The role of prosody in English sentence disambiguation

• Published in: The Journal of the Acoustical Society of America Vol. 136; no. 4_Supplement; p. 2176
• Main Author: Miller, Taylor L.
• Format: Journal Article
• Language: English
• Published: 01.10.2014
• ISSN: 0001-4966; 1520-8524
• DOI: 10.1121/1.4899881

Only certain ambiguous sentences are perceptually disambiguable. Some researchers argue that this is due to syntactic structure (Lehiste 1973, Price 1991, Kang & Speer 2001), while others argue prosodic structure is responsible (Nespor & Vogel 1986 = N&V, Hirshberg & Avesani 2000). The present study further tests the role of prosodic constituents in sentence disambiguation in English. Target sentences were recorded in disambiguating contexts; twenty subjects listened to the recordings and chose one of two meanings. Following N&V’s experimental design with Italian, the meanings of each target structurally corresponded to different syntactic constituents and varied with respect to phonological phrases (ϕ) and intonational phrases (I). The results confirm N&V’s Italian findings: listeners are only able to disambiguate sentences with different prosodic constituent structures (p < 0.05); those differing in (I) but not (ϕ) have the highest success rate—86% (e.g., [When danger threatens your children]I [call the police]I vs. [When danger threatens]I [your children call the police]I ). As reported elsewhere (e.g., Lehiste 1973), we also observed a meaning bias in some cases (e.g., in “Julie ordered some large knife sharpeners,” listeners preferred “large [knife sharpeners]” but in “Jill owned some gold fish tanks,” they preferred “[goldfish] tanks”).