Magnitude Estimation of Melodic Dissimilarity
Abstract
MAGNITUDE ESTIMATION OF MELODIC DISSIMILARITY RUDOLF E. RADOCY Background The traditional compositional techniques of retrograde, inversion, retrograde inversion, augmentation and diminution are well known to the student of music theory and musical form. Such techniques create variety while preserving a certain amount of stylistic unity. More extreme alterations of a melody or melodic fragment result from making the tonal sequence monotonic or changing the relative durations of sounds and silences. The structural melodic and rhythmic changes resulting from applications of such techniques are readily apparent in notation, but the correspond- ing auditory perceptual changes may be less obvious. Structural changes do not necessarily result in perceptual changes (Radocy and Boyle, 1979, p. 111). Just how unlike each other are the apparent perceptual changes which result from applying compositional techniques? Can listeners judge perceived melodic dissimi- larity on the basis of an initial impression, without analysis or recognition of the com- positional technique? There are various ways for people to indicate and quantify per- ceived differences. One of the writer's favourite techniques is magnitude estimation. In magnitude estimation, the observer matches one sensory continuum, often numerosity, with another. Stevens and his students used the technique for measure- ment of