COMMENT: Initial laryngeal is reconstructed on basis of PL pharyngealisation; however, it is not possible to establish its precise nature, because in most languages it was either lost (as part of the weak first syllable), or absorbed by class prefixes.
In PWC the root has a frequent labial prefix (probably originally a class marker). In EC the original structure of the root was best preserved in Av.-And. and Nakh; in most other languages, where the root was treated mainly as an adjective, it obtained a productive adjectival (*participial) suffix *-rV and underwent metatheses, usual in cases like this: *Hăʒ_ĔmV-rV- > *Hăʒ_mĔ-rV- > *Hămʒ_ĔrV- // *mHărʒ_Ĕ-. Cf. also Hurr. šēɣ-(a)lǝ 'ritually pure', šēɣǝ,šīɣǝ 'pure (water)', Urart. šēχa id. (?) - reflecting probably a variant *ʒ_ămHĔ- (see Diakonoff-Starostin 1986, 35).
COMMENT: Although the root is not widely spread in PEC, it seems both phonetically and semantically reliable. Some phonetic comments: both PN and PWC seem to point to a variant with an early loss of medial *-m- and transfer of the labialisation to the following vowel (otherwise the consonant in PWC would not be labialised, and in PN we would have a reflex *-mc̣- > *-ṭṭ-).
COMMENT: One of the cases of *ʒ́ > WC *ɣ in 'tense' words. The EC-WC comparison seems probable, although not indisputable. The EC reconstruction is more secure, although Lak. -r- and PL *-r- corresponding to PA *-n- should be explained: most probably -r- penetrated here from a suffixed form *Hānʒ́_V-rV- (cf., e.g., Cham. hanc̣i-ra-).
COMMENT: An interesting Tsez-Lak. parallel. In PC we have to assume an assimilation and metathesis (*ƛ̣iriṗ < *ʔirƛiṗ), not surprising in a complicated stem like this. The word, although not widely spread in modern languages, is probably archaic, because it corresponds very well to Hurr. *χuruppi > Hitt. χuruppi 'a k. of cake' (see Hиколаев 1985, 62).
COMMENT: For the phonetic development in PWC cf. *χ_ɦwĕje 'dog' (in this case the lateralisation in PWC may also have been favoured by the lateral resonant in the root).
The stem *HālχV (reflected in PN, Lak., PL and Khin. /with a former class prefix w-/) is a secondary development < *HāχɫV < *HāχuɫV. Several morphological notes: the stem is usually accompanied by class prefixes in the Western area (Nakh, Andian, Tsezian) but is prefixless in Eastern Daghestan (Lak., Darg., Lezghian). By now it is hard to judge which situation is original; however, it seems probable that the initial syllable *Hā- is historically a prefix, and that we should reconstruct the most archaic form as *χuɫV (cf. the rather frequent reflex of this stem in different areas: in Av. /χ:alá-ta-/, in PD /*χ:ʷala-/, in some Lezg. languages/Rut. χuläχ-dɨ, Tsakh. χɨli-na/. Cf. also Hurr. χel(-)di 'lofty, highborn' (see Diakonoff-Starostin 1986, 54-55).
COMMENT: A common NC temporal adverb. We must note that in most subgroups the root can be followed by some kind of affricate (although the affricates do not correspond to each other regularly).
The root is also attested in HU: Hurr. χennǝ, Ur. χenǝ 'now' (see Diakonoff-Starostin 1986, 32).
COMMENT: It seems probable that the Lak. word is related to Av.-And.-Tsez., although to explain it one needs to propose an extremely complicated reconstruction (*hćărnū) with a rather unique root structure (*HCVRRV). If the Lak. word is to be kept apart, the reconstruction will be simplier: *hănćū, and the etymology will be an Av.-And.-Tsezian isogloss.
COMMENT: The root denoted some light (yellowish or greenish) colour, whence all the different colour meanings in languages. Phonetically the comparison seems plausible, with several comments:
a) PL, PD and PWC suggest the reconstruction of *ī as main vowel; however, Av.-And. reflects are unclear;
b) The Darg. and PWC forms incorporate the former class prefix *b- (Darg. čibara < (with metathesis) *čirV-bV).
c) Most forms are deductable from the original root structure *HčVrV̆-; however, Av. čar- (borrowed in Kar. čaro-b) suggests rather *HčVrV̄-. Variations of this type can be sometimes met in adjectival stems.
COMMENT: Reconstructed for the PEC level. The vocalism is hard to establish (too few data). The PN and PL forms point to a root structure *HCVRV̄; however, the PA form reflects rather a metathesized structure *HRVCV̄. Cf. also Urart. c̣ir-ab- 'empty, uninhabited' (see Diakonoff-Starostin 1986, 44).
COMMENT: Reconstructed for the PEC level. A typical case of the HCVRV̆ structure; medial *-j- reconstructed to account for *-st- in PN. There is some doubt as to the reconstruction of the final resonant: the reconstruction *Hc̣_wējnǝ̆ (with assimilation *-n > -m in PL) is also possible (Aand-Cez languages are here not informative).
COMMENT: Reconstructed for the PEC level. One of the cases of exceptional preservation of a laryngeal in Av. (in the original *HCVCV structure). Since the stem has no class prefixes in PN, Av. and PL, the PA situation is probably secondary (the expected stem *č̣Vr- or *Hič̣Vr- began to have an adjectival inflection).
COMMENT: Reconstructed for the PEC level. In PTs we would rather expect *č̣V(r)-, but rules of root structure development are often violated in adjectives (because of active class prefixation / suffixation); otherwise all correspondences are regular. The root must have denoted some light colour (yellowish, greyish).
COMMENT: The original meaning in PEC was probably 'enclosure, yard' (for the semantics in Khin. cf. cases like Russ. дверь 'door' - двор 'yard' etc.). It is interesting to note a probable HU parallel: Hurr. ardǝ 'town', Ur. ardi-nǝ 'the Town' (name of Muṣaṣir, the capital of Urartu), with the root structure reflex similar to PN (see Diakonoff-Starostin 1986, 26). {It is also interesting to note Georg. χvedr- 'share, plot of arable land', perhaps reflecting an early PN form like *ɦudr- or *ɦwVdr-.}
COMMENT: An onomatopoeic Lezg.-WC isogloss; the correspondences, however, are regular, and the root may be old (unlike other words for "barking" in Caucasian languages, which are all heterogeneous and certainly recent onomatopoeia).
COMMENT: Correspondences are regular. The quality of the initial laryngeal is not quite clear (there is not enough evidence), but it must be reconstructed to account for the PWC pharyngealization.
COMMENT: Reconstructed for the PEC level (opposed to PWC *ṗƛ̣́ǝ 'four' q.v.). The final -ʔ in PN has a double nature: it also may be a reflex of the lost *-q̇-, but it has been anyway confused with the final -ʔ which in PN marks all numerals (*cħaʔ,*šiʔ,*qoʔ etc.).
Class prefixation in the EC numeral "four" is preserved only in PN and PA (but not Avar); PL *j- may also be a trace of the class marker.
