Developmental changes in the production of obstruents. I. An acoustic analysis of voiceless obstruents

Abstract

This is one of two papers measuring the acoustic structure of prevocalic obstruents in children and adults, examining age-related differences. This study looks at voiceless consonants. The data will be drawn in part from a large scale acoustic database developed in collaboration with ATR and include word initial voiceless obstruents from real words (in a range of phonetic contexts) as well as a series of syllable initial intervocalic obstruents drawn from nonwords (in the constrained context [a —a]). Of particular interest are the intervocalic consonants, as there is a relatively limited amount of knowledge on the acoustic nature of obstruents produced by children in this position. Utterances elicited from speakers ranging in age from 3–70 years were recorded online in a quiet room environment using high quality microphones and direct analog-to-digital conversion to computer disk. Acoustic measurements collected include the spectral moments [K. Forrest et al., J. Acoust. Soc. Am. 84, 115–123 (1988)], formant transitions, VOT, closure duration (for stops), noise duration (for fricatives), and preceding vowel duration. Acoustic differences obtained will be discussed as a function of age group, gender, and phonetic context. [Work supported by an INRS Award from NIH and research funding from ATR (Fox, P.I.).]