COMMENT: The comparison seems quite satisfactory both phonetically and semantically (for phonetic reasons Shagirov's proposal to compare PAK *pš:a- with Av. =ac- 'to knead' etc., see Shagirov 2, 35-36, should be declined). See Абдоков 1983, 170.
COMMENT: Cf. also Hurr. eššǝ 'horse', see Diakonoff-Starostin 1986, 34. The first (weak) syllable with the initial laryngeal was dropped in Av., Lak. and Khin. (and, of course, in PWC where it is the normal reflex), but is preserved in PD and PL. Medial -r- in PD does not represent the original *-n- (which was probably dropped); it is rather a trace of an oblique stem *ʔurči < *ʔuč-ri-.
One of the most secure common NC roots. See Чарая 1912, 48-49; Trubetzkoy 1930, 277; Балкаров 1964, 97; Шагиров 1977, 2, 141, Abdokov 1983, 124. Charaya notes also Kartvelian parallels: Georg. aču,ači 'interjection (addressing a horse)', ačua '(child.) horse' - most probably NC loanwords (cf. analogous loans of *jǝ̄mcō 'ox'). Abdokov (loc. cit.) suggests also an etymological connection of the root with *=ăč_wV 'to bring, carry' (q.v.), but there are phonetic problems with this solution.
COMMENT: A rather complicated case, but most of the reflexes can be more or less satisfactorily deduced from the protoform *ɦmVjć̣_wĂ (-j- to account for PN *-sṭ-). The Lak and PD form demonstrate the adjectival prefix *q̇V- (also present in a number of other adjectival stems); Lak. q̇urč̣i- < *q̇-Vmć̣-rV- (cf. the PL form; *-rV is a frequent suffixed morpheme in adjectives).
Note the frequent recurrence of reduplication (also typical for many adjectival stems): PTs *čača-, PL *č̣iIrč̣ʷV-m- (reflected in Rut. and Tsakh.), PAT *c̣ʷǝc̣ʷǝ-.
COMMENT: Reconstructed for the PEC level. Vocalism is hard to reconstruct (probably because of Ablaut), although the shift *m- > *n- in PN suggests an original front vowel. In PTs one has to suppose a metathesis *ƛam- < *mVƛ-.
COMMENT: Reconstructed for the PEC level. Correspondences are regular. The PL form ( > Rut. niggä) is a seeming exception: the Rut. form, reflecting the expected PL *neƛ̣Vj, was obviously modified under influence of Rut. niGGä 'strawberry' (q.v.).
COMMENT: A connection between PEC *ɦnǝ̄q̇_wǝ̄ and PWC *qʷǝnV (first suggested by Bouda 1948, 193-194) is very probable, although phonetically not quite regular: we should expect strong *q:ʷ in PWC, and reasons for its weakening are not yet clear.
COMMENT: Reconstructed for the PEC level. The root is best preserved in the Western area; however, its remnant in PL (*nVḳʷV-) makes the common EC reconstruction very probable. The Khin. form ḳä is possibly a reduction of original *HVnḳä; a trace of nasalisation may be possibly observed in the obl. base ḳi-mi-. Some problems are raised by the cluster *-ṭq̇- in PN; perhaps *ħāṭq̇a is an originally suffixed form ( < *ħāḲ-dV with a former plural suffix).
COMMENT: Except the metathesis in Lak. (rather usual in the HRVCV structure; regularly *naša-,*našu would be expected), correspondences are regular. In PWC there is a sporadic labial prefix (as in many other cases, possibly reflecting class prefixation).
COMMENT: PWC reflects the root with loss of labialisation; otherwise all the correspondences are regular. The root was obviously denoting a small bird in PNC; the shift to "eagle" in PD is a result of development "small bird" > "bird (in general)" (as in some Lezg. languages) > "bird par excellence" = "eagle". In any case the Darg. form is impossible to separate from this root.
COMMENT: Correspondences are regular. This etymology for PAK *ʡʷǝ 'mouth' is phonetically preferable to Trubetzkoy's comparison with Av. abize 'say' (see Trubetzkoy 1930) or Balkarov's comparison with Av. ḳal 'mouth' (Balkarov 1964, 101). Abdokov (1983, 76) correctly compares the WC form with Gin. haqu (misspelled as haq̇u), but adds also Av. ḳal and other unrelated EC forms.
COMMENT: A rather typical example of the *HCVRV̆ root structure. The Darg. form has a rather complicated history: it is probably a dissimilated plural form (*ʁIam-bV < *q̇em-bV with secondary pharyngealization before a front vowel), supplied additionally by a suffixed *-l.
COMMENT: A good common NC root with regular phonetic correspondences. See Trubetzkoy 1922, 241. Abdokov (1982, 82) compares the EC root (with a lot of confusion among EC reflexes) with WC *bǝ̃źʷa 'horn' - which is quite impossible for phonetic reasons.
COMMENT: Reconstructed for the PEC level. Correspondences are regular, except for the vocalism in Tsezian (*ɔ̃ or *õ would be expected): since the word is attested only in Gunz., the PTs reconstruction is not secure.
COMMENT: Reconstructed for the PEC level. The original meaning was probably "long hair, hair in tail" with the meaning "hair" preserved only in Darg. Correspondences are regular except for the vowel in PL (we would rather expect *ħirč:ʷ or *ħärč:ʷ); however, we must note that in this root most Lezghian languages lost the initial syllable, thus the PL vocalic reconstruction is not very certain.
PL has also another root, PL *(j)ic̣: (Arch. ic̣, erg. ic̣:i, Rut. dɨn < *(j)ic̣:ɨ-n) "goat's wool". It may reflect a separate PEC root which had merged with *ɦrĕʒ́_wǝ̆ in other languages (cf. the meaning "goat's wool" in Av. Chad. and PTs).
COMMENT: An Av.-And-Lezg. isogloss. The forms are comparable and point to a structure *HRVCV̆ (if we suppose a secondary loss of -r- in PL *hIic̣- < *hIirc̣-).
COMMENT: The root is best of all represented in the Av.-And.-Tsez. area, with a relic reflex also in Lak (we would have expected Lak. *t:VrVḳ, but the prothetic t:- probably did not appear in the word-medial position - since the root is present in Lak. only in a compound word). The etymology is certainly sound, although the vowels are hard to reconstruct.
COMMENT: Reconstructed for the PEC level. Correspondences are regular - except for the accent paradigm in Avar (paradigm C would be expected). This must be explained by supposing contamination of this root with PEC *ʔVms_wĕlʔē 'wild turkey' q.v. We must stress that these roots must be kept strictly distinct because of different reflexes of *sw (tense in *ʔVms_wĕlʔē, lax in *ɦswǟmV̆), although they tend to merge due to phonetic and semantic proximity.