Section 4.3.3 introduced data which reflect the restricted nature of the resultative phenomenon. This is not a fully productive phenomenon, in which any resultative phrase can be predicated of any noun phrase. This fact must be accounted for in any treatment of resultatives.
Levin and Rappaport Hovav (1995) do not address the issue of semantic restrictions on the resultative phrase at all, as they aim to give a syntactic account of resultatives. Their approach does not, however, exclude an interaction with pragmatics. In addition, the lack of productivity in this phenomenon creates particular problems for Wechsler's (1996) treatment, since he claims that raising resultatives impose no semantic constraints on the resultative phrases whatsoever. Goldberg (1995), on the other hand, approaches the issue by identifying semantic constraints which apply to the constructions she proposes, and assuming lexicalisation of highly idiosyncratic instances of the constructions. The problem of fully accounting for infelicitous substitutions remains, however.