Topic of this presentation will be our efforts to improve the results of Frank Wijnen and Hugo Quene (2000) presented in "Prosodic Phrasing and Relative Clause Attachment in a Three-Site Context". We, a small group of four students, therefore took a critical view on the original experiment and designed a new version of it, using new material and subjects. The course was supervised by Frank Wijnen himself, but most things were left to our decision.
So we can present some new results, and give some information about the caveats of designing, running and analyzing an experiment completely on your own, which has been a quite interesting experience to me.
The main question which we tried to answer was: Is it possible to transport relative clause attachment information in a globally ambigous context via prosody, and what means are used by the speaker for that? The experiment consists of three phases: Controlling preference effects in a pretest, a production task in latin square (*) design and finally analyzing both perceived attachment and prosodic properties of the recordings.
(*) not sure if this is the correct terminology