Bidirectional influences of pitch and time in auditory perception

Abstract

Auditory rhythms play a central role in human culture and communication, through both speech and music. The ability to track and predict the organization of events in time helps humans optimize attention, perceive emotion, coordinate actions, and understand social affiliations. The importance of these functions has inspired substantial efforts to model rhythm perception. However, despite a wealth of evidence that pitch influences rhythm perception, with higher speech and music perceived as faster, leading theories and models of rhythm perception have yet to incorporate these effects of pitch. This thesis addresses several empirical questions that have stood in the way of integrating pitch into these models. Specifically, 1) whether the perception of higher pitches as faster generalizes across more than two octaves and above 1000 Hz, 2) whether pitch influences synchronized motor tempo, and 3) whether pitch–timing interactions are bidirectional, such that tempo changes also influence perceived pitch. To answer these questions, we present data from ten experiments including subjective tempo ratings, sensorimotor timing, temporal discrimination, and pitch discrimination tasks. Our results suggest the existence of two separate effects of pitch on perceived timing. First, we present evidence in Chapters 2 and 3 for a unidirectional, negative quadratic effect of absolute pitch on perceived tempo. In this effect, both subjective and sensorimotor tempo rise with pitch between 110 and 440 Hz, peak somewhere between 440 and 1760 Hz, and decrease with pitch above that peak. In Chapters 4 and 5, we present evidence for a bidirectional and approximately linear bias to perceive higher pitches as faster and earlier sounds as higher. We propose that the former effect is most likely innate and a product of the structure of the auditory system, whereas the latter is learned from world structure and originates from cue integration at a later stage of processing.