Notes: In Av. the voiced -b- is obviously secondary (a normal result of *-p- > *-ṗ- /before a glottalised stop/ > -b-). Otherwise the correspondences are regular (*-l- disappearing in cluster in Av. and yielding nasalisation in PTs). The PD form probably comes from *hepq̇ul(i) with reduction of the medial vowel and cluster simplification.
There are some doubts whether the WC form belongs here (in this case a usual delabialisation *-Gʷ- > *-G- after a labial should be assumed) or to another root (*panG_VlhV 'leaf; bark' q.v.) - both possibilities should be considered.
Notes: Correspondences are regular. See Dumézil 1933, 21, Abdokov 1983, 132. The latter also adds PAK *śaʁá (Ad., Kab. śāʁa) 'braid, lace' - which, however, has an obscure morphological structure and does not correspond phonetically to other WC forms.
Notes: Cf. also Hurr. χinʒ-urǝ "apple" (borrowed in Arm. as χnʒor), see Diakonoff-Starostin 1986, 24. A very widely spread common NC root. The PWC form has a frequent prefixed *b(V)-. The original meaning could have been not "apple", but "medlar" (cf. the semantic correspondence between PN and PWC), with later semantic change (medlars look very much like small apples). Note that PN *ʡāǯ 'apple' should be quite definitely kept apart (PN *ǯ can not be derived from *ć in any possible way); see *HV̄ǯwĔ.
Despite Abdokov 1983, 110 the PEC forms can not be related to PWC *mIa 'apple'.
Notes: An Av.-Lezg. isogloss; since the PL form has a very archaic shape, the PEC reconstruction seems reliable.
It seems very probable that the same root serves as part of an old (PEC) compound which can be reconstructed as *dVHV-ʕwamħV. It means some small wild animal and can be traced in following languages:
a) PN *dʕaʡam 'badger', reflected in Chech. daʕam (RCD dʕama), Lev. daʕam, Itumk. dʕām, Ved. dʕaʕam, Shar. dʕām.
b) Av. ʕondó-ƛ̣: 'weasel' = PA *handa-riƛ̣:u (Akhv. Ratl. hindareƛ̣:u, And. (Khaid.) hanaroƛ̣:i, Tind. handareƛ̣:u, Cham. annaraƛ̣: (Gig. hannareƛ̣:u), God. ʕandiruƛ̣:i. This is a further compound with *riƛ̣:u 'weasel, squirrel' q.v. We must note here also a quite aberrant form Akhv. šĩda-reƛ̣:u (all other Andian languages point uniformly to *h-), which is probably a result of further contamination with šĩde- (obl. stem) 'bear'.
c) PTsKh *madu-ƛa 'weasel' q.v. ( < *madu-rVƛa = PA *handa-riƛ̣:u, Av. ʕondoƛ̣:).
It is hard to guess the original meaning of the first component *dVHV- (since it is very distorted phonetically in all daughter-languages); however, it can be (with deglottalisation) = PEC *ṭiHwV- 'small'; thus, the whole compound would mean something like 'small wolf'.
Notes: Reconstructed for the PEC level. In Darg. the root is preserved only within the derivate *ʕʷan-c̣(a) 'deaf' ( = "earless", with the privative suffix -c̣(a)). In PA and PC the root obtained suffixes (resp. *-di and *-χa); although their meaning is not quite clear, there is no doubt that the roots are comparable.
Notes: Reconstructed for the PEC level. Lak. raħ < *ʔarħ (it is clear, because original *r- > Lak. d-, as well as from comparing the Lak. and Darg. forms); otherwise correspondences are quite regular.
Notes: Reconstructed for the PEC level. The original form was trisyllabic (as is clearly seen from Lak. and Av. dialectal evidence). The medial vowel was lost in Akhv. and in PD (where *-n- regularly > -r- in a cluster with uvular).
Notes: PN *datta < *ramc̣:a (a regular reflex of the structure *HMVRCV̆). The precise meaning of the root in PNC is hard to determine, but it must have been a foliage tree or shrub bearing berries.
Notes: Reconstructed for the PEC level. The stem *HădV- (reduplicated in Darg.) may be somehow related to *HăṭV-, reflected in Akhv. aṭe-da and, possibly, PGB *š-aṭ- / *z-aṭ- (with expressive prefixes) 'naked' (Bezht. Tlad. šäṭṭö, Khosh. šäṭijo, Gunz. zaṭṭu).
Notes: An And.-Tsez. isogloss. There are no certain parallels in other languages. However, there is an interesting possibility of comparing PL *ʔala 'weed', possibly represented in Tsakh. ala id. The Tsakh. word seems to be, at first sight (together with Kryz. alaʁ and Khin. alaʁ id.) an Azerb. loan (Azerb. alaG 'weed'); however, within Turkic the root is purely Azeri (it is not attested even in Turkish), so it may itself be a Caucasian loanword.
Notes: An important common NC root, basically meaning 'sky' but with original religious and mystic connotations. There are no nominal reflexes in PD, but since the meanings 'breath, to breathe' and 'to get tired' often interchange, it seems tempting to compare also Darg. Ur. =amsVr / = ums- 'to be tired, get tired'.
Notes: Reconstructed for the PEC level. Although attested only in three languages (Chech., Darg. and Tab.), the correspondences are regular and the reconstruction seems reliable.
Notes: Cf. also HU: Hurr. edǝ 'body, thing', Ur. edi- 'thing' (see Diakonoff-Starostin 1986, 26). It is not quite clear which meaning is original - 'thing' or 'what', because both are rather widely reflected.
We should also mention a quite isolated form: Lak. ta 'when', which has an irregular devoiced t-, but possibly reflects the same EC root.
Notes: A common NC numeral (in WC it is reflected at least in Adygh languages, with less certainty - also in other subgroups). Reflexes in several EC subgroups reflect class prefixation (*u- or *j-); in PWC the initial labial also goes back to a class prefix (with assimilatory labialisation of the following affricate, as in a number of other cases). See Trubetzkoy 1930, 275; Абдоков 1983, 156.
In PEC the same root was also used in a contracted shape *-c̣ĕ- as part of names of tens, cf.:
Hurrian has replaced the numeral 'ten' by the word eman (of unknown origin), but has retained the root in ki-ži 'thirty' (ki- 'three'); names of other tens are unknown. See Diakonoff-Starostin 1986, 46.
