Use of temporal envelope cues for the recognition of words and phonologically significant contrasts

Abstract

The goals of this study were: (1) measure the recognition of words and speech pattern contrasts using temporal cues only, and (2) to evaluate the effect of training on the ability to perceive segmental, suprasegmental, and phoneme information from the speech waveform only. Fifteen subjects, 20–40 years of age, with normal hearing, were tested before and after extensive training. Test stimuli consisted of: (a) eight segmental and two suprasegmental contrasts of the SPAC test using a binary forced-choice paradigm, and (b) monosyllabic words from the isophonemic AB word lists presented in open set. Spectral information was eliminated by multiplying the speech waveform with white noise using commercially available array processing software. Testing and training were under computer control. Preliminary results suggest that: (1) intonation and stress are perceived well using temporal information only, (2) manner is perceived better than either place or voicing contrasts, (3) training improves the perception of speech using temporal cues only, and (4) there is a strong correlation between speech pattern contrasts as measured in open and closed set paradigms. The findings will be discussed in relation to the relative importance of the stimuli’s spectral and temporal cues and their relevance to clinical data of individuals with profound hearing loss.