Meaning:1 to roll (dough) 2 to wrap 3 to roll, turn round
Tabasaran:a-ḳ-uz 1
Rutul:ɨ-rḳ-as 2 (Khniukh.)
Tsakhur:al-i-ḳar-as 3
Comment: In Tab. cf. also ilč̣-i=ḳ- (Düb. ilṭ-i=ḳ-) 'to roll' (with an expressive preverb). The Tab. and Tsakh. forms reflect the PL Ablaut *a/*i; -ɨ- in Rut. is not quite clear.
Comment: The verb itself is preserved only in Lezg. and Tab. dialects (therefore the reconstruction is rather tentative: we can not establish whether there was or was no medial -l- and final -r). However, several languages reflect the old deverbative *ʔač̣i-l (or, with reduction, *č̣al / *č̣il) 'weed': cf. (beside the listed forms) also Lezg. Khl. hač̣al, Tab. ač̣al (Kand. ač̣il, Düb. č̣al), Ag. Fit. ač̣il 'weed'.
Comment: The Tab. and Ag. compound verbs contain the old deverbal noun from the stem *ʔilʁV- > *ʁV- (with contraction). Cf. also Ag. Bursh. ʁawur-di arki-s 'to understand'.
Comment: Cf. also Rut. (derived) a=lḳa-d 'lame', Tsakh. Gelm. ḳawa-na id. The Tsakh. and Ud. forms are metathesized.
Vinogradova-Klimov 1979 proposed to consider the Tsakh. and Ud. forms as loans from Armenian (Arm. kaɫ 'lame'); but the Armenian word has no secure IE etymology and should rather itself be regarded as one of the numerous Caucasian loanwords.
Comment: Cf. also Rut. s-ɨ=rq̇Ia- 'to be spoiled'. Phonetically and semantically similar forms are: Rut. l-i=q̇a- 'to ferment (of dough)', Ag. Bursh. jiq̇ʷa- 'to curdle, coagulate' - but these forms probably had been influenced by a productive PL root *ʔirq̇är- 'to reach, to become ready' (q.v.).
Comment: Obl. base *ʔalχIa- or *ʔalχIe- (Rut. arχIɨ-, Tsakh. arχIa-).4th class in Rut., but 3d class in Tsakh. Medial -r- before a uvular in Rut. and Tsakh. goes back only to PL *-l-.
Comment: The verbal stem is preserved only in Bud. As a nominal stem *ʔalχʷVr / *ʔilχIʷVr it is reflected in Ag., Tsakh. arχʷ (a back-formation from a form = Ag. plur. arχʷar with -r reinterpreted as a plural suffix), Kryz. ħür- (now only within a compound verb), Bud. ħure 'diarrhea'.
Variation between *χʷ / *χIʷ is probably explained by the root's expressive character.