Perception of middorsum palatal stops from the speech of three children with repaired cleft palate

Objective: The purpose of this study was to examine listeners' perception o f the middorsum palatal stop, a compensatory articulation used by individua ls with repaired cleft palates. Design: This study tested whether listeners could discriminate middorsum pa latal stops from matched alveolar (/t/) and velar (/k/) stops using a two-b utton "change/no-change" procedure. It also explored how listeners identifi ed the palatal stop by rating each sound on a scale of one to eight. Participants: Twenty listeners, 10 untrained and 10 trained in general phon etics (graduate students in speech-language pathology), participated in dis crimination and identification tasks during a 1-hour session in the Speech Perception Laboratory at the University of Buffalo. Measures: Discrimination was measured using d-prime, a score based on liste ners' hits, correct rejections, misses, and false alarms to the changes/no changes in the stimuli. Identification was measured by the mean rating scor e for each class of stops. Results: Listeners discriminated middorsum palatal stops from alveolar and velar stops, but their ratings for the middorsum palatal stops did not diff er from those for the regular stop consonants. The two groups differing in phonetic training did not perform differently. Conclusions: Listeners can discriminate middorsum palatal stops from other stop articulations, but they did not identify them differently from alveola r and velar stop consonants. The results suggest that considerable training listening to middorsum palatal stops is necessary for listeners to be able to reliably identify them.