conc▪ While [v1] is prob. a borrowing from Phl, and [v3] perh. derived from [v1], [v2] does not immediately seem related to any of the other two values (see however below s.v.).
▪ According to Rolland2014a, [v1] has as its origin the same Phl etymon from which also modPe ǧawḫ ʻtroupe, assembly of people or animals’ is derived. According to the DHDA, ǧawq ʻgroup, troop’ is the first item of the root to be attested (at the rather late date of 791 CE). Given also the fact that the Ar root does not exhibit any obvious Sem cognates (except for a Te word?), the probability of a borrowing is quite high.
▪ [v2] : related to Gz gʷəḥqʷa ʻto bend (intr.), be distorted, to stooped, bent, curved, bowed (because of old age), to become weak’ (accord. to DRS a relation considered by Dillmann)? – Accord. to DHDA, the first attestation of this value is the adj. †ʔaǧwaqᵘ ʻwry-mouthed’. As also this is from a very late date – 845 CE –, and as the Gz parallel is quite doubtful, one should not exclude the possibility of a development *ʻ[v1] troop, group > [v3] to call together a group, muster a crowd; to cry at s.o. > [v2] to have a wry mouth (like s.o. crying at a group/troop to make them assemble)’. But this is highly speculative.
▪ [v3] : prob. denom. from [v1].
▪ …