COMMENT: Correspondences are regular, and the comparison seems phonetically and semantically plausible. Cf. perhaps also Khin. ši-c̣- 'to write' ('to put down', cf. an analogous derivation in several Lezgian languages: Lezg. k-xi-, Ag. k-ix-, Rut. k-i=xe- 'to write' from PL *ʔeʎe- 'to put').
COMMENT: A reliable common NC root. Somewhat problematic is the relationship of PN *=aš- 'to melt, to thaw' (Chech., Ing., Bacb. =aš-): if it is cognate (which is not very certain), we should reconstruct PNC *=ĭrc̣wĂr; see also PEC *=is_Vn-. Perhaps related is Urart. c̣i- 'to flow, to rain' (see Diakonoff-Starostin 1986, 44) - although semantically not very secure.
COMMENT: The EC-WC comparison belongs to Trubetzkoy (1930, 278). Except for the loss of labialisation (which sometimes happens in verbal WC roots - possibly because of interaction with class prefixes) correspondences are regular, and the etymology seems plausible. Some forms (Av. orč̣-, PN *=erc̣-) reflect a variant with infixed (probably durative) *-r-; one can also compare Urart. erc̣- 'to return' (see Diakonoff-Starostin 1986, 45).
COMMENT: The etymology seems satisfactory, although it is sometimes hard to distinguish between the reflexes of this root and PNC *=ä̆nć̣Ĕ 'to crush, break' q.v.
COMMENT: Reconstructed for the PEC level. Correspondences are regular (in Khin. d- is a former preverb, while -n probably reflects the original *-l-conjugation). Ando-Tsezian forms (as well as some Lezgian) reflect a prefixless form with loss of *ʔi- in an unstressed first syllable.
COMMENT: Semantically the EC-WC parallel seems satisfactory ('bend' <-> 'twist, spin'); not quite clear is the reason for loss of labialisation in PWC.
COMMENT: A very stable common NC root; reflected in most languages as both a verbal and adjectival stem. Cf. also the nominal derivate *r-iG_wV / *G_wVrV 'draught' (PD *deʁʷ > Ak. deʁ, Chir. deʁʷ, Ur. diw, PTs > Gunz. (Nakh.) qor). Note a frequent use of the reduplicated stem *G_wiGwVr- (in PA, PC, Lak.). See Абдоков 1983, 144-145.
COMMENT: An Av.-Lezg. isogloss; acceptable, if we assume that the Avar word has a suffixed -t- or is an old compound (with a not quite clear second part).
COMMENT: A common NC root with regular correspondences. Class prefixation is retained in Andian languages and Darg.; labialisation in PWC is probably also a vestige of class prefixation (*ƛ̣́ʷa < *w-iƛ̣ilV with vocalic assimilation). However, a prefixless variant *ƛ̣ɨ̆lV (reflected in Av., Lak. and PL) may also have existed already in PNC or PEC.
The Lak. form has a pharyngealized vowel, which may reflect some lost laryngeal; but since other languages do not reflect it (see above about the secondary nature of pharyngealization in Tsakh.), it is probably new (expressive).
COMMENT: The etymology seems plausible both phonetically and semantically. Not quite clear is the status of medial -l- (the reconstruction depends on whether we consider the Lak. l- and Darg. -r- as historical affixes or as a part of the root).
COMMENT: Reconstructed for the PEC level. Although superficially the Darg. and Lezg. forms look quite alike, the correspondence between them is irregular. There are two possible explanations:
a) the original form was *-ilq_wV (which should have yielded PD *-ilq:ʷ- or *-ulq:-, and PA *-i(r)q:ʷ-), but in PD and PA there occurred a contamination with PEC *-iqʷV 'to run' - q.v. This hypothesis is possible, but not very likely, because we have to suppose an independent contamination in two branches (PA and PD);
b) the original form was *-ilqwV, which regularly yielded PA *-iqʷ- and PD *-ulq-; in PL we should expect *ʔilχʷe-, but this form was avoided because of its homonymousness with *ʔilχʷe- 'to die' (q.v.). For this reason the expected change *qʷ > *χʷ did not happen. This explanation seems more probable, because it involves irregularity only in one branch (PL).
COMMENT: Reconstructed for the PEC level. As well as its transitive counterpart (*=ilq_wV / *=ilχ_wV 'to slaughter' q.v.), the root demonstrates a non-standard variation of *q/*χ (the latter is reflected in Darg. and PTs, the former - in PA; the PL form may reflect either variant).
COMMENT: The "tense" stem *-ilq_wV- ( / *-ilχ_wV) is usually transitive ("to kill (plur. obj.), to slaughter") and opposed to the "lax" intransitive stem *-ilqwV- (/*-ilχwV) "to die (plur. subj.), to die out". This semantic opposition is rather obscured in Dargwa dialects, where the meanings are sometimes reversed, and sometimes organized in another way (e.g. in Chirag: -elχ:ʷ- 'to kill' vs. -elχʷ- 'to slaughter'), probably because of the intermixture of the two stems. WC languages have preserved only the "tense" stem, used only within compounds, thus it is difficult to formulate the precise meaning of the morpheme in PWC.
In both cases Av.-And. languages (and WC, although here only one variant of the stem was preserved) reflect affricates, while other languages reflect fricatives; a situation which is quite unique and does not occur elsewhere.