Octave equivalence and the incidental learning of absolute pitch.

Absolute pitch is the ability to label the pitch chroma of isolated tones. Recent work has demonstrated rapid improvements in pitch identification using an incidental contingency learning task. In the present work, we used a similar approach to assess whether knowledge about the pitch chroma (rather than pitch heights) of tones is possible in our task. In Experiment 1 with nonmusicians, seven notes in two octaves were played most often with their corresponding written note names and less often with the others. Tones from a third octave were played equally often with all note names (i.e., no contingency manipulation) to test for chroma-based generalization. We found significant but very small learning and generalization effects, revealing the relative difficulty of focusing on the pitch chroma during learning. The same task was used again in a sample of musicians in Experiment 2 with the thinking that musicians would have enhanced pitch chroma perception. As expected, we found more compelling evidence for learning and generalization, albeit only in explicit identification performance. In Experiment 3, we assessed whether it is fundamentally impossible to acquire knowledge based on pitch chroma by training participants with Shepard tones, which are largely devoid of pitch height cues. Interestingly, we found robust learning effects for these tones. However, we did not find evidence for transfer toward piano tones. We conclude that the major reason why absolute pitch is difficult to learn in adulthood is due to difficulty in perceiving pitch chroma due to the saliency of pitch height cues.