Are We Really Hearing in Our Heads What We Think We’re Hearing? The Role of Audiation in Musical Improvisation

An important and valued part of the skill of musical improvisation is to be able to play what we hear in our head (audiation). Improvisation is a cognitively demanding activity, involving the production of musical material in real time. This requires the simultaneous involvement and coordination of many different skills, and places demands on working memory, memory retrieval, auditory and sensory-motor systems. Some recent studies support a cognitive model of improvisation which posits the deployment of stored rhythmic and melodic patterns via motor programmes. According to the theory of event coding, actions and their perceptual consequences share the same cognitive representation and behavioural and fMRI studies have offered evidence supporting this theory. Since musical actions have sounds as perceptual consequences and sensorimotor coupling is bidirectional, this is compatible with improvisers imagining the sounds as they play them. However, phenomenological accounts and interview studies suggest musicians use different strategies to generate ideas in improvisation, such as musictheoretic ideas and motor patterns or ‘muscle memory’. So questions remain regarding the precise role of audiation in improvisation: what is musicians’ experience of musical imagery as they improvise? Is auditory imagery cognitively prior to action or post hoc? How accurate is auditory imagery? What proportion of musical output involves audiation and how sensitive is this to context? The aim of this paper is to offer a coherent explanatory framework for improvisation from the perspective of cognitive psychology and to propose experimental paradigms to begin to answer some of these questions. On the basis of a review of the literature, it is concluded that two approaches offer a way forward: altered auditory feedback (AAF) and a blocking paradigm in which interference conditions seek to disrupt the tonal loop in working memory.