Effect of Instruction in Appropriate Rubato Usage on the Onset Timings of Musicians in Performances of Bach

The purpose of this study was to examine the effect of instruction in the use of specific rhythmic nuances on the timings of a musical performance. Volunteer subjects were asked to perform Bach's Suite no. 3 for Violincello solo, Bouree no. 1, using a computer software program that controlled pitches, amplitude, and all other musical variables. The only aspect of the performance that the operator controlled was the rhythmic onset of each pitch. When the subjects had generated a performance with which they were pleased, they recorded it. Subjects in the experimental group were then given a model of rhythmic variations that represented the aggregate of the rubato variations used by 15 professional cellists in performances of this work. This model was presented visually in forms, including a version of the score that had been marked up to indicate incidences of rubato, graphs of these incidences, and a prose synopsis. Subjects were then given time to practice until they believed that they were prepared to make a second recording that incorporated these rhythmic variations. Subjects in the control group repeated the pretest procedures. All note-onset data were analyzed to determine the extent to which subjects imitated the model. Results indicated that both groups' pretests, as well as the control group's posttest, had almost no correlation with the model; while the results of the experimental group's posttest significantly correlated with the rhythmic model.