Sequential information in music as well as language does not form uniform streams, but can be subdivided into phrases of several levels. Prosodic phrase boundary markers (e.g. pauses) support a speedy and accurate processing. For auditory speech, a recent study gives direct electrophysiological evidence for the immediate use of prosodic markers for the resolution of syntactic ambiguities (Steinhauer et al. 1999). Each phrase boundary was indicated by a positive shift with a centro-parietal distribution, called Closure Positive Shift (CPS). The question arises whether a similar component can be found indicating the detection of phrase boundaries in music. We present evidence that this is indeed the case. In two experiments, musicians listened to unknown piano sequences consisting of 1, 2 or 3 phrases. ERP was measured at 64 or 32 electrodes and revealed a significant positive component in response to the phrase boundary with a peak latency of 550 ms after phrase boundary offset, an amplitude of about 2 mV, and a centro-parietal distribution. Moreover, parallel measurements with a whole head MEG system were used to gain insight into the underlying generator configurations. Further exploration showed that the presence of the component is influenced by various parameters concerning the structure of the music (phrase lengths, symmetry) and the interpretation (emphasis of phrasing, reflected by pre-final slowing and lengths of pauses). This finding indicates that the CPS might reflect a universal segmentation process which applies to musical perception as well.
Reference Steinhauer K, Alter K, Friederici AD. Nature Neuroscience 2(2), 191-196 (1999)