Re-enacting steps supports human path integration consistent with motor-corrected grid cell drift

Abstract

Efficient navigation, especially in the absence of vision, requires path integration -- the continuous updating of spatial position from self-motion cues. However, path integration is prone to cumulative error, which can be amplified when body-derived information is inconsistent between encoding and retrieval paths. Drawing on evidence from sensorimotor reactivation during memory retrieval, we hypothesized that re-enacting encoding-related movement patterns could serve as a body-derived mechanism to counteract such errors. In a novel virtual reality task with motion tracking, participants learned unique, irregular step sequences linked to specific target distances during an encoding phase. They later reproduced these distances in complete darkness under three conditions: self-paced free retrieval, retrieval with encoding-congruent movements, and retrieval with encoding-incongruent, regular gait. During free retrieval, participants naturally re-enacted encoding-related movements associated with improved distance reproduction. As predicted, distance estimation was significantly more accurate during congruent retrieval than during incongruent retrieval. These behavioral findings are consistent with a neural network model in which retrieved encoding-related motor patterns correct path integration errors in grid cells. Together, these results provide converging behavioral and computational evidence that body-derived, encoding-related motor patterns can enhance distance estimation, possibly by filtering grid cell error accumulation, offering new insights into embodied mechanisms of path integration.