Articulatory and Perceptual Correlates of Syllable Structure

Indiana University, Bloomington IN

Investigators

Abstract

Two projects will examine how it is that speakers organize the very large number of actions which make up speech into an extremely rapid orchestration. Specifically, this research seeks to establish a phonetic definition of syllable affiliation. While the linguistics and speech science literature makes reference to the syllable, exactly what this unit is and what it means for a consonant to belong in the onset or the coda of a syllable has yet to be determined. The data to be examined here is elicited using a controlled rate-scaling paradigm. In this paradigm, speakers produce syllables of specified structure repeatedly in time to a metronome which controls the rate of speech. The rate of the metronome is systematically varied in order to sample a broad range of speech rates. The research plan includes four experimental procedures grouped into two projects. The first project comprises two experiments which work with an articulatory database of approximately 2400 syllables, composed of various stop-vowel sequences produced by speakers who repeated syllables in time to a metronome. This database includes dynamic measurements of the location of oral speech articulators, recordings of the glottal opening, recordings of oral air pressure, as well as acoustic records. The first experiment will determine articulatory differences between consonants in onsets and codas, as well as how these differences evolve in the face of speech rate increases. Also obtained in this first experiment will be information on the articulatory strategies speakers use to change speech rates, and how these strategies apply to consonants made with different articulators. The second experiment will acquire perceptual responses to the same database, thus determining which aspects of the articulatory differences are relevant to syllabification from the listener's perspective. The two experiments combined will provide a coherent picture of basic syllabic organization from both articulatory and perceptual perspectives. The second project will extend the results to consonant clusters and to tense-lax vowel pairs. This project is particularly designed to examine how language-particular grammatical constraints on allowable segment sequences affect speakers' production patterns. This project forms the first step at generalizing the results of the first project to less constrained speech situations.