Stop and vowel identification from acoustic segments of VC and CV syllables

Abstract

Listen

Translate article

This study involved the identification by nineteen subjects of stops /p,t,k,b,d,g/ and vowels /i,u,ɜ̆/ from the acoustic cues of noise, transition, noise and transition, and transition and vowel which were edited from VC and CV syllables. The major findings are (1) vowels were identified from consonants better than consonants were identified from vowels (99% vs 52%); (2) vowels were identified from transitions better than were consonants (99. 3% vs 45. 6%); (3) R to L identifications of vowels from noise were greater than L to R identifications (73% vs 39%); (4) R to L and L to R identifications of vowels from transitions and from transitions plus noise were very high (at least 99%); (5) there was little difference in the identification of stops from noise cues in CV and VC syllables (97% vs 96%); (6) there was little difference in the R to L and L to R identification of consonants from transitions (46% vs 44%); and (7) R to L identification of consonants from transition and vowel was greater than L to R identification (66% vs 39%). The implications of the findings were considered with respect to perceptual models, in particular, those relating to vowel transitions, masking, and coarticulatory effects.