Implicit and explicit prosody in relative clause attachment

Using listening production, perception, and psycholinguistic tasks, this study investigates the interactive effects of prosodic prominence and phrasing on relative clause attachment in English. According to the Implicit Prosody Hypothesis (Fodor 1998/2002), the presence of a prosodic boundary before or after NP2 in NP1-NP2-RC constructions in English should prompt low or high attachment, respectively. We have found this to be systematically untrue for some individuals. We are currently exploring the possibility that the use of prosodic boundaries for syntactic closure in sentence processing is sensitive to individual differences in attention and memory for prominence structure. Another part of this project explores the role of prominence in RC attachment cross-linguistically.

Collaborators:
Sun-Ah Jun (UCLA)
Adam Chong (UCLA)

Related Publications:

 


Information structure and individual differences in the perception of prosody

This study consists of multiple experiments exploring the perception of prosodic events in English using a ‘rapid prosody transcription’ method (Cole, Mo, & Hasagawa-Johnson, 2010) and a similar task used in Bishop (2009). In these experiments, we attempt to predict naive listeners’ perception of phrasing and prominence using Intonational Phonological categories and sentence-level semantic/information structural context. Parts of this study are funded by PSC-CUNY Enhanced Grant #67842-00 45. to J. Bishop.

Collaborators:
Grace Kuo (Concordia University)

 


Perception of location in f0-range

Previous studies (Whalen & Honorof 2005; Whalen & Honorof 2010; Bishop & Keating 2012) have investigated whether listeners are able to identify where an isolated FO token falls in the individual range of an unfamiliar voice. While previous studies show a correlation between speakers’ ranges and listeners’ ratings of tokens within them, they utilize a small number of speakers, with fairly similar ranges. This study extends previous findings by testing a large number of voices with more varied ranges, while also probing for the role of voice quality and intrinsic pitch in F0 range estimation.

Collaborators:
Doug Whalen (Haskins & The CUNY Graduate Center)

Related Publications: