Interactive effects in memory for harmonic intervals.

Subjects compared the pitches of two temporally separated tones, which were both accompanied by tones of lower pitch. The standard ~S) and comparison ~C} tones were either identical in pitch or they differed by a semitone. However, the tone accompanying the S tone was always identical in pitch to the tone accompanying the C tone. Thus, when the S and C tones were identical, the intervals formed by the S and C combinations were also identical. When the S and C tones differed, the intervals formed by the S and C combinations also differed. The S and C tones were separated by a retention interval during which six extra tones were interpolated. The tones in the second and fourth serial positions of the interpolated sequence were also accompanied by tones of lower pitch. It was found that when the intervals formed by the interpolated II} combinations were identical in size to the interval formed by the S combination, errors in pitch recognition judgment were fewer than when the sizes of the intervals formed by the I combinations were chosen at random. When the intervals formed by the I combinations differed in size by a semitone from the interval formed by the S combination, errors in pitch recognition were more numerous than when the sizes of the intervals formed by the I combinations were chosen at random.