A Model of Performing Musicians' Melodic Perception through Transcription of Two-Voice Contrapuntal Music

Abstract

This article reports findings from two experiments examining strategies utilized by professional musicians who transcribed a Mozart duet, wherein we explored the possibility that their cognitive processes would be revealed through their transcription strategies on multiple sequential presentations of the same stimulus. We hypothesized that transcription accuracy would be influenced by four factors: Scale degrees, conjunct/disjunct melodic motion, vertical consonance/dissonance, and melody position. In Experiment 1, we reduced the foreground of Mozart's 1783 B-flat Major Duo for Violin and Viola, and used three taped repetitions of the reduction as the experimental stimulus. Participants transcribed the music as the tape played, adding to their transcription during an additional three minutes following each presentation on forms permitting them to see their previous transcriptions, but precluding change of their previous work. Experiment 2 inverted the melodic lines of the stimulus. Our results examined pitch accuracy as the unit of analysis, on each note of the multiple stimuli presentations. Data analysis confirmed the influence of scale degrees, melodic motion and melodic relationship on perceptual accuracy, but showed no influence of vertical consonance/dissonance.