Measurement of the Musical Interval Sense
Andrew G. Pikler, J. Donald Harris · The Journal of the Acoustical Society of America · 1961
Ten generally unmusical Ss were used in an effort to dissect and quantify the hierarchy of cues which enable the trained instrumentalist or singer to create a progression of musical intervals with great precision, speed, and consistency. Experiment A required Ss to create one at a time, by adjustments, the eight intervals of the major scale (C fixed at 262 cps). Errors were commonly half a semitone. The Octave was almost always significantly and seriously sharp. Evidently the intervals in isolation sustain an average error of up to a quartertone. Experiment B required Ss to set seven oscillators arranged in a rudimentary electric piano to create the scale. The intervals showed improvement in placement and variance, though deviations were still commonly a fourth of a semitone. Evidently the context of even a very primitive musical line is a distinct additional cue in pitch intonation. Experiment C was as B except that the tonic and octave were both fixed. With this additional anchor, placement of all tones was improved; for example, the fifth improved from 48 cents sharp to 21.5 cents sharp. Evidently errors increase with distance from an anchor. The further cue of harmony, not studied here, would suffice to bring intonation errors within limits of the DL.