Comments: Abaz. qʷadi and PAK *ħadá reflect a common WC compound *q́Iʷa-dV (with *dV 'die' q.v. in the second part). The Ubykh counterpart would be a morpheme like χʷa-, which is not attested. However, Ub. has dǝ-wá- 'to die' with the morpheme -wa- of unclear origin. Since Ub. w can go back to PWC *ʁ́Iʷ, it is probable that Ub. dǝ-wá- is a result of secondary assimilation: dǝwá- < *dǝ-ʁ́Iʷa- < *dǝ-q́Iʷa-, containing the same morphemes as Abaz. qʷadi and PAK *ħadá.
The original meaning of the morpheme *q́Iʷa (which occurs only within compounds) was probably 'to be killed, slaughtered' (cf. the external evidence).
Comments: PAT *qʷacV (cf. also Bzyb. á-χʷac). The Ubykh form is metathesized and partially reduplicated. Abdokov (1983, 106) cites also a dialectal Kab. form ħandǝr-qʷasa 'burdock', which may be actually borrowed from Abaz.
Comments: PAK *qʷǝ-ʒ́ǝ (a compound with *-ʒ́V 'grain'; cf. also Bz. á-χʷʒ́); PAK *ħa. The correspondence PAT *qʷ : PAK *ħ : Ub. χʷ points to PWC *q́Iʷ.
Comments: PAT *wa-qǝ 'night', *wa-qa 'tonight'. The element wa- is the same as in *wa-c̣ʷǝ 'tomorrow', *wA-žʷǝ 'now'. Etymology of *qA is not clear yet.
Comments: PAK *q:aráwǝ / *q:ǝrǝ́wǝ. The Ub. form may be an Ad. loan (although nothing contradicts its genuineness, too). Dumézil (1963,18) is correct in assuming that the Abaz. forms (Ashkh. q̇ru etc.) are borrowed from Kabardian.
Meaning:to be benumbed,dumbfounded (by cold,surprise etc.)
Adyghe:q:ǝ-n
Kabardian:q:ǝ-n
Ubykh:q(ǝ)-
Comments: PAK *q:ǝ- (see Kuipers 64); Ub. sǝ-qǝ́-n. In Shapsugh the verb also means 'to harden' (which may be a contamination with another WC root, *q:ʷa q.v.).
Comments: PAT *ʡa (used in conjunction with the verbal stem-suffix *-la-); PAK *ʡa in *š́ǝ-ʡa- 'to be' (Ad. š́ǝʡa-, Kab. śǝʡa- with a locative preverb *š́ǝ-), *jǝ-ʡa- 'to have' (Ad., Kab. jǝʡa- with a pronominal preverb *jǝ-); Ub. a-s-q̇á-ʁ "je l'ai" (a compound stem q̇a-ʁǝ-; etymology of the second part see under *ʁV). The correspondence PAT, PAK *ʡ : Ub. q̇ points only to PWC *q́:I.
See Shagirov 1,190; 2, 150 (with literature and with right dismissal of many false etymologies).
Comments: PAK *ʡa-tǝ-; Ub. a-s-q̇á-śa-n. Both PAK and Ub. reflect compounds with unclear second components. Since, despite Dumézil (1932, 107-108), PAK *t can not correspond to Ub. ś, we have to admit that the meaning 'to lift, raise' is here expressed by the component PAK *ʡa-, Ub. q̇a- < PWC *q́:Ia- (or *q̇́Ia-), which must be therefore distinguished from the homonymous PWC root for 'hand' (q.v.).
Proto-West-Caucasian:*-q́:Ia (?)
Meaning:lative
Abkhaz:-q̇a
Abaza:-ʡa
Ubykh:? -ʁa
Comments:Ub. -ʁa is irregular (-q̇a would be expected)
Comments: PAT *(a)ʕʷA-. In Abaz. a secondary delabialisation (probably under influence of ħa-ra,ʕa-ħa-ra 'be heard') had occurred; the original variant is preserved within the compound ʒǝr-ʕʷ-ra 'listen' (see under *ŁA 'ear').
PAK *ʡʷǝ- (there exists also a "centripetal" form *ʡʷa- > Ad., Kab. ʡʷa-n). Ub. 3 sg. a-q̇ʷǝ́-n.
The same root is also present in the common WC compound *ĺA-q:́IʷA- 'listen' (see under *ŁA 'ear').
Comments: PAK *q:ʷaLa (cf. also Ad. q:ʷaLa-bzǝw = Kab. q:ʷaLa-bzuw). Despite Shagirov (1,232) the Ub. form can not be regarded as an Adygh loan (because of pharyngealisation); see Mészáros 1934, 352, Abdokov 1973, 55. Although loss of labialisation in Ub. is not quite clear, other correspondences are quite regular.
The comparison of PAK *q:ʷaLa and Abkh. a-laħʷa 'raven' (Lomtatidze 1961, 121) is untenable; see Shagirov 1,232, Klimov 1967, 302.
