Neural Substrates of Interactive Musical Improvisation: An fMRI Study of ‘Trading Fours’ in Jazz

Abstract

Interactive generative musical performance provides a suitable model for communication because, like natural linguistic discourse, it involves an exchange of ideas that is unpredictable, collaborative, and emergent. Here we show that interactive improvisation between two musicians is characterized by activation of perisylvian language areas linked to processing of syntactic elements in music, including inferior frontal gyrus and posterior superior temporal gyrus, and deactivation of angular gyrus and supramarginal gyrus, brain structures directly implicated in semantic processing of language. These findings support the hypothesis that musical discourse engages language areas of the brain specialized for processing of syntax but in a manner that is not contingent upon semantic processing. Therefore, we argue that neural regions for syntactic processing are not domain-specific for language but instead may be domain-general for communication.

The neuroscience of musical improvisation

“…The results of Donnay et al (2014) provide an interesting contrast to the earlier work of Limb and Braun (2008). In Limb and Braun (2008), musicians showed deactivation of the DLPFC, suggesting less attention control during improvisation.…”

“…This notion was further reinforced by activation of other regions involved in executive processes, such as the left DLPFC, right ACC, and bilateral IFG. Activation of prefrontal regions were also reported in studies contrasting improvisation and memory retrieval (e.g., Bengtsson et al, 2007;Donnay et al, 2014). Berkowitz and Ansari (2008) provide support for several of Pressing's (1988Pressing's ( , 1998 theoretical predictions.…”

“…Several notable differences in task design are apparent in the literature: while some studies contrasted improvisation with memory retrieval (Bengtsson et al, 2007;Limb and Braun, 2008), others contrasted specific modes of improvisation, such as melodic and rhythmic (Berkowitz andAnsari, 2008, 2010;de Manzano and Ullén, 2012a). Moreover, some researchers have been interested in analyzing the effects of musical expertise (e.g., Berkowitz and Ansari, 2008;Pinho et al, 2014) and collaboration (Donnay et al, 2014) on brain activity during improvisation. Such methodological differences are important to consider when comparing results across studies.…”

“…However, this study differed from Limb and Braun (2008) in terms of the memory retrieval condition employed: instead of requiring participants to recall a single, overlearned melodic sequence, Bengtsson et al asked participants to retrieve self-generated melodies that were previously improvised in the scanner. In addition, improvisation was constrained to modifying simple, eight-bar melodic sequences presented, whereas other studies permitted largely unconstrained improvisation (e.g., Donnay et al, 2014;Limb and Braun, 2008;Liu et al, 2012). Bengtsson and colleagues reported improvisation-related activity in several prefrontal regions, including the DLPFC, right pre-SMA, and bilateral dorsal premotor cortex (PMD).…”

“…The improviser must therefore allocate considerable attentional resources to both internal (e.g., generative) and external (e.g., communicative) sources. Recently, Donnay and colleagues explored the neural correlates of collaborative improvisation in a sample of professional jazz pianists (Donnay et al, 2014). The experimental procedure involved "trading fours" with an experimenter (also a jazz musician) in a nearby control room; participants and the experimenter took turns performing four-bar musical segments-a common mode of improvisation in jazz ensembles.…”