Involuntary memory signals in the medial temporal lobe

Abstract

We highlight recent progress in neuroimaging and neuropsychological research on memory mechanisms in the medial temporal lobe that speaks to the involuntary nature of memory retrieval processes. We suggest that evidence form these studies supports Barzykowski and Moulin's proposal that memory signals involved in experiences of familiarity and déjà vu can be generated in the absence of retrieval intentionality.

We commend Barzykowski and Moulin (B&M) on presenting a theoretical model that considers links between mnemonic experiences that are typically not discussed together in the cognitive-psychology and cognitive-neuroscience literature. We agree with the emphasis on the shared involuntary nature of the memory experiences covered in the model, and the proposed central role that familiarity plays, as summarized in the schematic in Figure 1. At the same time, we feel that this model could be further developed at the mechanistic level through consideration of the functional properties of perirhinal cortex (PrC) in the medial temporal lobe; it is the brain region that has been most closely linked to familiarity assessment in extant research. To this end, we highlight several recent findings, not covered in the target article, that speak to the involuntary mode in which familiarity signals can arise.

PrC has been implicated in familiarity across many functional neuroimaging studies conducted in neurologically healthy participants and in research in individuals with brain damage (Köhler & Martin, 2020; Montaldi & Kafkas, 2022). Although the majority of reports comprising this literature establish a link between PrC and item familiarity in experimental contexts that involve intentional retrieval (e.g., Montaldi, Spencer, Roberts, & Mayes, 2006), a growing body of evidence suggests that this structure also supports involuntary effects of familiarity or fluency that can be observed in the absence of retrieval intentionality. For example, activity in PrC tracks experimental exposure history for verbal and non-verbal stimuli in tasks that show behavioural priming effects (Voss, Hauner, & Paller, 2009; Wang, Ranganath, & Yonelinas, 2014; Yang, McRae, & Köhler, 2023). In related functional magnetic resonance imaging (fMRI) work, we have recently shown that PrC not only tracks recent laboratory exposure, but also degree of judged lifetime exposure to object concepts outside the lab (Duke, Martin, Bowles, McRae, & Köhler, 2017; Yang et al., 2023). Specifically, using a paradigm in which participants made judgements that either required or did not require consideration of lifetime familiarity, we found that fMRI BOLD activity in PrC tracked this memory characteristic regardless of retrieval intentionality. Moreover, observed behavioural performance was also sensitive to lifetime familiarity under conditions in which the latter was task irrelevant (Yang et al., 2023). Overall, such evidence points to PrC as a structure that may support the cue-induced familiarity that is central to involuntary memory retrieval in B&M's model.

A role for PrC in involuntary memory has also been revealed in research conducted in neurological patients who experience déjà vu in the context of seizures with medial temporal-lobe origin (i.e., ictal déjà vu). These experiences have several phenomenological characteristics that mirror those highlighted by B&M for spontaneously occurring déjà vu in neurotypical individuals. At the core, ictal déjà vu is also characterized by an involuntarily arising feeling of familiarity that conflicts with a metacognitive awareness of situational novelty. Although ictal déjà vu in temporal-lobe epilepsy is a transient phenomenon, the eliciting seizure activity is tied to lasting structural and functional brain abnormalities. These abnormalities provide a window towards understanding what brain regions may be involved in triggering the familiarity experience that is central to déjà vu. Indeed, we have reported that ictal déjà vu is associated with quantifiable abnormal brain structure in PrC that also goes hand in hand with persistent familiarity-based memory impairments on experimental tasks of recognition memory (Martin et al., 2021).

While we suggest that the neural familiarity signals that are central to the involuntary familiarity and déjà vu experiences described in B&M's model are closely tied to PrC functions, we would not argue that activity in this region alone is sufficient to generate these mnemonic experiences. Rather, we agree with the authors' proposal that automatically generated familiarity signals may only serve as an initial trigger; depending on the outcome of downstream processes that include metacognitive evaluation of situational novelty, they may or may not lead to memory experiences of familiarity or déjà vu. As such, we do not want to leave the impression that we disagree with the authors' point of emphasis on large-scale brain networks, such as the default mode network, when attempting to explain involuntary memory retrieval. However, given the increasing appreciation of the complexity of memory-retrieval cascades in the cognitive neuroscience literature, we think that linking specific brain regions to specific component processes holds unique promise for further mechanistic and theoretical development of the present model.