Comparison of Vowel and Consonant Perception

Abstract

It has been suggested that there may be different mechanisms for vowel and consonant perception. A study of articulation-test-crew responses to minimal feature contrast vowel and consonant “rhyme” tests imbedded in various noises is reported here. Test-crew performance scores for vowels and consonants are compared with respect to noise spectral shape, test exposure, and specific test content by means of confusion matrices. Individual subject-response patterns for vowels and consonants are compared with respect to repeatability, predictability, variations from group means, and individual lingustic experience and history. It is found that perception of vowels relative to consonants varies among individuals, as does the absolute perception of each on a long-term average basis. This variation is not explicable in terms of individual pure-tone audiometric perception, or education, but may be related to individual linguistic experience and history as well as differences in perceptual processes. Occasional very large shifts in individual perception from trial to trial may be due to lapses in higher-order perceptual processes. No correlation was found between test scores and apparent subject fatigue to account for such lapses.