Kurumada Dissertation Oral Tuesday

We’re pleased to announce that Chigusa Kurumada is giving her dissertation oral presentation on Tuesday, March 19 from 9-10am in the Greenberg Room. The format for this open part of the oral exam is a 30-45 minute talk by the PhD candidate followed by questions by those attending, for a total of no more than one hour, so make sure to come promptly!

Navigating variability in the linguistic signal: Learning to interpret contrastive prosody
Prosody is a powerful tool for conveying information beyond that conveyed in the words and phrases constituting a sentence (e.g., nuclear accenting of contrasted items: YOU shouldn’t have done that vs. You shouldn’t have done THAT). Past studies have proposed links between prosodic representations (pitch accents and intonations) and the semantics and pragmatics of sentences (e.g., Jackendoff, 1972; Büring, 2003; Pierrehumbert & Hirschberg, 1990). However, no study has explicitly accounted for how listeners navigate variability across speakers and contexts in order to arrive at intended interpretations.

In this dissertation, I propose a psycholinguistic approach to this variability problem drawing on insights from recent work on phonetic categorization and adaptation. Specifically, I investigate English speakers’ pragmatic interpretations of contrastive prosody through large-scale judgment surveys and an eye-tracking experiment. In addition, I present some developmental data to investigate how young children may overcome the variability problem and acquire abstract prosodic representations.

The main finding is that listeners flexibly adjust their interpretations of prosodic contours by integrating high-level contextual expectations and low-level distributional information in the acoustic signal. In particular, probabilistic inferences based on contextual information play a key role both in children’s acquisition and in adults’ processing of contrastive prosody. Implications are discussed with respect to how listeners learn and maintain flexible mappings between the variable linguistic signal and underlying representations.

Dissertation committee: Eve Clark (chair), Dan Jurafsky, Meghan Sumner, Michael Frank (Psychology)
University oral exam chair: Yoshiko Matsumoto (East Asian Languages and Cultures)