UC Irvine

Music is capable of eliciting a wide variety of sensations and emotions. Current models of tonality perception assume that musical elements (tones, chords, phrases) activate multiple mechanisms in the brain that give rise to these sensations and emotions. Previous research has primarily used subjective judgments from trained musicians to derive the fundamental dimensions of tonality. Importantly, this research has used feedback-free experimental paradigms in which the listener must be trusted as the expert witness to ground truth. The current research, by contrast, uses feedback-driven tasks to measure the sensitivity of listeners to variations in the physical properties of musical stimuli. In this dissertation, I investigate several notions of dimensionality in tonality specifically looking at sensitivity in both experts and novices. We focus on the sensitivity of listeners to variations in the global statistics of brief, rapid, randomly-ordered sequences of musical tones called tone-scrambles. Previous research (Chubb et al., 2013) has shown that (1) most listeners (roughly 70%) cannot hear any difference between tone-scrambles that differ only in degree three of the major vs. minor diatonic scale, and (2) the other listeners are highly sensitive to this difference. In chapter 1, I show that degree six of the diatonic scale splits the population in the same way as scale degree three. In chapter two, a large sample of listeners is tested in each of the “3” and “6” tasks previously investigated along with three other similar tasks, in each of which the tones to be discriminated differ by a semitone. Performance in all five tasks is strongly correlated showing that a single mechanism predominates in enabling judgments of semitone differences. In chapter three, I take a much closer look at two listeners, who are high-performers in the semitone tasks used in Chapter 2, to investigate the number of dimensions of sensitivity that enable tonal discrimination. Multidimensional scaling applied to the results yields 3 dimensions for each of two listeners. In chapter four, I take a much closer look at the interaction between judgments of major and minor and the strength with which the tonic is established in a sequence. We find that tonic strength interacts with the strength of major and minor cues in the tone-scramble, influencing the judgments of listeners beyond what the major and minor tones on their own would elicit.