its derivation from the verbal root
*wly ‘to adhere, to be near’ advocated in
Voigt 1999:41 is highly unlikely
If the preposition is diachronically identical with the precative marker
*li (Huehnergard 2006:16), one may
trace it back to PS, assuming an independent loss of its prepositional function
in Akkadian and MSA (the l-precative is prominently attested in both). If this
relationship is rejected, the preposition *li may be regarded as a shared innovation
of the CS/EthS genealogical unity (Kogan 2015: 119)