Characteristics of phonation in women who produce voiced versus voiceless /h/

Abstract

Production of the consonant /h/ typically involves a vocal-fold abduction gesture. Although /h/ is traditionally described as voiceless in English, voiced allophones also occur. Our recent work has described the phonatory behavior of women who produce a mixture of voiced and voiceless /h/; our results suggested speaker-specific patterns of the factors that lead to devoicing. This study characterizes the behavior of normal women who produce exclusively voiced /h/ and compares it to those who produced both voiced and voiceless /h/. Measures of baseline (dc) airflow are used as an index of abduction degree; peak intraoral pressures during adjacent /p/ closures provide a measure of subglottal pressure; and pulse-by-pulse measures are made of f0 and pulse amplitude (ac flow) around the /h/ abduction gesture. Multidimensional analyses are carried out within speakers to determine how these vocal parameters are related. We expect that women who produce only voiced /h/ will differ from those who devoice some /h/ productions in the parameters that facilitate phonation, such as higher subglottal pressures, smaller abduction degrees, and reduced vocal-fold tension as reflected in f0 measures. Based on our previous findings, however, we also expect individual variation in the phonatory settings that produce voiced /h/. [Work supported by NIH.]