ABSTRACT

Observation of others’ actions activates motor representations in sensorimotor cortex. Although action observation in the real-world often involves multiple agents displaying varying degrees of action involvement, most lab studies on action observation studied individual actions. We recorded EEG-mu suppression over sensorimotor cortex to investigate how the multi-agent nature of observed hand/arm actions is incorporated in sensorimotor action representations. Hereto we manipulated the extent of agent involvement in dyadic interactions presented in videos. In all clips two agents were present, of which agent-1 always performed the same action, while the involvement of agent-2 differed along three levels: (1) passive and uninvolved, (2) passively involved, (3) actively involved. Additionally, a no-action condition was presented. The occurrence of these four conditions was predictable thanks to cues at the start of each trial, which allowed to study possible mu anticipation effects. Dyadic interactions in which agent-2 was actively involved resulted in increased power suppression of the mu rhythm compared to dyadic interactions in which agent-2 was passively involved. The latter did not differ from actions in which agent-2 was present but not involved. No anticipation effects were found. The results suggest that the sensorimotor representation of a dyadic interaction takes into account the simultaneously performed bodily articulations of both agents, but no evidence was found for incorporation of their static articulated postures.

摘要

观察他人的行为会激活感觉运动皮层中的运动表征。尽管现实世界中的动作观察通常涉及多个代理人表现出不同程度的动作参与，但大多数关于动作观察的实验室研究研究的是个体动作。我们记录了对感觉运动皮层的 EEG-mu 抑制，以研究观察到的手/手臂动作的多代理性质如何被纳入感觉运动动作表示。至此，我们操纵了代理参与视频中呈现的二元交互的程度。在所有剪辑中，有两个代理人，其中代理人 1 总是执行相同的动作，而代理人 2 的参与在三个层面上有所不同：(1) 被动和不参与，(2) 被动参与，(3) 主动参与。此外，还提出了不采取行动的条件。由于每次试验开始时的提示，这四种情况的发生是可以预测的，这允许研究可能的 mu 预期效果。与 agent-2 被动参与的二元交互相比，agent-2 积极参与的二元交互导致 mu 节律的功率抑制增加。后者与 agent-2 在场但未参与的行为没有区别。没有发现预期效果。结果表明，二元交互的感觉运动表示考虑了两个代理人同时执行的身体关节，但没有发现结合他们的静态关节姿势的证据。与 agent-2 被动参与的二元交互相比，agent-2 积极参与的二元交互导致 mu 节律的功率抑制增加。后者与 agent-2 在场但未参与的行为没有区别。没有发现预期效果。结果表明，二元交互的感觉运动表示考虑了两个代理人同时执行的身体关节，但没有发现结合他们的静态关节姿势的证据。与 agent-2 被动参与的二元交互相比，agent-2 积极参与的二元交互导致 mu 节律的功率抑制增加。后者与 agent-2 在场但未参与的行为没有区别。没有发现预期效果。结果表明，二元交互的感觉运动表示考虑了两个代理人同时执行的身体关节，但没有发现结合他们的静态关节姿势的证据。