In music development, smaller elements combine into larger structures, resulting in a hierarchy of linguistic units, such as motifs, sections, phrases and periods. Hierarchical structure is a key feature of musicand plays an important role in music theory.However, itremains puzzling how hierarchical structure influence music perception.The current three experiments were designed to explore how hierarchical structures influence musical syntacticintegration.
In experiment 1, we explore the perception of hierarchical boundariesin music, and test whether musical expertise modifies such processing. The ERPresults showed that CPS wasevoked at all boundaries in musicians, and the amplitude of the component increased as the hierarchical level became higher, suggesting that musicians could represent music events at different timescales in a hierarchical way. For non-musicians,the CPS was only elicited at the period boundary and undistinguishable negativities were induced at all the three boundaries. The results indicate that a different and less clear way was used by non-musicians in boundary perception. The comparison between musicians and non-musicians showed that the phrasing ability could be enhanced by years of musical training.
In experiment 2, chord sequences were presented to musically trained listeners, with the last critical chord either harmonic related or less related to its preceding context at phrase and periodlevel. ERPs data revealed an ERAN-like effect for phrase context in an early time window and a later interaction between the phrase and period contexts in a N5-like component.These results suggest that both the higher-level and lower-level contexts influence the processing of an incoming music event, and the lower-level process happens earlier than thehigher-level. Moreover, the interaction between the phrase context and period context in N5 may suggest that music syntactic integration at lower-level takes priority over integration at higher-level.
In experiment 3, chord sequences were also presented to musicians. Syntacticand temporal structure of chordsequences were manipulated, makingthe target chord either harmonic related or less related to its preceding phrase context, and also congruous or incongruous withits temporal structure. In this way, we aimed to examine weather and how musical temporal structure influence syntactic integration.ERPdata revealedan early right negativity(ERAN) and a later N5 for syntactic effect, and an early N150 and late negativities for temporal effect. Although there was no interaction between syntactic relatedness and temporal congruency, simple effect analysis revealed that less related harmonic context evoked larger N5 only when the temporal structure is congruous. The time-frequency analysis results further showed that there wan an interaction between temporal structure and syntactic integration in late stage. The less related harmonic context inducedlarger power in gamma frequency band(30-100 Hz)only when the temporal structure was congruous.The results indicate that when listening to music, listeners track both the pitch and temporal pattens. The two dimensions were independent in early stage,and in late stage(about 300-700 ms) syntactic integration was influenced by temporal structure.
The thesis confirmed the effectof musical hierarchical structures on music perception. Listeners not only perceived hierarchical structures, but also continuouslyintegrate current information into different structures.Moreover, syntactic integration is influenced by temporal dimension of musical structures.