Abstract
This paper argues for the significance of linear information in the computational component. In the grammatical model adopted here, syntactic and prosodic derivations proceed in parallel from left to right in the manner of Phillips’ (1996, 2003) incremental structure-building, and a syntactic object is spelled out as a prosodic object in multiple steps. The approach to linearization based on this model provides a prosodically-based account of the robust contrast between Japanese and English VP-internal, ditransitive idioms: obligatory adjacency in Japanese and optional and extensive disjointness in English. This approach reinforces the significance of linear information in the syntax, which is constrained by prosody (Zec and Inkelas (1990)).
