Research

Theme

This project has three components.

[A] The commonality of across-channel processing in time, space, and speech perception
Psychoacoustical experiments are conducted to explore common mechanisms that underlie across-channel processing in time, space and speech perception. We measures individual thresholds of between-frequency and between-ear gap detection, together with the perceptual boundaries of voice onset time (VOT) between voiced and voiceless consonants, e.g. /b/ and /p/, /d/ and /t/, and examine the relationship between the two. We also investigate the effects of perceptual learning. Can training on between-frequency gap detection improve the performance of between-ear gap detection as well as the identification of voiced and voiceless consonants? If there are indeed common mechanisms underlying these processes, we expect that perceptual learning will enhance the performance of the two latter tasks.
[B] Brain mechanisms of across-channel processing
We conduct ABR and MEG measurements to monitor brain activities at different stages of auditory processing during exposure to between-frequency/between-ear silent gaps and voiced/voiceless consonants to identify neurophysiological correlates of across-channel processing. Our previous study (Nakano et al., 2008) suggests the inferior colliculus as a possible site for human across-channel processing although further studies are required. MEG can provide a mapping of surface cortical activities in response to subtle differences between voiced and voiceless consonants which occur on the order of 10 msec. ABR provides responses from the brainstem and are analyzed by the method proposed by Willy Wong (Dajani et al., 2005) which allows the reconstruction of speech from the ABR signal. This approach provides one way by which the sounds “heard” by the inferior colliculus can be used to explore the role of IC in across-channel processing.
[C] Across-channel processing in perception of germinates and vision
The notion of across-channel processing is extended to the perception of geminates and to vision. Geminates, known also as “sokuon” in Japanese, are gaps or silent periods of 100 msec or more. They are much longer in length than VOT’s and longer than the gap threshold found for between-frequency gap detection. Geminates are bounded by phonemes of differing frequency components, drawing a parallel with between-frequency gap detection. Although the perception of geminates is often studied in terms of language comprehension and acquisition (Sadakata & Sekiyama, 2011), in this project we investigate it in context of across-channel processing. For visual across-channel processing, we employ gap detection between two stimuli of different channels in spatial frequency and spectral domains.
Page Top