Listening for Phi: Temporal Proportion in Recordings of Mozart's Piano Sonatas

摘要

The golden mean (the ratio of approximately 1:1.618) can be observed in the spatial proportions of much classical architecture and painting. There is some evidence to suggest that ratios of this kind have also been adopted to define the large-scale temporal proportions in musical compositions by composers as diverse as Dunstable and Bartok. Jane Perry-Camp [1] and John F. Putz [2] have suggested that Mozart used the golden mean in the sonata form movements of his 18 piano sonatas to define the respective lengths of the two repeated sections. These studies have shown that the number of bars in each of the two sections of every sonata form movement, often have a proportional relationship close to that of the golden mean. Yet the present paper will argue that counting the bars in a score is rather different from experiencing a piece of music as a sound in time. Although the temporal proportions of many of the movements of Mozart's piano sonatas do seem to fall within the ratio of the golden mean, it is questionable whether most listeners consciously appreciate these proportions. The present study analyses several complete sets of commercial audio recordings of Mozart's piano sonatas by a range of performers in order to explore whether the temporal proportions in the sonata form movements are preserved in recorded performance. The findings of this study are often at odds with those resulting from analysis of the scores.