Comment: Attested only in Archi, but having probable external parallels (although the Arch. pharyngealisation here is excessive, and probably secondary). Arch. qIʷeni presupposes a suffixed form *q:ʷänV-j; the suffixless form (with a transformed meaning) is also preserved in Archi: Arch. qIʷen 'rock, cliff'.
Meaning:1 behind 2 suffix of a locative case (series Post)
Lezghian:-q 2; -q-an 'suffix of profession'
Tabasaran:qa 1, -q 2, prev. q-
Agul:qa 1, -q 2, prev. q-
Rutul:quʔ 1, -χ/-q 2
Tsakhur:-qa 2, -q-än 'suffix of profession'
Kryz:-χ 2
Budukh:-χ 2
Archi:χi-r 1, -χu- loc. suffix
Udi:qo-š 1
Comment: As a case marker PL *-q(a) usually denotes location behind, but can be also used as a possessive case (the same development is characteristic also for Andian parallels). Somewhat problematic is relating here the Arch. translative case in -χ-ut ('through smth.') and the comparative case in -χ-ur. See Alekseyev 1985, 54.
PL *q(a)- is also used as a preverb ('back, behind': Tab., Ag., Rut., Tsakh. q-; possibly also Lezg. -χ- with the meaning of repetitiveness ('once more'), see Alekseyev 1985, 121).
Comment: External evidence speaks strongly in favour of *-t- in this root. The Lezg. form qʷarc may, in fact, reflect PL *qʷart, if we suppose that the oblique stem had been *qʷarte- (with a regular palatalisation -t- > -c- before a front vowel and expansion of -c to the direct base, as in several other PL roots). In this case similar forms in some other Lezgian languages (Tab. qʷarc, Rut. qʷarcalɨj 'house-spirit') are to be explained as loanwords from Lezg.
Meaning:1 bit,slice 2 notch, nick 3(biting part) > chin
Lezghian:q̇ac̣ 2
Tabasaran:q̇ac̣ 1
Agul:q̇ac̣ 1
Rutul:q̇ac̣ 3
Tsakhur:q̇ac̣ 3
Comment: Obl. base *q̇ac̣ɨ-, cf. Tab. q̇ac̣u-, Lezg. erg. q̇ac̣u. The Ag. form for "to bite" (q̇ac̣iḳas) goes back to this root. 4th class in Rut. and Tsakh.
Meaning:1 plait 2 forelock, mane 3 cap 4 parting (of hair)
Lezghian:q̇eč̣em 4
Rutul:q̇amč̣äl 1
Tsakhur:q̇amč̣ele 1
Budukh:q̇emelči 3
Archi:q̇am 2
Comment: The Lezg. form is a metathesis. The pure root is preserved only in Arch. (which also has an etymological doublet: q̇ama-t:u 'women's hair' - probably borrowed from Lak. or Darg.); all other languages reflect a suffixed form. The word belongs to the 3d class in all class-distinguishing languages.
Comment: 4th class in Rut. The PL paradigm can be reconstructed as *q̇ar,*q̇orV- (cf. Rut. q̇ɨra-) with a usual Ablaut type and further vowel levelling (direct base retained in Lezg. and Tab., obl. base - in Rut.).
Comment: Arch. has a compound, lit. "tortoise-frog". The relationship to this root of a Tsakh. form recorded in Khaidakov 1973 (qIɨrqIɨmɨṭ 'frog') is not quite clear.
