Katerina L. Haley
Ralph N. Ohde
Div. of Hear. and Speech Sci., Box 552, Sta. 17, Vanderbilt Univ. School of Medicine, Nashville, TN 37232
Previous research has noted a reduction in perceptual identification performance when the speaker varies from stimulus to stimulus. This reduction has been interpreted in the context of a speaker normalization process that separates speaker-specific information from invariant phonemic units. Alternatively, the effect may relate to stimulus uncertainty from a general psychoacoustic perspective. The present study addressed this possibility by examining the effect of variability not only in talker identity, but also in stimulus type. The stimuli were short segments taken from nasal consonent + vowel syllables produced by one male adult, one female adult, and two children. Segments of 25- and 50-ms duration were edited from the nasal murmur and the onset of the vowel. These stimuli were then grouped according to four variability conditions and presented to listeners for place of articulation identification. The results showed that both sources of variability reduced identification performance significantly. Separate analyses for the different segment types revealed that segment variability impaired perception of all segment types approximately equally, but that the 50-ms vowel segments were selectively spared in the speaker variability condition. Implications for competing models of speaker normalization are discussed, and consideration of stimulus uncertainty effects in speech perception research recommended. [Work supported by NIH, DC00464.]