Comments: PAT *ć̣ǝʕa- (cf. Bzyb. a-ć̣ā-rá); Ub. sǝ-ʒʁá-n. The word is obviously connected with EC forms like Av. c̣eχ:é-ze etc., thus a hissing Anlaut is probable for PWC. We may assume that Ub. ʒʁa- is a modification of *ʒ́ʁa- (since Ub. ʒ́ does not occur in any consonant clusters), which, in its turn, is assimilated from *ć̣(ǝ)ʁa-.
Comments: PAT *c̣ǝla (cf. also Bz. á-c̣la). Abdokov (1983, 133) proposes a connection with a-c̣íā 'vegetation', thus suggesting a root *c̣ǝ-; however, accentology speaks against this analysis.
Comments: PAT *ć̣ǝnḳ́ǝ-ra / *ḳ́anć̣ǝ-ra (cf. also Bzyb. a-ḳanć̣ǝć̣ǝ́r, Ashkh. ḳanc̣ǝra); PAK *-c̣ǝka / *-c̣ǝráḳa (meaning of the component: Ad. q:ām-, Kab. ħam- is unclear).
An expressive root with metatheses and some irregular changes; however, obviously archaic (cf. external data).
Comments: See Kuipers 1975, 21. The word is quite isolated among WC languages; it seems, however, worth mentioning Abkh. a-č̣̌aʁʷa 'line' (if the original meaning was 'caterpillar' - cf. the external evidence, the semantic development in Abkhaz would be just the same as in Lezghian).
Comments: PAT *ć̣ǝʒ́ǝ (cf. also Bzyb. a-ć̣ǝ́ʒ́ 'nit'); PAK c̣ǝzʷǝ́ (cf. also Ad. Tem. c̣ǝžǝ 'a k. of fly'). Kab. > Abaz. c̣ǝžʷ 'maggot of a gadfly'. In PAT we should expect *-ʒʷ-; there probably occurred an assimilation *ć̣ǝʒʷǝ > *ć̣ǝʒ́ǝ.
Comments: PAT *ć̣a- (cf. also Bz. a-ć̣a-rá); Ub. 1st p. a-s-ć̣á-n. PAK *c̣ʡa-: initial cluster is reconstructed on basis of Shaps. ṣq̇a-. The element -ʡa is treated (by Lomtatidze 1972, 130-133 and Shagirov 2,152) as an affirmative affix. Whatever it was, it influenced the development of the initial affricate (all affricates in such clusters become palatal in PAK). The original c̣- (which is the normal PAK reflex of PWC *ć̣) is probably retained in Kab. c̣ǝ-xʷǝ-n 'be acquainted (with someone)'.
Comments: PAT *c̣a (cf. also Bzyb forms recorded by Marr: á-d-c̣a-ra 'to heap, put on top', á-c-c̣a-ra 'to add'); PAK *ḳa-c̣ǝ- (*ḳa 'egg'). The original meaning of the root is 'to put' (or 'to give'): cf. with other preverbs: Abkh. á-kʷ-c̣a-ra, Abaz. (jǝ)-kʷ-c̣a-ra 'to put'; PAK *na-c̣a- "to put an eye [ = Russ. положить глаз] > to envy, to aspire" (Ad., Kab. nac̣a-n).
Comments: A long (trisyllabic root) with usual in such cases metatheses and dissimilations. The PAT form can be reconstructed as *ḳʷajǝcǝ ( ~ -ć-: the Bzyb form is not attested), cf. also Abkh. á-ḳʷejcej-ra 'to glitter'; Abaz. č̣ʷec is an irregular transformation of *ḳʷajc > *ḳʷec. PAK has a compound *pχa- 'wood' + *c̣aḳá 'fire-brand'.
Comments: The PAT form is a compound *jǝkʷa-c̣ʷa (for the first part cf. Abkh. a-jkʷá dark'). -c̣ʷa can not be treated here as the same root as PAT *jac̣ʷa 'blue, green' (q.v.) for accentological reasons: a-jkʷá(j)č̣a would be expected in Abkh.).
The PAK word *c̣ʷǝc̣a 'black' goes back (as proposed by Шагиров 2, 103) to *c̣ʷǝ 'black' + *ṗc̣a- 'thick, dense' with secondary elision of -ṗ-; the root *c̣ʷA 'black' is used also in a number of other compounds.
Since exactly the same correspondence is present in the word for 'fire' (q.v.) one can not separate the Ub. form from the rest, but problems remain: we have to account for deglottalisation and delabialisation of the root consonant. We can, however, suppose that the original form is still preserved in the word for 'coal': Ub. ǵá(n)-c̣ʷa, cf. PAK *c̣ʷa-mǝč̣́ǝ id.
Comments: PAT *c̣ʷa (in Abkh. - a compound with *gʷǝ 'heart'). Ad. < PAK *c̣ʷǝ. Ub. also has a compound with ǵǝ 'heart' (cf. Abkh.) + bIa 'grey'. Note the loss of glottalization in Ub. (a misspelling of Vogt?); the normal glottalized form of the root appears in c̣ʷa-cá 'gall' (-ca is probably 'burning, bitter').
Comments: PAT *ja-c̣ʷa; Ub. c̣ʷa(n)-ḳ́ with the old singulative suffix *-ḳǝ (q.v.). Vogt lists also the variant jacʷa (with secondary loss of glottalisation). In PAK the roots 'star' and 'sky' (*zʷA q.v.) had "changed places": *-sʷa (going back to *cʷa with loss of glottalization like in Ub.) meant originally 'star', but lost this meaning in the compound *wa-sʷá (where *wa- means 'sky' by itself, see *wa).
Comments: PAT *c̣ʷǝ. Abdokov (1983, 133) compares also Kab. janc̣ 'long pole, stick' (with possibly the same first component as in jan-ṭǝr 'log'), which is somewhat dubious (at any rate, we would expect -ṣʷ in Kab., and the reason for delabialisation is not clear).
Meaning:1 dirt, dirty; excrements 2 to become dirty
Adyghe:ṣʷa-jǝ 1
Kabardian:f̣e-j 1
Ubykh:q̇́ǝ-c̣ʷ- 2
Comments: PAK *c̣ʷa-jǝ́; Ub. sǝ-q̇́ǝ-c̣ʷǝ́-n. Both PAK and Ub. have compounds: in PAK with the stem *-ja "bad" (see Shagirov 1977, 2, 102); in Ub. the first component goes back to PWC *q̇́ǝ 'dirt, dung' q.v. We should relate here also the Ub. reduplicated stem c̣ʷac̣ʷ- (c̣ʷac̣ʷǝ́-n "cela fait mal", c̣ʷac̣ʷǝ́ "mal, douleur"). For the semantic shift "dirty" > "bad" cf. PAK *ʡa-jV "bad" ( < *q̇́ǝ-ja) where *ʡa = PAT *q̇́ǝ 'dung' and Ub. q̇́ǝ-.
Shagirov's supposition that PAK *c̣ʷa- here = PAK *c̣ʷǝ "good" is unacceptable.
Comments: PAT *c̣ʷǝ / *c̣ʷǝc̣ʷǝ (cf. also Abkh. a-r-č̣̌ǝ́,a-č̣̌ǝč̣̌ǝ́ 'sour', Abaz. č̣ʷ-q̇́á 'sour'); PAK *c̣ʷa-ʡʷǝ- ( = Ub. ʒʷa-q̇ǝ, Abaz. č̣ʷq̇́á). Ubykh has voiced ʒʷ- instead of *c̣ʷ as a result of dissimilation.
[NB: The common WC form *c̣ʷVq̇́ʷV (-q̇ʷV) may be analyzed as containing an adjectival suffix, but also may partly reflect another NC root, well reconstructed for PEC as *c̣_eḳwV (q.v.).; see Абдоков 1983, 151.]