In Brief

When a pianist hits a sour note, even the musically untrained can tell. How do they determine whether the note doesn't fit? New research on tonality, or the key of a piece of music, by psychologists at the University of Toronto at Scarborough may help unravel the mystery.

According to a study in the Journal of Experimental Psychology: Human Perception and Performance (Vol. 30, No. 2), listeners perceive a random sequence of tones as in key if certain notes are held longer than others. For instance, in the key of C, the C note must be held the longest. A key's next most important notes are called the major triad; in the C key's case, they would be E and G.

For listeners to discern a key of C, the major triad must be held for a shorter time than the C, but longer than the third most important set, known as the diatonic notes. Finally, the least important nondiatonic pitches must be held the shortest for listeners to correctly perceive a key. Researchers use the term "hierarchical structure" to define how well a piece adheres to this formula.

Nicholas Smith, PhD, now a postdoctoral fellow at Ontario's McMaster University, tested this hypothesis by having 100 undergraduate students familiar with Western music listen to randomly sequenced melodies played on a computer. After listening to the music, the participants heard a "probe tone" and judged how well it fit with the rest of the piece on a scale from 1 to 7.

When listening to a piece high in hierarchical structure, the students easily ranked probe tones--that is to say, they heard the key of the piece and had no trouble picking out the sour notes. However, when the duration of the melody's notes did not adhere to the tonal hierarchy, students tended to rate all the probe tones as equally appropriate. In other words, they did not hear any notes as sour.

Additionally, as a piece's hierarchical structure increased--as differences between important and unimportant notes became more exaggerated--students better discerned the melody's key.

According to Smith, these findings suggest that people listen for a piece of music's "goodness of fit" to a certain tonal profile. This tonal profile, he says, may be learned early in childhood, as is the underlying structure of language.

But, he adds, the specifics of musical structure will vary from culture to culture, as the specifics of language do. If this experiment were replicated with people familiar with, for example, Indian music, the participants might listen for a slightly different pattern.

"If we didn't have this template, notes would have less meaning," says Smith. "Part of the aesthetics of music is that some things sound stable, and some sound unstable."