral participants; (ii) a rich system of affixes often with quite concrete mean-
ings, especially locative ones; (iii) an intricate mixture of templatic and scopal 
5
27Jezikoslovni zapiski 28  2022  1
ordering; (iv) productive “nominal complexes” sharing the properties of words 
and phrases (Lander 2017).
Polypersonalism can be illustrated by example 25 from Abaza, featuring a 
verbal form with four person-number-gender prefixes each corresponding to a dis-
tinct participant:
Abaza (fieldwork data, textual example)
[25] j-ŝə-z-j-á-s-hʷ-p’
 3sg.n.abs-2pl.io-ben-3sg.m.io-dat-1sg.erg-say-npst.dcl
 ‘I will tell this to him about you all.’
Such quadripersonal forms are rare, but attested in natural texts, while forms ex-
pressing three participants like that in 5a above are fairly common. As already 
said, such an exuberant polypersonalism is made possible by the existence of a 
rich system of applicative prefixes such as the Benefactive z- and the Dative a- in 
25, adding indirect objects and the corresponding personal prefixes to both intran-
sitive and transitive verbs. Applicatives in NWC are very numerous (up to several 
dozens in Abaza and Abkhaz) and in terms of semantics range from very general, 
like the underspecified Dative in 25, to highly specific, as e.g. the locative applic-
atives, see 26 and Arkadiev et al. (to appear).
Abaza (fieldwork data, textual example)
[26] d-na-s-šjta-lə-n
 3sg.h.abs-trl-1sg.io-pvb:behind-go-pst
 ‘He followed me.’
Most of the locative applicatives originate diachronically from and often corre-
spond synchronically to body-part nouns (Ломтатидзе 1983; Arkadiev – Maisak 
2018: 121–125), cf. 27, which can be considered a type of noun-incorporation.
Abkhaz (Spruit 1986: 29; transcription and glossing adapted)
[27] a-maĉwaz   lə-mʁa-s-χə-jt’
 def-ring  3sg.f.io-pvb:finger-1sg.erg-take(aor)-dcl
 ‘I took the ring from her finger.’
Applicatives allow stacking and even limited recursion, as in 28 with two non-syn-
onymous instances of the benefactive.
West Circassian (Lander – Letuchiy 2010: 269)
[28] s-a-fə--f-e-txe
 1sg.abs-3pl.io-ben-3sg.io-ben-dyn-write
 ‘I write to him for their benefit.’
NWC languages also feature a number of non-applicative affixes with meanings 
comparable to those of independent lexical items; some of these are spatial and 
often correspond to nouns, cf. 29, where a nominal root occupies the same posi-
tion as other locative preverbs, while others rather express adverbial or predicative 
28 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
meanings. The latter are usually suffixes and allow variable ordering based on 
mutual scope (Korotkova – Lander 2010), cf. a minimal pair in 30.
Abaza (Клычев 1995: 67; transcription adapted, glossing added)
[29] a-sabəj	 	 d-gara-l-gʷa-n
 def-child  3sg.h.abs-pvb:cradle-3sg.erg-lay-pst
 ‘She laid the child into the cradle.’
West Circassian (Lander 2016: 3523)
[30] a. gʷəŝʷe-ŝwe-žjə-ʁ
  be.glad-sml-re-pst 
  ‘s/he pretended again that s/he was happy’  (refactive > similative) 
 b. gʷəŝʷe-žjə-ŝwa-ʁ
  be.glad-re-sml-pst
  ‘s/he pretended that s/he was happy again’  (similative > refactive) 
While the ordering of at least some suffixes in NWC is determined by their se-
mantic scope, the order of prefixes is mostly rigid and does not respect meaning 
relations. Thus, different parts of the NWC verb adhere to distinct principles of 
ordering and form-to-meaning mapping, sometimes fairly complex and involving 
elements belonging to distinct “zones” of the word (see e.g. Arkadiev – Letuchiy 
2011 for just one type of such phenomena). This complexity obviously reflects 
successive historical layers of grammaticalisation and morphologisation.
One of the most remarkable aspects of NWC morphology which has import-
ant consequences for the whole morphosyntactic organisation of these languages 
is relativisation (Hewitt 1979a; 1979b; Ландер 2012; Lander – Daniel 2019). In-
stead of the relative complementizers or relative pronouns found in the European 
languages and SC, or the participles used in relative clauses of NEC languages, 
NWC languages possess relative verbal prefixes that occupy the same slots in the 
verbal form as the corresponding person-number markers. Compare the Abaza 
finite clause in 31a with the relativisation of the absolutive argument expressed by 
the prefix j- (31b) and of the ergative argument by the prefix zǝ- in 31c.
Abaza (fieldwork data, elicited)
[31] a. a-phʷəspa	 a-čj’kʷən	 də-l-b-əj-t’
  def-girl def-boy 3sg.h.abs-3sg.f.erg-see-prs-dcl
  ‘The girl sees the boy.’
 b. [a-phʷəspa		 jǝ-l-ba-wa]	 	 	 	 a-čj’kʷən
  def-girl  rel.abs-3sg.f.erg-see-ipf  def-boy
  ‘the boy that the girl sees’
 c. [a-čj’kʷən	 dǝ-z-ba-wa]    a-phʷəspa
  def-boy 3sg.h.abs-rel.erg-see-ipf  def-girl
  ‘the girl who sees the boy’
Relativisation of non-arguments such as place, time and manner is achieved either 
by means of special prefixes, as in Abkhaz and Abaza (32), or by use of applica-
tives unattested in finite forms, as in Circassian (33).
29Jezikoslovni zapiski 28  2022  1
Abkhaz (Spruit 1986: 122; transcription adapted, glossing added)
[32] d-aχj-jə-z	 	 	 	 	 	 	 adgjəl
 3sg.h.abs-rel.loc-be.born-pst.nfin art+place
 ‘the place where he was born’
West Circassian (Ландер 2012: 288)
[33] [wəlape	 sə-z-čj’e-mə-kʷ’a-ʁe]    we-r
 Ulyap 1sg.abs-rel.io-pvb:under-neg-go-pst weather-abs
 ‘the (bad) weather because of which I did not go to Ulyap’
Relativisation in NWC is employed, beyond adnominal and headless relative 
clauses, also in the formation of different types of subordinate clauses (Caponi-
gro – Polinsky 2011), focus constructions and constituent questions (Сумбатова 
2009). Thus, Abaza and Abkhaz do not employ interrogative pronouns of the kind 
found in most languages of the world, but instead use special interrogative affixes 
attaching to relative verbal forms (Arkadiev 2020; Arkadiev – Caponigro 2021), 
cf. the examples in 34.
Abaza (fieldwork data, textual examples)
[34] a. j-wə́-c-kʷa-z-da
  rel.abs-2sg.m.io-be.with-pl-pst.nfin-q.h
  ‘Who was with you?’
 b. s-pnə		 wəs-ta	 j-wə-ma-ja?
  1sg.pr-at job-adv rel.abs-2sg.m.io-have-q.n
  ‘What are you doing at my place?’
 c. arə́j	 	 áχčja   [a]n-bá-ʕa-z-ʁəčj?
  this  def+money rel.tmp-q.adv-csl-1sg.erg-steal
  ‘When did I steal this money?’
This purely inflectional marking of constituent questions is typologically unique, 
but clearly well-motivated by the system of relative verbal forms as well as the 
overall propensity of West Caucasian languages towards expression of syntactic 
information inside polysynthetic verbs.
summaRy
As even such a short and incomplete survey as the above shows, the languages of 
the Caucasus present a wealth of non-trivial and typologically rare morphologi-
cal phenomena as well as a remarkable degree of diversity, attested both across 
language families and even between closely related varieties. All this makes the 
Caucasian languages ideal as a testing-ground for morphological theories and as 
a field of inquiry into macro- and microvariation in morphology. Among the the-
oretical and typological issues raised by the data surveyed here one can list the 
following ones.
The predominantly head-marking profile of NWC as opposed to the largely 
dependent-marking morphology of NEC poses interesting challenges to histor-
ical-comparative reconstruction under the assumption that the two families are 
6
30 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
related, see e.g. Chirikba 2016 for a hypothesis about the loss and subsequent 
renewal of morphology in the prehistory of NWC.
The coexistence of a variety of morphological techniques, both affixal and 
non-affixal, particularly characteristic of the various branches of NEC but also 
attested in the other two families, suggests a complex historical development of 
successive layers of morphology, as well as clearly contradicts the overly simplis-
tic conception of “flexive” vs. “agglutinating” types and affix-based models of 
morphology.
The complex relations between meaning and form, especially the various 
types of multiple and distributed exponence, whose different manifestations are 
attested in all three indigenous language families of the Caucasus, reveal the in-
adequacy of morpheme-based morphological frameworks that by their very archi-
tecture rule out such phenomena. These data call for more sophisticated models 
allowing for form and content of morphological expressions to be organised by 
distinct principles and to relate to each other by more complex mappings. Like-
wise, the diachronic study of these phenomena enriches our understanding of the 
pathways of change of morphological systems, which by no means always lead to 
more “economic” or “transparent” structures.
Finally, the exuberant polysynthetic morphology of NWC, unique in this 
part of the world and possessing a number of highly exceptional structures, 
broadens the horizons of linguistic typology and morphological theory by show-
ing how the often taken for granted boundaries between inflection, derivation 
and compounding can be largely blurred and how morphology can effectively 
take over from syntax such mechanisms as argument structure, relativisation and 
clause combining.
Since much of the diversity of the Caucasian languages is still insufficiently 
documented and given that most of the languages of the Caucasus are spoken 
by bilingual minorities and hence are endangered to different degrees, one of the 
goals of this article is to urge linguists to engage in their precise, sophisticated (i.e. 
typologically and theoretically informed) and simultaneously unbiased (in partic-
ular, free of Eurocentric preconceptions) description and documentation.
abbReviations
1 – 1st person; 2 – 2nd person; 3 – 3rd person; abs – absolutive; add – additive; adv – adverbial; 
aFRm – affirmative; antip – antipassive; aoR – aorist; aRt – article; ben – benefactive; Caus – 
causative; CiRCum – motion around; Cond – conditional; Csl – cislocative ‘hither’; Cvb – converb; 
dat – dative; dCl – declarative; deF – definite; diR – direct; dyn – dynamic; elat – elative; 
emph – emphatic; eRg – ergative; evid – evidential; F – feminine; FoC – focus; Fut – future; gen – 
genitive; gm – gender marker; h – human; hpl – human plural; imp – imperative; io – indirect object; 
ipF – imperfect; ipFv – imperfective; lat – lative; loC – locative; m – masculine; msd – masdar; 
n – non-human; neg – negation; nFin – non-finite; nom – nominative; npl – non-human plural; 
npst – non-past; obl – oblique; pFv – perfective; pl – plural; plsq – pluperfect; poss – possessive; 
post – localisation behind; pR – possessor; pRF – perfect; pRs – present; pst – past; pvb – preverb; 
31Jezikoslovni zapiski 28  2022  1
q – interrogative; Re – refactive; Rel – relativizer; RetRo – retrospective; RFl – reflexive; sbj – 
subject; sg – singular; sml – similative; supeR – localisation above; tmp – temporal; tR – transitive; 
tRl – translocative ‘thither’; up – motion upward; val – valency marker; witn – witnessed
ReFeRenCes
Anderson 2015 = Stephen R. Anderson, Dimensions of morphological complexity, in: Understand-
ing	and	Measuring	Morphological	Complexity, ed. Matthew Baerman – Dunstan Brown – Gre-
ville G. Corbett, Oxford: Oxford University Press, 2015, 11–26.
Arkadiev 2020 = Peter Arkadiev, Syntax in morphological guise: interrogative verbal morphology 
in Abaza, Linguistic	Typology 24.2 (2020), 211–251.
Arkadiev – Caponigro 2021 = Peter Arkadiev – Ivano Caponigro, Conveying content questions 
without wh-words: evidence from Abaza, in: Proceedings	of	Sinn	und	Bedeutung	25, ed. Patrick 
Grosz – Luisa Martí – Hazel Pearson – Yasutada Sudo – Sarah Zobel, 2021, 73–94, https://ojs.
ub.uni-konstanz.de/sub/index.php/sub/issue/view/29.
Arkadiev – Lander 2020 = Peter Arkadiev – Yury Lander, The Northwest Caucasian languages, in: 
The	Oxford	Handbook	of	the	Languages	of	the	Caucasus, ed. Maria Polinsky, Oxford: Oxford 
University Press, 2020, 369–446.
Arkadiev – Letuchiy 2011 = Peter Arkadiev – Alexander Letuchiy, Prefixes and suffixes in the 
Adyghe polysynthetic wordform: types of interaction, in: Languages	and	Cultures	in	the	Cau-
casus, ed. Vittorio Springfield Tomelleri – Manana Topadze – Anna Lukianowicz, München – 
Berlin: Otto Sagner, 2011, 495–514.
Arkadiev – Letuchiy 2021 = Peter Arkadiev – Alexander Letuchiy, Indirect antipassive in Circas-
sian, in: Antipassive:	Typology,	Diachrony,	and	Related	Constructions, ed. Katarzyna Janic – 
Alena Witzlack-Makarevich, Amsterdam – Philadelphia: John Benjamins, 2021, 483–514.
Arkadiev et al. (to appear) = Peter Arkadiev – Yury Lander – Irina Bagirokova, Applicative con-
structions in the Northwest Caucasian languages, in: Applicative	Constructions:	A	Comparative	
Handbook, ed. Denis Creissels – Fernando Zúñiga, Berlin – Boston: De Gruyter Mouton.
Arkadiev – Maisak 2018 = Peter Arkadiev – Timur Maisak, Grammaticalization in the North Cau-
casian languages, in: Grammaticalization	from	a	Typological	Perspective, ed. Heiko Narrog – 
Bernd Heine, Oxford: Oxford University Press, 2018, 116–145.
Authier 2009 = Gilles Authier, Development of introflexion (root-and-pattern morphology) in 
Budugh verbs, handout, https://docest.com/development-of-introflexion-root-and-pattern-
morphology-in-budugh-verbs. 
Authier 2010 = Gilles Authier, Finite and non-finite: prosodic distinctions on Budugh verb stems, 
in: Clause	Linking	and	Clause	Hierarchy:	Syntax	and	Pragmatics, ed. Isabelle Bril, Amster-
dam – Philadelphia: John Benjamins, 2010, 143–164.
Authier Ms. = Gilles Authier, Grammatical sketch of Budugh, manuscript, https://www.academia.
edu/66065917/. 
Baerman et al. 2017 = Matthew Baerman – Greville G. Corbett – Dunstan Brown, Morphological	
Complexity, Cambridge: Cambridge University Press, 2017.
Boeder 1967 = Winfried Boeder, Über die Versionen des georgischen Verbs, Folia	Linguistica 2 
(1967), 32–52.
Caballero – Harris 2012 = Gabriela Caballero – Alice C. Harris, A working typology of multiple 
exponence, in: Current	Issues	in	Morphological	Theory.	(Ir)regularity,	Analogy	and	Frequen-
cy. Selected	Papers	 from	the	14th	International	Morphology	Meeting,	Budapest,	13–16	May	
2010, ed. Ferenc Kiefer – Mária Ladányi – Péter Siptár, Amsterdam – Philadelphia: John Ben-
jamins, 2012, 163–188.
Caponigro – Polinsky 2011 = Ivano Caponigro – Maria Polinsky, Relative embeddings: a Circas-
sian puzzle for the syntax/semantics interface, Natural	Language	and	Linguistic	Theory 29.1 
(2011), 71–122.
Carstairs 1987 = Andrew Carstairs, Allomorphy	in	Inflection, London etc.: Croom Helm, 1987.
32 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
Chirikba 1999 = Vyacheslav Chirikba, The West Caucasian material in “The North Caucasian Ety-
mological Dictionary” by S. L. Nikolaev and S. A. Starostin, in: Studies	in	Caucasian	Linguis-
tics:	Selected	Papers	from	the	Eighth	Caucasian	Colloquium, ed. Helma van den Berg, Leiden: 
CNWS, 1999, 152–170.
Chirikba 2003 = Vyacheslav Chirikba, Abkhaz, München: LINCOM Europa, 2003.
Chirikba 2008 = Vyacheslav Chirikba, The problem of the Caucasian Sprachbund, in: From	Lin-
guistic	Areas	to	Areal	Linguistics, ed. Pieter Muysken, Amsterdam – Philadelphia: John Ben-
jamins, 2008, 25–93.
Chirikba 2016 = Vyacheslav Chirikba, From North to North West: How North-West Caucasian 
evolved from North Caucasian, Mother	Tongue:	 Journal	of	 the	Association	 for	 the	Study	of	
Language	in	Prehistory 21 (2016), 1–27.
Chumakina – Corbett 2015 = Marina Chumakina – Greville G. Corbett, Gender-number marking 
in Archi: small is complex, in: Understanding	and	Measuring	Morphological	Complexity, ed. 
Matthew Baerman – Greville G. Corbett – Dunstan Brown, Oxford: Oxford University Press, 
2015, 93–116.
Comrie 2008 = Bernard Comrie, Linguistic diversity in the Caucasus, Annual	Review	of	Anthropol-
ogy 37 (2008), 131–143.
Dahl 2004 = Östen Dahl, The	Growth	and	Maintenance	of	Linguistic	Complexity, Amsterdam – Phil-
adelphia: John Benjamins, 2004.
Daniel 2019 = Michael Daniel, Mehweb verb morphology, in: The	Mehweb	Language:	Essays	on	
Phonology,	Morphology,	and	Syntax, ed. Michael Daniel – Nina Dobrushina – Dmitry Ganen-
kov, Berlin: Language Science Press, 2019, 73–115.
Dobrushina et al. 2020 = Nina Dobrushina – Michael Daniel – Yury Koryakov, Languages and 
sociolinguistics of the Caucasus, in: The	Oxford	Handbook	of	the	Languages	of	the	Caucasus, 
ed. Maria Polinsky, Oxford: Oxford University Press, 2020, 27–66.
Dryer 2013 = Matthew S. Dryer, Prefixing vs. suffixing in inflectional morphology, in: The	World	
Atlas	of	Language	Structures	Online, ed. Matthew S. Dryer – Martin Haspelmath, Leipzig: Max 
Planck Institute for Evolutionary Anthropology, 2013, http://wals.info/chapter/26.
Dumézil – Esenç 1975 = Georges Dumézil – Tevfik Esenç, Le	verbe	oubykh:	études	descriptives	et	
comparatives, Paris: Klincksieck, 1975.
Elšík 2000 = Viktor Elšík, Romani nominal paradigms: their structure, diversity and development, 
in: Grammatical	Relations	in	Romani:	the	Noun	Phrase, ed. Viktor Elšík – Yaron Matras, Am-
sterdam – Philadelphia: John Benjamins, 2000, 9–30.
Erschler 2018 = David Erschler, Suspended affixation as morpheme ellipsis: Evidence from Ossetic 
alternative questions, Glossa 3.1 (2018), 1–41.
Fenwick 2011 = Rohan [Rhona] S. H. Fenwick, A	Grammar	of	Ubykh, München: LINCOM Europa, 
2011.
Forker 2013 = Diana Forker, A	Grammar	of	Hinuq, Berlin – Boston: De Gruyter Mouton, 2013.
Forker 2020 = Diana Forker, A	Grammar	of	Sanzhi	Dargwa, Berlin: Language Science Press, 2020.
Fortescue 2017 = Michael Fortescue, What are the limits of polysynthesis?, in: The	Oxford	Hand-
book	of	Polysynthesis, ed. Michael Fortescue – Marianne Mithun – Nicholas Evans, Oxford: 
Oxford University Press, 2017, 115–134.
Ganenkov 2010 = Dmitry Ganenkov, Topological relations in Daghestanian languages, Linguistics	
Special	issue	on	Spatial	Case 48.5 (2010), 1011–1041.
Gurevich 2003 = Olga Gurevich, The status of morpheme in Georgian verbal morphology, Pro-
ceedings	of	the	29th	Annual	Meeting	of	the	Berkeley	Linguistics	Society, Berkeley: University 
of California, Berkeley Linguistics Society, 2003, 161–172.
Gurevich 2006 = Olga Gurevich, Constructional	Morphology:	the	Georgian	Version, PhD disserta-
tion, University of California, Berkeley, 2006, https://escholarship.org/uc/item/1b93p0xs. 
Harris 2002a = Alice C. Harris, On the origins of circumfixes in Kartvelian, in: Philologie,	Typolo-
gie	und	Sprachstruktur:	Festschrift	für	Winfried	Boeder	zum	65.	Geburtstag, ed. Wolfram Bub-
litz – Manfred von Roncador – Heinz Vater, Frankfurt am Main: Peter Lang, 2002, 305–322.
33Jezikoslovni zapiski 28  2022  1
Harris 2002b = Alice C. Harris, Endoclitics	and	the	Origins	of	Udi	Morphosyntax, Oxford: Oxford 
University Press, 2002.
Harris 2009 = Alice C. Harris, Exuberant exponence in Batsbi, Natural	Language	and	Linguistic	
Theory 27 (2009), 267–303.
Harris 2017 = Alice C. Harris, Multiple	Exponence, Oxford: Oxford University Press, 2017.
Haspelmath 1993 = Martin Haspelmath,	A	Grammar	of	Lezgian, Berlin – New York: Mouton de 
Gruyter, 1993.
Haspelmath 2008 = Martin Haspelmath, An empirical test of the Agglutination Hypothesis, in: 
Universals	 of	 Language	 Today, ed. Sergio Scalise – Elisabetta Magni – Antonietta Bisetto, 
Dordrecht: Springer, 2008, 13–29.
Hewitt 1979a = B. George Hewitt, The relative clause in Abkhaz (Abžui dialect), Lingua 47 (1979), 
151–188.
Hewitt 1979b = B. George Hewitt, The relative clause in Adyghe (Temirgoi dialect), Annual of 
Ibero-Caucasian	Linguistics 6 (1979), 134–162.
Jespersen 1917 = Otto Jespersen,	Negation	in	English	and	Other	Languages, København: Bianco 
Lunos Bogtrykkeri, 1917.
Khvtisiashvili 2013 = Tamrika Khvtisiashvili,	Principal	Aspects	of	Xinaliq	Phonology	and	Mor-
phosyntax, PhD thesis, University of Utah, 2013, https://eric.ed.gov/?id=ED560426. 
Kibrik 1991 = Alexander E. Kibrik, Organising principles for nominal paradigms in Dagestanian 
languages: comparative and typological observations, in: Paradigms:	 the	Economy	of	Inflec-
tion, ed. Frans Plank, Berlin – New York: Mouton de Gruyter, 1991, 255–274.
Kibrik 1995 = Alexander E. Kibrik, Direct-oblique agreement of attributes in Daghestanian, in: 
Double	Case:	Agreement	by	Suffixaufname, ed. Frans Plank, New York – Oxford: Oxford 
University Press, 1995, 216–229.
Kibrik 2003 = Alexander E. Kibrik, Nominal inflection galore: Daghestanian, with side glances at 
Europe and the world, in: Noun	Phrase	Structure	in	the	Languages	of	Europe, ed. Frans Plank, 
Berlin – New York: Mouton de Gruyter, 2003, 37–112.
Korotkova – Lander 2010 = Natalia Korotkova – Yury Lander, Deriving suffix ordering in poly-
synthesis: evidence from Adyghe, Morphology 20 (2010), 299–319.
Lacroix 2009 = René Lacroix,	Description	du	dialecte	laze	d’Arhavi	(caucasique	du	sud,	Turquie):	
grammaire	et	textes, thèse du doctorat, Université Lumière de Lyon 2, 2009, http://theses.univ-
lyon2.fr/documents/lyon2/2009/lacroix_r.
Lander 2016 = Yury Lander, Word formation in Adyghe, in: Word-Formation.	An	International	
Handbook	of	the	Languages	of	Europe	5, ed. Peter O. Müller – Ingeborg Ohnheiser – Susan 
Olsen – Franz Rainer, Berlin: Mouton de Gruyter, 2016, 3508–3526.
Lander 2017 = Yury Lander, Nominal complex in West Circassian: between morphology and syn-
tax, Studies	in	Language 41.1 (2017), 76–98.
Lander – Daniel 2019 = Yury Lander – Michael Daniel, West Caucasian relative pronouns as re-
sumptives, Linguistics 57.6 (2019), 1239–1279.
Lander – Letuchiy 2010 = Yury Lander – Alexander Letuchiy, Kinds of recursion in Adyghe 
morphology, in: Recursion	in	Human	Language, ed. Harry van der Hulst, Berlin – New York: 
Mouton de Gruyter, 2010, 263–284.
Lander – Nichols 2020 = Yury Lander – Johanna Nichols, Head/dependent marking, in: 
Oxford	 Research	 Encyclopedia	 of	 Linguistics, 2020, https://doi.org/10.1093/acrefo
re/9780199384655.013.523. 
Lander – Testelets 2017 = Yury Lander – Yakov Testelets, Adyghe, in: The	Oxford	Handbook	
of	Polysynthesis, ed. Michael Fortescue – Marianne Mithun – Nicholas Evans, Oxford: Oxford 
University Press, 2017, 948–970.
Lomize 2011 = Grigorij Lomize, Fieldwork report on locative expressions in Besleney Kabardian, 
2011 (in Russian).
34 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
Maisak 2014 = Timur A. Maisak, Preverbs in Aghul: an elaborate system of locative prefixation, 
handout from the workshop East Caucasian preverbs and the compounding-derivation-inflec-
tion continuum, Pavia, September 2014.
Maisak 2021 = Timur A. Maisak, Endoclitics in Andi, Folia	Linguistica 55.1 (2021), 1–34.
Mattissen 2004 = Johanna Mattissen, A structural typology of polysynthesis, Word 55.2 (2004), 
189–216.
Mattissen 2006 = Johanna Mattissen, The ontology and diachrony of polysynthesis, in: Advances 
in	the	Theory	of	the	Lexicon, ed. Dieter Wunderlich, Berlin – New York: Mouton de Gruyter, 
2006, 287–353.
Mattissen 2017 = Johanna Mattissen, Sub-types of polysynthesis, in: The	Oxford	Handbook	 of	
Polysynthesis, ed. Michael Fortescue – Marianne Mithun – Nicholas Evans, Oxford: Oxford 
University Press, 2017, 70–98.
Mithun 1988 = Marianne Mithun, System-defining structural properties in polysynthetic languag-
es, Zeitschrift	 für	Phonetik,	 Sprachwissenschaft	 und	Kommunikationsforschung 41.4 (1988), 
442–452.
Mithun 1997 = Marianne Mithun, Lexical affixes and morphological typology, in: Essays	on	Lan-
guage	Function	and	Language	Type, ed. John Haiman – Joan Bybee – Sandra Thompson, Am-
sterdam – Philadelphia: John Benjamins, 1997, 357–372.
Nichols 1986 = Johanna Nichols, Head-marking and dependent-marking grammar, Language 62.1 
(1986), 56–119.
Nichols 1992 = Johanna Nichols, Linguistic	Diversity	in	Space	and	Time, Chicago – London: The 
University of Chicago Press, 1992.
Nichols 1994 = Johanna Nichols, Chechen, in: The	Indigenous	Languages	of	the	Caucasus 4, ed. 
Rieks Smeets, New York: Delmar, 1994, 2–78.
Nichols 1997 = Johanna Nichols, Modelling ancient population structure and movement in linguis-
tics, Annual	Review	of	Anthropology 26 (1997), 359–384.
Nichols 2011 = Johanna Nichols, Ingush	Grammar, Berkeley – Los Angeles: The University of 
California Press, 2011, https://escholarship.org/uc/item/3nn7z6w5. 
Nichols – Bickel 2013 = Johanna Nichols – Balthasar Bickel, Locus of marking: whole-language 
typology, in: The	World	Atlas	of	Language	Structures	Online, ed. Matthew S. Dryer – Martin 
Haspelmath, Leipzig: Max Planck Institute for Evolutionary Anthropology, 2013, http://wals.
info/chapter/25.
Nikolaev – Starostin 1994 = Sergei Nikolaev – Sergei Starostin, A	North	Caucasian	Etymological	
Dictionary, Moscow: Asterisk, 1994.
Plungian 2001 = Vladimir A. Plungian, Agglutination and flexion, in: Language	Typology	and	Lan-
guage	Universals:	An	International	Handbook 1, ed. Martin Haspelmath – Ekkehard König – 
Wulf Oesterreicher – Wolfgang Raible, Berlin – New York: Mouton de Gruyter, 2001, 669–678.
Polinsky 2020 = Maria Polinsky (ed.), The	Oxford	Handbook	of	the	Languages	of	the	Caucasus, 
Oxford: Oxford University Press, 2020.
Serdobolskaya 2011 = Natalia Serdobolskaya, Grammaticalization patterns of the Adyghe instru-
mental case, in: Languages	and	Cultures	in	the	Caucasus, ed. Vittorio Springfield Tomelleri – 
Manana Topadze – Anna Lukianowicz, München – Berlin: Otto Sagner, 2011, 515–539.
Smeets 1997 = Rieks Smeets, Suffixal marking of plural in Ubykh verb forms, in: Proceedings	of	
the	Conference	on	Northwest	Caucasian	Linguistics,	10–12	October	1994, ed. Sumru A. Özsoy, 
Oslo: Novus forlag – Instituttet for sammenlignende kulturforskning, 1997, 37–62.
Spruit 1986 = Arie Spruit,	Abkhaz	Studies, doctoral dissertation, Universiteit Leiden, 1986.
Stump 2016 = Gregory T. Stump, Inflectional	Paradigms.	Content	and	Form	at	the	Syntax-Mor-
phology	Interface, Cambridge: Cambridge University Press, 2016.
Testelets 2020 = Yakov G. Testelets, Kartvelian (South Caucasian) languages, in: The	Oxford	
Handbook	of	the	Languages	of	the	Caucasus, ed. Maria Polinsky, Oxford: Oxford University 
Press, 2020, 491–528.
Trubetzkoy 1926 = Nikolai S. Trubetzkoy, Studien auf dem Gebiete der vergleichenden Lautlehre 
der nordkaukasischen Sprachen I, Caucasica 3 (1926), 7–36.
35Jezikoslovni zapiski 28  2022  1
Tuite 1997 = Kevin Tuite,	Svan, München – Newcastle: LINCOM Europa, 1997.
Tuite 1999 = Kevin Tuite, The myth of the Caucasian Sprachbund: the case of ergativity, Lingua 
108 (1999), 1–26.
Tuite 2008 = Kevin Tuite, The rise and fall and revival of the Ibero-Caucasian hypothesis, Histo-
riographia	Linguistica 34.2–3 (2008), 23–82.
van Gelderen 2008 = Elly van Gelderen, Negative cycles, Linguistic	Typology 12.1 (2008), 195–
243.
Vogt 1971 = Hans Vogt, Grammaire	de	la	langue	géorgienne, Oslo: Universitetsforlaget, 1971.
Yu 2007 = Alan C. L. Yu, A	Natural	History	of	Infixation, Oxford: Oxford University Press, 2007.
Аркадьев 2014 = П. М. Аркадьев, О некоторых особенностях склонения в адыгских языках, 
в: Язык.	Константы.	Переменные:	памяти	Александра	Евгеньевича	Кибрика, отв. ред. 
В. А. Плунгян, Санкт-Петербург: «Алетейя», 2014, 552–563.
 [P. M. Arkad’ev, O nekotoryh osobennostjah sklonenija v adygskih jazykah, v: Jazyk.	
Konstanty.	 Peremennye:	 pamjati	 Aleksandra	Evgen’eviča	Kibrika, otv. red. V. A. Plungjan, 
Sankt-Peterburg: «Aletejja», 2014, 552–563.]
Аркадьев – Ландер 2020 = П. М. Аркадьев – Ю. А. Ландер, Амбификсы и другие звери, в: 
ВАПросы	 языкознания:	мегасборник	 наностатей	 к	юбилею	В.	А.	Плунгяна, ред. А. А. 
Кибрик и др., Москва: «Буки-Веди», 2020, 35–42.
 [P. M. Arkad’ev – Ju. A. Lander, Ambifiksy i drugie zveri, v: VAProsy	jazykoznanija:	mega-
sbornik	nanostatej	k	jubileju	V.	A.	Plungjana, red. A. A. Kibrik i dr., Moskva: «Buki-Vedi», 
2020, 35–42.]
Беляев 2014 = О. И. Беляев, Осетинский язык как язык с двухпадежной системой: групповая 
флексия и другие парадоксы падежного маркирования, Вопросы	 языкознания 2014, 6, 
31–65.
 [O. I. Beljaev, Osetinskij jazyk kak jazyk s dvuhpadežnoj sistemoj: gruppovaja fleksija i drugie 
paradoksy padežnogo markirovanija, Voprosy	jazykoznanija 2014, 6, 31–65.]
Кибрик 1977 = А. Е. Кибрик, Опыт структурного описания арчинского языка II:	таксономи-
ческая	грамматика, Москва: Изд-во Московского университета, 1977.
 [A. E. Kibrik, Opyt	strukturnogo	opisanija	arčinskogo	jazyka II: taksonomičeskaja	grammati-
ka, Moskva: Izd-vo Moskovskogo universiteta, 1977.]
Кибрик 2001 = А. Е. Кибрик (ред.), Багвалинский	 язык.	 Грамматика.	 Тексты.	 Словарь, 
Москва: «Наследие», 2001.
 [A. E. Kibrik (red.), Bagvalinskij	 jazyk.	Grammatika.	 Teksty.	 Slovar’, Moskva: «Nasledie», 
2001.]
Клычев 1995 = Р. Н. Клычев, Словарь	сочетаемости	локальных	превербов	с	суффиксоидами	
и	 глагольными	 корнями	 в	 абазинском	 языке, Черкесск: Карачаево-черкесское книжное 
издательство, 1995.
 [R. N. Klyčev, Slovar’	sočetaemosti	lokal’nyh	preverbov	s	suffiksoidami	i	glagol’nymi	kornjami	
v	abazinskom	jazyke, Čerkessk: Karačaevo-čerkesskoe knižnoe izdatel’stvo, 1995.]
Комри et al. 2015 = Б. Комри – М. Ш. Халилов – З. М. Халилова, Грамматика	бежтинского	
языка	I:	фонетика,	морфология, Лейпциг: Институт эволюционной антропологии им. 
Макса Планка – Махачкала: Институт языка, литературы и истории им. Г. Цадасы, 
2015.
 [B. Comrie – M. Š. Halilov – Z. M. Halilova, Grammatika	bežtinskogo	jazyka I: fonetika,	mor-
fologija, Lejpcig: Institut èvoljucionnoj antropologii im. Maksa Planka – Mahačkala: Institut 
jazyka, literatury i istorii im. G. Cadasy, 2015.]
Кумахов 1964 = М. А. Кумахов, Морфология	 адыгских	 языков:	 синхронно-диахронная	
характеристика I:	введение,	структура	слова,	словообразование	частей	речи, Нальчик: 
Кабардино-балкарское книжное издательство, 1964.
 [M. A. Kumahov, Morfologija	adygskih	jazykov:	sinhronno-diahronnaja	harakteristika I: vve-
denie,	 struktura	 slova,	 slovoobrazovanie	 častej	 reči, Nal’čik: Kabardino-balkarskoe knižnoe 
izdatel’stvo, 1964.]
36 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
Кумахов 1974 = М. А. Кумахов, К проблеме аблаута в абхазо-адыгских языках, Ежегодник	
иберийско-кавказского	языкознания	1 (1974), 80–90.
 [M. A. Kumahov, K probleme ablauta v abhazo-adygskih jazykah, Ežegodnik	iberijsko-kavkaz-
skogo	jazykoznanija 1 (1974), 80–90.]
Ландер 2012 = Ю. А. Ландер, Релятивизация	 в	 полисинтетическом	 языке:	 адыгейские	
относительные	конструкции	в	типологической	перспективе, диссертация ... кандидата 
филологических наук, Москва, РГГУ, 2012.
 [Ju. A. Lander, Reljativizacija	v	polisintetičeskom	jazyke:	adygejskie	otnositel’nye	konstrukcii	v	
tipologičeskoj	perspektive, dissertacija ... kandidata filologičeskih nauk, Moskva, RGGU, 2012.]
Ломтатидзе 1983 = К. В. Ломтатидзе, Основные типы локальных превербов в абхазском и 
абазинском языках, в: Система	превербов	и	послелогов	в	иберийско-кавказских	языках, 
отв. ред. Н. Т. Табулова – Р. Х. Темирова, Черкесск: Карачаево-черкесский НИИ истории, 
филологии и экономики, 1983, 10–13.
 [K. V. Lomtatidze, Osnovnye tipy lokal’nyh preverbov v abhazskom i abazinskom jazy kah, v: 
Sistema	preverbov	i	poslelogov	v	iberijsko-kavkazskih	jazykah, otv. red. N. T. Tabulova – R. H. 
Temirova, Čerkessk: Karačaevo-Čerkesskij NII istorii, filologii i èkonomiki, 1983, 10–13.]
Лютикова 2017 = Е. А. Лютикова, Падежная морфология, синтаксические категории и 
проблема классификации падежей, Acta	linguistica	Petropolitana	13.1 (2017), 650–679.
 [E. A. Ljutikova, Padežnaja morfologija, sintaksičeskie kategorii i problema klassifikacii pade-
žej, Acta	linguistica	Petropolitana 13.1 (2017), 650–679.]
Махмудова 2001 = С. М. Махмудова, Морфология	 рутульского	 языка, Москва: Институт 
языкознания РАН, 2001.
 [S. M. Mahmudova, Morfologija	 rutul’skogo	 jazyka, Moskva: Institut jazykoznanija RAN, 
2001.]
НКъ 2007 = Нарт	къэбархэр	=	Нартские	сказания, Майкоп: Качество, 2007.
 [Nart	qebarxer	=	Nartskie	skazanija, Majkop: Kačestvo, 2007.]
Пазов 2019 = С. У. Пазов, Усилительно-подтвердительная частица гь(ы)- и особенности её 
употребления в структуре отрицательного глагола в абазинском языке, в: Кавказская	фи-
лология:	история	и	перспективы:	к	90-летию	Мухадина	Абубекировича	Кумахова:	сбор-
ник	научных	статей, отв. ред. Б. Ч. Бижоев, Нальчик: Институт гуманитарных исследо-
ваний, 2019, 219–225.
 [S. U. Pazov, Usilitel’no-podtverditel’naja častica g’(y)-	i osobennosti eё upotreblenija v struk-
ture otricatel’nogo glagola v abazinskom jazyke, v: Kavkazskaja	filologija:	istorija	i	perspek-
tivy:	k	90-letiju	Muhadina	Abubekiroviča	Kumahova:	sbornik	naučnyh	statej, otv. red. B. Č. 
Bižoev, Nal’čik: Institut gumanitarnyh issledovanij, 2019, 219–225.]
Рогава – Керашева 1966 = Г. В. Рогава – З. И. Керашева,	Грамматика	адыгейского	языка, 
Краснодар – Майкоп: Краснодарское книжное издательство, 1966.
 [G. V. Rogava – Z. I. Keraševa, Grammatika	adygejskogo	jazyka, Krasnodar – Majkop: Kras-
nodarskoe knižnoe izdatel’stvo, 1966.]
Сулайбанов – Сумбатова 2022 = Г. Р. Сулайбанов – Н. Р. Сумбатова, Об одном типоло-
гическом раритете: каузатив в цугнинском говоре даргинского языка, Вопросы	языко-
знания	2022, 3, 109–131.
 [G. R. Sulajbanov – N. R. Sumbatova, Ob odnom tipolo gičeskom raritete: kauzativ v cugnin-
skom govore darginskogo jazyka, Voprosy	jazyko	znanija 2022, 3, 109–131.]
Сумбатова 2009 = Н. Р. Сумбатова, Коммуникативная структура адыгейского предложения: 
перспектива и фокус, в: Аспекты	полисинтетизма:	очерки	по	грамматике	адыгейского	
языка, отв. ред. Я. Г. Тестелец, Москва: Изд-во РГГУ, 2009, 559–611.
 [N. R. Sumbatova, Kommunikativnaja struktura adygejskogo predloženija: perspektiva i fokus, 
v: Aspekty	polisintetizma:	očerki	po	grammatike	adygejskogo	jazyka, otv. red. Ja. G. Testelec, 
Moskva: Izd-vo RGGU, 2009, 559–611.]
Тестелец 2019 = Я. Г. Тестелец, Грамматика	косвенности	в	дагестанских	языках, презента-
ция на международной конференции «Кавказские языки: типология и диахрония» памяти 
М. Е. Алексеева, Москва, 24 октября 2019 г.
37Jezikoslovni zapiski 28  2022  1
 [Ja. G. Testelec, Grammatika	kosvennosti	v	dagestanskih	jazykah, prezentacija na meždunarod-
noj konferencii «Kavkazskie jazyki: tipologija i diahronija» pamjati M. E. Alekseeva, Moskva, 
24 oktjabrja 2019 g.]
povzetek
Morfologija kavkaških jezikov: tipološki pregled
Kavkaz je področje, ki ga odlikuje največja jezikovna raznolikost v zahodnem delu 
Evrazije: na dokaj majhnem območju živijo govorci približno petdesetih jezikov, ki 
jih uvrščamo v pet jezikovnih družin. Na Kavkazu so avtohtone tri jezikovne družine: 
vzhodnokavkaška (naško-dagestanski jeziki), zahodnokavkaška (abhaško-adigejski je-
ziki) in južnokavkaška (kartvelski jeziki). Za kavkaške jezike velja ne samo to, da imajo 
vrsto redkih in enkratnih značilnosti, temveč tudi to, da so si med seboj občutno različni; 
med seboj se razlikujejo tako jezikovne družine kot tudi predstavniki posameznih dru-
žin. To posebej velja za oblikoslovje, ki je v tem kratkem tipološkem pregledu še zlasti 
izpostavljeno.
Bistvene razlike med kavkaškimi jeziki se kažejo že pri osnovnih tipoloških la-
stnostih. Poleg jezikov, za katere je značilna sufiksacija (avarski in lezginski jezik), so 
med kavkaškimi jeziki tudi tisti, ki pogosteje uporabljajo prefiksacijo (agulski, cezninski in 
posebej kartvelski ter zahodnokavkaški).
Če se v večjem delu vzhodnokavkaških jezikov skladenjska razmerja izražajo 
prek skladenjsko odvisnih enot znotraj klavze in samostalniške zveze (dependant-mar-
king language), se v abhaščini in abazinščini ta razmerja izražajo v skladenjskem jedru 
(head-marking language). V kartvelskih jezikih in adigejščini pa se skladenjska razmerja iz-
ražajo dvosmerno, tj. prek odvisnega ali jedrnega dela (double-marking language).
Podobne razlikovalne težnje lahko zaznamo pri izražanju prostorskih razmerij: v enem 
delu vzhodnokavkaških jezikov se prostorska razmerja izražajo predvsem s sistemom imen-
skih besed, v zahodnokavkaških jezikih pa s pomočjo glagola. V kartvelskih in nekaterih 
vzhodnokavkaških jezikih se prostor označuje tako z imenskimi kot z glagolskimi be-
sedami.
Navsezadnje pa je kavkaške jezike težko opredeliti s terminologijo tradicionalne-
ga razlikovanja med fleksivnimi in aglutinacijskimi jeziki. Za zahodnokavkaške jezike 
je značilna pretežno aglutinacijska morfologija, v kartvelskih in naško-dagestanskih je-
zikih pa obstajata oba morfološka principa, kar nakazuje različne stopnje razvoja mor-
foloških sistemov.
V kavkaških jezikih so predstavljeni praktično vsi tipi neobičajnih morfoloških 
kazalcev: t. i. ambifiksi oz. pone, ki imajo lahko pod vplivom različnih pogojev položaj 
predpone ali pripone, npr. negacijski označevalnik v abhaščini, abazinščini in ubišči-
ni, cirkumfiksi, ki so bistvena poteza kartvelskih jezikov, infiksi v vzhodnokavkaških 
jezikih, segmentne in nadsegmentne premene in celo transfiksi, ki so v buduškem jezi-
ku nastali kot posledica fonološkega vzajemnega vpliva med koreni in medponami oz. 
priponami.
Nič manj raznoliki niso tipi odnosov med morfološkimi pomeni in sredstvi njiho-
vega izražanja v kavkaških jezikih. Najzanimivejši tip tovrstnih odnosov je zmožnost t. i. 
večkratnega označevanja (multiple exponence) istega pomena znotraj ene besedne oblike.
V kavkaških jezikih se sistemsko pojavlja nekaj tipov že omenjenega večkra-
tnega označevanja, ki se razlikujejo po izvoru in strukturi. Tako je npr. v različnih 
naško-dagestanskih jezikih večkratno označevanje posebne vrste ujemanja nastalo kot 
po sledica gramatikalizacije zgradb s pomožnimi glagoli, pri katerih je vsak pomožni 
glagol prispeval svojo ujemalno pono. Drugi primer je možnost večkratnega označe-
vanja zanikanja v abazinskih osebnih glagolskih oblikah, ki je nastalo zaradi delovanja 
t. i. Jespersenovega cikla.
38 Peter Arkadiev  Morphology of the CauCasian languages: a typologiCal overview
Posebnost ubiškega jezika, ki zaznamuje celotni glagolski sistem, je večkratno 
označevanje števila absolutnega udeleženca, ki se dopolnjujoče uresničuje s katego-
rijami časa in povzročanja (kavzativa). Ločeno od pojava večkratnega označevanja je 
treba obravnavati deljeno izražanje različnih glagolskih kategorij v kartvelskih jezikih, 
pri katerem kazalniki niso specializirani afiksi, temveč kombinacije afiksov, kjer noben 
afiks nima svojega konkretnega pomena.
Neobičajna značilnost vzhodnokavkaških jezikov je t. i. dvodelno sklanjanje imen, 
ki se tvori v opoziciji imenovalniške osnove z osnovami stranskih sklonov. Razlike pri 
uresničitvi omenjene opozicije se v kavkaških jezikih kažejo v izraznih sredstvih (v neka-
terih jezikih gre lahko za več deset tovrstnih označevalnikov) in v različnih paradigmah. 
In ne samo to: v kar nekaj kavkaških jezikih opozicija med imenovalnikom in stranskimi 
skloni presega morfologijo in se kaže tudi pri ujemanju prilastkov in celo pri skladenjskih 
značilnostih celotnih samostalniških zvez. Zaradi tega lahko sklepamo, da v teh jezikih 
kot posebna slovnična kategorija deluje neimenovalniškost oz. pomen stranskih sklonov.
Najizrazitejša in najneobičajnejša poteza morfologije zahodnokavkaških jezikov, 
ki jih ločuje od vseh drugih kavkaških jezikov in jezikov Zahodne Evrazije, je polisin-
tetizem. Ta se kaže v naslednjih značilnostih: v t. i. polipersonalizmu (ko se v strukturo 
glagolske oblike vključijo do štirje afiksi, ustrezno povedkovim udeležencem); v velikem 
številu afiksov, ki imajo pogosto konkretne, večinoma prostorske pomene. Polisinteti-
zem se lahko kaže tudi v razširjenem združevanju samostalnikov s prilastki v morfološke 
imenske komplekse in tudi v izražanju praktično celotne skladenjske informacije klavze 
znotraj ene glagolske oblike.
Neobičajna značilnost polisintetizma v zahodnokavkaških jezikih je tudi prisotnost 
velikega števila aplikativnih predpon, ki uvajajo nepreme predmete in se pogosto uporab-
ljajo v leksikalnih in slovničnih procesih. V zahodnokavkaških jezikih zasledimo tudi 
združevanje zapovrstnega in izravnalnega načela pri razvrščanju morfemov; prav tako 
nenavadno je morfološko izražanje oziralnosti in posebnih zgradb z odprtimi vprašanji (v 
abhazijskem in abazinskem jeziku).