Cross -language studies of tonal perception: Hemodynamic, electrophysiological and behavioral evidence
"Overlearned" tonal processing in tone language speakers generates experience-dependent effects that were observed in neuroimaging and behavioral experiments. In a cross-language study using functional magnetic resonance imaging (fMRI), prelexical tonal processing engages the activation of the left planum temporale (PT), contrary to the commonly accepted view that pitch processing is an exclusive province of the right hemisphere. The left PT is hypothesized to play a critical role as an interface for accessing long-term categorical representations. A previous study using frequency-following response (FFR) reported that experience-dependent modification of pitch processing may extend down to the auditory brainstem. Based on this finding, we carried out another experiment that incorporates the comparisons of both native vs. non-native language groups and speech vs. nonspeech stimulus sets. It fails to replicate the cross-language difference but shows a significant difference in the harmonic-to-noise ratios of FFR spectral components between speech and nonspeech stimuli. Various factors in the experimental design may contribute to the negative result for the experience-dependent effect. A cross-language psychophysical study reveals that Mandarin tones are categorically perceived by native listeners in both speech and nonspeech. Non-native listeners exhibit no categorical effect when judging Mandarin tones in speech stimuli, but a quasi-categorical effect in nonspeech stimuli. This experiment not only demonstrates the effect of language experience on perceptual organization but also leads to a multistore model for explaining the contribution of different memory resources to categorical perception. A broader neural framework based on these findings indicates that language organization obeys the universal Hebbian learning rules. The observed experience-dependent effects in these experiments are related to the functional differentiation and integration of a complex network.
Gandour, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our