A common NC root with fairly good phonetic and semantic correspondences. PN and PL have forms with dental prefixes (aspectual? - cf. also the durative Ablaut grade *-i- in PL); other languages have only prefixed class markers. As for the phonetic side, there are some problems with the medial resonant (perhaps we should reconstruct *-l-, not *-r-, to account for its loss in PL), and with the Dargwa reflex: we should expect lax *-x- in PD. Violation of the phonetic rule (*ʎw > PD *x(ʷ) in tense words) may have been caused by early delabialisation (cf. an analogous delabialisation in PL). In some languages the root may have been influenced by PEC *-ăƛ_wVn 'to resemble, be similar' (or even merged with it), but the two roots certainly are to be kept separate.
The earliest form of the root could have been *ʎ_wVrĂ - this is suggested by the probable Hurr. match šir- 'to be equal, alike' (see Diakonoff-Starostin 1986, 49). The shape **=ă[r]ʎ_wĂ must have been a result of adding prefixes (*=a-ʎ_wVrĂ > *=a-ʎ_wrĂ > *=arʎ_wĂ).