Neural electrophysiological mechanism of voluntary forgetting

Abstract

We all have memories that we would rather not remember. Unwanted memories might be a bitter past, or simply outdated trivials. The ability to control the retrieval of unwanted memories thus could be adaptive and bears psychological benefits. Decades of research have shown that people could voluntarily suppress either encoding or retrieval of unwanted memories, inducing voluntary forgetting. Despite accumulative evidence delineating the behavioural and neural mechanisms of voluntary forgetting, unanswered questions and research gaps still await further exploration: 1) previous research usually adopted single isolated stimuli and rarely explored the complicated structures of episodic memory; 2) existing neural evidence has informed us the brain regions and their interactions involved in voluntary forgetting, however, the precise timing of this process remains poorly understood; 3) unwanted memories are often accompanied by negative emotions, but it is unclear whether voluntary forgetting could weaken the emotional charges. Therefore, this thesis aims to address these gaps in four studies using two widely used voluntary forgetting paradigms, Directed Forgetting (DF) and Think/No-think (TNT).

To start with, study 1 (Chapter 2) modified the item-method DF paradigm to examine voluntary forgetting of sequential memories, and applied EEG representational similarity analysis (RSA) to examine the underlying neural mechanism. Results suggested an asymmetric forgetting effect on the sequential memory, such that people preferentially forgot the recent items than earlier items. People also showed a concurrent forgetting of the temporal order. In chapter 3, study 2 follows the experimental procedure of study 1, with an additional frequency tagging manipulation during the encoding phase to track memory reactivation and its suppression. We replicated the asymmetric forgetting effect on the sequential memory, and found a reactivation of frequency tagged during items that were hard to forget.

In Study 3, the time course of voluntary forgetting was investigated in a TNT paradigm, by applying multivariate decoding on EEG. Results revealed a staged processing of memory suppression. Notably, inhibitory control suppressed the cue-target conversion processes at around 500 ms, followed by the suppression of targeted memories and their cortical reinstatement patterns that would lead to reminiscence.

Finally, study 4 examined the implicit emotional aftereffect of voluntary forgetting via the Affect Misattribution Procedure (AMP). Across four individual experiments, we observed the suppression induced forgetting of episodic memories, but no emotional benefits were found, even in an internal meta-analysis combining data across experiments.

This thesis provides novel evidence that voluntary forgetting of sequence memory is asymmetric and fragmented, with a concurrent loss of temporal order memory. Together with the suppression time course finding, our results suggested that successful voluntary forgetting warranted a rapid reaction to target at the unwanted memory. These findings provide inspiring perspectives for future study and may inform the translational application of voluntary forgetting.