Tension experience induced by tonal and melodic shift at music phrase boundaries

Abstract

Music tension is a link between music structures and emotions. As music unfolds, developmental patterns induce various emotional experiences, but the relationship between developmental patterns and tension experience remains unclear. The present study compared two developmental patterns of two successive phrases (tonal shift and melodic shift) with repetition condition to investigate the relationship with tension experience. Professional musicians rated on-line felt tension and EEG responses were recorded while listening to music sequences. Behavioral results showed that tension ratings under tonal and melodic shift conditions were higher than those under repetition conditions. ERP results showed larger potentials at early P300 and late positive component (LPC) time windows under tonal shift condition, and early right anterior negativity (ERAN) and LPC under melodic shift condition. ERSP results showed early beta and late gamma power increased under tonal shift condition, theta power decreased and alpha power increased under melodic shift condition. Our findings suggest that developmental patterns play a vital role in tension experiences; tonal shift affects tension by tonal shift detection and integration, while melodic shift affects tension by attentional processing and working memory integration. From the perspective of Event Structure Processing Model, solid evidence was given to specify the time-span segmentation and reduction.

Figure 1

(a) Examples of scores (right) and spectrums (left) of the music sequences. (b) Mean tension curves under three conditions. (c) Grand-average ERP waveforms at FC1 for three different developmental patterns. The map shows the distribution of the effect (MS minus RP and TS minus RP) in the time windows of 140–240 ms, 280–380 ms and 500–780 ms, respectively (top) and the topographies at corresponding time windows (bottom). (d) (1), Difference of ERSP of TS minus RP condition at 4–30 Hz (top) and 30–80 Hz (bottom) bands at F3 electrode, with topographies in the corresponding band frequencies and time windows (right); (2), Difference of ERSP of MS minus RP condition at 4–30 Hz (top) and 30–80 Hz (bottom) bands at C3 electrode, with topographies in the corresponding band frequencies and time windows (right). Black squares represent where difference was significant. (Note that no effect was found in 30–40 Hz during the time window of 760–950 ms, therefore no black square was labeled).