An evolutionary systems approach to language and music: linking predictive dynamics and multicomponent perspectives

Humans are unique among social animals in having both music and language. The evolution of these capacities can be studied through two main approaches, involving the identification and analysis either of features conserved across cultures, or traits conserved across species. Both are limited in their ability to contribute evidence towards long-standing debates, such as what features corresponded to which stages of ancestral communication systems, the order in which traits evolved, and whether and how music and language diverged. We propose adapting an evolutionary systems biology framework, originally intended to identify causal interactions in evolving systems and specify their possible evolutionary trajectories. This approach provides mechanistic explanations for phenotypic development and evolutionary transitions. In transferring it to evolutionary cognition, we propose as a target of investigation the domain of predictive dynamics, mediating between cognitive traits and spectrotemporal features. We reason that neurobiologically-implemented cognitive capacities (1) enable predictive dynamics that (2) shape acoustic features enabling prediction in social interactions, (3) are quantifiable via computational and dynamical systems modelling, and (4) inform evolutionary concepts to provide testable predictions. By proposing a way to measure and simulate these predictive dynamics, our framework offers a novel approach to analysing the evolvability of music and language.