Rapid and flexible creativity in musical improvisation: review and a model

Abstract

Creativity has been defined as the ability to produce output that is novel, useful, beneficial, and desired by an audience. But what is musical creativity, and relatedly, to what extent does creativity depend on domain-general or domain-specific neural and cognitive processes? To what extent can musical creativity be taught? To answer these questions from a reductionist scientific approach, we must attempt to isolate the creative process as it pertains to music. Recent work in the neuroscience of creativity has turned to musical improvisation as a window into real-time musical creative process in the brain. Here, I provide an overview of recent research in the neuroscience of musical improvisation, especially focusing on multimodal neuroimaging studies. This research informs a model of creativity as a combination of generative and reactive processes that coordinate their functions to give rise to perpetually novel and aesthetically rewarding improvised musical output.

Improvising at Rest: Differentiating Jazz and Classical Music Training with Resting State Functional Connectivity

“…This lack of tightly knit communities in improvisational musicians, coupled with more interactions between ECN and DMN, corresponds well with expected connectivity patterns of improvisational musicians: As improvisation requires constant retrieval of previously-stored musical knowledge, through which musical goals and semantic referents (e.g. musical themes) are filtered, in order to generate auditory-motor sequences that are novel and yet closely integrated with past knowledge (Pressing 1998, see also Loui, 2018).…”

“…and 3) Why do individuals differ in creative ability? In addressing the first of these questions within the study of musical improvisation, which is a real-time creative behavior (Loui, 2018), behavioral models suggest that musical improvisation requires a balance between automatic retrieval of candidate musical sequences from long term memory, and continuous appraisal of these sequences over the course of an improvisational performance (Pressing, 1988).…”

“…Given the alignment of behavioral and neural models of musical improvisation (Pressing, 1988;Loui, 2018), functional connectivity profiles associated with generalized creative behaviors (Beaty et al, 2015;2018), and task fMRI studies of musical improvisation (Limb & Braun 2008;Liu et al, 2012;Bengtsson et al, 2014;Pinho et al, 2014), we expect to see differences in ECN and DMN in improvising musicians. In the present study, we ask whether improvisationally-trained musicians show different functional connectivity patterns in the absence of task, as has been observed in populations of individuals who are highly creative as measured by the Torrance test of creative thinking (TTCT) (Beaty et al, 2018;Takeuchi et al, 2012).…”

Towards a standard model of musical improvisation

“…McPherson and Limb () provide a detailed overview of the challenges of studying improvisation, calling for a combination of methods to understand both creativity as a holistic entity and the quantification of its sub‐processes. Biasutti and Frezza () describe improvisation as an interconnected series of cognitive states (see Biasutti, ), whereas Loui () presents a tiered model of improvisation containing computational, algorithmic and implementation (neural) levels.…”

“…The components described by Pressing (), Beaty (), Biasutti and Frezza () and Loui () can be conceptualised as a collection of sub‐processes operating at variable timescales nested under the umbrella behaviour of improvisation. This review will describe the neural correlates of these sub‐processes and will review findings on the neural correlates of musical improvisation, music listening, domain‐general creativity and computational music analysis, presenting a conceptual model of music improvisation.…”