Notes: Reconstructed for the PEC level. The PL form is metathesized. The PN form reflects the original shape *ʎä̆mc̣_V (without labialisation of the first consonant, and with *-ṭṭ- pointing to a cluster with nasal); all other languages reflect a secondary form *ʎwä̆c̣_V with metathesis of labialisation. Despite these peculiarities (explained by the rare verbal *CVRCV- root structure), the etymology seems reliable.
Notes: Reconstructed for the PEC level. The root must have meant some cereal of inferior quality (cf. the meanings in Chech. and Av.). Correspondences are regular.
Notes: The root is not widely represented, but the regular correspondences between PN, Cham., Tab. and Ub. justify the reconstruction of PNC *ʎ_iHmV (in Tab. we must suppose a frequent metathesis of labialisation).
Some EC languages also have names for roe or mountain goat which are hard to separate from this root and are possibly derived from it:
a) *ʎ_im-kV ( > *ʎ_winkV or *ʎ_inkwV) reflected in Av. ʎ:inkʷá 'mountain goat', Cham. ʎũka (Gig. ʎunka) 'roe, chamois'. Note the same notation ʎ- (as in ʎama) in Cham. (only MSU recordings of the word are known to us; the strong ʎ:- in Av. presupposes the same consonant in Andian languages). A recent loan from Andian (Cham. or Tindi, where the root is not attested) is Inkh. ʎunka 'roe'. Less clear is the origin of PGB *ʎiga 'roe' (Bezht., Gunz. ʎiga) - is it an irregular development (with voicing -k- > -g-) from *ʎ_im-kV; or is it a Nakh loan (cf. Ing. lijg 'roe' with a diminutive suffix -g)?
b) *ʎ_im-tV ( > *ʎ_wintV) reflected in PD *xʷinta (Ak. hinta 'doe', Ur. hinta, Kub. xʷinta, Kait. šʷenta 'roe'). Borrowed from Darg. are Lak. hinta 'chamois, gazelle' and (from an Old Darg. form *finta) Av. pintá 'gazelle, antelope' (Chad. butá id.).
The proposed derivation of *ʎ_imkV and *ʎ_imtV seems probable to us; however, we must note that suffixes *-kV and *-tV are very rare, and more checking is needed to make a final decision on this etymology.
Notes: Reconstructed for the PEC level. Correspondences are regular, except for the PTs vowel (*ʎɨr would be expected): this is most probably explained by the influence of semantically close PTs *sar 'faeces' and *tar 'dung'.
Notes: Reconstructed for the PEC level.The root is also attested in HU: Urart. lutu 'women' (see Diakonoff-Starostin 1986, 27). The meaning of the Urartian form sheds light on the nature of the element *-dV in EC (present in Av., Darg. and Ur.): this is most probably an original plural formant.
See Trubetzkoy 1922, 239. Despite Abdokov (1983, 89), the EC root can not be related to WC *sʷǝ(mǝ)c̣V 'woman' q.v.
Notes: An interesting Av.-And.-WC isogloss; parallels in other languages are as yet unknown. Abdokov (1983, 95) cites also Chech. se-lχa-na 'yesterday' (the etymology of which see under *rVmƛ_Ă) and Kryz. išɨʁ 'light' (which is obviously a recent Turkic loanword). The Av.-And. form is compared by Abdokov with WC *ʎʷV 'light, white', which seems less probable for phonetic reasons.
Notes: Reconstructed for the PEC level. The initial consonant in PTs is not quite regular (one would expect a *ʎ:-) - probably because of the expressive nature of the root.
Notes: An And.-Tsez. isogloss, very poorly represented - thus the reconstruction is rather uncertain. This may be an expressive variant of PEC *lōmVʒ_ĭ 'dirt, pus' q.v.
Notes: An Aand-Tsez. isogloss. The semantic correlation 'wind' <> 'voice, shout' recurs in several cases and is explained by the common semantic component 'movement of air'.
Notes: An Av.-And.-Tsez.-WC isogloss. The etymology is possible, but not secure enough: the uncertainty is caused by problems of morphological analysis of Andian forms (unclear component -s:o(n)), Tsezian reduplication and the isolated character of the Shapsug form within WC (also with an etymologically unclear component -ħa).
Notes: The And-Lezg. name for `internal fat' seems to be a plausible phonetic match for the common WC word for `blood', although the comparison is still questionable (primarily because of scarce attestation within EC and semantic problems).
The Botl. and God. forms can not be separated from Kar. and Bagv.; thus their separate treatment as being related to PWC *š́:ʷa 'fat, oil' (see Abdokov 1983, 85) is absolutely unacceptable.
Notes: Some languages (PA, PD, PL) contain the final labial *-b- which is obviously an original plural suffix (there is no trace of it in the PC and PWC forms).
The strengthening *ʎ > *ʎ: in PWC is in this case possibly related to the loss of laryngeal (the presence of which is confirmed by the PC and PD forms); regularly we should expect a weak *ʎ (the vowel is short in PEC). Besides, the EC evidence points to a front *e vowel in the root, while in WC there is no palatalization. Maybe we should reconstruct some vowel variation in this root for PNC. In any case, the common origin of the word for 'three' in PEC and PWC is indisputable.
See Trubetzkoy 1922, 238, 243; 1930, 274; Абдоков 1983, 154.
Notes: A Laki-Andian isogloss. It should be distinguished from *xwɨmc̣_wɨ (q.v.), although partial contaminations have occurred in some languages. See Trubetzkoy 1922, 239-240.
Notes: Reconstructed for the PEC level. The reconstruction is not very reliable for several reasons: an isolated position of the Akhv. reflex within Andian; a unique reflex *ƛw- > *h- (through *x- ?) in PTs (although this is the only case and there are no contradicting examples); generally sparse reflexation of the root.
