Comments: The Abkh. and Ub. form are in perfect correspondence. In both cases there is a compound with PWC *ʎ́a 'foot' (in Abkh. also a connecting morpheme -r-). The second consonant can also be reconstructed as *-c(ʷ)-.
Comments: PAK *k:ʷaχʷá. See Shagirov 1, 109, Kuipers 58. The Shaps. dialect here has variants k:ʷāχʷa / gʷāχʷa (where the form with gʷ- is irregular), the Abadz. dialect - gʷāχʷa (instead of expected kʷāχʷa). Forms with secondary gʷ- can be easily explained as secondary loans from lit. Kabardian, thus Shagirov's (ibid.) hypothesis about the word being borrowed from Mong. goxo 'hook' seems to have little ground.
Comments: There are several competing forms in Ub. (see Vogt 1963, 98): cǝgʷǝ́,čǝ́gʷ,č́ǝgʷǝ́ (also listed is čǝdǝ́ which is a quite separate root). The form č́ǝgʷǝ́ corresponds best to Abkh. a-gʷǝ́ž́ (if we assume a rather common metathesis). In PWC we must reconstruct *č́: and either *gʷ or *k:ʷ.
Comments: PAT *ḳʷǝṭǝ; PAK *k:at:ǝ́. The AT and AK forms are obviously related; we can reconcile discrepancies if we assume an initial structure with different laryngeal features of the two stops (tense *k: and glottalised *ṭ) with subsequent assimilations (*k:ʷVṭǝ > PAT *ḳʷVṭǝ and *k:(ʷ)aṭǝ > PAK *k:at:ǝ́).
Comments: PAT *kǝtV; PAK *kátǝ. The semantic correlation ('village' - 'sheep-fold') points to the original meaning 'enclosure'. Osset. (Dig.) kät 'an open outhouse close to the sheep-fold', despite Abayev 1,590, is probably borrowed from Adygh.
Comments: Isolated in Kab. with obscure etymology (see Shagirov 1,213; his hypothesis of borrowing from Oss. ḳord 'group' does not seem phonetically plausible).
