Perception of Coarticulated Nasality

Abstract

CVC and CVVC syllables were prepared in which the final consonants were either nasal consonants (/m/, /n/) or non-nasal consonants. The entire final consonant along with its vowel-consonant transition was spliced away. The resulting CV and CVV syllables along with carrier phrases were spliced at random onto a tape for presentation to listeners who were asked to predict whether the missing consonant was nasal or nonnasal. Results indicate that the presence of nasal consonants can be predicted very well. It is hypothesized that listeners utilize this information to lighten the phoneme-processing load. Such perceptually significant coarticulation of velar opening across the vowels in CVVN-type sequences probably has counterparts in the coarticulation of other articulatory gestures. In this way, perception uses or follows production.