Notes: As in several other verbal roots, labialisation in PWC is here secondary (probably being a remnant of class prefixation). Otherwise the correspondence is perfectly satisfactory, both phonetically and semantically.
The initial laryngeal in PL (together with PD *ʔ- being derived from *ɦ-) should be probably treated as a historical preverb. Another preverb, *l-, is frequent with this root (cf. the exact parallelism between the Lak. and the PN forms).
There are no certain traces of this root in PTs; cf. perhaps Tsez. iχ-ju 'thick'. It is quite isolated within Tsezian, and may reflect the original adjective 'high' (the correlation 'high' - 'thick' is not unusual in NC languages).
Notes: In individual subgroups the root tends to contaminate with *=arq̇Vr 'to reach' or *=irq̇wV 'to ripen, grow' (q.v.); nevertheless, it differs from those roots in PN, and the precise parallelism of the *-n-conjugation in PA and PL leaves no doubt that it was a distinct root (one wonders, however, about its relationship to PEC *Hnäq̇wV̄ 'big' q.v.). See Абдоков 1983, 144.
Notes: One of the nominal roots with originally changing class prefixes. The labialized vowel reflex in PN and Lak. is probably due to the influence of labialized class prefixes.
The root should be kept distinct from the synonymous *ʔwăχ_wV q.v.
Notes: Reconstructed for the PEC level. In Nakh languages the root *=āq(q)- 'to pull out' has also a meaning 'to dig out' and thus may reflect a merger of two roots: PEC *Harq_wV 'dig' and *Hiq_V(r) 'take, pull out' (q.v.).
Notes: Reconstructed for the PEC level. Correspondences are regular. In [Diakonoff-Starostin 1986, 58] the root is compared with Urart. χarχar-š- 'to demolish' ( < *'sweep off'?).
Notes: Reconstructed for the PEC level. Both PL and PD reflect also the Ablaut grade *=Hirq̇_V(n). Reasons for conjugation variation (-0 and -n) are not yet clear. Cf. also Hurr. tiɣ-an- (causative) 'to show' ( < PHU *tiχ-), where t- reflects *r (see Diakonoff-Starostin 1986, 29).
Notes: The comparison seems probable phonetically and semantically (the semantic correlation 'to pour, wash' - 'to weep' is quite common). The PWC root, however, can belong either here or to *=imc̣_Vn q.v. [We should note that the Av.-and. forms reflect rather a form like *=Häć_Ăw with non-labialised -ć-; it is not clear, which is original - *=Häć_wĂ or *=Häć_Ăw.]
Notes: An expressive Av.-And.-Lezg. isogloss. The Archi form looks archaic, but since there are no parallels in other Lezgian languages, and since the correspondence between Av.-And. and Archi is not quite regular (we would rather expect PL *qIʷ > Arch. χIʷ), a borrowing is not excluded. Perhaps we should regard as a genuine parallel to the Av.-And. form not Arch. e=qIʷin-, but Rut. s-u=qIa- 'to shake, sway' (where -qI- regularly < PL *qIʷ).
Notes: Reconstructed for the PEC level. The Andian reflex is not quite secure (unclear is the -l-conjugation), but the Lak-Lezg. parallel is all right.
Notes: Reconstructed for the PEC level. "Lax" phonation (hence *-χ-) is suggested by PN, PA and PL reflexes; however, the PTs and Darg. forms point rather to *=Hēwχ_V(n). Reasons for this variation are not quite clear.
Notes: The original *-r-declension is preserved in PD; traces of it are perhaps also seen in PL (although the verb has a defective paradigm there). The verb may have had a second stem, *-HǝχVl- (or, *-HoχVl- with Ablaut) reflected in PD as *-uχVl- and in PA as *-iχʷVn-. The source of the PA labialisation is not quite clear - perhaps we should reconstruct *-HǝmχVr- or *-HǝwχVr- with a medial labial resonant.
Notes: Reconstructed for the PEC level. Correspondences are regular. The semantic correlation 'cold, coolness' : 'shadow' (occurring in PC, Lak. and PD) is quite usual in NC languages.
