Boundary-induced prosodic strengthening and its function in interactive and read speech

Abstract

This study investigates the functions of boundary-induced prosodic strengthening relative to speech style, focusing on English plosives and nasals in Intonational-Phrase(IP)-initial and IP-medial positions in interactive and read speech. Two accounts seek to explain the function of prosodic strengthening: syntagmatic contrast enhancement (SCE: enhancement of the contrast between neighboring consonants and vowels) (e.g., Cho & Keating, 2001) and paradigmatic contrast enhancement (PCE: enhancement of phonological contrasts) (e.g., Georgeton & Fougeron, 2014). For interactive speech, pairs of participants produced words while exchanging information about scenes; for read speech, each participant individually produced words in written sentences. For plosives, Voice-Onset-Time measurements showed SCE in read speech, with plosives being more consonant-like (less voiced) IP-initially than IP-medially, but PCE in interactive speech, with voiced and voiceless plosives being more distinct from each other and from other places of articulation IP-initially than IP-medially; spectral measures showed PCE in both interactive and read speech. These results suggest that speakers make plosives phonologically more distinct from one another in interactive speech to help listeners identify them better. Nasals were more consonant-like (less nasality) IP-initially than IP-medially in both interactive and read speech, showing SCE, attributable to nasals being phonetically and phonologically distinct from non-nasal consonants.