Notes: Reconstructed for PEC (however, a possibility should be considered of comparing this root with PWC *p(:)ǝƛ̣́ǝ 'four' q.v.). Lak. maIj- is probably a transformation of *maIlj- < *mali- (the source of pharyngealization is not clear). In Khin. the root has been transformed under the influence of the numeral "seven" (Khin. jiḳ). In Darg. k: is the normal reflex of *Ł; however, the reduction of initial labial and the source of *-ħ- is not yet clear (maybe *k:aħ- < *k:aI- is a result of a secondary transformation of *aIk:- which, in its turn, had been formed under analogy of *-arʁI- 'seven' from the original *mak:- ?).
A very probable parallel is Hurr. miri - a numeral from the first ten (it can be either 'five', 'six' or 'eight'). If it denotes 'eight', it matches very well with PEC *bǖnŁ_e ( > *mǖŁ_e > miri with a regular reflex -Ł- > -r-).
Notes: A Tsez-WC isogloss; not very reliable because of rarity of reflexes. It is not excluded that this is the same morpheme as the class marker *b- (q.v.).
Notes: Reconstructed for the PEC level. Not widely spread, and having a rare root structure (with the sequence -VHV- between two stops) - it is possibly an old compound of some kind, although it is hard now to make guesses about its components.
Notes: The vocalism is hard to reconstruct (since there are no Andian or Tsezian reflexes, while in other languages there's a strong tendency of vowel reduction and assimilation in the first syllable). The form *bV̆rŁVmV is reflected in PN and Lak. (buIrna < *buIrlVma with disappearance of -l- after -r- and assimilation *-m- > -n-; for the semantics cf. an analoguous shift 'ladder, stairs' > 'pillar, support' in the reflexes of PL *morʎ:ol 'stairs, ladder' q.v.; the source of Lak. pharyngealization, however, is not clear). The Av. form malí probably goes back to *malrí < *bV̆mŁVrē (with a simplification of the cluster -lr-).
Notes: Reconstructed for the PEC level. For the semantics in Laki cf. the meaning 'funeral stretcher' in some Andian languages ('funeral stretcher' > 'funeral device' > 'tomb-stone').
One could also compare Gunz. bɔḳi 'enclosure for cattle'; there is, however, a possibility of its being borrowed from Georgian: cf. Georg. baḳi, Zan. baḳi 'enclosure for cattle' (compared by Kuipers 1960 with Kab. baq 'enclosure', which is phonetically improbable). This Kartvelian root itself may be connected historically with PEC *bVrḳwV.
Notes: One of the common EC words probably early borrowed from Iranian. Cf. Osset. färät 'axe', presupposing an Old Persian source *paraɵu (whence also Tokh. peret,porat and a number of Turkic and Fenno-Ugric forms) - see Abayev 1,451. It is interesting to note that the original meaning 'axe' is preserved only in Dargwa, while Av.-And.-Tsezian languages changed it to 'stick' (having preserved their own words for 'axe', see *kwɨ̄rṭā,*ɦwālźV).