The ever-changing engram: towards an integrated understanding of long-term memory dynamics

We are witnessing exciting times in memory research – the traditional view of long-term memory as a stable entity is being replaced by the opposing idea that long-term memory is in fact fundamentally dynamic. In the last decade, basic mechanisms modulating the formation and dynamic restructuring of memory have been identified at multiple levels of analysis. This special issue reflects some of these developments.

Four papers in this Special Issue focus on intentional memory control and its consequences for mental health (Nørby, 2017), long-term memory retention (Hupbach, 2017), concurrent memory formation (Hulbert, Hirschstein, Brontë, & Broughton, 2017), and its boundary conditions when it comes to selectivity of forgetting (Akan & Sahakyan, 2017).

Ribot (1882) stated that, “Forgetfulness, except in certain cases, is not a disease of memory but a condition of health and life” (p. 61). Nørby (2017) provides examples for the functional contributions of forgetting to mental health, introducing the concept of “mnemonic emotion regulation”. Reviewing the literature, he argues that forgetting helps us maintaining a positive outlook on life, permitting us to approach new challenges unburdened by the weight of vivid, detailed memories of previous negative experiences and associated emotions. Furthermore, Nørby relates the feelings of fear and sadness and avoidance and withdrawal behaviours that accompany the mental disorders of anxiety and depression to a diminished ability to forget negative memories. He suggests that inhibition and attentional biases present two key mechanisms that moderate forgetting, and outlines potential interventions that could restore healthy forgetting in mental disorders.

Hupbach (2017) asks whether intentional forgetting can affect consolidation and thereby the long-term ease with which information can be subsequently accessed. She shows that instructing participants to forget a series of line drawings resulted in decreased recall of these drawings 24 h later when compared to participants who were instead instructed to remember them. In a second experiment, it was explored whether sleep promotes these effects. In comparison to 12 h of daytime wakefulness a single night of sleep significantly strengthened memory in the remember but not the forget group, suggesting that to-be-remembered information is preferentially strengthened during sleep. These findings add to the growing body of observations showing that the perceived significance of experiences moderates consolidation and thus their retention in memory (e.g., Carr, Jadhav, & Frank, 2011; Deuker et al., 2013; Gruber, Ritchey, Wang, Doss, & Ranganath, 2016; Stickgold & Walker, 2013).

The papers by Hulbert et al. (2017) and Akan and Sahakyan (2017) remind us that intentional memory control is not always as selective as we might wish. Hulbert et al. review recent work on the unintended side effects of memory suppression. Suppressing encoding or retrieval of long-term memory can cause a temporary systemic downregulation of the hippocampal memory system, as indicated by negative BOLD signals in fMRI compared to baseline conditions. Owing to the downregulated state, memory formation is impaired for items presented in close temporal proximity to the suppression attempts. Hulbert et al. carefully review the specific circumstances and moderating factors leading to this “amnesic shadow”, which has been observed in two paradigms, the hippocampal modulation paradigm (Hulbert, Henson, & Anderson, 2016) and the 2-back working memory task (Mullally & O’Mara, 2013). Hulbert et al.’s (2017) paper gives insight into neurocognitive mechanisms of memory suppression, explaining how target memories are dampened, and how “innocent bystanders” are affected. The unintended side effects should be carefully considered when implementing memory suppression as a therapeutic tool to alleviate symptoms associated with posttraumatic stress disorder, anxiety, or obsessive–compulsive disorder.

Akan and Sahakyan’s (2017) paper reveals an important limitation of one of the control methods, list-wise directed forgetting. Selectivity is one of the hallmarks of memory control in retrieval suppression and retrieval-induced forgetting. Yet, Akan and Sahakyan fail to find evidence for selective forgetting of particular aspects of an episode in the list-wise directed forgetting paradigm. In six experiments, using a variety of cueing procedures and stimuli ranging from word lists to sentences, the authors found no reliable evidence for selective directed forgetting. This finding suggests that different control strategies recruit different neurocognitive mechanisms, such as inhibition or context change (e.g., Anderson & Hanslmayr, 2014). These differences should be kept in mind when thinking about implementing these strategies in educational and clinical practice.

Together these four papers on intentional memory control show that the intention to forget and actively suppressing retrieval can shape the content of memory. On the one hand, reduced (long-term) accessibility of to-be-forgotten and suppressed memory contents can positively affect mental health; on the other hand, suppression operations might not be sufficiently precise and result in off-target effects, leading to forgetting of aspects that should have been retained. Further delineating the domain of directed forgetting, Akan and Sahakyan’s identify boundary conditions that will inform theoretical models as well as possible practical applications of these effects.

Three papers in this Special Issue target other aspects of memory dynamics that have recently become a central focus in human memory research: factors that affect memory consolidation during sleep (Seibold, Rasch, Born, & Diekelmann, 2017), the effects of retrieval on long-term accuracy of event memories (Marcotti & St. Jacques, 2017), and the immediate and delayed effects of stress on various memory processes (Quaedflieg & Schwabe, 2017).

Newly encoded memories are initially fragile and can be disrupted by interference learning. This effect was described by Müller and Pilzecker (1900) over 100 years ago and has since been replicated numerous times (McGaugh, 2000). Seibold et al. (2017) extend this line of research by asking whether cueing interfering information during sleep affects the consolidation of recently learned memories. Participants in their study were first trained to remember specific object locations (interference task) in the presence of an odour. Twenty-four hours later, they learned new target locations while no odour was present. Then participants were allowed to sleep for ∼40 min and either received the interference-related odour or vehicle stimulation during slow-wave sleep. After being awakened, memory for target locations was tested. Cueing interfering information during sleep did not negatively affect recall. The authors conclude that reactivation of interfering memories during sleep does not impair ongoing memory consolidation but might instead foster pattern separation and memory stabilisation.

Once consolidated, memories are not permanently fixed. Several lines of research have shown that human episodic memories can be updated and transformed upon their reactivation (for a recent meta-analysis, see Scully, Napper, & Hupbach, 2017). In this context, Marcotti and St. Jacques (2017) explore whether shifts in visual perspective during retrieval alter the content of event memories. In two experiments, the authors asked participants to deliberately shift from an “own eyes” visual perspective to that of an observer when recalling complex event memories. This perspective shift was accompanied by a decrease in subjective vividness of the evoked memories and resulted in impaired accuracy when these memories were recalled again at a later time point. Marcotti and St. Jacques’s conclude that visual perspective shifts can transform memories and outline the implications of their findings for forensic and clinical settings.

Quaedflieg and Schwabe (2017) review the immediate and delayed effects of stress on memory formation, modification and retrieval. Stressful events trigger the release of catecholamines and glucocorticoids which modulate processing in brain areas associated with memory functions, particularly the hippocampus and prefrontal cortex. These hormones have the immediate effect of “set[ting] the brain in a memory formation mode” (Quaedflieg & Schwabe, 2017, p. [367]), ensuring the long-term retention of stressful events. At the same time, rapid glucocorticoid actions cause a shift from a flexible to a more rigid and habitual control of memory. This allows the stressed individual to quickly respond to environmental threats with well-learned behavioural routines but impedes the encoding of detailed in favour of gist-based memories and hinders memory integration and updating. Although a transient shift to more habitual forms of memory is adaptive, Quaedflieg and Schwabe note that their chronic recruitment can lead to stress-related psychopathology and outline potential treatment approaches for the clinical practice.

Collectively, this Special Issue illustrates various facets of long-term memory dynamics, showing that memories are dynamic constructs, constantly being shaped and re-shaped by situational demands, physiological and intentional factors and the way we use them.

Disclosure statement

No potential conflict of interest was reported by the author.