VOT: Information

To control voice onset time (VOT) in measured increments, Arthur Abramson and Leigh Lisker used the Haskins Laboratories formant synthesizer. The basic pattern that they used was three steady-state formants for a vowel of the type [a]. Labial, apical, and velar stop releases were simulated by means of appropriate formant transitions. They synthesized VOT variants ranging from 150 ms before the release to 150 ms after it. For voicing before the release (voicing lead), they used only low-frequency harmonics of the buzz source. For voice onset after release (voicing lag), the interval between release and onset of the periodic source was excited by hiss alone. Three conditions of VOT for synthetic labial stops are shown in the figure, below.

Using such stimuli, Abramson and Lisker (1973) ran tests with native speakers of Latin American Spanish in which they identified stops synthesized with VOT varying in small steps. Some of the results of these tests can be seen in the figure, below. On the abscissa, negative numbers are assigned to voicing lead and positive numbers to lag, while the moment of stop release is labelled zero. The stimuli varied in 10-ms steps, except for the range of -10 to +50, which were made in 5-ms steps. Responses of the subjects showed a fairly good fit between production and perception. For each place of articulation, the identification curves are functions of VOT values. The synthetic patterns clearly provided enough information for two good perceptual categories at each place of articulation.

[From: Abramson, A., & Lisker, L. Voice timing perception in Spanish word-initial stops. Journal of Phonetics, 1973, 1, 1-8.]

VOT