Memory-based explanation
Research has associated déjà vu experiences with good memory functions, particularly long-term implicit memory. Recognition memory enables people to realize the event or activity that they are experiencing has happened before. When people experience déjà vu, they may have their recognition memory triggered by certain situations which they have never encountered.
The similarity between a déjà vu–eliciting stimulus and an existing, or non-existing but different, memory trace may lead to the sensation that an event or experience currently being experienced has already been experienced in the past. Thus, encountering something that evokes the implicit associations of an experience or sensation that cannot be remembered may lead to déjà vu. In an effort to reproduce the sensation experimentally, Banister and Zangwill (1941) used hypnosis to give participants posthypnotic amnesia for material they had already seen. When this was later re-encountered, the restricted activation caused thereafter by the posthypnotic amnesia resulted in three of the ten participants reporting what the authors termed "paramnesias".
Two approaches are used by researchers to study feelings of previous experience, with the process of recollection and familiarity. Recollection-based recognition is the ostensible realization that the current situation has occurred before. Familiarity-based recognition is the feeling of familiarity with the current situation without being able to identify any specific memory or previous event that could be associated with the sensation.
In 2010, O'Connor, Moulin, and Conway developed another laboratory analog of déjà vu based on two contrast groups of carefully selected participants, a group under posthypnotic amnesia condition (PHA) and a group under posthypnotic familiarity condition (PHF). The idea of PHA group was based on the work done by Banister and Zangwill (1941), and the PHF group was built on the research results of O'Connor, Moulin, and Conway (2007). They applied the same puzzle game for both groups, "Railroad Rush Hour", a game in which one aims to slide a red car through the exit by rearranging and shifting other blocking trucks and cars on the road. After completing the puzzle, each participant in the PHA group received a posthypnotic amnesia suggestion to forget the game in the hypnosis. Then, each participant in the PHF group was not given the puzzle but received a posthypnotic familiarity suggestion that they would feel familiar with this game during the hypnosis. After the hypnosis, all participants were asked to play the puzzle (the second time for PHA group) and reported the feelings of playing.
In the PHA condition, if a participant reported no memory of completing the puzzle game during hypnosis, researchers scored the participant as passing the suggestion. In the PHF condition, if participants reported that the puzzle game felt familiar, researchers scored the participant as passing the suggestion. It turned out that, both in the PHA and PHF conditions, five participants passed the suggestion and one did not, which is 83.33% of the total sample. More participants in PHF group felt a strong sense of familiarity, for instance, comments like "I think I have done this several years ago." Furthermore, more participants in PHF group experienced a strong déjà vu, for example, "I think I have done the exact puzzle before." Three out of six participants in the PHA group felt a sense of déjà vu, and none of them experienced a strong sense of it. These figures are consistent with Banister and Zangwill's findings. Some participants in PHA group related the familiarity when completing the puzzle with an exact event that happened before, which is more likely to be a phenomenon of source amnesia. Other participants started to realize that they may have completed the puzzle game during hypnosis, which is more akin to the phenomenon of breaching. In contrast, participants in the PHF group reported that they felt confused about the strong familiarity of this puzzle, with the feeling of playing it just sliding across their minds. Overall, the experiences of participants in the PHF group is more likely to be the déjà vu in life, while the experiences of participants in the PHA group is unlikely to be real déjà vu.
A 2012 study in the journal Consciousness and Cognition, that used virtual reality technology to study reported déjà vu experiences, supported this idea. This virtual reality investigation suggested that similarity between a new scene's spatial layout and the layout of a previously experienced scene in memory (but which fails to be recalled) may contribute to the déjà vu experience. When the previously experienced scene fails to come to mind in response to viewing the new scene, that previously experienced scene in memory can still exert an effect—that effect may be a feeling of familiarity with the new scene that is subjectively experienced as a feeling that an event or experience currently being experienced has already been experienced in the past, or of having been there before despite knowing otherwise.
In 2018 a study examined volunteers' brains under experimentally induced déjà vu through the use of fMRI brain scans. The induced "deja vu" state was created by getting them to look at a series of logically related and unrelated words. The researchers would then ask the participants how many words starting with a specific letter they saw. With related words such as "door, shutter, screen, breeze", the participants would be asked if they saw any words that started with "W" (i.e. Window, a term that was not presented to the participants). If they did note that they thought they saw a word that wasn't presented to them, then déjà vu was induced. The researchers would then examine the volunteers' brains at the moment of induced déjà vu. From these scans, they noticed that there was visible activity in regions of the brain associated with mnemonic conflict. This finding suggests that more research regarding memory conflict may be important in better understanding déjà vu.
Another possible explanation for the phenomenon of déjà vu is the occurrence of cryptomnesia, which is where information learned is forgotten but nevertheless stored in the brain, and similar occurrences invoke the contained knowledge, leading to a feeling of familiarity because the event or experience being experienced has already been experienced in the past, known as "déjà vu". Some experts suggest that memory is a process of reconstruction, rather than a recollection of fixed, established events. This reconstruction comes from stored components, involving emotions, distortions, and omissions. Each successive recall of an event is merely a recall of the last reconstruction. The proposed sense of recognition (déjà vu) involves achieving a good match between the present experience and the stored data. This reconstruction, however, may now differ so much from the original event it is as though it had never been experienced before, even though it seems similar.
In 1965, Robert Efron of Boston's Veterans Hospital proposed that déjà vu is caused by dual neurological processing caused by delayed signals. Efron found that the brain's sorting of incoming signals is done in the temporal lobe of the brain's left hemisphere. However, signals enter the temporal lobe twice before processing, once from each hemisphere of the brain, normally with a slight delay of milliseconds between them. Efron proposed that if the two signals were occasionally not synchronized properly, then they would be processed as two separate experiences, with the second seeming to be a re-living of the first.