(Ohio State University. Department of Linguistics, 2001) Winters, Steve
Jun (1995) and Hume (1998) incorporate perception into analysis of cross-linguistic trends in place assimilation and metathesis by claiming that the perceptual salience of specific segments motivates the ranking of relevant OT constraints. This study investigates the specific claims Jun and Hume make concerning the perceptual salience of cues for stop place of articulation to determine whether their salience actually could motivate the proposed OT rankings. Since both Jun and Hume based their proposals on a consideration of cues for stop place of articulation in the appropriate (VCCV) context for place assimilation and metathesis, this study only tested the salience of stops in this context. Listeners heard unreleased stops of three places of articulation (labial, coronal, dorsal) and two manners (oral, nasal) in two stress patterns preceding pre-vocalic oral stops of three other places of articulation. The perceptual salience (as measured in d') of stops in this context did not always bear out the predictions made by Jun and Hume. Interestingly, labials were generally the most salient place of articulation while dorsals were the worst. Nasal stops also turned out to be more salient than oral stops. Less surprisingly, pre-vocalic stops were more salient than post-vocalic stops, and place salience was highest for stops preceding coronals in pre-vocalic position. The variable success of Jun's and Hume's proposed hierarchies of place salience underscores the need to test the empirical validity of hypotheses concerning the interaction of phonology and perception.