Place: CS Conference Room in MUDD
Abstract:
This talk will describe a collection of recent studies examining the prediction and processing of intonational prominence in spoken language discourse. I will begin by offering a short tutorial on intonational structure and prominence, then describe efforts to predict prominence in English and also in Japanese (a language with a very different intonation system from English, and one that I have worked on extensively). I will then briefly summarize work specifically concerning the way in which discourse context and information status affects the distribution of prominences in these languages. Finally, I will talk in some detail about my very recent eye-tracking work on the processing of prominent pronominal forms in discourse context.