ABSTRACT
Two studies were performed on relations between the production and perception of sound contrasts in 19 speakers of American English. One study was of two vowel contrasts (/u-Ʊ/ and /α-^/); the other was of the sibilant contrast, /s-∫/. In both studies, measures of perceptual acuity for the contrasts were obtained in ABX discrimination tests. In the vowel study, measures were made of the articulatory and acoustic distinctiveness of the produced contrast, called “contrast distance.” In the sibilant study, measures were made of acoustic contrast distance and the speakers' use of a quantal biomechanical effect (called a “saturation effect”) while producing the sounds. The degree to which speakers used the saturation effect in producing /s/ and /∫/ was assessed with a custom-made sensor that measured contact of the under side of the tongue tip with the lower alveolar ridge; such contact normally occurs during the production of /s/ but not /∫/. Results from each of the studies showed that subjects with higher acuity produced the contrasts more distinctly. The most distinct sibilant productions were obtained from subjects who had high discrimination scores and who also used contact in producing /s/ but not /∫/. Subjects who did not use contact differentially when producing the two sibilants and who also discriminated the synthetic sibilants less well produced the least distinct sibilant contrasts. Intermediate degrees of sibilant contrast were found with subjects who used contact differentially or discriminated well. The combined findings from the two studies are compatible with the DIVA model of speech motor planning, in which goals for phonemic speech movements are in auditory and somatosensory spaces.
