The effects of speaking rate and amplitude variability on perceptual identification.
Mitchell S. Sommers, Lynne C. Nygaard, David B. Pisoni · The Journal of the Acoustical Society of America · 1992
Considerable evidence now exists [J. L. Miller, Phonetica 38, 159–180 (1981)] that the temporal properties of speech signals convey phonetic information in a rate-dependent manner. Although most recent models of speech perception have incorporated a stage at which listeners perceptually compensate for variations in articulation rate, as well as other linguistically important sources of variability, they have universally failed to specify whether there are perceptual costs associated with this ‘‘normalization’’ process. The present study was designed to determine whether open-set word identification is influenced by two sources of naturally occurring variability in the speech signal−differences in rate of articulation and overall stimulus amplitude. These two types of variability were chosen because the former is known to be important for phonetic judgments while the latter is generally considered phonetically irrelevant. The results demonstrated that identification of monosyllabic words was significantly poorer when the items were presented in lists containing three, as opposed to one, speaking rates. In contrast, identification accuracy was not significantly affected by the introduction of variability in overall amplitude. These findings extend the results of previous investigations [J. W. Mullenix and D. B. Pisoni, J. Acoust. Soc. Am. 85, 365–378 (1989)] that demonstrated detrimental effects of talker variability for word recognition and further suggest that only variations along phonetically relevant dimensions are likely to result in poorer identification performance. [Work supported by NIH.]