Comments: PAK *sákʷǝ. For PWC also *s may be reconstructed (since there are no Bzyb. data available).
Despite Shagirov (1977) PAK *sákʷǝ can not have been borrowed from Ub. ša-q̇Iʷǝ 'hair on head' (for phonetic reasons); however, he must be right in considering Ub. sakʷ as a loan from Ad. The Abkh. form can be compared if we assume a metathesis - which is rather common in roots of this type.
Comments: PAK *ḳa (in Kab. preserved only within the compound "hen's seed" = "hen's egg"). Comparison with Ub. ć̣ǝ- (present in several compounds) proposed by Shagirov (1,160; 2,132) is impossible phonetically.
Comments: PAK *ḳač̣́ǝ́. Identifying *ḳa as 'tail' and treating the PAK form as 'empty-tailed' (thus in Shagirov 2, 134) is obviously a folk etymology. The Abkh. word may be, but not necessarily is, an AK loanword (phonetic correspondences are quite regular here).
Meaning:1 shirt 2 quilted jacket 3 ancient women's dress
Abkhaz:a-ḳ́ač̣ 1
Adyghe:č̣́ač̣́ǝ 3
Kabardian:č̣aṣ́ 2
Comments: PAK *ḳ́ač̣́ǝ́. The formal coincidence of the root with PWC *ḳ́ač̣ʷǝ 'short' is probably fortuitous (cf. the external data). Despite Shagirov 2, 134 the Abkh. form is not a loan from Adygh (the meaning is different, and in the case of borrowing we would expect Abkh. č̣́, not č̣).
Comments: PAT *-ḳǝ : here this morpheme is productive. It is used also within the numeral 'one' (Abkh. a-ḳǝ́, Abaz. za-ḳǝ). In Ub. it survived only within some forms like ca-ḳ́ǝ́ 'tooth' (*'a single tooth' = PAT *cǝ-ḳǝ) etc. In AK cf. perhaps Ad. -č̣́a 'deverbative abstract' (?)
Comments: PAK *c̣ʷǝ(n)-ḳǝ́ / *ḳǝ-c̣ʷǝ́ (in Ad. Bzhed. both ṣ́ʷǝnč̣́ǝ and č̣́ǝṣ́ʷǝ exist; cf. also both variants in Kab. wǝ-n-č̣ǝf̣- / wǝ-f̣ǝnč̣- 'to go out (as light)'). The root *ḳǝ is used within these two compounds with *c̣ʷǝ 'dark, black'; cf. perhaps also PAK *ḳa-ʁʷa / *ṗḳa-ʁʷa 'soot' (Ad. ṗč̣́aʁʷā-sa, Kab. č̣āʁʷa,č̣aʁʷāsa) - a compound with *ʁʷa 'smoke'. See Kuipers 1975, 33, 52, 53.
Comments: PAK *ḳǝ-š́ǝ́ ( = Ub. ḳ́ǝ-ś 'smithy'); Ub. a-ḳ́ǝ́-n,a-sǝ-ḳ́ǝ́-n. In PAK the root is reflected also in *dǝśa-ḳǝ (Kab. dǝśač̣) 'goldsmith'. See Shagirov 2, 135 (with literature).
