This paper presents four judgment experiments that investigated the acceptability of sentences in which a comparative clause modified an NP embedded within a complex NP. The comparative clause occurred either adjacent to its head noun or in extraposed position. The first three experiments addressed the claim of Haider (1997) that comparative clause extraposition is subject to a c-command constraint on S-structure whereas relative clause extraposition is not. To test this claim, Experiments 1–3 compared sentences with comparative clauses to corresponding sentences with relative clauses. Extraposition was found to be as acceptable as non-extraposition when extraposition was across a verb but less acceptable when extraposition was across an adjective. This
held for comparative and relative extraposition alike. The results of the first three experiments do therefore not support the claim that comparative clause extraposition is subject to tighter syntactic constraints than relative clause extraposition. The fourth experiment tested the claim of Shannon (1992) that extraposition from a topic phrase is less acceptable than extraposition from a phrase in focus. This experiment investigated comparative clause extraposition from the subject across two types of predicates: individual-level predicates, for which the subject is the default topic, and stage-level predicates, which come without an (overt) topic by default. The results show that extraposition across a stage-level predicate is more acceptable than extraposition across an individual-level predicate and therefore support the claim that extraposition from a topic is disfavored.
