Proceedings abstract


Prof Richard Parncutt


Terhardt's theory and model of musical pitch perception, as further developed and applied by Parncutt, predicts the pitch-salience profiles of musical chords (i.e., the relative perceptual salience of 12 chroma following an isolated presentation of a chord of octave complex tones). Such profiles apply firstly to temporally isolated sounds. Research on streaming and auditory scene analysis by van Noorden and by Bregman, and on memory for melodic pitches by Deutsch, suggests that temporal context has a significant impact on such profiles, and raises doubts about the music-theoretical applicability of the approach.


We tested the effect of simple temporal contexts on the tone profiles of musical chords.


In each trial, music students heard a chord of octave complex tones followed by single tone, and rated how well the tone followed the chord. Different combinations of chord and probe tone were presented in random order and in random transposition. In a second experiment, the chord was either preceded or followed by a distractor tone, which listeners were instructed to ignore. When the distractor followed the chord, the distractor also followed probe tone.


We hypothesized that distractor tones would stream with nearby chord tones, attracting attention to them and increasing their perceptual salience. This was not confirmed. Instead, peaks were generally observed at both chord and distractor pitches. More importantly, when the chord and the distractor tone together created a more familiar tonal fragment or progression, the tone profile was more clearly structured.


Temporal context and streaming do not destroy pitches that are implied according to the pitch model, but not played. This seems to confirm the model's music-theoretic potential. However, responses were strongly influenced by listeners' familiarity with specific sound structures that occur frequently in western music and are not represented in the model.


 Back to index