Oscillatory Activity during the Processing of Musically Induced Emotions

The purpose of the present experiment was to investigate neuroelectric oscillatory activity during the processing of emotions. Drawing on studies showing that (a) music is capable of inducing strong emotions in listeners (Krumhansl, 1997) and (b) dissonance is associated with unpleasant experiences while consonance is perceived as pleasant (Blood et al., 1999), a newly developed experimental paradigm (Koelsch et al., 2004) was applied. Listeners were presented with excerpts (1min length each) of joyful (consonant) dance tunes of the past 4 centuries and electronically manipulated unpleasant (dissonant) counterparts. Participants were instructed to listen with closed eyes, tap the metre and rate how (un)pleasant they felt after each musical piece. Additionally, the experiment comprised 6 silence periods as baseline condition and the measurement of an ECG during the whole experimental session. An FFT (fast Fourier transform) was performed on EEG data and frequency bands were adjusted individually according to the methods of Doppelmayr et al. (1999). Instantaneous heart rate was calculated. Behavioral data and the typical triphasic pattern of heart rate (Bradley et al., 1999) indicate that the applied experimental paradigm induced pleasant and unpleasant emotions in listeners. Pleasant musical pieces evoked an increase of midfrontal theta power towards the end of the excerpts, whereas theta power remained constant during unpleasant pieces (Fig. 1). The comparison of both music conditions (consonant/dissonant) with rest revealed correlates of (a) attention-related processes (namely an overall power increase in the delta range during the listening to music) and (b) motor activity related to the tapping with the right index finger (left central power decrease in the alpha and beta range during tapping). In conclusion, our results show that the applied experimental paradigm using consonant/dissonant music and the analysis of neuroelectic oscillatory activity are suitable methods in the investigation of emotions. The findings suggest that power changes in midfrontal theta rhythm might relate to general emotional processes such as emotional self-regulation or decision-making.

Fig. 1 Theta power during the first and the second 22 s of pleasant and unpleasant musical excerpts (compared to rest).