right node raising

[Robert Levine]

Does anyone know of any recent work on Right Node Raising which
presents a formal account (or even an informal sketch) of how one
might account for properties of this construction without any appeal
to an extraction mechanism? Something using perhaps a
linearization-based treatment, where given a set of conjucts where one
or more constituents making up the right edge of each conjunct were
structure-shared across the whole set, the DOM specification of the
coordination linearized this structure- shared (sequence of)
constituent(s) to the right of all the coordinated elements in the
set? Or some other promising line of attack?

Thanks,

Bob