Notes: A very complicated case. All languages of Daghestan explicitly reflect the labialised *cw (correspondence Av. c : PA *s : Lak c(ʷ) : PD *s: : PL *-s). The Nakh and PWC forms, however, reflect a simple *c (otherwise PN would have *č, and PWC - *šʷ). The reason is obviously a secondary labialisation in EC, due to the labialising influence of *ʔw (labialisation is also indirectly reflected in PD -m-, assimilated from -n-). The root must have contained both a nasal (cf. forms like PN *nace, Lak. nuwca, PD *ʕems:, Arch. winisi) and a liquid resonant (cf. PA *risV and most Lezghian forms). However, a disyllabic protoform *ʔwnĭrcĔ or *ʔwrĭncĔ would give quite different reflexes in most languages (see phonetic tables). We must, therefore, suppose a protoform like *ʔwinĭrcĔ to account for all the complexity of development. The trisyllabic Archi form winisi thus appears to be an archaism, directly reflecting the PNC root structure.
The root as such is not reflected in PTs (we should expect something like *rizV or *rɨzV). There exists, however, a PC root *rɨcV(nV) (Tsez. recenjo, Gin. recen 'ant', Bezht. rica, Gunz. rɨcǝ 'tick'). It can be an indirect reflex of the same EC root, modified under the influence of other insect names like PTs *nɔcǝ 'louse' or *hũcV 'ant'.