Against the complex predicate analysis of secondary predication

This paper provides an empirical argument against the complex predicate analysis of the two types of secondary predication – depictive and resultative – put forth by Cormack & Smith (1999) and Rothstein (2003, 2004). The data coming from various languages present challenges to major assumptions on the syntactic structures of depictives and resultatives adopted by the complex predicate analysis. The issues disappear under the traditional approach to secondary predication, along with the novel argument that depictive secondary predicates are TP adjuncts. Finally, I show that the proposal makes a correct prediction on negated secondary predication in English and Japanese small clause constructions.