Comment: Also used in a compound verb: Lezg. zʷeri-z (zʷer iji-z) 'to spin; to run'; Tab. ž̌ar tuwu-z 'to spin, turn round'. Lezg. > Khin. zar- in zar-ḳʷi 'to spin, twist (a rope)'.
Phonetically a rather complicated case (aggravated by our ignorance of Lezg. and Tab. dialectal parallels for the root). The most obvious solution would be to reconstruct PL *zʷ- or *žʷ- (typical for expressive roots), but this contradicts the external evidence. PL *š:ʷ can, indeed, have voiced reflexes in Lezg. and Tab., but in the latter only in front of PL narrow vowels (*i, *ɨ) which is not the case. We can only suppose secondary voicing of an original voiceless fricative in an expressive word, or voicing in Lezg. and borrowing from Lezg. into Tab.
Comment: 3d class in Rut. Cf. also Lezg. Khl. šʷet, Ag. Fit., Burk. šit: (all reflexes uniformly pointing to *-c̣:). The root tends to contaminate with PL *š:uṭ 'bug' (q.v.); this may explain the Lezg. Akht. form feṭ (regularly fet would be expected).
Comment: The Lezg. vi (cf. also Nüt. ɣü) is used presently only in compounds (like lezgi-vi 'a Lezgian' etc.); also rare is it in Tabasaran (in compounds like murgu-ž̌i 'a brave man' etc.). The obl. vowel base had been probably *š:ʷija-, cf. Ag. šuj,šuwa-, Fit. xüj,xüja-. For initial *š:ʷ cf. also Ag. Burk. š:̌uj. Kryz. and Bud. have suffixed forms (originally probably plurals). See Хайдаков 1973, 99.
Comment: 3d class in Bud. Bud. > Khin. fit 'whistle'. With prothetic *ʔi- (*ʔiš:ʷṭ-) cf. Tab. uš̌t (coexisting with š̌uṭ / š̌iṭ, Düb. š̌iṭi), Lezg. üft 'whistle'. There are also irregular reflexes: Ag. ust, Bursh. š:̌uluṭ, Rut. sut, Tsakh. śut id. - easily explainable by expressive reasons.
Comment: Some irregularities in the reflexation of *-s: (which is probably an original case suffix) are caused by dissimilative / assimilative reasons.
Comment: This root should not be confused with another PL interrogative, *š:i q.v. It serves as a base for PL *šʷimV- 'how many' (Lezg. šumu-d, Tab. š̌nu-b, Rut. šumu-d, Arch. šume-jṭu; Ud. Vart. ema, Nidzh. hema is partially contaminated with another PL interrogative stem, *ha-/*hi- q.v.). A loan from Lezg. is probably Khin. sam 'how many'.
Comment: Tabasaran and Agul have preserved this relic numeral only within the compound *s(ʷ)imV-c̣u-r 'thirty' (with 'ten' as the second part). The initial consonant reveals a variation between *s(ʷ) and *š(ʷ)-;cf. also Tab. Kand. sumč̣ur, Düb. simič̣ur. In the Richa dialect of Agul the form is xim-c̣ur: this is obviously a contamination with the normal PL numeral 'three' *ʎep:ɨ- (q.v.); but neither Tab. simi-, nor Ag. Bursh. šin- in 'thirty' can not be traced back to *ʎep:ɨ-.
Comment: The Lezg. form can be explained as a metathesis šalam < *šamal < *šum-al (a form derived from an original *šum). In Ag. cf. (without the originally plural -ar-suffix) Bursh., Fit., Burk. šum id. If this etymology of Lezg. šalam is right, other similar forms (Rut. šalam, Tsakh., Khin. šalam, Kryz. šäläm, Bud. šalam / čalam) must be treated as loanwords from Lezghi.
Comment: The Tab. and Tsakh. forms are probably related (cf. perhaps also Rut. č̣ʷed 'jackdaw, rook'), although the correspondences between initial consonants and vowels are irregular (due to the root's expressive nature).
Meaning:1 head of a spindle 2 support for a spindle
Budukh:ṭunuḳ
Archi:taInḳuli 1
Udi:t:irik:al 2
Comment: The precise meaning in Bud. is "деталь прялки в виде железного стержня в кружке для наматывания ниток". The word was obviously influenced phonetically by PL *tinuk: (Bud. tunug / tunuk) "spindle".
Comment: The Bud. form contains a historical suffix *-aj and has underwent an assimilation (*taḳʷaj > ṭöḳe). Obl. stem unclear (Lezg. erg. taḳ-uni). 3d class in Budukh.
Comment: Since there are no reflexes in Tsakh., Ud. and Arch., we can not distinguish between PL *m: or *p:; external evidence suggests the former. Obl. base *tam:e- or *tam:ɨ- (cf. Ag. tabu-, Rut. tábɨ-). 3d class in all class-distinguishing languages.