The long-term objective is to investigate to what extent """"""""special"""""""" versus """"""""general"""""""" auditory mechanisms contribute to human speech perception. """"""""Special"""""""" mechanisms are those that are developed by human over the course of language learning and are presumed not be represented in the general primate auditory system. """"""""General"""""""" mechanisms are those that are shared by other primates and are presumed not be language- specific. The present experiments will examine how monkeys identify natural tokens of human consonant contrasts embedded in consonant-vowel syllables that vary as to talker, pitch, and vowel consonant. These experiments will determine when (if at all) the monkey """"""""breaks down"""""""", relative to the humans, in performing complex speech identifications. The first contrast to be tested will be the English voicing contrast /ba-p/h/a/, which has already been shown to be highly salient to animals, human infants, and non-native English listeners, and which will give us a baseline performance for the monkey relative to the human. Several other contrasts will then be tested that are known to be poorly differentiated by non-native human listeners: the Spanish prevoicing contrast /m/ba-pa/and the Hindi dental retroflex contrast/da-Da/, both difficult for English listeners, and the English liquid contrast/ra-la/, difficult for Japanese listeners. The monkey should provide an objective indication for the psychoacoustic salience of each contrast apart from linguistic experience. Comparisons of monkey performance with that of both native and non-native listeners will have implications for models of speech perception (Universal Model, Attunement Model or Perceptual Learning Model) that invoke facilitation, maintenance, loss or induction as processes by which human speech perception develops in infancy and in second-language learning.
Showing the most recent 10 out of 14 publications
