PAGE 10 OF 15
In analysing the data from her experiments with the tritone phenomenon, Deutsch found that the judgements of any particular listener were consistent. Each time a given pitch class was heard, the listener invariably heard it the same way. For instance, a listener might, whenever she hears C#, tend to always hear it as higher, whether it is the first in a pair or the second. If it was the first tone in a sequence, then she would hear the sequence as going down; if it were the second tone in a sequence, the sequence would always be heard as going up. Deutsch understood this in terms of the tonal circle we just now considered. Tones that were always heard as higher appear at the top of the circle while those that are always heard as lower appear at the bottom of the circle.
In the following chart, the results for one subject are shown. The x-axis indicates the percentage of times the listener perceived the sequence as descending. As is shown, this listener consistently heard the pairs C#-G, D-G#, and D#-A as ascending. Meanwhile, she consistently heard F#-C, G-C#, A-D#, A#-E, and B-F as ascending. The other pairs were, for this listener, more ambiguous.