The role of syntax in the nuclear stress rule

How directly the operation of phrasal stress placement refers to syntax is theory-dependent: directly in some (e.g., Truckenbrodt 1995, Kahnemuyipour 2009), indirectly in others (e.g., Chomsky & Halle 1968, Halle & Vergnaud 1987). Adequately evaluating this issue requires knowing both the relevant syntactic structure(s) as well as how syntax interacts with phonology – neither is trivial. This paper argues that syntax transparently feeds prosody at regular sub-intervals of structure building (e.g., Uriagereka 1999, Chomsky 2001), and the phrasal stress placement operation refers directly to syntactic hierarchy, without exception. As such, wherever the NSR’s predictions are “incorrect”, the syntactic representation must be amended (e.g., Steedman 2000, Wagner 2005). The major contribution of this work is not in its specific findings, but rather in its demonstration of a methodology by which phrasal stress data are used to understand syntax and not vice versa.