Abstract
Using short and long contexts, the present study investigated musical priming effects that are based on chord repetition and harmonic relatedness. A musical target (a chord) was preceded by either an identical prime or a different but harmonically related prime. In contrast to words, pictures, and environmental sounds, chord processing was not facilitated by repetition. Experiments 1 and 2 using single-chord primes showed either no significant difference between chord repetition and harmonic relatedness or facilitated processing for harmonically related targets. Experiment 3 using longer prime contexts showed that musical priming depended more on the musical function of the target in the preceding context than on target repetition. The effect of musical function was decreased, but not qualitatively changed, by chord repetition. The outcome of this study challenges predictions of sensory approaches and supports a cognitive approach of musical priming.
Acknowledgements
This research was supported by the CNRS programme “Cognition and Information Processing” and the International Foundation for Music Research.
Notes
However, it should be emphasized that forward masking usually involves long maskers (200 ms) and short signals (less than 50 ms). Experiments 1 and 2 displayed the opposite relation with short primes (i.e., 50 ms to 750 ms) and long targets (i.e., 1,330 and 2,000 ms).
An additional contrast decomposing the interaction between target repetition and musical function confirmed that response times in the repeated condition were significantly shorter for tonic targets than for nontonic targets, F(1, 39) = 13.22, p < .001, MSE = 10,311.90.
This explanation is likely to be modulated by the length of the musical piece, notably when we consider that it is very usual for symphonic masterworks to end with numerous repetitions of the tonic (as in Beethoven's symphonies, for example).
Simulations were run with an implementation of Bharucha's model on Matlab. Given that the ISI between chords was set to 0 and that all chords were played with the same duration, the time transpired since the last offset (t) was identical for each chord and was set to 1. The rate at which activation decays (d) was set to.02. represents the activation level of target chord units averaged over the sequence set. When the decay was set to.04 (i.e., representing a stronger decay over time), the activation pattern of target units was similar to , except that the difference between tonic and nontonic targets was further decreased for the repetition condition only.