Rethinking the role of motor simulation in perceptual decisions.

Abstract

For decades, motor simulation has been invoked as a mechanism that utilizes covert output of the motor system to solve cognitive problems, including perceptual decisions. Evidence for motor simulation is based on indirect behavioral attributes including biomechanically constrained reaction times, the influence of the physical position of the participant’s limbs on performance, and a subjective sense of motion. Here, we propose that a large set of perceptual judgment experiments, where images of hands are used as stimuli and that have these behavioral attributes, are not based on motor simulation. We propose a fundamentally different mechanism for the effects observed in hand judgment: the multisensory binding of a mismatched seen hand with an unseen felt hand. Both hands are interpreted as belonging to the participant and obligatory coregistration of their positions into a single internal representation of body state causes the response delay, irrespective of the perceptual discrimination of hand laterality. Support for this alternative view is based on novel variations of the laterality task that manipulate local perceptual features to control binding. Multisensory integration is automatic, orientation dependent, and biomechanically constrained. The multisensory model eliminates conceptual inconsistencies that arise when motor simulation is invoked in this task. Finally, the multisensory binding model predicts the observed reaction time delays during hand judgments in a broad range of patients who have a distorted body schema. The new model questions the role, if any, of motor simulation in perceptual decision making and emphasizes the need for stronger experimental evidence to establish that motor simulation is occurring.