Browsing by Subject "auditory adaptation"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Perceptual Categorization and Neural Representations of the Human Voice
    (2024-08) Gao, Zi
    The human voice is a highly relevant and social auditory stimulus. The questions of whether the voice forms a special category in auditory perception and, if so, how the brain’s responses to voice are different from those to non-voice sounds have been a central topic of voice perception research. This dissertation explores the distinction between voice and non-voice processing in different domains and via different modalities. In the first study, both musicians with absolute pitch ability and average listeners were shown to perform worse with voice than non-voice stimuli in tasks that required fine-tuned pitch processing. This voice-disadvantage effect stands in contrast to several previously reported voice advantages and suggests that differences in processing are unlikely to be due to an overall enhancement in the processing of voice as compared to non-voice. Instead, the direction and degree of voice-related enhancement may be feature- or task-dependent. While the existing literature on voice perception has focused on comparing the responses to predefined categories of voice and non-voice stimuli, it remains unknown whether the perceptual boundary between these categories is fixed or flexible. The second study investigated the existence and extent of perceptual shifts in voice/non-voice categorization following either voice or non-voice contexts. Contrastive adaptation effects were observed following repeated exposure to either voice or musical instrument sounds, such that ambiguous sounds, created by morphing vowels with instrumental tones, were more likely to be reported as a voice after exposure to musical instruments, and vice versa. The third study demonstrated that the contrastive adaptation effects generalized to the more ecologically valid contexts of spoken sentences and melodies, but not to contexts of non-speech human vocalizations or non-pitched percussion sounds. The fourth and final study of this thesis explored neural correlates of voice and non-voice categorization using electroencephalography (EEG) under passive (unattended) listening conditions. Average differences in the morphology of the EEG waveforms were observed between vowel and instrument sounds. Using an oddball paradigm, a mismatch-negativity (MMN) was observed for rare instrument sounds in the context of vowel sounds, but not vice versa, providing only weak evidence for pre-attentive categorization. Finally, although the mean EEG waveform in response to ambiguous morphed voice-instrument sounds was not substantially affected by whether the morphed sound was presented in the context of voice or instrument sounds, trial-by-trial categorization based on logistic regression models performed at above-chance levels, suggesting that the EEG responses to the target sounds were affected by the context. Overall, the studies in this thesis provide us with a new understanding of how voice and non-voice sounds are perceived and processed by the human auditory system.