Notes: One of the cases of the non-trivial correspondence PEC *χw (with tense phonation) : PWC *Łʷ (cf. also *χ_Hwĕje 'dog'); note in both cases the existence of an oblique stem or derivates in -r-. See Абдоков 1983, 181-182.
For PEC several nominal derivates can be reconstructed:
2) *Hĕmχ_wV-rV, (with Ablaut and/or reduction) *Hĭmχ_wV-rV > *χ_HwVrV 'mill, mill-stone' > PN *ħēr, PA *χʷor-, PC *ʔaʁ:ʷǝr, Lak. hara-qalu.
3) *Hĭmχ_wV-rV > *χ_HwVrV 'meal, flour' > PN *ħāru, Av. χ:ara-b, PL *χ:Iur. This form probably differed from the second one in vocalism (cf. the opposition in PN), which is rather hard to reconstruct in this case.
4) *Hĭmχ_wV-wV̄ > *χ_HwVwV̄ 'mill, mill-stone' > Av. hobó, PA *ʔiχʷobu / *χʷobu, PC *ʁ:ʷǝbǝ,*ʁ:ʷǝbǝ-r, PL *χ:Iuw(a).
It is rather difficult to attribute the PD form *ʔurχ:ʷab / *ʔurχ:ʷam : the latter variant probably reflects (with metathesis) the second stem (PEC *Hĕmχ_wVrV 'mill'); the former variant may actually represent a contamination of the second and the fourth stems.
A NC source is probable for Osset. kʷɨroj / kurojnä 'mill' (the Indo-European etymology of it is hardly possible, see Abayev 1958, 612).
Notes: Reconstructed for the PEC level. Class prefixation in Lak. and Tsezian is probably secondary (cf. the prefixless structure *ħirc̣- in PA). PN *-b- is a former aspective (plural) marker.
The root *Hĕrc̣V- has a parallel in HU *asχ- ( < *arc- with a typical HU development): Hurr. ašχ- 'to raise; endow', ašχu/o 'high, upper', Ur. ašχ- 'to endow' (see Diakonoff-Starostin 1986, 47).
Notes: A widely represented common NC verbal root. It reveals a constant variation between -0 and -n-conjugations (in PA, probably also in PL); it is, therefore, not excluded that we deal here with a merger of two (or more) individual verbal roots (see above in the Andian part), which are now almost impossible to separate from each other. A possible old nominal derivate see under *ħG_unu.
Notes: The Bezht. form ƛ̣oq exactly parallels PA *ƛ̣olq̇:ʷV (an old compound, see above), thus this compound with *ƛ̣a- 'above' must be dated by the Av.-And.-Tsez. period. The PWC form has a frequent labial prefix (with a usual in this case delabialisation of the following consonant). The comparison seems reliable, both phonetically and semantically (an alternative comparison of Adygh forms with PEC *mōnqī 'breast' - see Abdokov 1983, 79 - is improbable for phonetic reasons).
It seems possible to compare also Lak. Khosr. q̇alaq̇u 'lid, cover' which is either a reduplication or a reflex of the same compound with *ƛ̣a- 'above' (if q̇alaq̇u < *ḳalaq̇u through assimilation).
Notes:Cf. also Hurr. -ae, Ur. -ai, -ae, -ajǝ 'adverbial, adjectival, gerundial or participial suffix'. Some of the modern forms without -j may actually reflect some other case (perhaps essive), or just a pure oblique stem.
Notes: Cf. also HU: Urart. zil(i)b- 'seed; descendants, kin' (see Diakonoff-Starostin 1986, 48).
An interesting common NC root. An original polysemy ("root, seed" : "kin, clan") can be reconstructed, because both meanings keep recurring in different subbranches (Nakh, Avar, WC, HU). The meaning 'leaf' in PC is obviously secondary - although phonetically PC *ƛib(u) goes back directly to *Hĭƛ̣_īwV, semantically it has merged with PNC *ƛ̣ăpi 'leaf' q.v.
Notes: Reconstructed for the PEC level. Vowel length is not clear (PN probably reflects an alternation of long/short root vowels); otherwise regular. Cf. also Hurr. χill- 'to speak' (see Diakonoff-Starostin 1986, 51).
See Trubetzkoy 1922, 241 (with incorrect Lezg. and Darg. parallels).
Notes: Reconstructed for the PEC level. Correspondences are regular. It is interesting to note that both in Chech. and in Inkh. the word (Chech. huma, Inkh. himon) also means "thing". In Chech. this is probably just a case of homonymy (huma 'thing' < PN *fu-ma is actually derived from an interrogative pronominal stem 'what'). The meaning of the word in Inkhokvari apparently reflects Nakh influence (and may be also affected by the presence of a similar interrogative stem - hi-bo 'which').
Notes: A common Av.-And.-Tsezian morpheme, with only sporadic parallels from other languages. Besides Khinalug, possible traces of this root can be perhaps found in Arch. inž 'self' (where -n- is absolutely superfluous if we talk about reflexes of *ʒ́ĭ q.v. - thus it may be an old compound of *HinV- + *ʒ́ĭ). If this morpheme is not an Av.-And.-Tsez. innovation, it must have played the role of the oblique base of one of the PEC reflexive pronouns (see *ʒ́ĭ and *č[ŭ]).
Notes: This root (originally meaning 'person, man') is seldom used independently (only in PA and some Tsezian languages). In EC languages it is most frequently used in a compound with *ćwijo 'man' ('man-person'); in PWC it became a morpheme of Nomina Agentis ("person doing smth."). The etymology seems phonetically and semantically quite plausible.
Notes: Despite the root's expressive meaning, correspondences are regular (except for the Darg. form, where one should rather expect -q:ʷ-) and the reconstruction seems probable. See Abdokov 1983, 136.