This thesis explores the way we perceive correspondences between music and body movement. This is done by analysing similarities between music and movement in terms of specific features in music and movement. The analyses have been performed on recordings where participants were asked to make a movement that they thought matched a short excerpt of musical sound, first in a collection of ‘sound-tracings’ where the participants made a movement on a digital tablet, and second, in a collection of ‘free dance movements’ where the participants were asked to move freely to music. The approach to analysis has been developed on the basis of ecological theory of perception and motor theories of perception that address the perception-movement link. Included in the theoretical framework is also research on multisensory perception. As a point of departure it is assumed that correspondences is perceived on the basis of similar changes in features in music and movement i.e. that correspondence emerges when music and movement co-evolve and change similarly within the same time window. Changes in features are thus analysed on what we call a chunk level in both music and movement.
The analyses of the ‘sound-tracings’ demonstrate how participants are sensitive to correspondences on the basis of changes in pitch, timbre, mode of production (e.g. sustained vs. iterative) and onset density. The analyses of the ‘free dance-movements’ material show that correspondences emerge when overall activation change similarly in music and movement. On more detailed level correspondences are observed in terms of changes in onset density, articulation, speed, as well as how music and movement is subtly performed as ‘pulling’/’stretching’ or ‘hurried’/’jerky’. These findings suggest that there is a certain consistency in the way we perceive music-movement correspondences. However, the analyses also clearly demonstrate that correspondences emerge flexibly, i.e. that the same musical excerpt may correspond to different variants of movement. It may therefore be difficult to analyse consistency within a traditional quantitative research paradigm and it is proposed that music-movement relations are best examined by detailed qualitative analysis.