Annotation for Palmer, Caroline

Mapping Musical Thought to Musical Performance

Annotation (by Stephen C. Simms):

The goal of Palmer's work is to gain better understanding of the relationship between a performer's musical intention and his musical realization. Toward this end, she has designed and written about two scientific experiments.
The first experiment involves the performance of an excerpt of music in a "musical" fashion and an "unmusical" (mechanical) fashion by several performers. The data were recorded on an IBM personal computer attached to a weighted 88-key MIDI keyboard. The pianists were instructed to play in a musical fashion and then in an unmusical fashion.
Palmer examined the performances in terms of three rules: chord asynchrony, rubato, and note overlap. Palmer defines chord asynchrony as "the difference between note onsets that are notated in the musical score as synchronous." Rubato is usually perceived as a tempo change but can also be thought of as the shortening or lengthing of notes in deviation from the score. Note overlap is the intersection of one notes decay with an adjacent one's attack. For example, a legato passage will feature overlap while a staccato passage will attempt to have none.
Palmer's work showed first that unmusical performances contained less of these three elements than musical performances. Furthermore, she found that the musical performances of experts featured more of these rules than the musical performances of students.
In the second experiment, performers played a piece and then notated their interpretation on an unedited musical score. In the article she shows how one of the performers marked phrase boundaries and circled the melodic line that was emphasized. It is this specificity on the performer's part that then permits the experimentor to examine a relationship between intention (the marks on the score) and performance (the MIDI data in the computer). From this study Palmer discovered that the notated melody is played earlier than the other elements appearing simultaneously. She also found that the line notated by the performer features the greatest amount of note overlap and that the notated phrase boundaries feature the largest amount of rubato.