Three experiments examined the perceptual validity and parsimony of 2 musical segmentation modelsthe perceived phrase structure (PPS) model (Temperley, 2001) and the generative structural grammar of music (GSGM) model (Lerdahl & Jackendoff, 1983) . Each of these models is composed of a set of rules that specifies where the perception of a natural break in musical phraseology is most likely to occur. In 3 experiments assessing the models from different perspectives, participants listened to

more »

... ies and then drew a vertical line in a row of equally spaced dots corresponding to where a natural break in the stimulus was heard, with each dot representing 1 note regardless of the duration of the note (cf. Deliège, 1987) . The PPS and GSGM models were similar in their overall predictive accuracy; however, the PPS model gap rule, in which large temporal intervals between notes yield the perception of a musical phrase boundary, explained nuances arising within the broader GSGM model. The PPS gap rule is therefore suggested as a strong factor in musical segmentation, adding to our understanding of the perception of musical phrases.