Open Conference Systems, Acoustics Week in Canada 2015

Font Size: 
Startle-induced evidence for multi-dimensional information in speech plans
Chenhao Chiu, Bryan Gick

Last modified: 2015-10-17


Mouthed and vocalized speech tasks differ not only in kinematic motor control, such as shorter word duration and hypoarticulation in lip movements for mouthed speech (Crevier-Buchman et al., 2011), but also in breathing pattern (Murphy et al., 1997) and EMG amplitude (Wand et al. 2009). While a vocalized speech plan must account for the built-up of intraoral pressure, a mouthed speech plan involves no or limited intraoral pressure and less significant perioral muscle activations.


Using the design of startling auditory stimulus (SAS, cf. Stevenson et al. 2014), the current study examines SAS-induced vocalized and mouthed syllable /ba/ production. As vocalized speech and mouthed speech require different degrees of intraoral pressure and muscle activation, it is predicted that kinematic distinctions between vocalized and mouthed speech are preserved in SAS-elicited responses. Results show that while lip compression appeared obligatory in vocalized trials in both control (91% of testing trials) and SAS-induced responses (87%), lip compression only occurred sporadically in mouthed trials (63% control responses vs. 24% SAS-induced responses). More significant lip compression displacement was also observed in vocalized speech than mouthed speech. These findings suggest that multi-dimensional information, such as intraoral pressure and correspondent muscular coordination, is included in speech plans.