Effects on speech of introducing aer

Certain contextually-predictable phonetic variations have been hypothesized to be due to differing aerodynamic factors: (1) the F0 perturbation on vowels following voiced and voiceless obstruents, (2) the slight differences in Voice Onset Time, and (3) devoicing of back-articulated stops such as [ ]. To test these hypotheses we introduce two techniques that create artificial perturbations in pharyngeal pressure thus causing changes in the magnitude of the pressure drop across the glottis and the pressure drop across the supraglottal constriction. One method involves passively venting oral air through a tube inserted between the cheeks and behind the upper molars. The other method involves connecting a vacuum source to a catheter inserted into the pharynx via the nasal passage. Preliminary data suggest that the F0 perturbation characteristic of voiced / voiceless obstruents is not significantly affected by the aerodynamic perturbation. The other two effects are.