Melodic Perception and Cognition: the Influence of Tonality.
William Eastman Lake · Deep Blue (University of Michigan) · 1987
The central purpose of this study was to examine the role played by an underst and ing of the tonal system shared by a community of musicians in shaping the implications for melodic continuation of different pairs of tones. Ancillary purposes were: (1) to examine the data-collection process to identify other factors that may have contributed to implications for melodic continuation, (2) to investigate the problem-solving strategies of participants, and (3) to make some comparisons of the results of this study to extant musical theory. The sample of twenty-two undergraduate music majors participated in the study. These musicians sang melodic continuations for various tone-pairs, each of which was placed in a specific location in a major key by a tonic-establishing context. They were instructed to use their intuitive underst and ing of what is normative in tonal music to guide their melodic continuations. The first tones of these continuations were then analyzed to discover how they were constrained by features of the stimuli, such as scale-degree function, implied harmony, and interval. The major findings of the study were: (1) A normative tonal melodic continuation is strongly constrained by the tones immediately preceding it and the relationship of those tones to the tonic. (2) Proximity in pitch is an important norm of tonal melodic progression. (3) Nondiatonic harmony implied by melody tones has a strong influence on tonal melodic continuation in many cases. (4) Melodic continuations tend to form certain commonplace diatonic interval patterns (melodic gestalts) in a limited number of cases. (5) The tendency to move to a tone of relatively greater structural importance in the tonal hierarchy constrains tonal melodic continuations in many cases. (6) Circle-of-fifths distance has an influence on the normative direction of resolution of nondiatonic tones. Secondary findings included: (1) The nontransposing (NT) and transposing (T) versions of the data-collection test did not significantly effect participants' performance. (2) Certain differences among participants' responses suggest a possible dichotomy of musical cognitive style. (3) Some musical theories can be modified, or at least qualified, on the basis of the results of this study.