Комментарии: PAK *ja-zǝ. The first part goes back to the demonstrative pronoun *ja (q.v.). The Ad. form is ingeniously explained by Shagirov (1, 172) as a contraction from the erg.-obl. case *jaz-š́ < *jazǝ-š́ (-š́ is the normal ending of this case in Ad. pronouns, but jaž́ in the Bzhedugh dialect has no ending either in the erg.-obl. case or in the nominative case).
Because of an obvious connection in EC, Yakovlev's etymology of *-zǝ ( = *zǝ 'one', see Yakovlev 1948, 69, 143, Shagirov 1, 173) should be abandoned.
Комментарии: PAT *-ḳamǝśǝ (cf. also Bzyb. a-śǝ́ś-ḳamś,a-śǝ́ś-ḳlamś). The first component (Abkh. šǝš-, Bzyb. śǝś-, Abaz. č́-) is not clear. The PAT form can be related to Ubykh if we assume a metathesis (*ḳamǝśǝ < *śVmVḳV) and assimilative / dissimilative processes - which is not surprising in an expressive root like this (cf., e.g., numerous irregularities in the IE root for 'ant' - *morm- / *morw- / *bhorm- etc.).
Комментарии: PAT *zVʕʷV; PAK *wǝ-zǝ (*wǝ- is a frequent verbal prefix). For PWC we can also reconstruct a compound *źV-ʁʷV (besides the listed forms cf. also Kab. wǝzǝʁʷa = wǝzǝ); the second part (*-ʁʷV) is not quite clear etymologically (it can be attributed to several PWC roots).
Комментарии: PAT *zʷǝ- (cf. Bz. á-ž-ra); PAK *ja-sʷa- (intr.), *jǝ-sʷǝ- (tr. > Ad. jǝšǝ-n, Kab. jǝfǝ-n); Ub. 1 sg. a-z-ʒʷá-n. The PAK form has an irregular unvoiced stop (*-zʷ- should be expected). This is probably due to the influence of *psǝ 'water' (very frequent in combination with the verb 'to drink'). The original meaning of the root - as it is clearly seen on external evidence - was 'to milk, (to drink milk)'. Derivates from this root mean 'milch cow, cow' in many EC languages. Hence it is natural to relate to this root also PWC *zʷǝ / *mǝ-zʷǝ 'cow' (with *mǝ- as a probable deverbative prefix): PAT *zʷǝ (Abkh. á-ž̌, Bzyb. á-ź̌, Abaz. žʷǝ) 'cow', Ub. mǝzʷ-ʁǝ́ 'heifer'.
Комментарии: PAT *ʒ́ǝ-śǝ, pl. *ʒ́a-ra: cf. Abkh. pl. á-ʒa-ra, Bzyb. á-ʒ́(ǝ)ś, á-ʒ́a-ra, Abaz. pl. ʒa-ra. PAK *źa-jǝ (used only as an element of composita). The comparison belongs to Abdokov (1983, 118).
Комментарии: PAT *zʷǝ-ʕʷa-nǝ (the final -d in Abaz. is secondary). -nǝ here is a common nominal suffix, and *zʷǝ-ʕʷa corresponds exactly to PAK *zʷa-ʁʷá. Ub. def. á-zʷa.
The meaning 'star' of PAK *zʷa-ʁʷá is obviously secondary (it was acquired after the original 'star' morpheme, *-cʷa, shifted to 'sky' in the compound *wa-cʷa - see *c̣ʷA; the two roots have thus "changed places" in PAK). Abdokov (1983, 93) compares the AT and AK forms with PAK *ž́ǝ 'wind, air', which is improbable for phonetic reasons (see PWC *ɣ́ǝ).
Комментарии: PAT *ʒ́a-rV (-rV is originally a collective suffix), cf. also the Abzh. variant a-bgǝ́-ʒǝr, Bzyb. a-bgǝ́-ʒ́ar (the compound means literally "fox's cornel"); PAK *za (Ad. reflects only the compound *za-jǝ́, where -jǝ is the "tree suffix", = Kab. zej 'cornel-tree').
Reason for spirantization ʒ > z in Abaz. is not quite clear; perhaps it is due to Adygh influence, although the word as such can not be regarded as an Adygh loan (despite Lomtatidze 1944,209, Kondzharia 1969,8).
The correspondence PAT *ʒ́ : PAT *z points to PWC *ʒ́ or *ǯ.
Комментарии: PAT *ʒ́ǝ-śa- (cf. also Bz. á-ʒ́-śa-ra). The morpheme is present only within a compound (see *śV-). It must be kept separate from *p-ʒǝ 'water' q.v.
Комментарии: PAT *na-ʒa-,*ʕa-ʒa- (with different locative preverbs; the second form is reflected in Abkh. ā́-ʒa-ra, Abaz. ʕa-ʒa-ra; cf. also Bz. á-naʒa-ra,ā́ʒa-ra).
The Ub. forms q̇a-ʒǝ- 'approach', q̇a-ʒ- 'near' contain the same root; it is also present in the postposition -ō-n-ʒa 'to, up to, towards' ( = Abkh. (n)ʒa, Abaz. -ʒa id.; see Dumézil 1932, 75, 250-251). However, the element q̇a- is the Ub. preverb q̇a- 'in hand(s), at hand' (so q̇a-ʒǝ- means something like 'near at hand') and does not correspond directly to PAT *ʕa-, PAK *q:a-.
See also under PWC *šǝ ( ~ *ś-) about possible AK parallels.
Комментарии: PAT *ʒ́a- (cf. also Bzyb. a-ʒ́a-rá), *ma-ʒ́a 'secret' (Abkh. á-maʒa, Bzyb. á-maʒ́a, Abaz. maʒa). In Ub. variants wa(n)ʒ́á and wIa(n)ǯá are attested; wa- here can be treated as a prefix meaning 'to do, execute'. The Kab. word means literally 'to make (-ṣ́ǝ-n) a secret (bzǝ-)'. See Shagirov 1,97.
PAT *ʒ́a- can, in fact, reflect a merger of two WC roots: cf. also *ć:ǝ- > Ub. wǝ-ćǝ- 'to steal' (perhaps also -pś- in Kab. ʁa-pś-ḳʷǝ-n 'to hide'?).
Комментарии: PAT *bǝža (b- may be secondary under the influence of *bǝž́a 'between' q.v.). In PAK we have only compounds *ʒ́a-gʷǝ 'hearth' (*gʷǝ 'heart, middle'), *ʒ́a-nǝq:ʷǝ (*nǝq:ʷǝ 'middle') > Ad. ǯ́a-nǝq:ʷ 'place before the hearth'.
See Shagirov 1977, 199 for the comparison of PAK and Ub. He reconstructs for PAK the meaning "back part"; however, it must be secondary, cf. the older meaning 'middle, half' in PAT and Ub. za(n)-ǯ́á 'middle, half' (za- "one"?). The correspondence between PAK *ʒ́ (Ad. ǯ́, Kab. ź), PAT *ž and Ub. ǯ́ points to a rare PWC phoneme *ʒ́ʷ.
Комментарии: PAT *ʒʷa (with emphatic gemination). Phonetically the Ad.-Abkh. comparison is very probable; the semantic change must be postulated as "to pour water (enough) for washing" > "to cut off (water), to dam".
Комментарии: PAT *ǯʷanǝza (~ -ź- - because the Bzyb. form is not known); cf. also Abaz. Ashkh. vanǝza. The Ad. form ǯ́anaź points to PAK *ʒ́anaźǝ. It is possible that we should compare also Kab. źanfa-n 'intestines (of animal); entrails'. This will support the reconstruction *ʒ́ (Ad. ǯ́ : Kab. ź), but we will have to reconstruct PAK variants *ʒ́anaźǝ / *ʒ́anǝsʷa ( < *ʒ́anǝzʷa with different dissimilational processes).
A regular correspondence for PAK *ʒ́ would be PAT *ǯ (*ž) < PWC *ʒ́ʷ. Therefore for PAT we must suppose a metathesis of labialisation: *ǯʷanǝza < *ǯanǝzʷa. The correlation PAT *ǯanǝzʷa : PAK *ʒ́anǝzʷa would unequivocally point to PWC *ʒ́ʷanVzʷV. However, the Ad. form ǯ́anaź allows for still another possibility of reconstruction: PWC *ʒ́ʷanVźʷV (regularly > PAK *ʒ́anaźǝ > Ad. ǯ́anaź; > PAT *ǯʷanǝźa - since /see above/ *-ź- is also possible in PAT).
Pace Shagirov (1977, 277), it is absolutely impossible to analyze the PAT form as *ǯʷa + *nǝza.
