Comment: Lezg. and Tab. have an expressive dental preverb. Loss of -n-conjugation in Tsakh. is not clear; it may be caused by an influence of other similar roots (cf. e.g. Tsakh. -o=ʁI- 'to drink, swallow' (with different preverbs)). We may note that an analogical influence occurred in Tab. Düb. - we have there qIuṭ-u=q:I- 'to swallow', opposed to Tab. Kand., lit. qIuṭ-u=q̇I- and obviously influenced by u=q:- 'to drink' (see *HVq̇_Vr).
Comment: Labialisation in Archi is probably secondary (automatic in the infinitive eq̇ʷmus < *eq̇n-bos and analogical in other forms). The root is attested only in Archi, but has possible external parallels.
Comment: Cf. also Arch. dur. =erχ:u-r (preserving old labialisation). There are no Tab. or Ag. data (these languages preserve well the old Inlaut resonants), thus reconstruction of Inlaut *-r- is also possible. See Лексика 1971, 258.
Meaning:1 to comb 2 to scratch, scrape 3 comb (n.)
Lezghian:čuχʷa- 2
Tabasaran:ǯ̌-u=χ- 1
Agul:ž-irχ- 1
Tsakhur:giw-a=χa- 1
Archi:=eχ:a- 1
Udi:oχ 3
Comment: Cf. also Tab. Düb. ǯi=χ-, Ag. Bursh. ǯurχa-, Arch. dur. =erχ:u-r. All reflexes point uniformly to *-χʷ-, except Arch. -χ:-. Tense -χ:- in Archi should be probably explained by the influence of gobχ:a-s 'to scratch, scrape' (or the Lak. form =uχ:Ia-n ?). Lezg., Tab. and Ag. reflect a stem with an expressive preverb (*č:- or *č-).
Comment: In most languages the stem is used in a compound with *jirḳʷ 'heart' in the oblique form (*jirḳʷe- or *jirḳʷe-le-); in Tsakh. this form had become a preverb (ḳile-). The verb belongs to the strong conjugation class.
Comment: In Tab., Ag., Kryz., Arch. and Ud. we witness a case of interaction between the root laryngeal (normally reflected as -h-) and the class infix -b- (PL *p:). Cf. the paradigm in Kryz. (inf. lɨpɨ-ǯ < *l-i-b-ʔIʷɨ-, dur. lu-ri, prohib. m-uʔu), Bud. (juʔu,jɨpǝ-ǯi), Arch. (bo-s < *ʔi-p:-ʔIʷɨ-s, dur. wa-r < *ʔiʔIʷa-r without the class infix). In Lezg. and Kryz. there is a *l- preverb; other languages reflect a prefixless form.
See Trubetzkoy 1930, 88; Гигинейшвили 1977, 138; Талибов 1980, 277.
Comment: Complicated reflexes in modern languages are due to combinations of class infixes with the Inlaut laryngeal: cf. Tab. iṗu-b (*ʔi-b-ʔ-) / iṭu-b (*ʔi-d-ʔ-); Lezg. praet. ṭü-na (*ʔi-d-ʔu-na) as opposed to the infinitive ne-z (*ʔin-Vs: < *ʔiʔ(ʷ)l-Vs:). The root-final *-l- is well preserved in Rut., Kryz., Bud.; cf. also Ag. Bursh. ʕüṭal-as. On the other hand, the Rut. terminative form li-ʔi-r and the Arch. lah- reflect the metathesized variant *liʔʷä-.
Comment: The Rut., Kryz., Bud. and Ud. forms are historical plurals (with -Vr and -uχ). The Archi form is also a result of a secondary reinterpretation of the form *jam (< *ʔIam:) as a plural; a new singular form *ja had been built and later yielded (quite regularly) Arch. oj. Note that an analogous thing happened also in Udi, where there exists a noun i 'hearing' (and the verb i-baksun 'to hear').The old "plural" form (but already refigured according to the singular) is preserved in Arch. ojom 'ears'.
See Лексика 1971, 126; Гигинейшвили 1977, 116; Талибов 1980, 274.
Comment: Several forms (Rut., Tsakh., Ud.) go back to a suffixed form *ʔIam:u-l (or *ʔIum:u-l with assimilation/Ablaut); the suffixless stem is preserved in Shakhdagh languages and Arch. 3d class in all class-distinguishing languages.
