Chord Discrimination by Pigeons

PIGEONS WERE TRAINED IN A GO/NO-GO TASK TO discriminate a major triad from four other chord triads. These latter chords were constructed by altering the third or fifth of the triad by one semitone, creating minor, suspended fourth, augmented, or flat five chords. Experiment 1 used a C root to test these chords. Experiment 2 examined the discrimination of these chords using the D root. Pigeons could discriminate these harmonically complex triads, with manipulations of the fifth supporting better discrimination than manipulations of the third. The augmented chord was perceived as the most dissimilar from the major chord and the suspended fourth chord as the most similar. A combination of attention to sensory consonance and absolute properties of the chords is suggested to account for the results. Comparable human results suggested some overlap in their perception of harmonic elements, but a considerable difference in their flexibility to use this information across different contexts.

Sensory Consonance

THE AIM OF THIS SERIES OF experiments was to determine whether consonant and dissonant chords elicit similar or different electrophysiological effects out of a musical context and whether these effects are similar or different for musicians and nonmusicians. To this end, w e recorded t he e vent-related b rain potentials (ERPs) elicited by the different intervals of the chromatic scale that were classified into three categories: perfect consonances, imperfect consonances, and dissonances. Participants were to decide, on a six-point scale, whether the intervals evoked pleasant or unpleasant feelings. To test the hypothesis that the perception of dissonance results from the superposition of the partials of close frequencies (Helmholtz, 1877), two notes were either played together (harmonic intervals) or successively (melodic intervals). Since, in this latter case, the two notes are played at different points in time, the perception of roughness, if any, should be weaker than for harmonic intervals. In line with Helmholtz's hypothesis, results showed larger differences for harmonic than for melodic intervals, which were mainly found on the N1-P2 complex for musicians, on the N2 component for nonmusicians, and on a later negative component for both musicians and nonmusicians. However, these results also point to the influence of expertise and cultural factors, since different results were obtained when ERPs were averaged as a function of music theory and according to the participants' responses.

Preference for Sensory Consonance in 2- and 4-Month-Old Infants

The preferences of 2- and 4-month-old infants for consonant versus dissonant two-tone intervals was tested by using a looking-time preference procedure. Infants of both ages preferred to listen to consonant over dissonant intervals and found it difficult to recover interest after a sequence of dissonant trials. Thus, sensitivity to consonance and dissonance is found before knowledge of scale structure and may be based on the innate structure of the inner ear and the firing characteristics of the auditory nerve. It is likely that consonance perception provides a bootstrap into the task of learning the pitch structure of the musical system to which the infant is exposed.

Different Brain Mechanisms Mediate Sensitivity to Sensory Consonance and Harmonic Context: Evidence from Auditory Event-Related Brain Potentials

The goal of this study was to analyze the time-course of sensory (bottom-up) and cognitive (top-down) processes that govern musical harmonic expectancy. Eight-chord sequences were presented to 12 musicians and 12 nonmusicians. Expectations for the last chord were manipulated both at the sensory level (i.e., the last chord was sensory consonant or dissonant) and at the cognitive level (the harmonic function of the target was varied by manipulating the harmonic context built up by the first six chords of the sequence). Changes in the harmonic function of the target chord mainly modulate the amplitude of a positive component peaking around 300 msec (P3) after target onset, reflecting top-down influences on the perceptual stages of processing. In contrast, changes in the acoustic structure of the target chord (sensory consonance) mainly modulate the amplitude of a late positive component that develops between 300 and 800 msec after target onset. Most importantly, the effects of sensory consonance and harmonic context on the event-related brain potentials associated with the target chords were found to be independent, thus suggesting that two separate processors contribute to the building up of musical expectancy.