COMMENT: Reconstructed for the PEC level. Although not widely represented outside the Av.-And. area, the root seems quite reliable both phonetically and semantically.
It is interesting to note the root's full homonymousness with PNC *nHäƛ̣_wV 'blue' (q.v.): perhaps it had meant originally "a blue (or green) insect".
COMMENT: A Nakh-Tsez isogloss (with metathesis in PN: *χēn < *nē(r)χ). Medial -r- is to be reconstructed because of the PTs reflex *-χ- (otherwise *-χ:- would be expected).
COMMENT: A Lak-Darg. isogloss. The etymology is not very reliable, because the forms are very isolated, and do not fit into the paradigm of the 1st person pronoun reconstructed on other evidence (see *zō).
If we compare the paradigm of the 1st p. pr. in Dargwa dialects, it becomes probable that PD *nu is a secondary development < *du (or, possibly, *du-n, cf. PL *zon): cf. Ak. nom. nu, erg. nuni, dat. nab vs. Chir. nom. du, erg. di-c:e, dat. damī. Still, PD *nu- in the 1st p. plural *nu-š:a stays unexplained, as well as Lak. na (also < *da-n or *t:a-n, cf. the obl. stem t:u- ?).
An etymological possibility would be to consider the morpheme *nɨ̆ as originally a collective plural pronoun: cf. its use in PD *nu-s:a 'we' (exclusive), *nu-x:a 'we' (inclusive), *nu-š:a 'you (plur.)', quite parallel to Av. ni-ž,ni-ƛ: and nu-ž. Its usage in the 1st person plural could have influenced the 1st person singular in Lak. and Darg., and this influence was strengthened by a chance of secondary nasalisation (e. g. PD *nu < *du-(n)).
COMMENT: The original meaning was probably 'chief, ruler' ( > PWC 'god', Av. 'prince'). The semantic change 'ruler' > 'bride-groom' in most EC languages is usual (cf. the borrowed Turk. *bäg in Ag. beǵ, Kryz. bäg, Bud. bǝg, Khin. bäg, Ud. bäj 'bride-groom', the borrowed Pers. padšah 'king' in Av. Chad. parcaħ 'bride-groom' etc.).
In PL (Tsakh.) the root underwent metathesis: *nɨwca > *muca > *cuma. With analogous metathesis we can possibly relate here also Darg. ucm-ij 'ruler, prince'.
COMMENT: In PEC there occurred a labial assimilation (*nŭlxV > *mŭlxV); we must reconstruct *n- for PNC, because *mŭlxV would have yielded PWC *PVxV. Although not very widely spread, the root seems safely reconstructable (since the meaning 'asp' is constantly recurring in different subgroups, it may have been the name of the asp-tree).
COMMENT: Correspondences are regular, and this is undoubtedly a common NC kinship term. Cf. also HU *nas- > Hurr. naž-ardǝ 'concubines' (see Diakonoff-Starostin 1986, 37). The root has interesting parallels in other linguistic families: Semitic (Arab. niswa 'women', nisāʔ id. - whence Turk. nisa); IE (*snus- 'daughter-in-law') - thus being an old Eurasiatic "Wanderwort".
COMMENT:Also widely used for building denominative attributes. Cf. Hurr. -nnǝ 'suffix of substantivized relative adjectives', as well as Hurr., Urart. -nǝ 'ablative, ablative-instrumental'. [Lak. Dat. -n, PN Dat. *-na and PL directive *-na may go back to a separate directive case *-na].
COMMENT: The root is well attested in PN and PWC, but has left no traces in most EC languages. Khaidakov (1973) cites two other forms: Lak. maža 'oak-tree' and And. noži id., but if they are genuine (the source is not very reliable), they must be ultimately borrowed from Nakh (see above about the reconstruction of *č̣ in PN, which later changed to -ž- (-ǯ-) in Chech. and Ing.).
Further connections of the root are not quite clear. Borrowing from the supposed Proto-Iranian *nauča- 'some coniferous tree' (see about this root under *ɦwmĭc̣_Ĕ) is not very probable - both semantically (the root *nV̆č̣ē means only leaf-bearing trees in NC languages) and phonetically (because a glottalised *č̣ is reconstructed). Thus for the time being we prefer to regard this root as genuine - until further research is done.
COMMENT: The root is attested only in Nakh, but cf. Urart. nek- 'to flow, flow out' (see Diakonoff-Starostin 1986, 31).
The Urartian parallel makes a loan from Iranian improbable: thus Osset. nakä 'swimming' is probably a loan from Nakh, not vice versa (pace Abayev 1973, 152) and hardly goes back to *snāka-.
COMMENT: Lak. and And. have suffixed forms (in Andian reinterpreted under the influence of folk-etymology, see above); nevertheless, all the forms correspond rather well to each other, and the PNC antiquity of the root seems quite probable. The root is widely represented in other Caucasian languages (Georg. neswi 'melon' - later borrowed in Gunz. neso id., Balk. naša,narša 'cucumber', Osset. nas / nasä 'pumpkin') - but all those forms are most probably North Caucasian loanwords (none of them can explain the appearance of -r- in PN). See the discussion in Shagirov 1, 935, Abayev 1973, 161.
COMMENT: The PEC-PWC comparison belongs to Trubetzkoy (1930, 278; see also comments on p. 460; see further Абдоков 1983, 169), and seems quite satisfactory.