Velopharyngeal aerodynamics of /m/ and /p/ in tracheoesophageal speech.

The purpose of this study was to compare oral pressure (P(o)), nasal airflow (V(n)), and velopharyngeal (VP) orifice area estimates from 12 tracheoesophageal (TE) and 12 laryngeal speakers as they produced /p/ and /m/ in syllable series. The findings were as follows: (1) TE speakers produced greater P(o) than the laryngeal speakers; (2) for /p/, TE speakers generated V(n), and VP orifice area estimates comparable with, or less than, the laryngeal speakers; and (3) for /m/, TE speakers had V(n) and VP orifice area estimates greater than the laryngeal speakers. The elevated P(o) could be the result of several factors such as high source driving pressures and vocal tract volume changes postlaryngectomy. Attempts at more precise articulation, and subsequently less coarticulation, by the TE speakers may explain the V(n) and VP orifice area estimates for /p/ and /m/. TE speakers may be limiting the oral-nasal cavity coupling for /p/ (smaller VP gap, less V(n)) in an attempt to produce a very precise oral /p/. For /m/, TE speakers may be attempting to overtly mark the consonant as a nasal (greater V(n), larger VP gap). Further studies are needed to confirm/refute the explanations postulated here regarding the VP aerodynamic differences that were identified.