Comment: 3d class in Kryz. and Bud. Obl. stem *pirpe- or *pirpɨ- (Tab. pirpu- points rather to *pirpɨ-, but this can be a recent innovation due to the wide spread of -u- bases).
Comment: Cf. also Lezg. Khl. p:ac̣i 'kid' (reflecting *p:-). The root contains a common PL diminutive morpheme *-j. Also derived from this root may be expressive forms: Bud. bic̣el, Kryz. mec̣ej 'penis (of a boy)'.
In Ag. Bursh. the adjective 'small' does not have the initial b-: Ag. Bursh. ic̣í-r. There exist also probably related forms in Tab.: Tab. c̣i-b 'few', Düb. c̣u-w id. This may be an indication that the initial *p:- in PL is a former class prefix (although loss of b- may be a secondary development in the Tab.-Ag. dialect area).
Comment: PL obl. base *p:aIƛ:e- (cf.Ag. Bursh. baIži-; the second front vowel conditioned the palatal reflexes of *ƛ: in all languages). Cf. also Tab. Kand. baž 'son, boy'.
Comment: Cf. also Rut. Khn. baIl, Arch. pl. bat:ur 'horns', bat-karṭi 'sinciput'. Obl. base *p:aIl:a- (cf. Lezg. p:ele-, Rut. bäli-). 4th class in Rut., Kryz. and Bud., but 3d class in Arch.
Comment: Cf. also Tsakh. Tsakh.baIǯe 'nostril; hole (in general)', (pl.) baIǯe-bɨ 'groin'. /We should also mention Kryz. päčer, Bud. pǝčǝ 'groin' - although their phonetic relationship to the Tsakh. form is not clear./
Comment: Oblique base (in *-ä) may be (not very certainly) reconstructed on basis of Lezg. Khl. p:eše- and Ag. Burk. ṗaIǯ,ṗaIǯu-. The Ag. form has been influenced by ḳeǯ 'leaf' (PL *ƛ̣ač̣a q.v.). 4th class in Kryz. and Bud.
Comment: Cf. also Tab. Düb. baIṭur 'dirt', Ag. Bursh. (with reduplication) baIṭaṭaj "sheep's dung" (both forms point to PL pharyngealisation). Tab., Ag. and Rut. reflect a suffixed form with PL *-Vr (plural suffix). Initial p- in Rut. and Bud. is irregular: probably a result of assimilation.
The vowel variation probably reflects the old Ablaut *p:aIṭ (dir. stem), *p:oIṭV- (obl. stem).
Comment: Attested only in Ag., but having possible external parallels. Laryngeal features are not quite clear: handwritten MSU records have baIṭi in Bursh., but baIti in Burk.; in the printed version there appears baIti in both dialects - this still is to be checked.
Comment: Cf. also Lezgh. p:ak:a-ma, Ag. bagajmi 'morning, dawn'. The semantic correlation 'morning/evening' ('dawn/dusk') is rather usual. The Rut. form bɨga may reflect the old oblique stem *p:ok:a-.
Comment: Obl. vowel stem *p:ak:ʷa-, cf. Tab. baga-x (with delabialisation), Ag. bagʷa-. For *-k:- cf. Lezg. erg. p:ak:uni, Arch. pl. bak:ur. The irregular front vowel and loss of labialisation in Kryz. and Rut. are probably due to contamination with another PL root *p:eƛ:V- q.v.
Comment: Cf. also Lezg. Khl., Akht. p:alk:an. The PL form is somewhat hard to reconstruct: Lezg. and Kryz. forms would point to *p:alƛ̣:an (with a unique combination *-lƛ̣:-); in this case the Tsakh. form should be regarded as a borrowing from Shakh-Dagh languages. On the other hand, peculiar reflexes could be due to the very peculiar cluster *-ltk- in PEC.
Comment: Obl. stem *p:alḳʷa- (cf. Lezg. Khl. p:arḳʷa-, Tsakh. Gelm. barḳʷa-). Medial *-l- is reconstructed because of the liquid's preservation before a velar (original *-r- would be lost). 3d class in Tsakh., 4th class in Kryz. The root must have denoted some specific bone; to judge from external evidence, originally it must have been some bone of leg.
Comment: The initial *p:- (*b-) is a former class prefix (to judge from external evidence). We reconstruct medial *(m) which should have been there if the EC etymology is right; all the available languages (Ag., Rut. and Tsakh.) regularly drop *-m- before dentals in consonant clusters.
Comment: Oblique vowel base *p:op:a- (or *p:op:ä-), cf. Ag. buwa-, Arch. buwa, (with vowel levelling) Lezg. p:ap:a-, Tab. baba-, Ag. Fit. baba-. In Arch. there occurred an irregular shift -p:- > -w- (paralleled by -t:- > -j- in dija 'father'). A "nursery" word, but undoubtedly of PL origin.
