Testing the relative salience of audio and visual cues for stop place of articulation

Abstract

Recent OT analyses have appealed to putatively universal rankings of perceptual entities in order to motivate analyses of phonological phenomena. J. Jun [WCCFL 14, 221–237 (1995)] bases a universal constraint ranking for susceptibility of stops to place assimilation on the inherent ‘‘salience’’ of place cues. E. Hume [WCCFL 17 (1998)] appeals to the ‘‘perceptual vulnerability’’ of labial stops in order to account for their cross-linguistic tendency to undergo metathesis. Both phonologists motivate their analyses with speculative and unquantified assumptions about place cue salience. This study used an objective definition of salience to empirically establish the relative strengths of audio and visual cues for stop place of articulation. Comparing results between audio-visual and audio-only groups showed that visual cues are strongest for labials but are also significantly higher for coronals than dorsals. Adding acoustic information between speech reception threshold and comfortable listening level conditions had weaker effects than adding visual information, but contributed most strongly to the intelligibility of dorsals. Overall results suggested that labial stops have the strongest cues for place regardless of modality, contrary to what Hume, Jun and other researchers would predict. [This material is based upon work supported by a National Science Foundation Graduate Fellowship.]