Vowel duration change and its underlying physiological mechanisms.

Abstract

Two explanations have been proposed for the relationship between vowel-target formant frequency and articulatory stress. The first, the “extra energy” hypothesis, suggests that stressing is accompanied by larger signals to the articulators, so that stressed syllables are longer and have more extreme formant values. The second, the “undershoot” hypothesis, suggests that the signals sent to the articulators are of constant magnitude, but that changes in timing result in differences in formant frequency. This view leads to a prediction that the relationship between target formant frequency and duration is fixed, whatever the cause of the duration variation. Acoustic and electromyographic measures were made of productions of nonsense syllables with varying stress and speaking rate, by three speakers. Results fail to support the undershoot hypothesis, since syllable duration and vowel target frequency are independent. While speaking rate variations are accomplished in a different manner by the three speakers, the “extra energy” model for stressing seems to be supported.