Item

Abstract

Motor resonance (MR) has been a prominent idea to explain online motor control strategies. To date there is little evidence for this idea in online motor control tasks using realistic social interactions. Here we set out to test one important prediction of MR in realistic social interactions, namely that the visual human-likeness of the interaction partner should modulate online motor control. We used a novel virtual reality set up in which participants naturally interacted with a life-sized virtual avatar, who looked either like a human or like a robot (between subjects, 14 per group). Participants' task was to high-five this avatar, whose hand position (on 50% of the trials) was randomly moved to one of four locations during motor execution (online motor control task; identical kinematics of both avatars). We tracked participants' hand positions with optical tracking. Hand trajectories showed that participants were faster in carrying out the high-five movements with humans than with robots. However, there was little evidence for a profound effect of the human-likeness on corrective movements during online motor control. This is first evidence that - as predicted by MR – motor control in social interactions is different for different human vs. non-human like interaction partners.