Biased self-other attribution changes feedback control: link between the sense of agency and sensorimotor control

Agents must correctly attribute the source of sensory inputs to the self or other for successful interaction with the environment. Many previous research works have revealed that self-other attribution depends on the discrepancy between sensory prediction and actual sensory feedback (sensory prediction error). Self-other attribution is assumed to determine how agents reflect sensory feedback in their feedback control. However, the direct effect of self-other attribution on feedback control is still unclear. In the current study, participants’ self-other attribution was biased through induction training in the selfward or otherward direction at a cognitive-judgment level. Before and after the induction, participants controlled a morphed cursor between self and other movements and evaluated the self-other attribution of the cursor movement. We first confirmed that the induction biased their self-other attributions toward the intended direction (i.e., selfward or otherward). Next, we examined whether the respective bias changed the feedback control of the cursor. For the self-biased group, the more participants attributed their visual feedback to themselves, the more they tried to control the morphed cursor in a specific condition. In contrast, for the other-biased group, the more participants attributed their visual feedback to others, the less they tried to control the morphed visual feedback in a specific condition. These results suggest that self-other attribution is not simply a cognitive process reflecting the sensory feedback of motor control but also a factor affecting motor control in a top-down manner.