Comments: The Ub. 1st p. pl. pr. is the only one which has reliable EC parallels (the EC exclusive 1st p. pl. pr.); despite Dumézil 1932, PAT *ħa and PAK *t:a have no connection with it.
Comments: PAT *śǝ, attested only with preverbs, cf. Abkh. ája-s-ra 'to pass', a-ḳǝ́l-s-ra 'to pass through, crawl through', a-bž́ǝ́-s-ra 'to pass by'; Abaz. ḳǝl-s-ra,bž́ǝ-s-ra id.; Bzyb. ā́-ś-ḳ́a-ra 'to move (here), come (here)', a-na-ś-ḳ́a-ra 'to move (there), get away'. PAK *na-sǝ- 'come, reach (there)', *q:a-sǝ- (Ad., Kab. q:a-sǝ-n) 'come, reach (here)'.
We must notice that in PAK probably a contamination of two PWC roots took place: on one hand, PAK *na-sǝ- and *q:a-sǝ- are exact formal matches of Bzyb. a-na-ś- and ā-ś- (see above); on the other hand, semantically they also match perfectly Abkh. a-na-ʒa- 'to reach (there)' and ā-ʒa- 'to reach (here)' (see under PWC *ʒV). The derived postposition PAK *-nasǝ 'until' is also exactly matching Abkh. -nʒa and Ub. -ōnʒa. See Dumézil 1932, 75, 250-251; Shagirov 1, 279-280.
Comments: PAT *ša,*ša-qa (with an unclear component -qa). PAT *ša and Ub. -ša correspond regularly and go back to *š:ʷa; there is, however, some uncertainty, because the meaning of the first component in Ub. is not clear.
Comments: Ub. def. á-šʷa. PAT *šǝ-kʷǝśǝ : probably from earlier *šʷǝ-kʷǝśǝ with dissimilation (cf. also different dissimilations and metatheses in modern languages: Bz. á-śkʷś, Abaz. skʷšǝ). The meaning of the component -kʷǝśǝ is not clear. Abdokov (1973, 54) compared the Abkh. word with AK *jǝʎasǝ 'year' which is improbable.
Comments: PAK *śa-wá. We can discover the same root (with the original meaning "to bear") in Ad. śā-ʎ(a) 'placenta', χʷǝ-śa-n 'childbirth fits'. PAK *ś regularly goes back to PWC *š́ʷ.
Comments: PAK *śa: Ad. ǝ-śa jǝč̣́ǝn 'to get furious', lit. 'to go out of one's place' - cf. Kab. jǝ-ṗa jǝč̣́ǝn id. (where ṗa = 'place'); Kab. naṣ́a-śa 'place under one's eyes'. These and some other idioms allow to reconstruct the original meaning "place".
In Ub. cf. forms like ǯ́aʁʷa-šʷa 'ford, place for wading' (ǯ́aʁʷa- 'to wade') etc.
See Шагиров 1977, 282. Since there is no parallel in AT, it is impossible to choose between PWC *š́ʷ or *śʷ.
Comments: The AK root is attested within the compound *pćǝħá-śħa 'evening' (Ad. pč́ǝħā-śħ, Kab. pśǝħa-śħa), where the first component by itself means probably 'dusk; sunset/sunrise' (see *bǝćʷǝ). Despite Shagirov (2, 27) the component śħa here means 'evening' by itself and has nothing to do with PAK *śħa 'head'.
The correspondence PAK *śħa = Ub. šʷǝwá is exact and points unambiguously to PWC *śʷǝ-Łʷa. This stem itself is probably an old compound, with probable external parallels for both parts).
Comments: PAT *sʷǝrǝ (cf. also Bzyb. a-ś̌ǝ́r). PAT *sʷ can go back to PWC *šʷ, *sʷ or *x́ʷ, thus the PWC reconstruction is rather tentative (*šʷ is the best match for PEC *cʷ).
