Comment: -r- in Lezg. warz probably comes from the oblique base PL *woc:-ra- / *wac:-ra-, cf. Lezg. wac:ra, Arch. bocro, Kryz. wuzur-. The Ablaut *-a-/-o- in the root is witnessed by Kryz. wäz : wuzur- and Tsakh. Gelm. waz : wɨza-. Belongs uniformly to the 3d class in all class-distinguishing languages. See Лексика 1971, 193; Гигинейшвили 1977, 75.
Comment: Cf. also Tab. Düb. warža, Ag. Bursh., Burk. waIrš. The correlation Ag., Tab. -r- : -0- in other languages points to PL *-l- (*-r- would have been preserved in this cluster by all languages).
Comment: The final -f in lit. Tab. (also in Düb.: warf) is quite unique and may reflect an occasional variant *warhʷ. Cf. the forms of other dialects: Kand., Khiv. warʔ; Ag. Fit., Burk. warw, Bursh. war.
Comment: Obl. base *worc̣:e- (cf. Ag. /with vowel levelling/ wart:u-, Rut. wɨtɨ- /with consonantal levelling: -t:- would be regularly expected/, Tsakh. wut:e-). The vocalism of the oblique stem in Rut. and Tsakh. penetrated into the direct stem. 3d class in Rut. and Tsakh.
Comment: 3d class in Rut. and Arch. In Lezg., Ag. and Rut. there occurred a metathesis of nasalisation (medial -n-, having lost nasalisation, changed to -r- which was regularly lost in Rut.).
Comment: Cf. also Lezg. Khl. wärχI, Tsakh. obl. woχIa-, Gelm., Tsakh. prop. wɨχI. 3d class in Rut. and Tsakh. Within the Lezghian group the roots *merχIʷ (q.v.) and *werχIʷ are in complementary distribution (the former - in Tab., Kryz. and Bud.; the latter - in Lezg., Rut. and Tsakh.) and appear to be one root with a strange variation *m-/*w- in Anlaut. Historically, however, they are two different roots having undergone contamination (the regular reflex of PEC *wēqwV in PL would have been *weχʷ which changed to *werχIʷ under the influence of PL *merχIʷ < PEC *mħĕrqwĕ q.v.).
