Comments: Abkh. ā́nda goes back to PAT *ʕanǝda < *ʁ́Ianǝda with delabialisation; Ub. wǝdʷá presupposes PWC *ʁ́Iʷǝ(n)dʷa. We should probably reconstruct a compromising *ʁ́IʷV(n)da with labialising/delabialising assimilations in Ub. and Abkh.
Comments: PAT *ʕʷa-ž́ǝ. Shagirov 1, 133 considers it as borrowed from AK; we would rather follow Dumézil 1932, 124 and regard it as a genuine cognate - but possibly influenced by PAK *ʁʷa-ž́ǝ́. In PAK there exists an independent form *ʁʷa (Ad., Kab. ʁʷa 'yellow, reddish').
Comments: Ub. a-z-ʁʷá-n. The Abaz. verb is used in expressions like sgʷǝ ʕʷ-iṭ "it seems to me" (lit. "my heart thinks"). The PAK form *gʷǝ-ʁa- has a delabialised *ʁ as a result of contamination with another PWC stem, *ǵʷǝ-Ga- 'to hope' q.v. (PAK *gʷǝ-ʁa- means both 'to think, suppose' and 'to hope'). See Shagirov 1, 115.
Comments: PAK *ʁʷǝmǝ́. Shagirov's (1,137) attempt to analyse the PAK form as 'board' (PWC *Gʷǝ, not retained in AK) + 'thick' (*mǝ, a root which does not exist) was certainly unsuccessful.
Comments: Isolated in Ub. with the only possible parallels in EC (despite Dumézil 1932,124 the root has nothing to do with PAK *wǝna, PAT *ʕʷǝnǝ 'house').
Comments: PAT *ʕʷa-, PAK *ʁʷǝ-. Not quite clear is the loss of labialization in Ub. The root serves also as a basis for the adjective 'dry' in all languages: Abkh. a-ʕʷá, Abaz. jǝ-r-ʕʷu, Ad. ʁʷǝ-ʁa, Kab. ʁʷǝ-r, Ub. ʁ-q̇a.
Comments: PAT *ʕʷǝ (in Abkh. - only within a compound with *wa- q.v.); Ub. def. a-ʁʷá. It is not quite clear how this morpheme in PAT correlates with its plural counterpart - *-ʕa in *wa-ʕa 'people' (Abkh. a-wā́, Abaz. waʕa).
Comments: PAT *lǝ-ʕʷa (for the etymology of *lǝ- see *Ła). PAK *ʡʷǝ́-ʁʷa - a compound with an unclear first component. Resemblance to PWC *ʁʷa 'yellow' is of course quite coincidental (despite Шагиров 2, 161).
Comments: In PAT used as the 3d p. pronoun *ja-ra 'he'; cf. also *jǝ (Abkh. j(ǝ)-, Abaz. j(ǝ)- the verbal prefix of the 3d person) and *-jǝ within complex demonstrative stems (*a-nǝ-jǝ 'that', *a-rǝ-jǝ 'this' > Abkh., Abaz. ani,ari etc.).
In AK the root functions as the 3d p. possessive prefix (jǝ- 'his'), see Шагиров 1, 180.
In Ubykh jǝ- is the demonstrative pronoun expressing near deixis ('this'); the same morpheme is probably used as a personal prefix of the 3d p. pl.
Comments: PAK *jaž́á. The word has no PAT or Ub. parallels, but reliable correspondences in PEC - thus the inner PAK etymology proposed in Шагиров 1, 169 (ja- - same as in *jaṭa 'clay, dirt', -ž́a - related to PAK *ž́ǝ 'wind, air') seems less convincing.