Charles H. Sears, “A Contribution to the Psychology of Rhythm,” The American Journal of Psychology, Vol. 13, No. 1 (Jan., 1902), pp. 28-61. Stable URL : http://www.jstor.org/stable/1412203
The paper here presented is a report of an experimental study of the time values given by competent players to the notes of several simple selections of music. When the musician begins his studies certain statements are made to him with regard to the relative values of notes. He is told that a half note should be given half the time of a whole note, a quarter note half the time of a half note, an eighth note half as much time as quarter, and so on, and that a dot placed after a note adds one half to its length. It is implied that all notes of the same kind should receive equal amounts of time unless a change of tempo is indicated, that a triplet should divide into three equal parts the time usually given to two like notes, that, except for purposes of expression, all measures are of the same length, etc. Toward fulfilling these requirements he strives with the metronome as an assistant. How far the trained musician accomplishes what the notes set before him indicate and what he sets out to do is an interesting question not only to the psychologist, but also to the musician. To answer this question for a few simple cases by direct measurement of the lengths of the notes played has been the aim of this study. The problem was suggested by Meumann, but so far as I know, no investigator has offered a contribution of facts gathered directly from the records of performers, although some work has been done in closely related fields as will presently be shown. [...]