COMMENT: The original meaning of the root may have been 'young goat', 'kid' (preserved in PWC; shifted to 'young small cattle' > 'young sheep' in PN; shifted to 'goat in general' in the majority of EC languages). Phonetic correspondences are regular.
This root is not preserved in Lak, being superseded by c̣uku < PEC *ʒĭkV̆ 'kid'. It is interesting to note, however, that in the Khosr. dialect the plural form from c̣uku is c̣u-rt:u, where the root *c̣_ü̆hnV̄ may be preserved (as c̣u-); the lit. Lak has already a new form c̣uk-ri.
There is some evidence pointing to a PEC oblique stem *c̣_ü̆hnV̄-rV- (cf. Akhv. c̣:ir(i), PL *c̣eh-rV-, perhaps also Lak. c̣u-r- in the plur. Khosr. c̣u-r-t:u).
Abdokov (1983, 118) compares the WC root with PEC *ƛ̣_äɦɫǝ̄ 'lamb', which is impossible for phonetic reasons. The same author (1983, 124) compares the EC forms with Kab. c̣āqʷa 'antelope' (which he explains as a compound: c̣a + qʷa 'goat' + 'pig'). We were unable to find this exact form in dictionaries; only in Nogma 217 there is a word ц̂ак"h [ = /c̣āq:a/] 'a k. of antelope, джейран', which is interesting (see Turchaninov's notes ibid., as well as Abayev 1979, 58-59 on Osset. säʁ 'goat' = Old Ind. chāga-, IE *sk'ago-), but isolated and has obviously nothing to do with EC *c̣ü̆hnV̄.
COMMENT: An expressive root; not very widely spread (within EC present only in Av.-And.-Tsez.) and liable to irregular changes. Nevertheless, the protoform can be more or less securely reconstructed. [For metathesized variants in WC cf. some similar EC forms: Gunz. ḳonč̣u 'tumour', Akhv. q̇:ʷač̣a 'abscess'; there is, however, hardly enough evidence to reconstruct a separate root].
COMMENT: An Av.-And.-Lezg. isogloss. Correspondences are regular, but there is not enough evidence to reconstruct the precise vocalism. EC > Osset. c̣äx 'grey, blue' (see Abayev 1958, 333-334).
COMMENT: The root is attested also in HU: Ur. arc̣ib/wǝ 'name of a horse = Eagle' (see Diakonoff-Starostin 1986, 45).
There is little doubt that all the listed forms belong together, although there are some problems. Not quite clear is the status of the prefixed syllable *ʔār-, attested in PN, HU, Lak., PD and PL. It should be noted that the Lak., PD (and possibly PL > Ag) forms reflect a variant with metathesized labialisation (*ʔār-c̣_wämʔV̆ > *ʔwār-c̣_ämʔV̆ ); in PL additionally the nasalisation was also shifted to the initial position. Quite irregular (and as yet unexplained) is -č̣- (instead of expected -c̣-) in PD (Chir.); however, there is no reason to suspect a borrowing. We should note also an interesting form: Cham. ercim 'kite'; since it is quite isolated within Av.-And., it must be an old Nakh borrowing (before the denasalisation *-mʔ- > -w-).
COMMENT: The original form is probably *c̣_ĕmhV (the WC form shows no trace of labialisation), but most EC languages reflect a form with shifted labialisation. The PN form *tuχe is interesting: it may be related here only if we assume the suffixed nature of *-χe (which is semantically not clear); but in any case the development *tuχe < *cwVn-χV within PN is proved by an early Ossetian loan: cäχχ / cänχä 'salt'.
The EC forms had been compared by several authors with PWC *ǯ́ǝ (/*č́ǝ) 'salt' (see Mészáros 1934, 287, Klimov 1969, 292, Abdokov 1983, 104); this is, however, not plausible for phonetic reasons, and the comparison with PWC *ć̣ǝ 'salty' seems preferable.
COMMENT: Reconstructed for the PEC level. In PD one would rather expect *c:-; preservation of glottalized *c̣ is due to early delabialisation. A reconstruction *c̣_wǝfɨ̄ is not excluded: in this case we can follow Abdokov (1983, 185) and compare PAT *c̣ʷaxǝ- 'to conceal, hide' (Abkh. á-ć̣̌aχ-ra, Abaz. č̣ʷaχ-ra).
COMMENT: The etymology seems reliable, despite some assimilative/dissimilative processes and metatheses, usual in roots of this structure. The PN reflex *sṭ- is explained by the influence of *-j- (preserved in the WC form).
In some languages there is confusion between this root and the diminutive form of the PNC word for star, *ʒ_hwăr-ḳV (/*ʒ_hwăr-kV). The latter is certainly reflected in Ub. c̣ʷa(n)-ḳ́ and Lak. c̣uku (cf. plur. c̣ur-t:i). The Cham. (Gig.) form c̣aka 'star' reflects probably a merger of the two roots (opposed, e.g., in the closely related Tindi language: c̣akʷa 'spark' vs. c:aru 'star').
COMMENT: The WC form has a frequent labial prefix (which conditioned loss of labialisation *pǝzʷV > *pǝzV). The EC-WC comparison was suggested by Shagirov (1977). There exist also interesting Kartvelian parallels: cf. Kartv. *ʒ̥1̇u- 'female, bitch' (compared with Lak. in Klimov 1963); Chan. bozo 'girl' (cf. PAK *bzǝ), Svan. zural 'woman' (see Lomtatidze 1961). Untenable is Trubetzkoy's (1930) comparison of the WC root with Darg. gʷaza, Lak. k:ʷac:a etc. 'mare'.
In Nakh languages, besides *psṭuw, there is a series of similar forms: Chech. stē 'woman; female' (pl. steš, Usl. stij), stē-n 'female', Akk. sēwã 'female', Ing. se id., pl. istij ( = Chech. stij) 'women, wives'. There are two possible solutions: to consider these forms as derived from *psṭuw (from the oblique base *psṭawV-), or to compare them separately with an isolated Darg. form - Chir. cade 'female' (reconstructing PEC *cVjdV (~ ć-)). This problem can not be definitely resolved without Bacbi evidence.
COMMENT: The correspondences are basically regular. However, there is a number of strange forms in Eastern Daghestan: Darg. Chir. č̣aIp(a), Arch. č̣eIp 'cradle', Bud. č̣af id. which are quite irregular phonetically and must have been borrowed from some related language - but it is not clear, from which one.
COMMENT: Reconstructed for the PEC level. An expressive stem (often reduplicated), but nevertheless well reconstructable. In PN glottalisation was lost (probably as a result of assimilation). Note the identical suffixation (*-Vm) in PN and PA.
COMMENT: Reconstructed for the PEC level. Except for dissimilatory deglottalisation in Darg. (before a historical diminutive suffix -iḳ-), correspondences are regular.
