Hearing the Moment: Measures and Models of the Perceptual Centre

The perceptual centre (P-centre) is the hypothetical specific moment at which a brief event is perceived to occur. Several P-centre models are described in the literature and the first collective implementation and rigorous evaluation of these models using a common corpus is described in this thesis, thus addressing a significant open question: which model should one use? The results indicate that none of the models reliably handles all sound types. Possibly this is because the data for model development are too sparse, because inconsistent measurement methods have been used, or because the assumptions underlying the measurement methods are untested. To address this, measurement methods are reviewed and two of them, rhythm adjustment and tap asynchrony, are evaluated alongside a new method based on the phase correction response (PCR) in a synchronized tapping task. Rhythm adjustment and the PCR method yielded consistent P-centre estimates and showed no evidence of P-centre context dependence. Moreover, the PCR method appears most time efficient for generating accurate P-centre estimates. Additionally, the magnitude of the PCR is shown to vary systematically with the onset complexity of speech sounds, which presumably reflects the perceived clarity of a sound’s P-centre. The ideal outcome of any P-centre measurement technique is to detect the true moment of perceived event occurrence. To this end a novel P-centre measurement method, based on auditory evoked potentials, is explored as a possible objective alternative to the conventional approaches examined earlier. The results are encouraging and suggest that a neuroelectric correlate of the P-centre does exist, thus opening up a new avenue of P-centre research. Finally, an up to date and comprehensive review of the P-centre is included, integrating recent findings and reappraising previous research. The main open questions are identified, particularly those most relevant to P-centre modelling.