In their article, “Are involuntary autobiographical memory and déjà vu natural products of memory retrieval?” Barzykowski and Moulin (B&M) bring two fascinating and usually disparate fields together in a way that we believe will be fruitful for future research and theory. The idea that déjà vu should be considered a form of involuntary memory makes great sense given that déjà vu interrupts ongoing processing and pulls attention toward the experience of déjà vu itself; also, both déjà vu and involuntary autobiographical memory (IAM) are arguably phenomenologically surprising in nature (Christoff, Irving, Fox, Spreng, & Andrews-Hanna, Reference Christoff, Irving, Fox, Spreng and Andrews-Hanna2016; Mills, Zamani, White, & Christoff, Reference Mills, Zamani, White and Christoff2021).

Although the linkage made in the article is already exciting, a potentially helpful extension for the field will be to specify the common mechanism(s) that might be shared between IAM and déjà vu, as well as what distinguishes them phenomenologically (e.g., Neisser et al., Reference Neisser, Abreu, Drane, Pedersen, Parsons, McNeely-White and Cleary2023). We propose that IAM and déjà vu may not share a placement on the same gradient, per se, but rather the mechanism of cue familiarity detection. Upon detecting familiarity with a cue, an inward directed attentional search of memory may be launched that can manifest in the form of attempts at recollecting potentially relevant candidate information (Carlaw, Huebert, McNeely-White, Rhodes, & Cleary, Reference Carlaw, Huebert, McNeely-White, Rhodes and Cleary2022; Huebert, McNeely-White, & Cleary, Reference Huebert, McNeely-White and Cleary2023). We propose that cue familiarity detection can be triggered involuntarily in the presence of a cue or set of cues, which can be internally or externally driven. The familiarity “signal” that emerges varies according to the degree of feature overlap between the cue(s) and memory representations (Clark & Gronlund, Reference Clark and Gronlund1996; McNeely-White, McNeely-White, & Cleary, Reference McNeely-White, McNeely-White and Cleary2021; McNeely-White, McNeely-White, Huebert, Carlaw, & Cleary, Reference McNeely-White, McNeely-White, Huebert, Carlaw and Cleary2022); if it is above a critical threshold, attention may be directed inward toward a search of memory to attempt to retrieve candidate relevant information to the current situation.

This can take at least two different forms, which are similar as noted by B&M. Sometimes, the search launched by initial cue familiarity detection may result in successful retrieval of a relevant prior experience from memory, leading to an experience of an IAM. Notably, although most IAMs have identifiable cues (Ball & Little, Reference Ball and Little2006; Schlagman & Kvavilashvili, Reference Schlagman and Kvavilashvili2008), it is also possible that the cue cannot be identified (the cue-elicited content simply comes to mind in response to the cue, even if the cue itself was not identified by the experiencer). Other times, the search that is launched by initial cue familiarity detection fails. When it fails, we propose that attention is drawn to the sense of familiarity that prompted the inward directed attentional search of memory. In the absence of any retrieved memory content, the experience might be that of either a general familiarity feeling (if the familiarity seems like it is explainable) or déjà vu (if the familiarity cannot easily be explained away). From this perspective, one possible common memory-based mechanism between IAM and déjà vu is cue familiarity detection, and a major differentiating factor between the two is whether the inward directed search of memory results in retrieved content or not.

With this in mind, one aspect that may phenomenologically distinguish IAM and déjà vu is the presence versus the absence of content (see Neisser et al., Reference Neisser, Abreu, Drane, Pedersen, Parsons, McNeely-White and Cleary2023). In the case of IAM, content is retrieved that can usually be attributed to a cue or cues that elicited it roughly 85% of the time (Ball & Little, Reference Ball and Little2006; Schlagman & Kvavilashvili, Reference Schlagman and Kvavilashvili2008). In the case of déjà vu, no content is retrieved, leaving a person in a state of contentless inward focus and a mere sensation of a memory; this sensation can be difficult to explain away when there is no content identified and it is unclear what elicited it, which can be off-putting. The inability to explain the feeling may contribute to a more salient phenomenological experience of surprise or spontaneity in the thought stream (Mills et al., Reference Mills, Zamani, White and Christoff2021).

In this regard, déjà vu may fit within Whittlesea and Williams's (Reference Whittlesea and Williams2000, Reference Whittlesea and Williams2001a, Reference Whittlesea and Williams2001b) discrepancy attribution framework in that it may be when the familiarity feeling is surprising that déjà vu occurs; however, as B&M note, no laboratory paradigm has yet successfully operationalized “surprise” in this context, and existing attempts at doing so have been unsuccessful (see Cleary, Morris, & Langley, Reference Cleary, Morris and Langley2007; Karpicke, McCabe, & Roediger, Reference Karpicke, McCabe and Roediger2008). We propose that it is the inexplicable contentless sensation of memory that characterizes déjà vu. When people experience a strong sensation of familiarity, they have an inherent need to explain why the feeling is occurring, and may do so by conjuring up details that are inaccurate (Carlaw et al., Reference Carlaw, Huebert, McNeely-White, Rhodes and Cleary2022; Huebert et al., Reference Huebert, McNeely-White and Cleary2023) or even confabulating details (Moulin, Reference Moulin2013, Reference Moulin2018) in an effort to explain away the feeling. The more difficult it is to do this, the more likely the experience might be labeled as déjà vu as opposed to just a feeling of familiarity.

Some evidence for this possible mechanistic link can be found in other shared memory-based characteristics of IAM and déjà vu. For example, a shared characteristic between IAM and déjà vu is that both are associated with medial temporal lobe (MTL) processes (Christoff et al., Reference Christoff, Irving, Fox, Spreng and Andrews-Hanna2016; Cleary & Brown, Reference Cleary and Brown2022). It is plausible that this commonality is due to MTL processes being responsible for the initial familiarity detection that launches the search of memory; some researchers have indeed proposed that familiarity detection results from processing occurring in the MTL (e.g., Elfman & Yonelinas, Reference Elfman and Yonelinas2015). From a phenomenological perspective, both IAM and déjà vu are more commonly reported in younger adults compared to older adults (e.g., Cleary & Brown, Reference Cleary and Brown2022; Schlagman, Kvavilashvili, & Schulz, Reference Schlagman, Kvavilashvili, Schulz and Mace2007). It is plausible that this commonality is due to the common mechanism of cue familiarity detection prompting initial memory search, and that this mechanism peaks in young adulthood and then diminishes with age.

In sum, we were delighted to see the interesting links made by B&M across two interesting involuntary and spontaneous thought processes. We think this work will inspire new theoretical and empirical work, and it has certainly prompted us to think more about their shared links, as discussed in this commentary.