Abstract
Joint action has increasingly become a key topic to understand the emergence of the human mind. The phenomenon is closely linked to several theoretical concepts, such as shared intentionality, which are difficult to operationalize empirically. We therefore employ a paradigm-driven, bottom-up approach, and as such discuss co-representing the partner’s and one’s own actions as key mechanism for joint action. After embedding co-representation in the broader landscape of related theoretical concepts, we review neurobiological, ontogenetic, and phylogenetic studies, with a focus on whether co-representation and its flexible deployment should be construed as a low- or high-level cognitive process. The empirical findings convergently suggest that co-representation does not require strong inhibitory skills or mentalistic understanding and occurs automatically. Moreover, more cooperative species are better at flexibly suppressing co-representation when required for cooperation success, and frequently rely on cooperation markers, such as mutual gaze. We thus contribute to closing the current gap between theoretical concepts related to joint action research and their empirical investigation, and end by highlighting additional approaches for doing so.