Do So and the Structure of VP

Do So and the Structure of VP

The do so construction in English is analyzed differently by Culicover and Jackendoff (2005) and Sobin (2008). The former analyzes it on the basis of do so as a non-proform, while the latter analyzes it on the basis of do so as a pro-form. The non-proform analysis accounts for this construction with externalized arguments relatively well. However, it fails to capture the relationship between this construction with an externalized argument and a “plain” do so construction. Moreover, it does not account for why argument externalization occurs and how it occurs. In this respect, it has difficulty in accounting for various characteristics of do so constructions. On the contrary, the pro-form analysis may capture some common characteristics of do so constructions, since it regards do so as a pro-form. Despite this merit, it makes the structure of the VP in do so constructions unnecessarily complex by assigning each VP to each adjunct. This seems to disagree with a speaker's recognition of the structure of do so constructions. Therefore, I propose a new analysis that will solve all these problems and capture common characteristics of do so constructions.