Goldberg's Construction Grammar directly addresses this issue by defining constructions as components of grammar in their own right -- certain syntactic constructions, such as the resultative construction, exist because there is a form-meaning pair in the grammar which licenses the construction. Similarly, the Wechsler () approach licenses the construction via a lexical rule which adds the resultative phrases and ensures appropriate semantic integration, and the more specific Raising Rule which also adds a complement NP to the subcategorisation lists of the verbs which appear in the construction.
The analysis, in contrast, provides a post-hoc explanation of the construction without explaining why it occurs in the first place. They suggest (p. 55) that the resultative phrase is licensed by virtue of the identification of an event position in the argument structure of the verb and an event position in the argument structure of the head of the resultative phrase. How this identification could also account for the addition of a post-verbal NP, however, is unclear -- these NPs seem to be required on the analysis only due to the semantic role they play in the resultative construction. According to the Change of State Linking Rule (p. 51), these NPs must refer to an entity undergoing a change of state, but, the verbs in this construction do not independently encode a change of state and so the post-verbal NPs cannot independently be licensed via this linking rule. These NPs only play this semantic role by virtue of the interpretation they receive in the context of the resultative construction. Thus one of the core syntactic properties of the resultative construction cannot be accounted for in terms of independent syntactic or semantic factors.