Musical coordination in a large group without plans nor leaders

Abstract

A widespread belief is that large groups engaged in joint actions that require a high level of flexibility are unable to coordinate without the introduction of additional resources such as shared plans or hierarchical organizations. Here, we put this belief to a test, by empirically investigating coordination within a large group of 16 musicians performing collective free improvisation—a genre in which improvisers aim at creating music that is as complex and unprecedented as possible without relying on shared plans or on an external conductor. We show that musicians freely improvising within a large ensemble can achieve significant levels of coordination, both at the level of their musical actions (i.e., their individual decisions to play or to stop playing) and at the level of their directional intentions (i.e., their intentions to change or to support the music produced by the group). Taken together, these results invite us to reconsider the range and scope of actions achievable by large groups, and to explore alternative organizational models that emphasize decentralized and unscripted forms of collective behavior.

Coordination dynamics of multi-agent interaction in a musical ensemble

“…Thus, RQA applied to this model system provides an indication of the possible expected distributions for what recurrence dynamics to expect in truly emergent coordination in multi-agent human interaction in the wild—perhaps in less orchestrated (i.e. more improvised) forms of musical interaction such as leaderless interaction in free jazz improvisation 41 , and even day-to-day social dynamics extending beyond musical interaction, such as walking in groups 42 or interacting in large crowds.…”

Brain-to-brain communication during musical improvisation: a performance case study

“…It has also been noted from recent musical improvisation studies that the familiarity between musicians predicts stronger coordination of intentions during the performance. 75 Increased neural synchrony between two participating individuals may indicate mutual, efficient, and effective social interaction 76 and can be modulated by the degree to which the participating individuals feel socially connected, the activity they are engaging in, and the interaction setting. 21,[76][77][78] An interpretation of the aforementioned temporal bispectrum changes is that during bispectrum decreases, participants were not communicating effectively, and the "closeness" between each other had an important role in this communication.…”

What it is like to improvise together? Investigating the phenomenology of joint action through improvised musical performance

“…The former would be more intuitive, because it seems easier for performers to produce similar behaviour when they all intend to maintain the current musical outcome of the ensemble, compared to situations when they all intend to change it, for it is less likely that they intend to change it in the same direction. As a matter of fact, we observed in a companion study dealing specifically with the interaction dynamics in the same ONCEIM performance, that at the level of the group, musical intentions tended to be more aligned on the maintain-side than on the change-side of the continuum (Goupil et al, 2020, see in particular Fig. 7E).…”

“…For each of these two dimensions, this metric reflects the extent to which each musician's values forward predict the values of the other musician. Over the past few years, Granger causality has become a standardised way to assess information flow between pairs of agents engaged in collective music making (e.g., see Aucouturier & Canonne, 2017;Chang et al, 2019;D'Ausilio et al, 2012;Goupil et al, 2020;Hilt et al, 2019). For each musician and each measure (sonic activities and musical intentions), Granger causality between their values and values of each of the other musicians was computed using the stats model toolbox in Python (Seabold & Perktold, 2010).…”