Comment: The PL paradigm must be reconstructed as *kac̣a ( > Ag. ḳac̣ with assimilation), obl. *koc̣ä- ( > Tsakh. obl. kɨc̣a-, spread also to dir. kɨc̣a). The root must have denoted some small bone.
Comment: The root reveals a strange vacillation of initial *ḳ- (Tab., Ag.), *k:- (Tsakh.) and *k- (all other languages). The oblique base must be reconstructed as *ḳalče- (cf. Ag. ḳarči-, Rut. kači-, Tsakh. gače-); medial *-l- is postulated because of the liquid's preservation in Kryz. and Bud. (*-r- should have vanished in this position). The word belongs to the 3d class in Rut. and Tsakh., but to the 4th in Kryz. and Bud.
Comment: Cf. also Ag. Bursh. ḳam 'sour cream'. The Tab. form reflects a *-j-suffix (*ḳamVj). -u- in Ag. (Rich.) probably reflects the old oblique base vocalism in the paradigm *ḳam(a), *ḳomV- (cf. Ag. ḳuma-), later transferred to the direct base.
Comment: The Ag. form is irregularly transformed (note that Shaumyan often records glottalised stops as tense, thus the real form is probably gimiḳ < *ḳamiḳ, with insertion of epenthetic -i-).
Comment: 4th class in Archi. The root tends to contaminate with PL *ṭäInḳ 'drop' q.v. (which explains the irregular -e-vowel in Archi). They must be, however, reconstructed separately (they are opposed, e.g., in Ag. Bursh.: ṭinḳ vs. ḳanṭ, and have different external parallels).
Meaning:1 group, heap, crowd 2 grove 3 small sheep-flock
Lezghian:ḳaṗal 1
Tabasaran:ḳaṗal 2
Kryz:ḳäbäl 3
Budukh:ḳǝṗǝl 3
Comment: In Bud. ḳǝṗǝl is the MSU recording; Meilanova has ḳapal (either another variant or a misspelling). 3d class in Kryz. (the Kryz. word also means 'cloud'), 4th class in Bud.
Meaning:1 rolling pin 2 battledore, roller 3 stick
Lezghian:ḳar 1
Tabasaran:ḳar 2
Agul:ḳar 3
Rutul:ḳar 3
Comment: Obl. stem *ḳarä- (cf. Tab. ḳari-, Rut. ḳara-). 4th class in Rut. In Lezg. there also exists a word ḳaras (Khl. ḳaras) 'wood, firewood'; one wonders, if it is not a compound of *ḳara + *ʔIʷes: 'wood, firewood' (q.v.).
Comment: Cf. also Lezg. Khl. ḳaš 'hammer', Ag. Burk. ḳaš 'stick'. Obl. stem *ḳašɨ- (cf. Tsakh. ḳašɨ-, Ag. Burk. ḳašu-). 4th class in Tsakh., 3d class in Shakh-Dagh. Perhaps we should also add Ud. k:oč:I 'handle' (possibly reflecting an early variant *ḳač̣ʷ, cf. the external evidence) - although Ag. ḳuč̣ 'handle' should be best considered as a loan from Dargwa.
Comment: 4th class in Archi. Cf. also Ag. Bursh. ḳiṭa id. Agul reflects a form with vocalic metathesis (*ḳiṭaj < *ḳaṭij, with a regular shift -i- > -e- before -a- in the next syllable); the same in Lezg., but with another suffix (kṭɨ-r < *ḳiṭV-r). In Ud. -ṭ- > -d- as a result of dissimilation.
Comment: Cf. also Tab. Kand., Düb. ḳaḳul. Tab. has a suffix, belonging originally to the oblique stem (cf. Rut. ḳäʔäl-). Final -ʔ in Rut. is a result of dissimilation (cf. similarly ḳäʔ 'top' < PL *ḳäḳʷ). 4th class in Rut.
Meaning:1 small stone, gravel 2 sheep's dung 3 dry clay
Lezghian:kḳal 1
Tabasaran:ḳeḳel 1 (Düb.)
Rutul:ḳäḳäl 2
Tsakhur:ḳaḳalaj 1
Kryz:ḳɨḳäl 1
Budukh:ḳɨḳǝl 3
Comment: Cf. also Rut. Khn. ḳɨḳal 'small stone' (the meaning "sheep's dung" in Luch. is obviously a result of contamination with PL *ḳälḳäṭ q.v.).
See Лексика 1971, 187; Хайдаков 1973, 72; Талибов 1980, 289.
Comment: Both in Lezg. and Ud. suffixed forms are observed (probably original pluralia). Ag. ḳeḳ should be kept distinct from the reflexes of PL *q̇Iäq̇:Iʷ 'throat; top of boot' q.v. (in the Burk. dialect both ḳeḳ and q̇eq̇ are used for "top of boot", but in the Richa dialect ḳeḳ is 'top of boot', while q̇Iaq̇ is 'throat').
Comment: The meanings vary greatly, but generally imply some small thing: Lezg. "a small piece of wood"; Tab. "trigger (of a gun)"; Rut. "dung (of sheep)" (cf. also Rut. Khniukh. ḳerḳel id.); Tsakh. "ankle"; Kryz. "dung (of a donkey, horse)"; Bud. "small lumps".
An expressive reduplicated stem (*-ṭ should probably be treated as an expressive suffix); all the listed forms are quite impossible to separate from each other, thus the supposition that Lezg. ḳerḳeṭ and Tsakh. ḳerḳeṭ are borrowed from Arm. k:ǝrk:it: "hard" (Виноградова-Климов 1979, 158) is obviously wrong.
