Reactivation protects human motor memories against interference from competing inputs

Abstract

Many newly encoded memories are labile when acquired but then consolidate to more stable states. Reconsolidation theory posits that reactivating a consolidated memory again destabilizes it, increasing its vulnerability to interference from competing memories. In a series of 3-day experiments, we investigated the fate of a motor memory when it is reactivated and challenged with a competing one. We pursued a modular design in which humans adapted to a visuomotor rotation A (day 1), then an opposite rotation B (day 2), followed by a retest on A (day 3). We first found that reactivating A before learning B (A-AB-A) caused no greater impairment in A retention than non-reactivation (A-B-A). That is, while interference occurred, it appeared to be uninfluenced by reactivation, contradicting reconsolidation predictions. We then tested an alternate idea, that reactivation might serve to protect the original memory from interference. In subsequent experiments, we introduced no rotation (N) trials either prior to A relearning (A-AB-NA and A-B-NA groups), or immediately after B learning (A-ABN-A and A-BN-A groups). Here, we observed that reactivation served a protective function, but only when B was washed out immediately, preventing its consolidation (A-ABN-A group). Collectively, our results show that reactivation does not necessarily increase the susceptibility of a motor memory to interference but may rather shield it from degradation by competing learning. Our findings align with theories positing memory transitions between active and inactive states, and hold implications for strategies focused on improving memory retention in rehabilitation, sports and skills training.