When memories are reactivated, they enter a period of instability in which they can be affected by a variety of follow-up manipulations. The existence of this type of memory reconsolidation offers the potential for clinical interventions of maladaptive memory. However, such potential cannot be fully exploited until the internal mechanisms of memory changes via reconsolidation are better understood. In the current study, we used a three-day AB-AC paradigm that included self-referential simulation processing and employed electroencephalogram (EEG) techniques to explore how post-retrieval updates of episodic memory come about. Behaviorally, we found that reactivation alongside interference learning (ReI-L, AB-AC, n = 52) can produce much more false memories compared to no reactivation new learning (New-L, AB-DC, n = 31) and reactivation repetitive learning (Rep-L, AB-AB. n = 30). More importantly, ERP results revealed that trials from ReI-L in which memory distortions subsequently occurred showed an observable (compared to the new-learning without memory reactivation) but attenuated (compared to trials associating with later intact memory) amplitude of frontal N400, indicating a moderate level of early conflict reactivation is necessary to trigger crucial memory instability. In addition, to promote optimal distortion of the original memory, a sufficient later constructional processing is also required, reflecting in these intrusive/later false trials showed a larger amplitude of late posterior negativity (LPN). A linear classifier employing neuro features of FN400 and LPN during the reconsolidation phase could predict the original memory retention with 72% accuracy. The present findings indicate that nuance in post-retrieval interference, moderate conflict with protracted construction can lead to optimal alterations of episodic memories.
Keywords: Episodic memory; False memory; Post-retrieval updates; Reconsolidation; Single-trial ERP.
Copyright © 2021 Elsevier Inc. All rights reserved.