Comment: The Archi word is an old nominal derivate and points to original *χ:ʷ (cf. plur. anχ:-um). The Lezg. and Ag. forms, however, favour the reconstruction of weak *-χʷ-. The reason for this is not quite clear (influence of other verbs of "beating", e.g. PL *t:-arχV-,*jirχIa- q.v. ? - The latter would explain also the irregular pharyngealisation in the Agul form).
Meaning:1 to plough, to sow 2 tillage, arable field
Lezghian:c:a- 1
Tabasaran:u=z- 1
Agul:uz- 1
Rutul:=äza- 1
Tsakhur:eza- 1
Kryz:=iz- 1
Budukh:s-uz- 1
Archi:=ac:a- 1
Udi:ez 2
Comment: The verb belongs to the weak verb class in PL. Cf. also Lezg. pret. c:a-na, Tab. Düb. u=ʒ-, Ag. Bursh. uza-, Bud. dur. s-ize-ǯi, and the derivate PL *ʔic:an 'tillage' (Lezg. c:an, Tab. izan, Rut. Mishl. jiʒan, Bud. jizen). The root reveals a rare PL Ablaut type *ä/*i (as opposed to the much more often *i/*ä Ablaut).
Comment: Cf. also Lezg. k-iče-z 'pour, sprinkle', Bud. s-eχči 'jump', q:-eči 'vomit'; Rut. s-a=če- 'to pour, strew'; Tab. h-e=č- 'to winnow'. Semantic shifts ('flow,pour' > 'jump, fly' or > 'vomit') are quite usual. It is interesting to note an expressive reduplicated stem in Lezgian (čarčar 'waterfall'), possibly derived from this root and identic to PA *čʷarčʷa ( ~ o) id.
Comment: The root normally does not have class prefixes, thus the Ag. form with r- is probably secondary (the same with Tab. Düb. rič̣̌ri id.). In Ud. cf. also aIjiI-n 'ferment, yeast'.
Meaning:1 to soak 2 to urinate 3 to melt 4 liquid 5 green
Tabasaran:u=x- 2 (Khiv.)
Agul:d-ixe-f 4
Rutul:s-ä=xa- 1
Tsakhur:al-e=xa- 3
Budukh:s-oxr- 2
Udi:dχi (Vart.) 5
Comment: The verb belongs to the weak conjugation class. Cf. also Rut. s-ä=xɨ-d 'wet' (participial form), q-i=xe- 'to gulp, drink greedily', Ud. Nidzh. däj(i) 'raw, green' ( < däx(i)ji, cf. the Vart. form). The Tab. and Ag. vowels should be rather explained by assimilation (Tab. u=x-úz < *i=x-úz) and reduction (Ag. dixé-f < *dexé-f) than by Ablaut.
Comment: Cf. Arch. dur. =arq̇Iu-r. In Tsakh. (Mikik dialect) the MSU recordings distinguish between ǵ-o=q̇Ia- 'to rumple; knead' and g-o=q̇Iʷa- 'to rub'; it seems probable that these are only transcription differences.
Comment: Cf. also Lezg. erg. wiḳ-ini, Khl. wek,wek:-eni, Tab. Düb. jirk:aʁa, Ag. Bursh. jark:ʷaχ, Burk. urk:aʁ (the Tab. and Ag. forms go back to a suffixed *ʔärƛ̣:ʷ-aχ:). 4th class in all class-distinguishing languages. Obl. stem not clear (there is only the Tsakh. form oḱɨ- / oḱe- with variation).
Comment: Cf. also Lezg. pret. g-ata-na, Tab. Düb. j-at- 'beat', k-at- 'touch' ( > 'smear' in literary Tab.), Ag. Tp. ata-s 'beat', Bud. q:-ut- 'to hit (repeatedly)', Rut. k-ä=ta- 'to touch' ( = Tab. k-at-), s-ä=ta= 'to throw'. Labialisation of PL *-tʷ- is suggested by the Arch. durative form =artu-r. Not clear are the reasons for *a/ä vowel alternation. The verb belongs to the weak conjugation in PL.
Comment: Cf. also Arch. pres. babc̣u-r. The Kryz. morphological structure is not quite clear (other languages do not point to an *-r-conjugation), otherwise correspondences are regular. Cf. also derivates: Arch. abc̣on 'shavings', Lezg. ruc̣u-gul 'chisel', Ag. (Bursh.) ruc̣el, (Fit.) c̣ela, (Burk.) c̣ala 'chip, sliver', Kryz. c̣ɨl-äṗ (with an originally plural suffix *-Vp:) id., possibly also PL *q:-uc̣a (*q:-Vwc̣a) id., reflected in Lezg. Khl. q:uc̣, Rut. q:uc̣, Tsakh. ʁuc̣a. A derivative like *ʔV(w)c̣Vl- can perhaps also be observed in Kryz. c̣ɨl-ḳäm (and, with irregular transformations, also in Bud. c̣iniḳ < *c̣ilnVḳ < *c̣il-ḳVn) 'chip', although the second part of the compound is not clear.
Comment: Arch. dur. dabʎu-r, Tab. Düb. q-i=x-. In Ag. the root merged with -ix- 'to put' (see PL *ʔeʎe-), but the two roots are still distinguished in the Burshag dialect (iši- 'to put' vs. -(i)xi- in alt-xi-s 'to unlock', the difference in reflex being conditioned by the following PL vowel).
Comment: Cf. Lezg. pret. χa-na, Tsakh. q-a=χa- 'to chop, split', qɨta=χa- 'to tear off, break off', ǵ-o=χa- 'to mow'. The verb belongs to the strong class in PL. In Rut. and Tsakh. the stem is used only with preverbs (the notation ǵa=χ:-,qɨta=χ:- in the MSU edition is faulty: cf. ibid. qa-χ-).
A derivate from this root is PL *(ʔä)χV-r(a) 'wound', reflected in Lezg. χer, Kryz. χɨr.
Comment: The Rut. verb is used also with different preverbs, cf. ʔe=ʔe- 'to push in, lock', eʁ-e=ʔe- 'to take out', laʔl-e=ʔe- 'to lift, raise', keʁ-i=ʔe- 'to choose', q-i=ʔi- (dur.) 'to return' etc. The Bud. form is probably an exact parallel of Rut. ʔe=ʔe- (with the ʔ- preverb).