Hst.723 Neural Coding and Perception of Sound Theme 4: Pitch and Temporal Coding

Pitch is often referred to as a primary parameter in music, a basic concept upon which other musical categories, such as pitch intervals and harmony, can be built (Snyder, 2000). While ANSI defines pitch primarily on a uni-dimensional space, Shepard (1982) postulated a multi-dimensional spatial model, such that the Euclidean distances between the musical tones (in the Western Tonal Tradition) reflect their perceived relations. This complex percept is important for human not only to appreciate melody and harmony in music (Dowling, 1999), but it is also used to convey prosody in speech (Winter et al, 2001); to segregate multi speakers talking simultaneously (Darwin, 1997); and in tonal languages, such as Cantonese, to carry lexical meanings (Fok, 1974). However, our understanding of the coding mechanism of pitch is only at a basic level. There are two competing theories of how frequency, a physical quantity that pitch percept is dependent on, is coded. The first one is the “place” theory, based on the spectral analysis performed in the inner ear and the fact that different frequencies excite different places along the basilar membrane (Moore, 2003). The other is the temporal coding theory, in which the frequency is related to the time pattern of the neural impulses evoked by the stimulus. There are also different mechanisms of complex pitch perception that have been postulated in the last few decades, ranging from Schouten’s temporal theory, and pattern recognition theory proposed by investigators such as Goldstein and Terhardt, to variations on Licklider’s autocorrelation model, which incorporates both the place and timing information in their pitch extraction algorithms. In this theme report, a sample of the vast and diverse research in pitch perception is presented. Literature reporting elements that influence pitch judgment, such as spectral weighting and auditory grouping mechanisms, are reviewed. Neurophysiological studies are surveyed to ascertain whether there are anatomical correlates that support the aforementioned pitch extraction models. Finally, a brief sample of studies in pitch perception at the cortical level is summarized.