Adaptation to an electropalatograph palate: Acoustic, impressionistic, and perceptual data

The purpose of this study was to evaluate adaptation to the electropalatograph (EPG) from the perspective of consonant acoustics, listener perceptions, and speaker ratings. METHOD: Seven adults with typical speech wore an EPG and pseudo-EPG palate over 2 days and produced syllables, read a passage, counted, and rated their adaptation to the palate. Consonant acoustics, listener ratings, and speaker ratings were analyzed. RESULTS: The spectral mean for the burst (/t/) and frication (/s/) was reduced for the first 60'120 min of wearing the pseudo-EPG palate. Temporal features (stop gap, frication, and syllable duration) were unaffected by wearing the pseudo-EPG palate. The EPG palate had a similar effect on consonant acoustics as the pseudo-EPG palate. Expert listener ratings indicated minimal to no change in speech naturalness or distortion from the pseudo-EPG or EPG palate. The sounds /t, d, , s, z, / were most likely to be affected. Speaker self-ratings related to oral comfort, speech, tongue movement, appearance, and oral sensation were negatively affected by the presence of the palatal devices. CONCLUSIONS: Speakers detected a substantial difference when wearing a palatal device, but the effects on speech were minimal based on listener ratings. Spectral features of consonants were initially affected, although adaptation occurred. Wearing an EPG or pseudo-EPG palate for approximately 2 hr results in relatively normal-sounding speech with acoustic features similar to a no-palate condition.