Perception of stop consonants produced by esophageal and tracheoesophageal speakers

Abstract

This preliminary study investigated the perceptual characteristics of stop consonants produced by two esophageal and two tracheoesophageal (TE) speakers, and one normal speaker. Speech stimuli consisting of CVCVC syllables were presented in the sound-field to nine naive, normal-hearing listeners who transcribed their perceptions of each stimulus. Listener responses for each speaker were collapsed into confusion matrices and analyzed for overall intelligibility, place, and voicing features. Error data were also quantified and analyzed. Results indicate that overall, stop consonants produced by TE speakers were significantly more intelligible than those produced by esophageal speakers. For all six stops investigated, intelligibility was greater for TE speakers. Findings are discussed in relation to existing aerodynamic, acoustic, and speech production data for alaryngeal speakers.