The term coarticulation designates the fusion of small-scale events, such as single sounds and single sound-producing actions, into larger units of combined sound and body motion, resulting in qualitative new features at what we call the chunk timescale in music, typically in the 0.5.–5 s duration range. Coarticulation has been extensively studied in linguistics and to a certain extent in other domains of human body motion as well as in robotics, but so far not so much in music, so the main aim of this paper is to provide a background for how we can explore coarticulation in both the production and perception of music. The contention is that coarticulation in music should be understood as based on a number of physical, biomechanical and cognitive constraints, and that coarticulation is an essential factor in the shaping of several perceptually salient features of music.