There are several unresolved semantic issues in Wechsler's treatment as well. As discussed in Section 4.3.1, there are subtle differences in the meaning of directed manner of motion constructions and that of other resultative constructions. Furthermore, there are differences in meaning between resultative constructions and the standard syntactic frames in which the verbs appear. In particular, as pointed out by Levin and Rappaport Hovav (1995, p. 54), the resultative construction must incorporate a causal relation. In Wechsler's approach, resultative phrases are added to a verb's subcat list, and the content of that phrase is unified with the telos value, as a result giving the sentence a telic interpretation, but no elements of meaning are added. So the meaning of We ran our Nikes threadbare is represented as, essentially, We ran and at the end our Nikes were threadbare, rather than We caused our Nikes to be threadbare by running. This problem is remedied by Wechsler (p.c.) through reinterpretation of the telos feature as involving causation. Even when the causative component is incorporated, however, the problem of distinguishing between the semantics of verbs of manner of motion on a directed motion use and the semantics of other resultatives remains, due to the uniform semantic treatment of all resultatives on this account. That is, if I am correct in asserting that the directed manner of motion verbs do not induce causation while standard resultatives do, Wechsler cannot account for the semantic distinction between them.
It is furthermore unclear that all instances of raising resultatives are actually truly raising, in the sense that they do not assign a semantic role to their NP complement. Consider again w4, repeated here, and contrast these examples with w15.
[4.48] Olof painted himself into a corner. [Wechsler , p. 2, (7b)] We laughed ourselves silly. [Wechsler , p. 2, (7d)]
Olof painted the dog into a corner. *We laughed Billy/the guests silly.
The variant of w4a, w15a, is acceptable, while the variant of w4b, w15b, is not. In the latter example, the referential identity of the NP complement and the subject is crucial to the acceptability of the example, and hence there is clearly something controlling the nature of the unsubcategorised complement. I suggest that the form of this resultative is conventionalised, and this conventionalisation contrains the general applicability of the mechanisms proposed by Wechsler.