Both terms, šahatː ~ šavatː and šel, are of unknown origin. The relative antiquity of Vartashen šel is proven by Caucasian Albanian šel-ihesun 'to be good, apt for smth., to suit'. Schulze [Schulze 2001: 320] suspects an Azerbaijani loan in šel (< Azerbaijani gözäl 'beautiful, nice'; incorrectly quoted as güzel 'good, nice' by Schulze), which is unjustified.
Caucasian Albanian: ey [Gippert et al. 2008: IV-14]; perhaps an important archaism (thus [Gippert et al. 2008]). Cf. also another root in šel-ihesun 'to be good, apt for smth., to suit' ("it (the salt) is thenceforth good for nothing") [Gippert et al. 2008: IV-33] with the generic verb ih-esun 'to be, become' [Gippert et al. 2008: IV-19].
LEDb; Saadiev 1994: 418; Comrie & Khalilov 2010: 715. In [Comrie & Khalilov 2010] the variant ɣälä is also quoted. Looks like a loanword (the fluctuation ʁ ~ ɣ is particularly irregular, cf. also the Alyk Kryts form ɣala), but the source in unidentified.
The same in the other subdialects: Khudig, Arsug iǯe-d 'good' [Magometov 1970: 48, 92, 231 sentence 10, 234 sentence 27]. Cf. Magometov's examples: Khudig "A good man has killed the wolf", "Ibragim is a good boy", Arsug "Nowadays, roads are good".
Suleymanov 2003: 86; Shaumyan 1941: 143. Cf. the examples: "with a good figure" (said of a person) [Suleymanov 2003: 43], "Guseyni is my good friend" [Suleymanov 2003: 118], "I have a good (= interesting) book" [Suleymanov 2003: 178], "good horse" [Magometov 1970: 205 sentences 17-18].
The same in the other subdialects: Tsirkhe, Kurag iǯe-f 'good' [Magometov 1970: 49, 94, 122] (cf. Magometov's example for Tsirkhe: "I know that you are a good man").
Common Aghul notes:
Final -d, -t, -f, -r are adjectival suffixes (fossilized class exponents) [Magometov 1970: 92; Shaumyan 1941: 45].
Uslar 1979: 729, 1009; Dirr 1905: 175, 246. The form is actually from the Khanag subdialect; the proper Kondik term for 'good' is unknown. Polysemy: 'good / nice'. Cf. Uslar's and Dirr's examples: "Make this cart good (= repaired)!", "He is a good mullah", "good man", "good water". The same in the Khyuryuk subdialect: iǯˈi {ижжи} 'good' [Genko 2005: 74].
Distinct from the more specific Khanag term χuš 'good, nice, pleasant' [Uslar 1979: 955, 1009; Dirr 1905: 215, 246] ("a very nice man", "this food is pleasant for me", "nice weather", "welcome!"), borrowed from Azerbaijani xoš 'nice'.
Uslar 1896: 479, 638. Distinct from the more specific Gyune term iyˈer 'nice, beautiful; good' [Uslar 1896: 439] (according to Uslar's examples, 'nice' is the basic meaning).
The same in Literary Lezgi: qsa-n {хъсан} 'good' [Talibov & Gadzhiev 1966: 353; Gadzhiev 1950: 917; Haspelmath 1993: 504, 520]. Distinct from literary iyˈer {иер} 'nice, beautiful; good' [Talibov & Gadzhiev 1966: 134; Gadzhiev 1950: 917; Haspelmath 1993: 492] and qːenˈi {къени} 'kind, good (of person); fully functional, operational' [Talibov & Gadzhiev 1966: 186] (incorrectly glossed simply as 'good' in [Haspelmath 1993: 502, 520]).
Final -n in qisˈe-n ~ qsa-n is an adjectival suffix, see [Gaydarov et al. 2009: 139 f.]; historically it is a genitive exponent that modifies the substantive stem.
Proto-Lezgian:*yukːɨ-6
NCED: 643. Distribution: A highly unstable word; the Proto-Lezgian term cannot be reconstructed with certainty. The root *yukːɨ- [NCED: 643] denotes 'good' in Tsakhur and 'right (spatial)' in Shinaz Rutul (according to [NCED: 643], based on the unpublished MSU recordings), having been lost in the rest of Lezgian. We prefer *yukːɨ- as the Proto-Lezgian equivalent for 'good' for two reasons. First, it finds reliable North Caucasian cognates with the meaning 'good'. Second, the Caucasian Albanian form ey could originate from *yukːɨ-, if we assume the development *-kː- > -y-, as in modern Nidzh Udi (but not in Proto-Udi) [NCED: 125]. The latter argument is, however, weak, since Caucasian Albanian historical phonetics requires additional investigation.
In Rutul, the meaning 'good' is expressed by the root *HVχːʷV [NCED: 620], whose original meaning must have been something like 'kind, beautiful' (this follows from its cognates in other Lezgian languages: 'kind' in Archi, 'handsome, beautiful' in Aghul).
In Lezgi, 'good' is derived from the substantive root *qɨs(a) 'part, property' [LEDb: # 48] (cf. its meaning 'goods, possessions' in Archi, 'part' in Aghul).
In Archi, superseded by *pːVhˤV- ~ *hˤVpːV- 'big' (see notes on 'big').
In Vartashen Udi, šel 'good' apparently originates from the meaning 'good for smth., apt for smth.', as suggested by its Caucasian Albanian cognate.
In Aghul-Tabasaran, 'good' is expressed by *ʔičːV-, which is missing from the rest of Lezgian (some hypothetical external North Caucasian comparanda are proposed in [NCED: 248]).
Reconstruction shape: Tsakhur-Rutul correspondences are regular.
Semantics and structure: Primary stative verbal root 'to be good'.
Number:35
Word:green
Nidzh Udi:yäšil {йаьшил}-1
Gukasyan 1974: 126; Kibrik & Kodzasov 1990: 235; Comrie & Khalilov 2010: 675; Mobili 2010: 293. Cf. also the variant yešil in [Kibrik & Kodzasov 1990: 235]. Borrowed from Azerbaijani yašɨl 'green'.
Distinct from däy(i) {даьй(и)} 'unripe; raw; green' [Gukasyan 1974: 109; Mobili 2010: 97] (this is, however, not the basic Nidzh term for 'green', according to data in [Kibrik & Kodzasov 1990; Comrie & Khalilov 2010]).
Vartashen Udi:gog-in-1
Dzheiranishvili 1971: 220; Fähnrich 1999: 16; Schiefner 1863: 86; Starchevskiy 1891: 492, 493. Polysemy: 'green / blue' (although in [Dzheiranishvili 1971] glossed only as 'green'). Quoted as gög 'green / blue / sky' in [Schulze 2001: 279]. According to Yu. Lander's field records from the Zinobiani (Oktomberi) village on 2011, this is currently the default word for 'green'. Borrowed from Azerbaijani *gög (> Modern göy) 'green / blue / sky'.
Two other (inherited) terms for 'green' found in the sources are:
The exact difference between gog-in, dχi and däy is unclear.
Common Udi notes:
Not reconstructible.
Caucasian Albanian: Unattested. Cf. dai 'wet, marshy (vel sim.)' attested in dai-χunʸ 'marsh-meadow' [Gippert et al. 2008: II-21, IV-13], which is an etymological cognate of the modern Udi term discussed above. Further postulation of the meaning 'green' for dai on the basis of the word daizde 'gold' [Gippert et al. 2008: II-73, IV-13] is uncertain.
Kibrik & Kodzasov 1990: 235; Comrie & Khalilov 2010: 675. In [NCED: 333] the variant čʼuk-nu is also quoted. The suffix -nV is attested in some other nominal stems, but its synchronic semantics and function are unclear.
Kibrik et al. 1999: 871, 894; Ibragimov & Nurmamedov 2010: 402. In [Comrie & Khalilov 2010: 675], erroneously quoted as čɨwa-n {чыван}. Polysemy: 'green / wet'.
In [Comrie & Khalilov 2010: 675], the word kʼatɨpe-n {кIатыпен} is also quoted as a synonym - an error for the presumed kʼatɨlʸe-n {кIатылен}, cf. Tsakhur-Kum Tsakhur below.
Kibrik & Kodzasov 1990: 235. In [Comrie & Khalilov 2010: 675], the modern depharyngealized variant is quoted: čiwa-nʸ {чиванʹ}. Polysemy: 'green / wet'.
Common Tsakhur notes:
It is unclear how the Proto-Tsakhur term for 'green' should be reconstructed. Tsakhur-Kum kʼatlʸe-n and literary kʼatɨlʸe-n can represent an archaism, because the development 'wet' > 'green' is attested cross-linguistically, whereas 'green' > 'wet' seems less normal (in this case čiˤwa-n is the Proto-Tsakhur word for 'wet'). On the other hand, kʼatlʸe-n and kʼatɨlʸe-n are isolated and etymologically obscure forms; therefore, they could reflect a loanword of unknown origin.
Kibrik & Kodzasov 1990: 235; Suleymanov 2003: 53; Shaumyan 1941: 191. It must be noted that in both [Suleymanov 2003] and [Shaumyan 1941], this word is quoted as ʁaze-r {гъазер}.
The same in the Khanag subdialect: čirˈi 'green' [Uslar 1979: 968, 994; Dirr 1905: 218, 229]. The same in the Khyuryuk subdialect: čirˈi {чири} 'green' [Genko 2005: 181].
The same in the Khiv subdialect: čru {чру} 'green' [Genko 2005: 181]. The same in Literary Tabasaran: čurˈu {чуру} 'green' [Khanmagomedov & Shalbuzov 2001: 339].
The same in Literary Lezgi: qːacːˈu {къацу} 'green' [Talibov & Gadzhiev 1966: 183; Gadzhiev 1950: 245; Haspelmath 1993: 501, 520]. Distinct from literary kal {кал} 'unripe' [Talibov & Gadzhiev 1966: 148; Gadzhiev 1950: 245], borrowed from Azerbaijani kal 'unripe'.
The same in the Akhty dialect: Khlyut qːacːˈɨ 'green' [Kibrik & Kodzasov 1990: 235].
Derived from an old substantive, attested as Gyune qːaʒ [abs.] / qːacː-ˈu- [obl.] 'corn shoots' [Uslar 1896: 484] and Literary Lezgi qːaz [abs.] / qːacː-u- ~ qːacː-adi- [obl.] {къаз} 'green corn shoots' [Talibov & Gadzhiev 1966: 175; Gyulmagomedov 2004, 1: 384] (there is also a literary substantive qːaz [abs.] / qːaz-di- [obl.] 'green color; green dye; green yarn' [Talibov & Gadzhiev 1966: 175; Gyulmagomedov 2004, 1: 384], which possesses more generic semantics, but nevertheless looks like a late back-formation due to secondary -z- in the oblique stem).
Proto-Lezgian:*čʷiˤlä-3
NCED: 532. Distribution: An unstable word. Several equivalent (from the distributional point of view) candidates enter into competition. Out of these, we choose *čʷiˤlä- [NCED: 532] as the most likely Proto-Lezgian root for 'green'. It retains the basic meaning 'green' in West Lezgian (Tsakhur, Rutul). Note the polysemy 'green / wet' in Tsakhur (implying a recent shift 'green' > 'wet'). In Udi, *čʷiˤlä- has produced the basic term for 'fish' (see notes on 'fish'). In Budukh, *čʷiˤlä- has shifted to the meaning 'dim, muddy'; in Aghul, to the meaning 'blue'. An important advantage of *čʷiˤlä- over its competitors is that *čʷiˤlä- possesses good North Caucasian cognates with the meaning 'green'.
In archaic Udi, dχi 'green' and däy 'green / blue' are likely to have been derived from the verbal root *ʔäɬar- 'to be wet, soaked' [NCED: 277]. Note the additional shift of däy to 'unripe; raw' in Nidzh Udi.
In Kryts, 'green' is a suffixal derivation from the root *čʼeƛʼ [NCED: 333], whose original meaning is likely to have been 'a k. of reptile', see notes on 'fish'.
In Aghul and Lezgi (this could be either a Proto-East Lezgian feature or a late areal isogloss) 'green' is derived from the substantive *qːac: (oblique *qːacːɨ-) [NCED: 464], whose original Proto-Lezgian or at least Proto-East Lezgian meaning is unclear: in Aghul, it denotes 'green color; green dye; green yarn', but specifically 'corn shoots' in Lezgi, whereas external North Caucasian comparison suggests the initial meaning 'dirt'.
In Tabasaran, the old root was superseded with *čirV- [NCED: 554], which originally denoted 'a k. of light color' (cf. its meaning 'blond, red-haired' in Aghul and 'variegated', 'yellow', 'grey' and so on in other groups of the North Caucasian family).
Etymologically unclear forms include: Archi oˤlˈow ~ uˤlˈuw 'to be green' (cf. [NCED: 537]), Tsakhur-Kum Tsakhur kʼatlʸe-n 'green'.
In modern Udi and Budukh the old word is superseded with the Azerbaijani loanwords.
Chumakina et al. 2007; Kibrik & Kodzasov 1990: 42; Kibrik et al. 1977b: 214, 353; Comrie & Khalilov 2010: 178; Mikailov 1967: 205; Dirr 1908: 194, 205. Polysemy: 'hair / a single hair'.
Distinct from qʼamˈa-tːu 'woman's hair, long hair' [Kibrik et al. 1977b: 301; Chumakina et al. 2007] from qʼam 'forelock; mane' [Kibrik et al. 1977b: 301].
Dirr 1912: 145 (sub irdɨ); Ibragimov 1978: 114. Polysemy: 'hair / a single hair'.
In [Comrie & Khalilov 2010: 178], 'head hair' is glossed with two synonyms: čʼar, qʼulid {чIар, кьулид}. The second word is, in fact, the genitive form of qʼul 'head'.
Kibrik & Kodzasov 1990: 42. Polysemy: 'head hair / a single hair / (goat's) fur'. Distinct from Dyubek kuš 'long woman hair (sg., pl.)' [Kibrik & Kodzasov 1990: 42].
The same in the Khanag subdialect: čʼar 'hair' [Uslar 1979: 972, 990; Dirr 1905: 218, 225]. The same in the Khyuryuk subdialect: čʼar {чIар} 'hair' [Genko 2005: 185].
Kibrik & Kodzasov 1990: 42. Polysemy: 'head hair / a single hair / (goat's) fur'. Distinct from Kondik kuš 'long woman hair (sg., pl.)' [Kibrik & Kodzasov 1990: 42].
The same in the Khiv subdialect: čʼar {чIар} 'hair' [Genko 2005: 185]. The same in Literary Tabasaran: čʼar {чIар} 'hair' [Khanmagomedov & Shalbuzov 2001: 341].
Uslar 1896: 596. Polysemy: 'head hair / a single hair'.
The same in Literary Lezgi: čʼar {чIар} 'hair' [Talibov & Gadzhiev 1966: 377; Gadzhiev 1950: 98; Haspelmath 1993: 485, 520].
The same in the Akhty dialect: Khlyut čʼar with polysemy: 'hair / goats's fur (pl.)' [Kibrik & Kodzasov 1990: 41, 42].
Proto-Lezgian:*čʼaˤr2
NCED: 378. Distribution: This root is retained as the basic expression for 'head hair / a single hair' in Archi, on the one hand, and in all Nuclear Lezgian lects, on the other. Apparently in all the aforementioned languages, this word also denotes 'human body hair'. The polysemy 'human hair / goat's fur' is either a Proto-Nuclear Lezgian feature or a late areal isogloss. External North Caucasian comparison confirms *čʼaˤr as the Proto-Lezgian term for 'head hair'.
In Udi, *čʼaˤr was superseded with *pVpːV-, whose original meaning was something like 'soft, fluffy hair' [NCED: 865]. On the contrary, in Caucasian Albanian, 'hair' is expressed by the etymologically obscure stem ikuˤ.
Gukasyan 1974: 137; Kibrik & Kodzasov 1990: 26; Comrie & Khalilov 2010: 207; Mobili 2010: 168. It is not clear from the gloss in [Kibrik & Kodzasov 1990] whether this term denotes 'hand' only, or 'hand / arm'. Quoted only as 'hand' in [Comrie & Khalilov 2010: 207]. Glossed as ambivalent 'рука (= hand + arm)' in [Gukasyan 1974; Mobili 2010].
There also exists a separate term aˤm {аъм} 'arm; wing' [Gukasyan 1974: 58; Kibrik & Kodzasov 1990: 26; Comrie & Khalilov 2010: 204; Mobili 2010: 115].
Distinct from maχˤa {маъхаъ} with polysemy ‘handful / palm of hand’ [Gukasyan 1974: 172].
Gukasyan 1974: 137; Fähnrich 1999: 19; Dirr 1903: 15, 22, 28; Schiefner 1863: 84; Schulze 2001: 290; Starchevskiy 1891: 506. Polysemy: 'hand / arm'. Explicitly glossed only as 'hand' in [Schiefner 1863; Fähnrich 1999] (there is no separate term for 'arm' in these two sources), [Schulze 2001] (although the texts from [Bezhanov & Bezhanov 1902] demonstrate the generic meaning 'hand / arm'), and as ambivalent 'рука (= hand + arm)' in [Gukasyan 1974; Dirr 1903].
Distinct from aˤm, which is translated as 'wing; shoulder; side' in [Schiefner 1863: 76] and incorrectly as 'arm; pole, thills; door wing' in [Schulze 2001: 251] (in fact, the meaning 'arm' is unattested in [Bezhanov & Bezhanov 1902], the only anatomic meaning of aˤm attested in this source is 'shoulder').
Common Udi notes:
Common Udi *kul with an irregular paradigm in both dialects: kul [abs.] / k- [obl.] (the oblique stem is explained by the historically normal loss of -l- in the intervocalic position, [NCED: 130]). Cf. Nidzh köyi {коьйи}, Vartashen koy {кой} ‘sleeve’ [Gukasyan 1974: 136], historically derived from Proto-Udi *kul.
Caucasian Albanian: kul [abs.] / kul- ~kuy- [obl.] [Gippert et al. 2008: IV-42]; no expression for 'arm' is attested.
Chumakina et al. 2007; Kibrik & Kodzasov 1990: 27; Kibrik et al. 1977b: 259, 379; Comrie & Khalilov 2010: 207; Mikailov 1967: 185; Dirr 1908: 159, 220. Suppletive paradigm: kˈul- [abs., erg.] / kurˈa [loc.] / kˈur-ʁul [pl.]. Explicitly glossed as 'hand' in the aforementioned sources (with polysemy: 'hand / door handle').
Distinct from χol 'arm / branch (of tree)' [Chumakina et al. 2007; Dirr 1908: 190, 220]. The latter is explicitly glossed as 'arm' in the aforementioned sources. In turn, in [Kibrik et al. 1977b: 333, 379; Mikailov 1967: 200] χol is translated as ambivalent 'рука' (i.e. 'hand + arm'). It is interesting that in [Kibrik & Kodzasov 1990: 26] χol is explicitly quoted with the meaning 'arm + hand'.
Note also that in [Comrie & Khalilov 2010: 204, 207] it is kul which is proposed both for 'hand' and 'arm' (χol is not quoted in [Comrie & Khalilov 2010] at all).
Browsing through available texts clearly suggests that kul is the default term for 'hand' in Archi (e.g., "Don't touch it with your hand" [Kibrik et al. 1977b: 36] and so on). The only found context for 'arm' contains the word χol: "I have broken an arm (χol) and leg" [Dirr 1908: 126].
Kibrik & Kodzasov 1990: 26, 27. Actually, two terms enter into competition: χäb and kɨl, and Kibrik & Kodzasov's data is somewhat equivocal. Both χäb and kɨl are explicitly treated as generic terms for 'hand + arm' in [Kibrik & Kodzasov 1990: 26], but in the specific entry 'hand (Russian: кисть руки)' only Kryts χäb is quoted [Kibrik & Kodzasov 1990: 27]. Keeping in mind some peculiarities of this dictionary concerning the discussed anatomic terms (cf., e.g., notes on Archi), we provisionally prefer to assume χäb as the default Kryts word for 'hand'. In [Comrie & Khalilov 2010: 204, 207] only χäb is proposed both for 'arm' and 'hand' (kɨl is not quoted in [Comrie & Khalilov 2010] at all).
Authier 2009: 33, 55, 68, 82, 118, 120, etc. This is opposed to kil 'arm' [Authier 2009: 34, 59, 106, 119], although there are some contexts, where χab demonstrates the meaning 'arm': "to break an arm" [Authier 2009: 344].
According to [Kibrik & Kodzasov 1990: 26], another word for 'hand' is kɨl {кыл} with polysemy: 'arm / hand / sleeve / handle'. An example in [Meylanova 1984: 80] confirms the meaning 'hand' for kɨl: "to take one's hand", literally "getting hand in hand". It seems, however, that kɨl is a statistically less frequent expression for 'hand' than χab. Cf. several examples for χab 'hand' in [Meylanova 1984: 143] as well as in other sources, e.g., "The stick hurt my hand" [Talibov 2007: 76], "The human right hand is bigger/stronger/longer than the left one" [Talibov 2007: 97, 118, 185], "Mother pulled her child's hand" [Talibov 2007: 181], "A human looks at the face, an animal looks at the hand" [Talibov 2007: 276], "Hand of an arrogant man creates nothing" [Talibov 2007: 283].
In [Comrie & Khalilov 2010: 204, 207], χab is erroneously quoted as 'arm', whereas 'hand' is erroneously glossed by two "synonyms": kʼurkʼučʼ and kɨl. In fact kʼurkʼučʼ means 'brush, tassel' [Meylanova 1984: 99], and the underlying expression of Comrie & Khalilov's gloss "kʼurkʼučʼ, kɨl" was apparently a genitive construction 'tassel of arm' - a mechanical translation of Russian кисть руки 'tassel of arm', which is the designation of 'hand' in scientific Russian.
In [Comrie & Khalilov 2010: 207], 'hand' is quoted as patak {патак} - an error for pataχˤ {паIтах} or pataqˤ {паIтахъ} (cf. data from other dialects). Apparently the same term is reflected as partʸaqˤ {паIртяхъ} (sic!) 'paw' in [Ibragimov & Nurmamedov 2010: 285].
There also exists a specific term for 'forearm': guč [Kibrik et al. 1999: 875; Ibragimov & Nurmamedov 2010: 124].
Tsakhur-Kum Tsakhur: χɨlʸ [Kibrik & Kodzasov 1990: 26]. The specific term for 'hand' is pataχˤ, with the polysemy: 'paw / hand' [Kibrik & Kodzasov 1990: 27]. There also exists a specific term for 'arm': guč [Kibrik & Kodzasov 1990: 26].
The specific term for 'hand' is pataχˤ, with the polysemy: 'paw / hand' [Kibrik & Kodzasov 1990: 27] (in [Dirr 1913: 193] only with the meaning 'paw').
There also exists a specific term for 'arm': guč [Kibrik & Kodzasov 1990: 26]. According to [Dirr 1913], however, guč denotes 'forearm' [Dirr 1913: 149], whereas 'upper arm' is expressed as kʼɨr [Dirr 1913: 178] (the exact phonetic shape of the latter word is unknown).
The specific term for 'hand' is pataqˤ, with the polysemy: 'paw / hand' [Kibrik & Kodzasov 1990: 27]. There also exists a specific term for 'arm': guč [Kibrik & Kodzasov 1990: 26].
Distinct from perx {перхь} with polysemy 'hand / paw' [Dzhamalov & Semedov 2006: 207] (this is quoted in [Comrie & Khalilov 2010: 207] as the only term for 'hand').
Kibrik & Kodzasov 1990: 26; Magometov 1970: 23; Suleymanov 2003: 55; Shaumyan 1941: 191. In [Magometov 1970], transcribed as χːil - an important archaism. Polysemy: 'hand / arm'.
There also exists a more specific term for 'hand': gap [Kibrik & Kodzasov 1990: 27], which etymologically corresponds to the words for 'palm of hand' in some other Aghul dialects: Gequn gap, Fite gap [Kibrik & Kodzasov 1990: 28]. External Lezgian etymology could confirm that 'palm of hand' is the primary meaning of this word. In Burshag, the meaning 'palm of hand' is expressed analytically as kːalan yiqʼˤ, literally 'back (anatomic) of kːal' [Kibrik & Kodzasov 1990: 28].
Suleymanov 2003: 55; Shaumyan 1941: 191. Tpig ʁil is glossed by Suleymanov and Shaumyan as Russian "рука", which can mean 'hand', 'arm' or 'hand + arm'. The standard Aghul polysemy 'hand / arm' of ʁil is proved by the following examples: "to run one's hand over the horse" [Suleymanov 2003: 21], "I have five fingers on my hand" [Shaumyan 1941: 54], "There is a bracelet on woman's arm" [Shaumyan 1941: 35].
The same in the Khanag subdialect: χːil 'hand, arm' [Uslar 1979: 957, 1004; Dirr 1905: 216, 241]. The same in other subdialects: Khyuryuk, Kumi χːil {ххил} 'hand, arm' [Genko 2005: 168].
The same in the Khiv subdialect: χil {хил} 'hand, arm' [Genko 2005: 164]. The same in Literary Tabasaran: χil {хил} 'hand, arm' [Khanmagomedov & Shalbuzov 2001: 316].
Common Tabasaran notes:
Note the retention of tense fricative χː in the Northern subdialects.
Uslar 1896: 386, 630. Polysemy: 'hand / arm'. 'Palm' is expressed analytically as ʁilin čːin 'face of ʁil' [Uslar 1896: 386].
The same in Literary Lezgi: ʁil {гъил} 'hand / arm' [Talibov & Gadzhiev 1966: 94; Gadzhiev 1950: 735; Haspelmath 1993: 490, 520]. There are also two specific literary terms for 'hand': kːap with polysemy: 'hand / handful / chunk of bread' [Talibov & Gadzhiev 1966: 150] (inaccurately glossed as 'palm of the hand' in [Haspelmath 1993: 494]) and paronymous kːap-aš 'hand / handful' [Talibov & Gadzhiev 1966: 150]. 'Palm' is expressed analytically as kːapan yuqʼ 'centre of kːap' [Talibov & Gadzhiev 1966: 150].
The same in the Akhty dialect: Khlyut χil 'hand / arm' [Kibrik & Kodzasov 1990: 26]. There is also a specific Khlyut term kːapː-ˈač with polysemy: 'paw / hand' [Kibrik & Kodzasov 1990: 27].
Proto-Lezgian:*kʷil1
NCED: 706. Distribution: There are three main roots attested with the meaning 'hand' in Lezgian languages. The data can be summarized as follows:
'HAND'
Udi
Archi
Kryts
Budukh
Tsakhur
Rutul
Aghul
Tabasaran
Lezgi
*kʷil [NCED: 706]
hand/arm
hand
hand/arm
arm / hand / sleeve / handle
elbow (suffixed)
branch, cluster
twig, vine
branch, cluster
*χːɨl [NCED: 895]
arm
hand / arm
hand / arm
hand / arm
hand / arm
hand / arm
*mːaχː [NCED: 819]
handful / palm of hand
hand/arm
hand
armful
armful
handful
Although there can hardly be any doubt about *kʷil as the main Proto-Lezgian root for 'hand', the exact details are not entirely clear.
The easiest solution is to propose the lexical opposition 'hand' / 'arm' for Proto-Lezgian, despite the fact that such an opposition is atypical for the attested Lezgian lects and that the reconstructed syncretism 'foot / leg' (see notes on 'foot') could contradict the opposition 'hand' / 'arm'. The aforementioned roots can be reconstructed with the following meanings:
1) *kʷil 'hand';
2) *χːɨl 'arm';
3) *mːaχː (metathesized *χːamː in Proto-Nuclear Lezgian) 'handful'.
The root *χːɨl was lost in Udi, where *kʷil acquired the additional meaning 'arm' ('hand' > 'arm'); in Caucasian Albanian, *kʷil denotes 'hand', but no word for 'arm' is attested.
In the second outlier, Archi, the opposition *kʷil 'hand' / *χːɨl 'arm' is retained (the third root *mːaχː was lost).
Apparently, the opposition *kʷil 'hand' / *χːɨl 'arm' / *χːamː 'handful' was still retained in Proto-Nuclear Lezgian as well, but was subsequently eliminated in individual subgroups in different ways.
In Proto-South Lezgian (Kryts, Budukh), *kʷil acquired the additional meaning 'arm' ('hand' > 'arm'), whereas the old root for 'arm', *χːɨl, got lost - the same process as in the Udi branch. The root *χːamː in the meaning 'hand' seems to have been a relatively recent introduction ('handful' > 'hand'); we suppose that it is an areal isogloss, which is currently affecting Kryts and Budukh dialects. It is very likely that the new term *χːamː 'hand' tends to completely supersede the old term *kʷil in modern Kryts and Budukh. As a result, *χːamː acquires the additional meaning 'arm' ('hand' > 'arm'). Note also the development 'arm / hand' > 'sleeve' and 'arm / hand' > 'handle' in Budukh.
In Proto-West Lezgian (Tsakhur, Rutul) and Proto-East Lezgian (Aghul, Tabasaran, Lezgi), the old opposition *kʷil 'hand' / *χːɨl 'arm' was eliminated in favor of the latter root, i.e., *χːɨl started to denote 'hand / arm' ('arm' > 'hand'). It must be noted, however, that the tree topology and certain semantic shifts (for which see below) suggest that such an elimination was an independent development in Proto-West Lezgian and Proto-East Lezgian or, rather, an areal isogloss, which affected both protolanguages.
The old root *kʷil 'hand' survived in West Lezgian as the suffixed form 'elbow' (Rutul); such a semantic shift seems, however, somewhat surprising. The anatomic semantics of *kʷil was lost in Proto-East Lezgian, where this root acquired the meaning 'branch, cluster' (with the further shift > 'twig, vine' in Tabasaran).
External North Caucasian comparison confirms the Proto-Lezgian reconstructions *kʷil 'hand' [NCED: 706] and *mːaχː 'handful' [NCED: 819]. As for Proto-Lezgian *χːɨl 'arm', its Lak cognate denotes 'wing' [NCED: 896]; it seems that typologically the semantic shift between 'arm' and 'wing' can occur in both directions.
Finally, some peculiarities should be noted. In Shinaz Rutul and Koshan Aghul, 'hand' can be expressed by the root *kːap ~ *kːapː, whose Proto-Lezgian meaning is likely to have been 'palm of hand' [NCED: 298]. In Tsakhur and Rutul dialects, the words for 'paw' can acquire the specific meaning 'hand'. Udi köyi ~ koy 'sleeve' provides an additional instance for the development 'hand / arm' > 'sleeve'.
Common Udi *bul with an irregular paradigm in both dialects: bul [abs.] / b- [obl.] (the oblique stem is explained by the historically normal loss of -l- in the intervocalic position, [NCED: 130]).
Caucasian Albanian: bul [abs.] / bi(y)- [obl.] 'head / top / beginning / self' [Gippert et al. 2008: IV-12].
Chumakina et al. 2007; Kibrik & Kodzasov 1990: 10; Kibrik et al. 1977b: 257, 355; Comrie & Khalilov 2010: 185; Mikailov 1967: 184; Dirr 1908: 157, 206. 'Head of man'.
Distinct from oˤnt 'head (of woman or animal); head (of onion etc.); top (of mountain); chief, ringleader' [Chumakina et al. 2007; Kibrik & Kodzasov 1990: 10; Kibrik et al. 1977b: 290, 355; Dirr 1908: 171].
The old root for 'head' is retained in the adverb ʼil-lˈi-ʼ 'under one's head' [Chumakina et al. 2007; Kibrik et al. 1977b: 269] (-lˈi- is the frequent oblique stem marker [Kibrik et al. 1977a 2: 16], - is the sublative ending [Kibrik et al. 1977a 2: 52]).
Kibrik et al. 1999: 880; Ibragimov & Nurmamedov 2010: 185; Comrie & Khalilov 2010: 185. Borrowed from Azerbaijani källä 'head' (ultimately from Persian kalla 'head'). According to examples in [Kibrik et al. 1999], this is currently the default word for 'human head' in Mishlesh.
Additional synonyms include the inherited wukʼulʸ {вукIулʹ} [Kibrik et al. 1999: 889, 892; Ibragimov & Nurmamedov 2010: 107] and the borrowed baš 'head' [Kibrik et al. 1999: 870] (< Azerbaijani baš 'head').
Distinct from kalːe 'head (of large cattle)' [Kibrik & Kodzasov 1990: 10], borrowed from Azerbaijani källä 'head' (ultimately from Persian kalla 'head').
Borch-Khnov dialect: yuqʼul 'head' [Ibragimov 1978: 234, 239, 281]. For the Borch-Khnov dialect, an unclear word gɨʁˤ {гыIгъ} 'head' is also quoted in [Ibragimov 1978: 228].
All the dialectal forms - qʼul, huqʼul, wɨqʼɨl, yuqʼul - are etymologically related.
The same in the Khanag subdialect: kʼul with polysemy: 'head / ear (of cereals) / nipple, teat / hill, top (of mountain)' [Uslar 1979: 800, 992; Dirr 1905: 190, 226]. The same in the Khyuryuk subdialect: kʼul {кIул} 'head' [Genko 2005: 112].
The same in the Khiv subdialect: kʼul {кIул} with polysemy: 'head / ear (of cereals) / nipple, teat / top (of mountain)' [Genko 2005: 112]. The same in Literary Tabasaran: kʼul {кIул} with polysemy: 'head / chief / ear (of cereals) / top (of mountain)' [Khanmagomedov & Shalbuzov 2001: 214].
The same in Literary Lezgi: qʼil {кьил} with polysemy: 'head / chief / ear (of cereals)' [Talibov & Gadzhiev 1966: 203; Gadzhiev 1950: 144; Haspelmath 1993: 503, 521]. Distinct from the rude word kelːe {келле} 'head' [Talibov & Gadzhiev 1966: 155; Gadzhiev 1950: 144; Haspelmath 1993: 521], borrowed from Azerbaijani källä 'head' (ultimately from Persian kalla 'head').
The same in the Akhty dialect: Khlyut qʼil with polysemy: 'head / ear (of cereals)' [Kibrik & Kodzasov 1990: 10].
Proto-Lezgian:*woƛʼul1
NCED: 1041. Distribution: This stem is attested as the basic term in Caucasian Albanian-Udi, on the one hand, and in all Nuclear Lezgian languages, on the other. The locative form of *woƛʼul is also retained as the Archi adverb 'under one's head'.
In Archi, *woƛʼul was superseded by *kʷɨltʼ- ~ *kʼʷɨlt- [NCED: 695], which is a good candidate for the status of the Proto-Lezgian term for 'temple'. This word means 'temple' or 'cheek' in Nuclear Lezgian (thus 'temple' > 'cheek'); external North Caucasian comparison could also point to the meaning 'temple'.
Note some specific semantic shifts of *woƛʼul 'head', reflected as synchronic polysemy in individual lects: 'point, spike' (Nidzh Udi), 'beginning', 'button', 'north' (Vartashen Udi), 'top', 'self' (Caucasian Albanian), 'ear (of cereals)' (Kryts Proper, Northern Tabasaran, Lezgi), 'lid, cover' (Budukh), 'hill' (Gequn Aghul, Tabasaran), 'nipple, teat' (Tabasaran).
The inherited term tends to be superseded with Azerbaijani or Persian loanwords in some Tsakhur, Rutul and Lezgi dialects.
Reconstruction shape: Basic correspondences seem regular, except for the sporadic syncope of the first syllable in some languages.
Semantics and structure: Primary substantive root. Historically *woƛʼu-l with the common nominal l-suffix. If Udi ber 'pillow' does indeed originate from *woƛʼV-rV (thus in [NCED: 1041]), the suffix -l can be singled out on the Proto-Lezgian level.
Common Udi *i-bak-esun; derived from the root i plus the light verb -bak- 'to be(come)' [Schulze 2005: 561 f. (3.4.2.2 #10); Schulze-Fürhoff 1994: 474]. The Udi verb corresponds to Caucasian Albanian ih-esun 'to hear'. Both CA ih and Udi i go back to Proto-Caucasian Albanian-Udi *iχ(i)- with the shift χ > h > 0 in the intervocalic position before front vowels or before consonants. The Udi morpheme i is quoted as a separate word with the meaning 'hearing, ability to hear (Russian: слух)' in [Gukasyan 1974: 127; Mobili 2010: 152], although it is unclear whether this i can function as an independent item or if it has been extracted from the verb 'to hear' by Gukasyan.
Expressions for 'to hear' and 'to listen' are clearly opposed in modern Udi, as well as in Caucasian Albanian. Verbs for 'to hear' contain the old verbal root *ʔeɬɨ- (~ -ɬː-): Udi i-bak-sun, CA ih-esun, both forms regularly without pharyngealization. Expressions for 'to listen' are based on the word for 'ear': Udi umuχˤ / imuχ laχsun, lit. 'to put the ear' [Comrie & Khalilov 2010: 665; Gukasyan 1974: 166; Starchevskiy 1891: 488], CA ʕi-biqʼ-esun, lit. 'to take the ear', normally with pharyngealization (for the Caucasian Albanian secondary ʕi- instead of expected **ʕim- see notes on 'ear').
Caucasian Albanian: ih-esun [Gippert et al. 2008: IV-21], a cognate of the Udi term. Distinct from ʕi-biqʼ-esun 'to listen to / to obey / to follow, observe, endure, take on' [Gippert et al. 2008: IV-22], which is based on the secondary morpheme ʕi 'ear' (q.v.) plus the light verb -biqʼ- 'to seize' [Gippert et al. 2008: II-43, IV-10].
Distinct from oy ˈača- 'to listen; to obey', lit. 'to use the ear' [Chumakina et al. 2007; Kibrik & Kodzasov 1988: 173; Kibrik et al. 1977b: 287; Comrie & Khalilov 2010: 665] and ˈoɬːa- 'to be silent; to listen' [Chumakina et al. 2007; Kibrik et al. 1977b: 290].
Kibrik & Kodzasov 1988: 172. Missing from [Meylanova 1984] as a separate entry, but attested in examples, e.g., [Meylanova 1984: 132 sub tarp, 133 sub telfun, etc.].
Distinct from ibir qːusu- {ибир къусу} 'to listen', literally 'to put ear, to cover with ears' [Kibrik & Kodzasov 1988: 173; Meylanova 1984: 63; Comrie & Khalilov 2010: 665].
In [Comrie & Khalilov 2010: 664], 'to hear' is incorrectly glossed as ibira qːusu {ибира къусу} - an error for ibir qːusu 'to listen'.
Kibrik et al. 1999: 875, 899; Ibragimov & Nurmamedov 2010: 210; Comrie & Khalilov 2010: 664. Ablaut paradigm: qː=iːxʸ-e < *qː=iyxʸ-e [imperf.] / qː=ayx-ɨ [perf.] / qː=ayxʸ-es [fut.]. In [Kibrik et al. 1999: 875], the imperf. stem is incorrectly quoted as qː=ixʸ-e [imperf.] with short -i-; the correct form with -iː- can be found in examples in [Kibrik et al. 1999: 572, 692].
Distinct from kʼɨr gʸaqː- 'to listen; to obey', literally 'ear + to show' [Kibrik et al. 1999: 881; Ibragimov & Nurmamedov 2010: 127].
Dirr 1912: 174, 200; Makhmudova 2001: 149, 167, 251. Literally 'sound happens to X' with un 'sound' + the suppletive verb class=ik- / class=iš- / class=ruʔ- 'to become' (y= is the class 4 exponent). In [Comrie & Khalilov 2010: 664], erroneously quoted as ul y=iš- {ул йишин} (literally ul 'eye' + 'to become').
Distinct from un y=ɨχˤ- {ун йыхIыс} 'to hear about, find out (Russian: прослышать)' [Makhmudova 2001: 99], literally un 'sound' + the verb y=ɨχˤ- '?'.
Dzhamalov & Semedov 2006: 255, 394; Comrie & Khalilov 2010: 664. Literally un 'call, appeal' + the verb y=ɨχˤ- '?' (the exact syntactic construction is not documented).
Kibrik & Kodzasov 1988: 172. Literally un 'sound' + the verb ečʼʷ- '?' (the exact syntactic construction is not documented).
Distinct from q=as-ɨ- 'to listen' [Kibrik & Kodzasov 1988: 173].
Common Rutul notes:
All the dialects demonstrate an analytic construction with un 'sound' and an auxiliary verb. In Mukhad this verb is the common 'to become' ('sound happens to X'), whereas the meanings of Ixrek y=ɨχˤ- and Luchek ečʼʷ- are unknown. Ixrek y=ɨχˤ- formally coincides with the verb y=ɨχˤ- 'to strike, hit' (see notes on 'to kill'), where y= is a prefix with general semantics, but the semantics of hearing is strange in this case; on the other hand, Ixrek y=ɨχˤ- may consist of the class 4 exponent y= attached to the verb ɨχˤ- '?'.
Initial q= in the verbs for 'to listen' is a prefix with general semantics [Ibragimov 1978: 95; Alekseev 1994a: 227; Makhmudova 2001: 165]. Mukhad q=ac-u- and Ixrek q=äs-u-, Luchek q=as-ɨ- are obviously related (for the deaffricativization in Rutul dialects cf. [NCED: 138], although this problem still awaits more detailed research). The Mukhad and Ixrek data point to labialized *cʷ, thus the link between this Rutul root and Proto-Lezgian *ʔasV 'to be silent' seems unlikely, pace [NCED: 262].
Koshan Aghul:daχ xi-4
Kibrik & Kodzasov 1988: 172; Shaumyan 1941: 144. Literally daχ 'sound' + the verb xi- 'to become' [Kibrik & Kodzasov 1988: 62] (the exact syntactic construction is not documented).
Distinct from yirkʼʷ alaši- 'to listen', literally 'heart + to put' [Kibrik & Kodzasov 1988: 173].
Kibrik & Kodzasov 1988: 172. Literally un 'sound' + the verb xa- 'to become' [Kibrik & Kodzasov 1988: 62]. A second synonym for 'to hear', quoted in [Kibrik & Kodzasov 1988], is un baˤ-, literally 'sound' + 'to go' (q.v.). The exact syntactic constructions are not documented.
Distinct from yabur qixa- 'to listen', literally 'ear + to overtake' [Kibrik & Kodzasov 1988: 173].
Dirr 1907: 146, 184; Shaumyan 1941: 144. Literally un 'sound' + the verb xa- 'to become' [Dirr 1907: 119] (the exact syntactic construction is not documented).
Suleymanov 2003: 161; Shaumyan 1941: 144. Literally 'sound happens to X' with un 'sound' + the verb xa- 'to become' (cf. examples in [Shaumyan 1941: 116], [Magometov 1970: 201 sentence 2]). The same in the Tsirkhe and Kurag subdialects: un xa- 'to hear' [Shaumyan 1941: 144], [Magometov 1970: 211 sentence 42, 45].
Distinct from two Tpig expressions for 'to listen': ibur aqixa-, literally 'ear + to overtake' [Suleymanov 2003: 86], yirkʼʷ aliyana-, literally 'heart + to put' [Suleymanov 2003: 26].
Common Aghul notes:
Both of the Aghul analytical expressions for 'to hear' (daχ xi-, un xa-, literally 'sound happens to X') are innovations of areal origin.
The same in the Khanag subdialect: yik-ˈ 'to hear' [Uslar 1979: 751, 1006; Dirr 1905: 166, 242]. The same in other subdialects: Khyuryuk yik-ˈ {йибкув}, Kumi yik-ˈ {йикув}, Arkit ik-ˈ {ибкув} 'to hear' [Genko 2005: 73, 79].
Differently in the Chuvek subdialect: yix-ˈ {йибхьув} 'to hear' [Genko 2005: 79].
Distinct from the verbs for 'to listen': Khanag q-iw-iqː-ˈ 'to listen' [Uslar 1979: 809, 1006; Dirr 1905: 192, 242], Khyuryuk q-iw-iqː-ˈ {хъивикъкъув} 'to listen' [Genko 2005: 171]. These forms look like a compound of iw 'ear' q.v. and the verb iqː- '?' plus the spatial prefix q= 'behind', which modifies the whole expression. In Dyubek, 'to listen' is simply q=iqː-ˈ [Kibrik & Kodzasov 1988: 173] (i.e. the spatial prefix + the verb), which can be a compression of the aforementioned compound.
The same in the Khiv subdialect: yex-ˈ {ерхьуб} 'to hear' [Genko 2005: 64]. The same in Literary Tabasaran: yex-ˈ {ебхьуб} 'to hear' [Khanmagomedov & Shalbuzov 2001: 154].
Distinct from the verbs for 'to listen': Khiv q-eb-eq-ˈ {хъебехъуб} 'to listen' [Genko 2005: 170], Literary Tabasaran q-p-eq-ˈ {хъпехъуб} 'to hear' [Khanmagomedov & Shalbuzov 2001: 324], Kondik q-eb-eq-ˈ 'to listen' [Kibrik & Kodzasov 1988: 173] - apparently a compound of eb 'ear' + the verb eq- '?' plus the spatial prefix q= 'behind' (cf. similar expressions in Northern Tabasaran, although the used verbs differ).
More transparent are the analytic expressions for 'to listen', which literally mean 'to put the ear behind': Chara ib q=ib-ˈ 'to listen' [Genko 2005: 73] (ib 'ear' + the Chara verb ib-ˈ 'to put in' [Genko 2005: 73] + q= 'behind'); Literary Tabasaran ib q=ˈiw- {иб хъивуб} 'to hear' [Khanmagomedov & Shalbuzov 2001: 172] (ib 'ear' + the Literary Tabasaran verb ˈiw- {ивуб} 'to put in' [Khanmagomedov & Shalbuzov 2001: 172] + q= 'behind').
Common Tabasaran notes:
Two verbs with the meaning 'to hear' enter into competition here: yik- (Northern Tabasaran, except for the Chuvek subdialect, which is located on the border between two dialects, see the map in [Alekseev & Shikhalieva 2003: 12]) and yex- (Southern Tabasaran). The external Lezgian comparison speaks in favour of Southern yex- as Proto-Tabasaran 'to hear', rather than Northern yik-.
Uslar 1896: 364, 632. Literally 'sound happens to X' with wan 'sound; voice' + the suppletive verb že- / xa- 'to become' [Uslar 1896: 435]. Distinct from Gyune yab aɣˈal- 'to listen', literally 'to string the ear' [Uslar 1896: 328, 443].
Several similar expressions for 'to hear' are documented for Literary Lezgi: wan že- / xa- {ван хьун}, literally 'sound (wan) happens to X' [Gadzhiev 1950: 784; Haspelmath 1993: 510, 521], wan qːʷe- / atːa- {ван атун}, literally 'sound (wan) comes to X' [Gadzhiev 1950: 784], ses qːʷe- / atːa-, literally 'sound (ses) comes to X' [Talibov & Gadzhiev 1966: 291].
Distinct from the literary expressions for 'to listen': yab gu- {яб гун}, literally 'to give ear', and yab akali- {яб акалун}, literally 'to string/attach the ear' [Talibov & Gadzhiev 1966: 35, 399; Gadzhiev 1950: 784; Haspelmath 1993: 480, 522].
Proto-Lezgian:*ʔeɬɨ- ~ *ʔeɬːɨ-1
NCED: 411. Distribution: This stem is retained as the basic verb for 'to hear' in Caucasian Albanian-Udi, on the one hand, and in some Nuclear Lezgian lects, on the other. More precisely, *ʔeɬ(ː)ɨ- is to be reconstructed as the Proto-Nuclear Lezgian verb 'to hear': it means 'to hear' in South Lezgian (Kryts, Budukh), Tsakhur and Southern Tabasaran, but shifted to the meaning 'to keep silence' in Rutul. External North Caucasian comparison confirms the meaning 'to hear' for Proto-Lezgian *ʔeɬ(ː)ɨ-.
The second candidate is *ʔi(r)kɨ(r)- [NCED: 650], which means 'to hear' in Archi and, surprisingly, in Northern Tabasaran, but got lost in the rest of languages. The exact Proto-Lezgian meaning of this root cannot be established (it should be noted that some of its external North Caucasian cognates also demonstrate the meaning 'to hear'). The Tabasaran situation, when two main dialects possess different verbs for 'to hear' (*ʔeɬ(ː)ɨ- vs. *ʔi(r)kɨ(r)-) can only be explained as an independent semantic development *ʔi(r)kɨ(r)- '?' > 'to hear' in Archi and Northern Tabasaran.
In Rutul, Aghul, Lezgi, 'to hear' is expressed by analytic constructions 'sound happens to X' with different words for 'sound' and different auxiliary verbs. This is a late areal isogloss that affected the central part of the Lezgian territory.
Caucasian Albanian: huˤkʼ [Gippert et al. 2008: IV-27]. The alphabetical sign for a pharyngealized vowel uˤ apparently reflects the influence of the following ejective or the front vowel ü.
Kibrik et al. 1999: 879, 899; Ibragimov & Nurmamedov 2010: 181; Comrie & Khalilov 2010: 219. Polysemy: 'heart / stone (of fruit)' (the two meanings formally differ in pl. forms).
Distinct from baʁɨr 'heart (figurative)', baʁrɨ 'heart, breast (figurative), soul' [Ibragimov & Nurmamedov 2010: 66], borrowed from Azerbaijani baɣɨr 'liver (anatomic); breast, heart (figurative)'.
The same in the Khanag subdialect, but with retention of the irregular paradigm: yukʼ [abs.] / kʼ-a- [obl.] 'heart' [Uslar 1979: 756, 1005; Dirr 1905: 180, 241]. The same in the Khyuryuk subdialect: yukʼ [abs.] / kʼ-a- [obl.] {юкI} 'heart' [Genko 2005: 197].
The same in Literary Lezgi: rikʼ {рикI} 'heart' [Talibov & Gadzhiev 1966: 278; Gadzhiev 1950: 762; Haspelmath 1993: 505, 521].
The same in the Akhty dialect: Khlyut rikʼ 'heart' [Kibrik & Kodzasov 1990: 34].
Proto-Lezgian:*yirkʼʷ1
NCED: 678. Distribution: One of the most stable Lezgian roots, retained in its basic meaning in all Lezgian languages.
The following semantic shifts, reflected as synchronic polysemy in individual lects, can be noted: 'stone (of fruit)' (Tsakhur, Rutul), 'soul' (Rutul).
Dzheiranishvili 1971: 205, 247; Schiefner 1863: 105; Schulze 2001: 301; Starchevskiy 1891: 506. The word is written with tense qː as {му̇kы̇нах} in [Bezhanov & Bezhanov 1902: Lk. 1.69], although this Bežanovs' form was mistranscribed with plain q in [Schulze 2001: 301] (with further speculations about such a deglottalization).
The sources vary in this case. Modern dictionaries [Gukasyan 1974: 157; Mobili 2010: 190] quote qːˤancː {къаънцI} as the Vartashen term for 'horn', whereas sources of the 19th century show muqːˤa. Apparently there has been a lexical replacement over the course of the last century (note that [Dzheiranishvili 1971] reflects the archaic norm). Vartashen qːˤancː originates from Lezgian *qːˤa[n]cʼ(a) 'hook' [NCED: 462]. We treat both words as synonyms.
Common Udi notes:
Common Udi *muqːˤa, derived from muqːˤ 'deer' [Gukasyan 1974: 177; Mobili 2010: 217; Schiefner 1863: 104]. Cf. [Schulze 2001: 301] for some dubious etymological solutions for *muqːˤa (in particular, the author unjustifiedly suspects an Arabic loanword here).
The same in the Khanag subdialect: kʼarč with polysemy: 'horn / woman's plait / handle, grip' [Uslar 1979: 798, 1004; Dirr 1905: 190, 241]. The same in the Khyuryuk subdialect: kʼarč {кIарч} 'horn' [Genko 2005: 111].
The same in the Khiv subdialect: kʼarč {кIарч} with polysemy: 'horn / woman's plait / handle, grip' [Genko 2005: 111]. The same in Literary Tabasaran: kʼarč {кIарч} 'horn' [Khanmagomedov & Shalbuzov 2001: 213].
The same in Literary Lezgi: karč [abs.] / kr̥čː-ˈuni- [obl.] {карч} 'horn' [Talibov & Gadzhiev 1966: 152; Gadzhiev 1950: 730; Haspelmath 1993: 35, 494, 521]. The tense non-aspirated čː in the Gyune and Literary oblique stems is explained by the synchronic rule, according to which the lax aspirated T yields Tː after a voiceless segment, see [Haspelmath 1993: 47, 55] (such an interesting example as karč / kr̥čː-ˈ should be added to Haspelmath's illustrative lists).
The same in the Akhty dialect: Khlyut karč [abs.] / kr̥č-a- [obl.] 'horn' [Kibrik & Kodzasov 1990: 46].
Proto-Lezgian:*kalč ~ *kʼalč ~ *kːalč4
NCED: 723. Distribution: This word has been retained in its basic meaning in all Nuclear Lezgian lects, but has been completely lost in both outliers (Udi, Archi). The following semantic shifts, reflected as synchronic polysemy, should be noted: 'corner, edge' (Tsakhur), 'woman plait; handle, grip' (Tabasaran).
A second candidate is *pːaˤlː [NCED: 285], which means 'horn / top of the head' in Archi and 'forehead' in Nuclear Lezgian. Since the shift 'top of the head' > 'horn' is typologically more normal than vice versa, we assume that the Proto-Archi meaning was 'top of the head' (correspondingly, the Proto-Lezgian meaning of *pːaˤlː should be 'top of the head' or 'forehead').
The Proto-Udi substantive 'horn' was derived from *meˤrƛ 'deer' [NCED: 300]. In modern Vartashen Udi, it was superseded with *qːˤa[n]cʼ(a), whose original Proto-Lezgian meaning was 'hook' [NCED: 462].
Reconstruction shape: Correspondences seem regular except for the unclear fluctuation of the initial stop: *k- in Kryts, Budukh, Rutul, Lezgi, *kʼ- in Aghul, Tabasaran, *kː- in Tsakhur. External North Caucasian comparison proposed in [NCED: 723] speaks in favor of *kʼ-.
Semantics and structure: Primary substantive root. The oblique stem is *Kalče-.
Gukasyan 1974: 119, 274, 277; Kibrik & Kodzasov 1990: 221; Comrie & Khalilov 2010: 116; Mobili 2010: 300; Schulze-Fürhoff 1994: 467. Suppletive paradigm: zu [abs., erg.] / bez-i ~ bez [gen.] / za- [obl.]. The genitive form bez-i is from [Gukasyan 1974]; bez is from [Kibrik & Kodzasov 1990], probably a recent syncope of the final vowel.
The suppletive paradigm coincides in both dialects. The genetive form is to be analyzed as *b=ez-i with the fossilized class prefix and the nominal genitive ending -i.
Caucasian Albanian: zu [abs., erg.] / bezi [gen.] / za- [obl.] [Gippert et al. 2008: II-37, IV-16].
The historical shape Vz of the genitive form is retained only in Koshan; in other dialects the genitive has been levelled after the rest of the paradigm.
Uslar 1896: 59. Paradigm: zu-n [abs.] / za [erg., obl.] / zi ~ zi-n [gen.].
The same in Literary Lezgi: zu-n [abs.] / za [erg., obl.] / zi [gen.] {зун} 'I' [Gaydarov et al. 2009: 150; Haspelmath 1993: 184].
The same in the Akhty dialect: Khlyut zɨ-n [abs.] / za [erg., obl.] / zɨ [gen.] 'I' [Kibrik & Kodzasov 1990: 221].
Proto-Lezgian:*zo-n ~ *zo1
NCED: 1084. Distribution: Retained as the independent personal pronoun for the 1st p. sg. in all lects except for Borch-Khnov Rutul, where 'I' originates from the personal pronoun 'we (incl.)'.
The absolutive form is to be reconstructed as *zo-n ~ *zo. The status of the suffix -n is unclear. It is attested in Archi and in most Nuclear Lezgian languages: South Lezgian (Kryts, Budukh) and East Lezgian (Aghul, Lezgi; it must be noted that in Tabasaran *zo(-n) has not survived, having been levelled after the oblique stem). On the contrary, -n is absent from Caucasian Albanian-Udi and West Lezgian (Tsakhur, Rutul). Finally, Alyk Kryts shows synchronic doublets zi-n ~ zi. In all these cases the nasal suffix could be explained as influence on the part of the 2nd p. sg. pronoun *u̯o-n 'thou' q.v., although it must be noted that the external North Caucasian comparanda also demonstrate fluctuation between forms with and without -n.
The oblique stem can be safely reconstructed as *za- (retained in Caucasian Albanian-Udi, Archi and Proto-Nuclear Lezgian).
It is unclear how we should reconstruct the Proto-Lezgian ergative form. In Caucasian Albanian-Udi and many Nuclear Lezgian lects (Kryts, Budukh, Mishlesh Tsakhur, Ixrek Rutul, Koshan Aghul, Gequn Aghul, Fite Aghul, Proper Aghul), the ergative form coincides with the absolutive one (*zo-n ~ *zo). On the contrary, in Archi and the rest of Nuclear Lezgian (Mikik Tsakhur, Gelmets Tsakhur, Mukhad Rutul, Luchek Rutul, Keren Aghul, Lezgi), the ergative form is based on the oblique stem *za- (implying the Proto-Lezgian suffix-free ergative form *za).
The genitive stem is to be reconstructed as *class=iz. The class prefixation has been retained as a living pattern in Archi and as a fossilized prefix in Caucasian Albanian-Udi, but has been lost in Nuclear Lezgian. In some lects, the old genitive form can be additionally modified with the synchronic genitive suffix: Caucasian Albanian-Udi, Tsakhur (cf. also Rutul). In South Lezgian (Kryts, Budukh), non-Koshan Aghul, Lezgi, the old genitive form was eliminated, having been levelled after the zV- pattern of the rest of the paradigm. The irregular voiceless fricative in Archi class=is is inexplicable.
Additionally, a specific dative form *class=ez could be theoretically reconstructed, based on Archi class=ez (no traces in other Lezgian languages).
Gukasyan 1974: 78; Mobili 2010: 54. In [Comrie & Khalilov 2010: 248] 'to kill' is translated as be-pː-sun {бепIсун} - apparently a typo for be-s-pː-sun {беспIсун}, the latter originally from syncopated *be-s-b-sun with the regular development bs > pːs (for which see [Maisak 2008a: 149]).
Gukasyan 1974: 78; Fähnrich 1999: 9; Dirr 1903: 2, 70, 96; Schiefner 1863: 101; Schulze 2001: 255; Starchevskiy 1891: 489. In [Dzheiranishvili 1971: 265] 'to kill' is translated as be-s-pː-esun ~ be-s-b-esun (the former variant apparently stays for syncopated be-s-pː-sun with bs > pːs).
Common Udi notes:
Common Udi *be-s-b-esun. A transitive/causative from be-s-, formed with the light verb -b- 'to do' [Schulze 2005: 569 ff. (3.4.2.2 #22 ff.)]. As accepted in [NCED: 662], be-s- is apparently a contracted form of the infinitive bi-es from the verb bi-esun ~ bi-sun 'to die' q.v. The change i > e is not entirely clear, however.
Caucasian Albanian: A labile verb with the suppletive paradigm: bilʸ-a- (present, imperative, future) / pʼur-i- (past) and polysemy: 'to die / to kill' [Gippert et al. 2008: II-44, IV-35]. See notes on 'to die'.
Chumakina et al. 2007; Kibrik & Kodzasov 1988: 170; Kibrik et al. 1977b: 185; Kibrik et al. 1977a 3: 234; Comrie & Khalilov 2010: 248; Mikailov 1967: 173; Dirr 1908: 131, 224. The main meaning of this frequent verb is generic: 'to perform an action most typically associated with the given object in the current situation', an additional specific meaning is 'to kill', obj. = human or animal, sg. or pl. For the pl. object cf. examples like "Who can kill them (= two lovers) in the presence of the people?" [Kibrik et al. 1977b: 11].
More marginal are two complex causative verbs, which literally mean 'to make to die':
1) kʼis-a- from the verb =kʼa- 'to die (sg.)' q.v. Found in some texts, e.g., [Kibrik et al. 1977b: 61, 85, 88]. Apparently used with sg. obj. only.
2) χʷis-a- with polysemy 'to kill / to beat up (obj. = people) / to scold severely, condemn / to wear out' [Chumakina et al. 2007] from the verb =χʷi- 'to die (pl. subj.)'. This is not specified in [Chumakina et al. 2007], but apparently χʷis-as is applied to pl. obj. only.
Kibrik & Kodzasov 1988: 171. Missing from [Meylanova 1984] as a separate entry, but attested in examples, e.g., [Meylanova 1984: 26 sub barut, etc.].
In [Comrie & Khalilov 2010: 248] 'to kill' is glossed as ülum siʔi- {уьлум сиъи}, literally 'death' + 'to do', which could be a neologism on the authors' part (note that ülum is an error for ülüm 'death' [Meylanova 1984: 140]).
Comrie & Khalilov 2010: 248. In [Comrie & Khalilov 2010], erroneously quoted as gʸ=iɢ- {гикъаз}.
Common Tsakhur notes:
Initial gʸ= is a prefix with general semantics [Ibragimov 1990: 124; Kibrik & Kodzasov 1988: 41]. The verb gʸ=ikʼ- 'to kill (sg. obj.)' contains the same root as 'to die (sg. subj.)' q.v., modified with another prefix; the verb gʸ=atʼ- 'to kill (pl. obj.)' contains the same root as 'to die (pl. subj.)', modified with another prefix.
Dirr 1912: 144, 202; Ibragimov 1978: 121; Makhmudova 2001: 71, 116, 252; Comrie & Khalilov 2010: 248. Labile verb with polysemy: 'to kill / to die (q.v.)', applied to sg. obj./subj.
Distinct from class=ɨrqʼ- / class=qʼɨrqʼ- 'to kill (pl. obj.) / to die (pl. subj.)' [Dirr 1912: 144, 166].
A second candidate is y=ɨχˤ-, glossed as 'to kill' in [Makhmudova 2001: 107] with the example: "Matsay killed his own chicken" [Makhmudova 2001: 176-177]. But the main meaning of y=ɨχˤ- is 'to strike, hit; to wound', as it is glossed in [Dirr 1912: 163] (with examples) and [Ibragimov 1978: 121]; cf also two additional examples: "The raising sun touched (lit. struck) the mountain top" [Makhmudova 2001: 73], "Beat the drum!" [Makhmudova 2001: 210].
Dzhamalov & Semedov 2006: 405. Labile verb with polysemy: 'to kill / to die (q.v.)'. It must be noted that in the main section of the dictionary [Dzhamalov & Semedov 2006: 131] this verb is only glossed as 'to die', whereas in [Comrie & Khalilov 2010: 248] 'to kill' is translated as y=iqʼ-e h=aʔ- - causative from class=iqʼ-i-r.
A second candidate is y=ɨχˤ-ɨ-r {йыIхыIн} 'to strike, hit; to wound (with a weapon), kill; to push, shove' [Dzhamalov & Semedov 2006: 136, 405], with the example: "We have killed a bear" [Ibragimov 1978: 213].
Kibrik & Kodzasov 1988: 171. Synchronically irregular verb with three parallel imperfective stems: y=ɨ=l=χˤ-a-r- ~ r=ɨχˤ-a-r-~ y=ɨ=r=χʕ-a-r- [imperf.] / ɨχˤ-ɨ-r [perf.]. Note the rare imperfective infix -l- and the relict imperfective prefix r-; the third imperfective stem is regular. Used both with sg. and pl. obj.
Common Rutul notes:
The Proto-Rutul verb for 'to kill (sg. obj.)' was the labile =iqʼ- 'to die (sg. subj.) / to kill (sg. obj.)'. The verb for 'to die (pl. subj.) / to kill (pl. obj.)', attested in Mukhad and Luchek, is etymologically related, formed with reduplication.
There is a tendency in Rutul dialects to restrict =iqʼ- to the meaning 'to die' and ascribe the meaning 'to kill' to the verb y=ɨχˤ- 'to strike, hit', which is derived from ɨχˤ- 'to strike, hit' [Kibrik & Kodzasov 1988: 102] with the prefix y=. This process is currently finished in the Luchek dialect. It must be noted that the semantic development 'to hit' > 'to kill' is normal, whereas vice versa is typologically odd.
Uslar 1896: 529, 636, 637. Imperfective stem. Labile verb with polysemy: 'to die / to kill'; for further notes, see 'to die'.
Proto-Lezgian:*ʔiƛʼe1
NCED: 661. Distribution: Preserved in the majority of Lezgian languages. The complete Proto-Lezgian system is reconstructed as *ʔiƛʼe 'to die / to kill' (sg.), *ʔilχʷe 'to die / to kill' (pl.). Further see notes on 'to die'.
Replacements: {'to perform an action most typically associated with the given object in the current situation' > 'to kill'} (Archi), {'to cut' > 'to kill'} (Budukh), {'to cut' > 'to die / to kill'} (Tsakhur), {'to strike, hit' > 'to kill'} (Rutul).
Gukasyan 1974: 141; Fähnrich 1999: 20; Mobili 2010: 174; Schulze 2001: 292. In [Fähnrich 1999: 20] the variant kːaˤkːapː is also quoted. In [Schiefner 1863: 82] quoted as kːäkːäp.
Common Udi notes:
Common Udi *kːaˤkːa-pː with a fossilized plural suffix.
Caucasian Albanian: lʸek, attested once in Is. 35.3 [Gippert et al. 2008: VII-23] (missing from the dictionary in [Gippert et al. 2008: IV]). Possesses reliable Lezgian cognates (Lezgian *läk 'a part of the leg' [NCED: 755]).
Chumakina et al. 2007; Kibrik & Kodzasov 1990: 32; Kibrik et al. 1977b: 294, 363; Comrie & Khalilov 2010: 212; Mikailov 1967: 196; Dirr 1908: 175, 211. In [Comrie & Khalilov 2010] incorrectly quoted with the variant poˤmr {поIмр} due to misunderstanding of a peculiarity of the Russian alphabet. Suppletive paradigm: poˤmp [abs. sg.] / pˈoˤmp-li- [obl.] / poˤt [abs. pl.] / pˈoˤ-r-čay [erg. pl.]. As proposed in [Kibrik et al. 1977a 2: 45], could be analyzed as the root poˤ with nasal reduplication in sg. poˤ-m-p (a unique pattern for the noun system) and regular plural forms poˤ-t, pˈoˤ-r-čay.
Kibrik & Kodzasov 1990: 32; Dirr 1907: 138, 174; Shaumyan 1941: 186. The more archaic variant qʼʷaqʼʷ is from [Dirr 1907]; in [Shaumyan 1941], both forms are quoted; the modern source, [Kibrik & Kodzasov 1990], gives only qʼuqʼ.
The same in the Khanag subdialect: qʼamqʼ with polysemy: 'knee / stalk (of cereal)' [Uslar 1979: 880, 995; Dirr 1905: 201, 231]. The same in the Khyuryuk subdialect: qʼamqʼ {кьамкь} 'knee' [Genko 2005: 105].
The same in the Khiv subdialect: qʼamqʼ {кьамкь} 'knee' [Genko 2005: 105]. The same in Literary Tabasaran: qʼamqʼ {кьамкь} 'knee' [Khanmagomedov & Shalbuzov 2001: 208].
Uslar 1896: 506, 615. Paradigm: met [abs.] / metʼ-ˈi- [obl.].
The same in Literary Lezgi: met [abs.] / metʼ-ˈi- [obl.] {мет} 'knee' [Talibov & Gadzhiev 1966: 237; Gadzhiev 1950: 296; Haspelmath 1993: 498, 522].
The same in the Akhty dialect: Khlyut met [abs.] / metː-ˈa- [obl.] 'knee' [Kibrik & Kodzasov 1990: 32].
Proto-Lezgian:*pɨˤmp ~ *pʼɨˤmpʼ2
LEDb: #230. Distribution: An unstable word. From the distributional point of view, the best candidate is *pɨˤmp / *pʼɨˤmpʼ which denotes 'knee' in Archi, on the one hand, and in South Lezgian (Kryts, Budukh), on the other. The meaning shifted to 'angle, corner' in Tabasaran and Lezgi. The Proto-Lezgian reconstruction is not entirely clear phonetically and morphophonologically (see below); external North Caucasian etymology of this root is unknown.
The second candidate is *qʼamqʼ [NCED: 907], which is attested in the meaning 'knee' in some West Lezgian (Rutul) and some East Lezgian lects (Aghul, Tabasaran). Formally, *qʼamqʼ and *pɨˤmp occur with criss-crossing distribution within Nuclear Lezgian. As in some other cases, we suppose that *qʼamqʼ in the meaning 'knee' is an areal introduction, shared by some languages (including Proto-Dargi *qʼʷaqʼʷa ‘knee’ in the adjacent Dargi lects!) after the split of Proto-Nuclear Lezgian.
Various replacements occurred in individual languages.
Udi: *kʼälkʼ- (~ -e-, -r-) [NCED: 720], modified with the fossilized plural suffix. The exact meaning of Proto-Lezgian *kʼälkʼ- cannot be established with certainty. Outside Udi, this root is attested as 'calf of leg' (Lezgi) and 'top of boot' (Aghul). The attested meanings suggest something like 'a part of leg between knee and ankle'.
Caucasian Albanian: *läk [NCED: 755], whose original meaning seems to have been 'leg bone' (shifted to 'bone' q.v. in Archi and into 'foot, leg' q.v. in Tabasaran-Aghul).
Tsakhur: *qʼaraCay / *Caraqʼay [LEDb: #204], this root means 'shin' in Rutul. Unattested outside West Lezgian.
Lezgi: *wenčʼː [NCED: 1042]; this root denotes 'lower corner of sack or bag' in Archi. External North Caucasian comparison points to the primary anatomic semantics ('a part of leg').
Replacements: {'a part of leg between knee and ankle' > 'knee'} (Udi), {'leg bone' > 'knee'} (Caucasian Albanian), {'shin' > 'knee'(?)} (Tsakhur), {'knee' > 'angle, corner'} (Tabasaran, Lezgi), {'knee' > 'stalk (of cereal)'} (Northern Tabasaran).
Reconstruction shape: Correspondences are not quite regular due to fluctuation of the ejective ~ plain stops: *p-p in Archi, Kryts, Budukh ('knee') vs. *pʼ-pʼ in Tabasaran, Lezgi ('angle, corner').
Semantics and structure: Primary substantive root. The irregular Archi paradigm suggests that the original shape of the root could be *pɨ (~ *pʼɨ) with further reduplication.
Common Udi *a-ba-bak-esun, derived from *a-ba (Nidzh ava, Vartashen aba) 'knowing, knowledgeable, competent' [Gukasyan 1974: 31] with the light verb -bak- 'to be(come)' [Schulze 2005: 561 f. (3.4.2.2 #10); Schulze-Fürhoff 1994: 474]. The adjective *a-ba contains the common adjectival suffix -ba, on which see [Gukasyan 1974: 272; Schulze 2005: 222 f. (3.2.9.1 #7)]; for the sporadic fricativization b > v in the Nidzh dialect see [Dzheiranishvili 1971: 277; Maisak 2008a: 150 f.]. In [Gippert et al. 2008: II-76] Nidzh ava ~ Vartashen aba are not segmented, but incorrectly treated as primary roots originating from Lezgian *ʔacʼa- (with *cʼ > v/b).
Caucasian Albanian: aa- [Gippert et al. 2008: IV-4]. Phonetically, apparently aʔa- with ʔ in the place of the lost Lezgian *cʼ. Distinct (although in some contexts very close semantically) from the less frequent verb čal-χ-esun 'to know, realize' [Gippert et al. 2008: IV-36], which corresponds to modern Udi čal-χ-esun 'to recognize, know, experience, make the acquaintance of smb.' [Gukasyan 1974: 236; Schulze 2001: 265].
Chumakina et al. 2007; Kibrik & Kodzasov 1988: 177; Kibrik et al. 1977b: 314, 360; Comrie & Khalilov 2010: 727. In [Mikailov 1967: 197; Dirr 1908: 181, 209] and also [Comrie & Khalilov 2010: 727] 'to know' is quoted as sˈin-ke-, although the actual meaning of this paronymous complex verb is 'to find out, learn; to recognize; to feel (pain)' [Chumakina et al. 2007; Kibrik et al. 1977b: 314].
Distinct from bˈecʼːa- 'to be able to; to know how' [Chumakina et al. 2007; Kibrik et al. 1977b: 201].
Saadiev 1994: 423, 440, 442. In [Kibrik & Kodzasov 1988: 177; Comrie & Khalilov 2010: 727] quoted as a complex expression with the verb xi- 'to become': class=äcʼä-r xi-. Apparently used in perfective (aorist) form only.
Kibrik & Kodzasov 1988: 177. Polysemy: 'to know / to be able'. A complex expression hacʼǝ-r yɨxǝ-r- can also be used (participle plus the verb yɨxǝ- / sǝxǝ- 'to be(come)'). Missing from [Meylanova 1984: 40], where only the participle hacʼǝr 'knowledgeable, understanding' is quoted.
In [Comrie & Khalilov 2010: 727] 'to know' is incorrectly glossed as baladu {баладу}, which means something like 'to be acquainted (with a person)', cf. [Meylanova 1984: 25].
Kibrik & Kodzasov 1988: 177; Suleymanov 2003: 92; Shaumyan 1941: 146. Polysemy: 'to know / to get to know'. An analytic construction with the verb xi- 'to become' [Kibrik & Kodzasov 1988: 62]. The variant aħa-r is from [Suleymanov 2003].
Kibrik & Kodzasov 1988: 177. Irregular verb with the synthetic present form ħuye; the rest of the paradigm is formed analytically as ħa-r xi-. The auxiliary verb xi- in Kibrik & Kodzasov's gloss may be an error for xa- = the verb xa- 'to become' [Kibrik & Kodzasov 1988: 62].
The same root in the Usug subdialect: Ha-r waˤ- 'to know' [Shaumyan 1941: 146, 197]; the auxiliary verb waˤ- means 'to go' q.v.
Suleymanov 2003: 92; Magometov 1970: 202 (sentence 10); Shaumyan 1941: 146. Polysemy: 'to know / to get to know'. An analytic construction with the verb xa- 'to become'.
The same root in the Kurag subdialect: aHa-y-e ~ y=aʁˤä-y-e [imperf.] / ʡ=aHa-r xa- ~ aHa-r xa- [other forms] 'to know / to get to know' [Magometov 1970: 167, 181, 188, 209 sentence 9]. Magometov's inconsistent transcription of the root fricative is obviously wrong, but the picture is similar to the Keren (Richa) dialect: the synthetic imperfect (presence) and the analytic rest of the paradigm.
The analytic construction in other subdialects:Duldug aHa-r xa-, Tsirkhe aχˤa-r xa- 'to know; to get to know' [Shaumyan 1941: 146].
Initial y= and ʡ= look like desemanticized preverbs.
Common Aghul notes:
A poorly documented verb, both phonetically and paradigmatically. In all the dialects, the proper verbal forms tend to be superseded with the analytic construction aHa-r + the auxiliary verbs 'to become' or 'to go'. The nominal form aHa-r ~ Ha-r represents the old participle or gerund.
The same in the Khiv subdialect: aʁˤa x-ˈ ~ aʁˤu x-ˈ {аьгъяхьуб} 'to know' [Genko 2005: 23].
Common Tabasaran notes:
In all dialects the equivalent for 'know' is an analytic construction: the participle from the lost verb aʁˤ- 'to know' + the suppletive auxiliary verb 'to become', for which see [Kibrik & Kodzasov 1988: 62].
Uslar 1896: 191, 193, 594. Literally 'to be-known to X'. The analytic construction with the participle čːi-r že-/xa- or simple čːi že-/xa- can also be used, literally 'to become (že- / xa-) known to X'.
Similarly in Literary Lezgi: čːi- with polysemy: 'to know / to know how', literally 'to be-known to X' [Haspelmath 1993: 139, 484, 522]. Used in the imperfective; for the perfective, the analytic construction with the participle čːi-r xa- {чирхьун} is used, literally 'to become (xa-) known to X' [Gadzhiev 1950: 248; Haspelmath 1993: 139]. Not attested in [Talibov & Gadzhiev 1966].
Proto-Lezgian:*ʔacʼa-1
NCED: 262. Distribution: Retained as the basic root for 'to know' in Caucasian Albanian-Udi, on the one hand, and in most Nuclear Lezgian lects, on the other: West Lezgian (Tsakhur, Rutul), South Lezgian (Kryts, Budukh). It must be noted that in Udi, the verb got lost, whereas the root survived in the suffixed adjective as an element of the complex verb (literally 'to be knowing'). Similarly, the old verb tends to be superseded with the analytic construction participle + 'to be(come)' in Nuclear Lezgian. The original meaning of *ʔacʼa- shifted to 'to be able to; to know how' in Archi.
In Aghul and Tabasaran, the old root was lost, superseded with *ʔaχːˤa- [NCED: 565] (as in the case of *ʔacʼa-, the most frequent construction is participle + 'to be(come)'). The exact Proto-Lezgian meaning of *ʔaχːˤa- is not clear, apparently something like 'to get to know, to learn' (cf. its meanings 'teaching, lesson' in Archi, 'to be able to' in Tsakhur).
In some lects, the basic meaning 'to know' is expressed with etymologically obscure roots: Archi(sˈini), Lezgi (čːi-).
Replacements: {'to know' > 'to be able to; to know how'} (Archi, Budukh).
As correctly proposed in [Schulze 2001: 333], an Azerbaijani loanword. Borrowed from Azerbaijani dialectal (Zaqatala) xazal 'leaf', literary xäzäl 'fallen dry leaves' (ultimately from Persian).
Caucasian Albanian: ʒeeup [Gippert et al. 2008: IV-36]. The form is not fully reliable and etymologically obscure; it is attested only in the plural meaning 'leaves' = 'foliage'.
Kibrik & Kodzasov 1990: 96; Suleymanov 2003: 194. Applied to both trees and herbs.
Distinct from kʼiǯ 'paper, sheet of paper; letter (message)' [Shaumyan 1941: 183; Suleymanov 2003: 125; Magometov 1970: 152].
It is surprising, however, that in the only Burshag example found for the meaning 'leaf' it is the word kʼiǯ that is used: "In autumn, leaves fall down from the trees" [Shaumyan 1941: 40].
Kibrik & Kodzasov 1990: 96. Applied to trees. Distinct from cʼab 'leaf of herb' [Kibrik & Kodzasov 1990: 96] (in [Magometov 1970: 36], apparently erroneously glossed as 'tree leaf').
The same in the Usug subdialect: pʼaˤš 'tree leaf' [Shaumyan 1941: 153]. Distinct from Usug kʼež 'paper, sheet of paper; letter (message)' [Shaumyan 1941: 183].
Kibrik & Kodzasov 1990: 96; Dirr 1907: 137, 176; Shaumyan 1941: 153. The variant pʼaˤž is from [Dirr 1907]. Applied to trees.
Distinct from cʼaw 'leaf of herb' [Kibrik & Kodzasov 1990: 96].
Distinct from kʼeǯ ~ kʼiž 'paper, sheet of paper; letter (message)' [Dirr 1907: 130; Shaumyan 1941: 183] (the latter form is from [Dirr 1907]).
Fite Aghul:cʼab5
Kibrik & Kodzasov 1990: 96. Applied to both trees and herbs.
Aghul (proper):cʼab5
Suleymanov 2003: 194. According to Suleymanov's examples, Tpig cʼab is applied to both trees and herbs. In [Shaumyan 1941: 153], however, the Tpig word for 'tree leaf' is quoted as paǯ.
Distinct from kʼeǯ 'paper, sheet of paper; letter (message)' [Suleymanov 2003: 125; Shaumyan 1941: 183]. It is interesting, however, that in the only Tpig example found in Shaumyan's work for the meaning 'leaf' the word kʼeǯ is used: "In autumn, leaves fall down from the trees" [Shaumyan 1941: 40].
Common Aghul notes:
A rather unclear situation here, with three candidates for the meaning 'leaf': pʼaˤž, cʼab and kʼeǯ. The word cʼab (cʼaw) has a cognate in Tabasaran (the closest relative of Aghul): cʼab ~ cʼaw 'herb leaf', suggesting that it is possible to reconstruct the opposition 'tree leaf' vs. 'herb leaf' (cʼab) for Proto-Aghul; in such a case the former term would be superseded with the latter one (cʼab) in some dialects, where this typologically rare semantic opposition was eliminated.
It is uncertain how the Proto-Aghul word for 'tree leaf' should be reconstructed. Upon first sight, pʼaˤž is the best candidate ('tree leaf' in Keren and Gequn), but first, it should be noted that kʼeǯ, which everywhere means 'sheet of paper', might be attested in the meaning 'tree leaf' in Koshan and Proper Aghul (if Shaumyan's examples are valid); second, the semantic development 'leaf' > 'sheet (of paper)' is typologically possible, whereas vice versa looks odd. These facts could point to kʼeǯ as the Proto-Aghul term for 'tree leaf'. Both kʼeǯ and pʼaˤž possess external Lezgian cognates with the meaning 'leaf', but kʼeǯ seems a more preferable candidate for the status of this basic Proto-Lezgian term. It should be noted that, as plausibly proposed in [NCED: 298], pʼaˤž acquires the shape pʼaˤǯ in some Aghul dialects under the influence of kʼeǯ. This can additionally confirm that the main Proto-Aghul word for 'tree leaf' was kʼeǯ, whereas pʼaˤž denoted some specific kind of leaves.
Kibrik & Kodzasov 1990: 96. Apparently applied to both trees and herbs.
The same in the Khanag subdialect: kʼaǯ with polysemy: 'leaf of tree / sheet of paper / letter (message)' [Uslar 1979: 793, 997; Dirr 1905: 189, 232], opposed to cʼaw 'plant tops (Russian: ботва)' [Uslar 1979: 959].
The same in the Khyuryuk subdialect: kʼaǯ {кIажж} 'leaf of tree' [Genko 2005: 110], as opposed to cʼaw {цIав} 'leaf of herb, plant tops' [Genko 2005: 176] (the dialectal origin of cʼaw is not specified by Genko, but phonetically the form is Northern Tabasaran; Khyuryuk is the most probable variant).
Kibrik & Kodzasov 1990: 96. Apparently applied to both trees and herbs.
Differently in the Khiv subdialect: kʼaǯ {кIажж} with polysemy: 'leaf of tree / paper, sheet of paper / letter (message) / amulet' [Genko 2005: 110], as opposed to cʼab {цIаб} 'leaf of herb, plant tops' [Genko 2005: 176].
In Literary Tabasaran: kʼaǯ {кIаж} with polysemy: 'leaf / paper, sheet of paper / letter (message) / amulet' [Khanmagomedov & Shalbuzov 2001: 212] - apparently applied to both trees and herbs; the literary word cʼab {цIаб} means 'petal' [Khanmagomedov & Shalbuzov 2001: 331].
Common Tabasaran notes:
The typologically rare opposition kʼaǯ 'tree leaf' / cʼab 'herb leaf' is to be reconstructed for Proto-Tabasaran.
Uslar 1896: 359, 616. Applied to trees (at least).
The same in Literary Lezgi: pːeš {пеш} 'leaf' [Talibov & Gadzhiev 1966: 264; Gadzhiev 1950: 336; Haspelmath 1993: 500, 522] (applied to both trees and herbs).
Differently in the Akhty dialect: Khlyut riqʼ-ˈäy 'leaf' [Kibrik & Kodzasov 1990: 96] (applied to both trees and herbs). Final -Vy is a rare desemanticized suffix.
In is unclear how the Proto-Lezgi term for 'leaf' should be reconstructed. Note the different treatment of Lezgian *pː in Gyune (b) and Literary Lezgi (pː).
Proto-Lezgian:*ƛʼačʼa1
NCED: 773. Distribution: An unstable word. Three roots enter into competition with each other in this "criss-crossed" situation. Tentatively, we fill the slot with *ƛʼačʼa [NCED: 773], which has the best distribution among the candidates. The root *ƛʼačʼa means 'leaf' in Archi, on the one hand, and in Tabasaran (specifically 'tree leaf') and apparently Proto-Aghul, on the other. Additionally, this root means 'grain' in Udi (< *'husk'?), 'stem, stalk' in Lezgi, Kryts, 'pod, seedpod' in Tsakhur, 'straw' in Budukh (cf. also Rutul qʼačʼ 'grain' quoted in [NCED], not found in other sources). However, external North Caucasian cognates of Lezgian *ƛʼačʼa are fairly scant and dubious.
The second candidate is *rɨƛʼʷ (reduplicated *ƛʼʷɨrɨƛʼʷ) [NCED: 784]. It is attested as 'leaf' in Rutul and Akhty Lezgi, but has been lost in the rest of languages (in [NCED], cf. also Budukh kʼurukʼ 'bud' is quoted, not found in primary sources). Lezgian *rɨƛʼʷ ~ *ƛʼʷɨrɨƛʼʷ has a weak distribution, and there are no internal reasons to reconstruct this root as the basic Proto-Lezgian term for 'leaf'. Nevertheless, *rɨƛʼʷ ~ *ƛʼʷɨrɨƛʼʷ has very good external North Caucasian cognates that point to the meaning 'leaf'.
The third candidate is *pːaˤša [NCED: 297], attested as 'leaf' in Kryts, Gyune Lezgi, some Aghul dialects (but probably not Proto-Aghul). In Budukh, this stem means 'bud, gemma'. It is unclear how the exact meaning of Proto-Lezgian *pːaˤša should be reconstructed.
In Tsakhur, 'leaf' is expressed by *tʼʷela [NCED: 1006], whose original Proto-Nuclear Lezgian (and Proto-Lezgian?) meaning was no doubt 'twig, rod' (with a further shift to 'rib' in some lects).
Cf. also cʼab, which means specifically 'herb leaf' in Proto-Aghul-Tabasaran (without further etymology?).
Inherited terms for 'leaf' were superseded with Azerbaijani loanwords in Udi, Alyk Kryts, Budukh.
Gukasyan 1974: 71, 93 (sub бул); Comrie & Khalilov 2010: 561; Mobili 2010: 49, 74. Polysemy: 'to lie / to lie down / to sleep / to fall asleep'. This currently serves as the default expression for 'to sleep' q.v. in Nidzh.
Gukasyan 1974: 71, 93 (sub бул); Fähnrich 1999: 8; Schiefner 1863: 100; Schulze 2001: 254; Starchevskiy 1891: 486. Polysemy: 'to lie / to lie down'. According to texts in [Bezhanov & Bezhanov 1902], in the past tenses this verb acquires an additional meaning 'to sleep' q.v.; in [Dirr 1903: 42, 43] the masdar is also quoted with the meaning 'to sleep'.
Common Udi notes:
Common Udi *bas-kː-esun 'to lie; to lie down', formed with the aid of the light verb -kː- 'to let(?)' [Schulze 2005: 561 f. (3.4.2.2 #14, 21); Schulze-Fürhoff 1994: 474]. The synchronic root bas- is to be analyzed as *b=as- with a fossilized class prefix ([NCED: 1038], followed by [Gippert et al. 2008: II-71]). Alternatively (thus [Schulze 2001: 254]) and less likely - as *ba=s- with the locative preverb ba(y)- 'in', for which see [Maisak 2008a: 156; Harris 2002: 69; Schulze 2005: 580 ff. (3.4.3)] (for unknown reasons, this preverb is glossed as 'down' in [Schulze 2001: 254]).
Caucasian Albanian: The verb bas-kʼ- is, in fact, attested only in the past tense with the meaning 'to sleep, to fall asleep' [Gippert et al. 2008: IV-8] (the same situation as in archaic Vartashen, see above). This is, however, probably not the basic expression for 'to sleep' q.v.
Chumakina et al. 2007; Kibrik & Kodzasov 1988: 88; Kibrik et al. 1977b: 194, 365; Kibrik et al. 1977a 3: 235; Comrie & Khalilov 2010: 561; Dirr 1908: 129, 212. Perfective stem: ˈaχu-. Polysemy: 'to lie / to lie down / to sleep'; applied to animated subj. The meaning 'to sleep', proposed in [Mikailov 1967: 172] and [Comrie & Khalilov 2010: 235], is proven by the example "He is still asleep (lit. lying down)" [Chumakina et al. 2007], although this is probably not the default expression for 'to sleep' q.v. In [Mikailov 1967: 172] =ˈaχa- is incorrectly translated as simply 'to sleep'. The same verb is used in the expression for 'to sleep' q.v. (ˈaχu-ke-).
Distinct from =ˈeɬːa- 'to lie (inanimate subj.); to put' [Chumakina et al. 2007; Kibrik et al. 1977b: 229].
Kibrik & Kodzasov 1988: 88; Comrie & Khalilov 2010: 561. Ablaut paradigm: c=u-l-qʼäl- [imperf.] / c=uqʼul- [perf., imv.]. Initial c= is the preverb of general semantics, -l- is the imperfective infix. Polysemy: 'to lie / to lie down' (in the aforementioned sources quoted for 'to lie down'; the stative meaning 'to lie' is given in [NCED: 265]).
Meylanova 1984: 82, 219; Kibrik & Kodzasov 1988: 88. In [Comrie & Khalilov 2010: 561] the imperfective l-infixed form is quoted with a typo: qːalqːal {къалкъал} for qːalqʼal {къалкьал}. Polysemy: 'to fall, go sprawling / to lie / to lie down'. Initial qː= is the preverb 'out' [Alekseev 1994: 271].
Kibrik et al. 1999: 68, 875; Ibragimov & Nurmamedov 2010: 211; Comrie & Khalilov 2010: 561. Ablaut paradigm: qːilʸ=ex-a [imperf.] / qːalʸ=ix-u [perf.] / qːalʸ=ix-as [fut.]. Thematic -u- in perf. can hardly serve as an unambiguous indication of labialized -xʷ- in the root. More important evidence for -xʷ- is the negative masdar class 1/2 qːil-dʸ-e-r-xʷu-y, quoted in [Ibragimov & Nurmamedov 2010: 211]; it is not clear, however, whether this form is reliable. Cf. the positive masdar 1/2 with plain -x- in the same source: qːalʸ-i-r-xɨ-y. It is possible theoretically that -x- has been levelled across the paradigm after forms of class 3, where -pxʷ- > -px- or -oːxʷ- > -oːx- (a regular dissimilative process), but such forms are not numerous, thus it is strange to regard them as the source of levelling.
Polysemy: 'to lie / to lie down' (for the stative meaning cf., e.g., an example in [Kibrik et al. 1999: 220]). Applied to sg. subj.
Distinct from qːalʸ=akʼʷ- 'to lie; to lie down' (pl. subj.) [Kibrik et al. 1999: 68, 875; Ibragimov & Nurmamedov 2010: 211].
In [Ibragimov & Nurmamedov 2010], qːilʸ=ex- and qːalʸ=akʼʷ- are quoted with polysemy: 'to lie / to sleep', which frequently occurs in the Tsakhur dialects, but this is not the Mishlesh case, where 'to sleep' is expressed by a specific verb (q.v.).
Kibrik & Kodzasov 1988: 88; Dirr 1913: 151, 229. Ablaut paradigm: ʁilʸ=ex-a [imperf.] / ʁalʸ=ix-u [perf.] / ʁalʸ=ix-as [fut.]; note the thematic -u- in perf. Polysemy: 'to lie / to lie down / to sleep'. Applied to sg. subj.
Distinct from ʁalʸ=ekʼʷ- 'to lie; to lie down; to sleep' (pl. subj.) [Kibrik & Kodzasov 1988: 88].
Gelmets Tsakhur:
Not attested. In [Comrie & Khalilov 2010: 561], 'to lie down' is glossed as sowkamiš-x- {совкамишхьез}, which looks like an Azerbaijani loanword: the Azerbaijani verbal root sowka- + the Azerbaijani perfect suffix -miš + the Tsakhur verb ɨx- 'to become', although the actual source of sowka- has not been identified (the possible candidate is Azerbaijani söykä-mäk 'to lean against, rest against', if one assumes a dialectal development in Azerbaijani into **sövkä- 'to lie (down)').
Cf. ʁalʸ=ix- 'to sleep' q.v.
Common Tsakhur notes:
Initial ʁ=Vlʸ= (Mishlesh qː=Vlʸ=) is a double prefix with general semantics [Ibragimov 1990: 125; Kibrik & Kodzasov 1988: 41].
Ibragimov 1978: 32, 120; Makhmudova 2001: 71, 96, 243. In [Comrie & Khalilov 2010: 561], erroneously quoted as l=uχ- {лухун}. Polysemy: 'to fall (subj. = only human?), go sprawling / to lie'. Cf. an example: "Anuts is lying" [Makhmudova 2001: 71]. In [Dirr 1912: 158], only attested with the meaning 'to fall' (no expressions for 'to lie' are provided by Dirr). Regular paradigm: l=uk-a-r- [imperf.] / l=uk-u-r [perf.].
In [Comrie & Khalilov 2010: 561], a second term for 'to lie down' is also quoted: k=utʼ- {кутIун} (not found in other sources), on which see below.
Dzhamalov & Semedov 2006: 181, 203, 352. A collocation ('to lie, rest'), which consists of the adverb ow 'down' and the verb luk- 'to fall (subj. = only human?)' with the regular paradigm: l=uk-ä-r- [imperf.] / l=uk-u-r [perf.].
A second candidate is k=utʼ- {кутIун} 'to lie / to be ill' [Dzhamalov & Semedov 2006: 148], with the examples: "He (or it?) is lying on the ground", "I have been ill for many days". In [Comrie & Khalilov 2010: 561], 'to lie down' is quoted as ow k=utʼ- with the same adverb ow 'down'.
The difference between ow l=uk- and k=utʼ- is unclear; we treat both as synonyms.
Kibrik & Kodzasov 1988: 88. Regular paradigm: l=u=r=k-a-r- [imperf.] / l=uk-u-r [perf.]. Polysemy: 'to lie / to lie down'. This verb is quoted in the meaning 'to lie down' in [Kibrik & Kodzasov 1988], but the lexical opposition 'to lie' : 'to lie down' is atypical for this region; hence, we assume the aforementioned polysemy for Luchek l=uk-.
Common Rutul notes:
Shinaz dialect: in [Ibragimov 1978: 163] the verb k=utʼ- is quoted with the meaning 'to get into bed, lie down into bed'.
It is unclear how the Proto-Rutul verb for 'to lie' should be reconstructed, because the available lexicographic information is very scarce. The widespread verb l=uk- rather represents the Proto-Rutul term for 'to fall, go sprawling', which has latter acquired the meaning 'to lie' in some dialects (cf. the analytic Ixrek construction "down + to fall"). On the contrary, k=utʼ- 'to lie' may be an archaism.
Initial l=, k= are prefixes with general semantics [Ibragimov 1978: 95; Alekseev 1994a: 227; Makhmudova 2001: 165].
Pace [NCED: 644], no direct traces of labialized -kʷ- in l=uk- are observed in the available Rutul data (due to the dissimilative delabialization ukʷ > uk).
In [Kibrik & Kodzasov 1988], three verbs are quoted as synonyms for 'to lie down' (scil. 'to lie'): aχ-a-, utː=ark-i- and ʁ=ark-i-; semantic and pragmatic nuances are unknown.
In [Kibrik & Kodzasov 1988], two verbs are quoted as synonyms for 'to lie down' (scil. 'to lie'): fa=tː=ix-a- and a=q=ux-a-; semantic and pragmatic nuances are unknown.
A different root in the Usug subdialect: qa=d=ark-a- 'to lie; to lie down' [Shaumyan 1941: 141] ("He lay on the bed").
Dirr 1907: 104, 175. Polysemy: 'to lie / to lie down / to sleep (q.v.)'. Cf. the example for the stative meaning 'to lie': "We will lie near this she-donkey" [Dirr 1907: 77].
Shaumyan 1941: 141. Polysemy: 'to lie / to lie down'. Shaumyan's example for the stative meaning: "He lay on the bed". It must be noted that in [Suleymanov 2003: 35], a=q=ux-a- is more specifically translated as 'to lie down for a short time (Russian: прилечь)'. Probably no expressions for the generic 'to lie (animated subj.)' in [Suleymanov 2003].
Differently in the Tsirkhe subdialect: q=arx-a- 'to lie; to lie down' [Shaumyan 1941: 141].
Common Aghul notes:
We presume that Koshan and Keren verbs, which are quoted in [Kibrik & Kodzasov 1988] for the concept 'to lie down', actually possess the polysemy: 'to lie / to lie down'.
In sum, four or five verbal roots, sometimes modified with various spatial prefixes, are attested in Aghul dialects with the meaning 'to lie (down)': 1) aχ-a- 'to lie (down); to sleep'; 2) =ark-i- 'to lie (down); to sleep'; 3) =arx-a- 'to lie (down); to sleep'; 4) =ix-a- 'to lie (down)'; 5) (a)qux-a- 'to lie (down)'. The areal isoglosses of polysemy 'to lie / to sleep' and derivation 'to fall' → 'to lie (down); to sleep' are rather strong in Aghul, which makes the reconstruction of the Proto-Aghul verbs for 'to lie' and 'to sleep' difficult.
It seems that the prefixed =arx-a- has secondarily acquired the meanings 'to lie' (Tsirkhe subdialect of Proper Aghul) and 'to sleep' (Keren, Proper Aghul, Fite), because the main synchronic meaning of (=)arx-a- is 'to fall' [Magometov 1970: 164], and external Lezgian comparison confirms this [NCED: 602].
The root =ark-i-, modified with various spatial prefixes, seems to be a recent introduction in the generic meanings 'to lie (down)' (Koshan, Usug Keren) and 'to sleep' (Burshag Koshan). The basic meaning of the prefixless stem ark-i- is retained in Burshag Koshan as 'to fall down' [Kibrik & Kodzasov 1988: 84]; various verbal prefixed stems from this root in Aghul dialects also demonstrate the semantics of 'falling' [Kibrik & Kodzasov 1988: 84; Shaumyan 1941: 137, 138]. This Aghul root originates from Proto-Lezgian *ʔarkɨr- / *ʔerkɨr- 'to fall; to let fall' [NCED: 266].
Similarly, Keren (Richa) fa=tː=ix-a- is secondary in the generic meaning 'to lie (animated subj.)', because in other Aghul dialects the standard meaning of this prefixed stem is 'to throw, let fall; to be ill in bed, be laid up' [Shaumyan 1941: 149; Suleymanov 2003: 165]. External Lezgian comparison suggests that the primary meaning of Proto-Lezgian *ʔeɬːʷɨ- (from which Aghul =ix-a- originates) was 'to put; to lie (inanimate subj.)' [NCED: 279].
The expressions aqux-a- (Richa Keren, Tpig), qux-a- (Gequn) 'to lie (down)' must apparently be analyzed as prefixed a=q=ux-a-, q=ux-a-, where a= is the spatial preverb ʔa= (ʔ is often dropped in modern Aghul dialects, T. Maisak, p.c.). If so, the verbal root =ux-a- must be regarded as an ablaut variant of =ix-a-, discussed above. Theoretically, however, one can treat (a)quxa- as the analytic construction aqu xa-, where aqu is the regular past participle from the verb aq- '?' and xa- is the common auxiliary verb 'to become'. In this case, Gequn qu(-)xa- is the result of sporadic vowel reduction, on which see [Suleymanov 1993: 42 f.]. Indeed, the Tabasaran (closest relative of Aghul) data can speak in favour of the postulation of the Aghul verb aq- 'to lie' (thus [NCED: 264]), but all Aghul sources (including [Kibrik & Kodzasov 1988] and [Suleymanov 2003]) quote (a)qux-a- as one synthetic stem. Because of this, we prefer to follow the prefixal analysis (following [NCED: 279]).
No verb for 'to lie (animated subj.)' can be assuredly reconstructed for Proto-Aghul, but the prefixless verb aχ-a- is safely reconstructible as the Proto-Aghul term for 'to sleep' q.v. (aχ-a- 'to sleep' is retained in Burshag Koshan, Gequn and the Tsirkhe subdialect of Proper Aghul). Maybe the Burshag Koshan and Gequn situation is primary (aχ-a- with polysemy: 'to lie; to sleep'), and aχ-a- was also the basic Proto-Aghul verb for 'to lie (animated subj.)'.
Kibrik & Kodzasov 1988: 88. Apparently applied to both sg. and pl. subject.
The system in the Khanag subdialect is more complicated: d=ˈaq- 'to lie; to lie down' (sg. subj.) [Uslar 1979: 656, 997; Dirr 1905: 163, 232], as opposed to d=ˈax- 'to lie; to lie down' (pl. subj.) [Uslar 1979: 659, 997; Dirr 1905: 162, 232].
The same in the Khiv subdialect: d=ˈaq- {дахъуб} 'to lie; to lie down' [Genko 2005: 59]. The same in Literary Tabasaran: d=ˈaq- {дабхъуб} 'to lie; to lie down' [Khanmagomedov & Shalbuzov 2001: 137].
Apparently all Southern verbs are applied to both sg. and pl. subject.
Common Tabasaran notes:
We presume that Tabasaran verbs which are quoted in [Kibrik & Kodzasov 1988] for the concept 'to lie down' actually possess polysemy: 'to lie / to lie down'.
The Khanag opposition d=ˈaq- 'to lie (sg. subj.)' / d=ˈax- 'to lie (pl. subj.)' can be either a Proto-Tabasaran archaism, lost in other subdialects, or a local introduction.
The same in Literary Lezgi: qːat=kːˈi- [imperf.] / qːat=kːˈa- [perf.] {къаткун} 'to lie / to lie down' [Talibov & Gadzhiev 1966: 182; Gadzhiev 1950: 331; Gaydarov et al. 2009: 181; Haspelmath 1993: 42, 501, 522; Gyulmagomedov 2004, 1: 399] (it must be noted that clusters like tk are prohibited in the literary language [Haspelmath 1993: 47]).
Cf. in the dialects of the Samur group: Migrakh (subdialect of the Doquzpara dialect) qːatː=xˈa- [perf.] 'to lie' [Meylanova 1964: 249], Jaba qːat=ka- [perf.] 'to lie' [Ganieva 2007: 196 sentence 68], Qurush qːat=ka- [perf.] 'to lie' [Ganieva 2008: 256 sentence 128, 257 sentence 143].
The fluctuation k(ː) ~ x seems strange, but these verbs should hardly be kept apart from each other. One possibility might be a sporadic dissimilative fricativization tk > tx in the cluster.
Initial qːat(ː)= (i.e. qːa=t(ː)=?) are desemanticized spatial prefixes.
Proto-Lezgian:*ʔaχär-2
NCED: 273. Distribution: The verbs for 'to lie' and 'to sleep' must be analyzed together with each other. The basic data can be summarized as follows:
'TO LIE / TO SLEEP'
Proto-CA-Udi
Archi
Kryts
Budukh
Tsakhur
Rutul
Aghul
Tabasaran
Lezgi
*ʔ[a]s(ː)ʷɨn- [NCED: 1037]
lie/sleep
sleep
sleep
*ʔaχär- [NCED: 273]
lie, sleep (complex verb)
sleep
sleep
sleep
lie/sleep
sleep
*näwƛʼ [NCED: 619]
sleep (complex verb)
*ʔaqʼel- [NCED: 264]
lie
lie
*ʔeɬːʷɨ- [NCED: 278]
lie
k=utʼ-lie
*ʔaqɨ-(?)[NCED: 264]
lie
*ʔikʷän- (~ -l-) [NCED: 644]
lie
First of all, attention should be paid to two semantic isoglosses, which seriously obscure the picture. The first isogloss is the polysemy 'to lie / to sleep' (usually with the shift 'to lie' > 'to sleep', but not obligatory). It seems ancient, since it affects all of Lezgian (including Caucasian Albanian) as well as some other Dagestanian languages. The second isogloss is the derivation 'to fall' > 'to lie' (> 'to sleep'), which affects the Samur territory (Nuclear Lezgian) and seems relatively recent.
Two main candidates for the Proto-Lezgian meanings 'to lie' and 'to sleep' are *ʔ[a]s(ː)ʷɨn- and *ʔaχär-. Before discussing them in details, some clearly innovative formations should be ruled out.
In the Caucasian Albanian-Udi branch, 'to sleep' can be expressed analytically as 'to be in sleeping', in conjunction with the Common Proto-Lezgian noun *näwƛʼ 'dream, sleeping' [NCED: 619]. This formation competes with the verb bas-kʼ-esun (< *ʔ[a]s(ː)ʷɨn-) 'to lie / to sleep' in both Caucasian Albanian and modern Udi, but apparently such an analytical construction is a relatively early innovation of the Caucasian Albanian-Udi branch.
In South Lezgian (Kryts, Budukh), the meaning 'to lie' originates from 'to fall, go sprawling' (cf. synchronic polysemy in Budukh). The original meaning of *ʔaqʼel- [NCED: 264] was something like 'to dangle, shake', as follows from the same semantics in Aghul-Tabasaran, on the one hand, and in some other North Caucasian groups, on the other.
In Tsakhur, 'to lie (animated subj.)' is expressed by *ʔeɬːʷɨ- [NCED: 278], whose original labile meaning was 'to put / to lie (inanimate subj.)', as follows from its Lezgian cognates. Additionally, this root can acquire the polysemy 'to lie / to sleep' in some Tsakhur dialects.
In Rutul, the original verb for 'to lie' seems to be k=utʼ-, without further etymology. This stem tends to be superseded with the root *ʔikʷän- (~ -l-) [NCED: 644] 'to fall, go sprawling' in modern Rutul dialects.
In Aghul dialects, the most complicated situation is observed with four or five verbs for 'to lie', three of which are also attested in the meaning 'to sleep'. It seems that the most economic scenario is to reconstruct *ʔaχär- with Proto-Aghul polysemy 'to lie / to sleep'. Other verbs are recent dialectal introductions: *ʔarɬɨ- [NCED: 602] 'to fall' > 'to lie', 'to sleep'; *ʔarkɨr- / *ʔerkɨr- [NCED: 266] 'to fall down' > 'to lie', 'to sleep'; *ʔeɬːʷɨ- [NCED: 278] > 'to lie'.
In Tabasaran, 'to lie' is expressed with *ʔaqɨ- [NCED: 264], whose original meaning could be 'to fall' vel sim., this root looks rather problematic etymologically: pace [NCED], Archi =ˈaχa- 'to lie' can be satisfactorily etymologized as *ʔaχär-, whereas Aghul aqu xa- 'to lie' should rather be analyzed as prefixed a=q=ux-a-.
In Lezgi, *ʔikʷän- (~ -l-) [NCED: 644] 'to fall, go sprawling' shifted to the meaning 'to lie' (the same development as in Rutul).
Finally, we can return to *ʔ[a]s(ː)ʷɨn- [NCED: 1037] and *ʔaχär- [NCED: 273]. The first one, *ʔ[a]s(ː)ʷɨn-, denotes both 'to lie' and 'to sleep' in Caucasian Albanian and Udi. It also survived in two Nuclear Lezgian languages (Tsakhur, Lezgi), where it means 'to sleep'.
The second one, *ʔaχär-, means 'to lie' in Archi (with the synchronic derivative 'to sleep'), but 'to sleep' in the bulk of Nuclear Lezgian.
In such a mirror situation, external North Caucasian comparison should be involved. The external data point to the primary meaning 'to sleep' or 'to dream' for Lezgian *ʔ[a]s(ː)ʷɨn-, thus we postulate this root as the Proto-Lezgian term for 'to sleep'. This stem acquired the polysemy 'to lie / to sleep' in the Caucasian Albanian-Udi branch (due to the common areal isogloss), but survived as 'to sleep' in some West Lezgian (Tsakhur) and East Lezgian (Lezgi) languages.
On the contrary, Lezgian *ʔaχär- originates from the Proto-North Caucasian root with the meaning 'to fall'. Thus, we postulate *ʔaχär- as the Proto-Lezgian root for 'to lie', assuming the shift 'to fall' > 'to lie' in Proto-Lezgian. It was lost as a verb in the Udi branch after the verb 'to sleep' acquired the polysemy 'to sleep / to lie'. In most Nuclear Lezgian lects, *ʔaχär- primarily meant both 'to sleep / to lie' (the isogloss of polysemy), but currently 'to lie' is normally expressed by various verbs for 'to fall' (a more recent semantic isogloss). Additional evidence for the original meaning 'to lie' is the Vartashen Udi adjective b=arχi 'transversal, horizontal' (< *'lying') [Gukasyan 1974: 71] with the fossilized class prefix. Such a scenario is not straightforward, but seems the most economic one.
Kibrik & Kodzasov 1988: 88; Shaumyan 1941: 141. The two former stems are from [Kibrik & Kodzasov 1988]; the latter one q=ark-i- is from [Shaumyan 1941: 141] (with explicit polysemy: 'to lie / to lie down'). Note the Burshag stem ʢ=ark-i- 'to sleep' q.v., modified with another prefix.
The same root in the Khudig subdialect: H=ark-i- 'to lie' [Shaumyan 1941: 133 f., 149] ("He [the third son of the king] used to lie in ashes") and q=ark-i- 'to lie; to lie down' [Shaumyan 1941: 141] ("He lay on the bed").
Build tree