New research provides experimental evidence that the sensation of déjà vu can trigger an illusory feeling of being able to predict the future. The findings suggest that when people detect familiarity in a novel scene, it creates a subjective sense that they know what will happen next, even when they cannot actually predict the outcome. This study was published in the journal Consciousness and Cognition.
Psychologists have studied déjà vu for decades to understand why a situation can feel intensely familiar despite the individual knowing they have never encountered it before. A common feature associated with déjà vu is a “feeling of prediction.” This is the strong sensation that a person knows exactly how a moment will unfold or what is around the next corner.
Previous studies have identified a correlation between the intensity of the familiarity and the strength of this predictive feeling. When people report stronger déjà vu, they typically report a stronger sense of knowing the future.
But prior research was primarily correlational, meaning it could not definitively prove that familiarity causes the feeling of prediction. It was possible that the association arose because participants were explicitly asked about déjà vu, leading them to answer in ways that aligned with popular conceptions of the phenomenon.
The authors of the new study—Andrew Huebert of the University of Southern Mississippi, Sarah Myers of Minnesota State University: Mankato, and Anne Cleary of Colorado State University—aimed to determine if familiarity detection itself plays a causal role.
They sought to manipulate the level of familiarity a person experiences to see if it directly increases the sensation of being able to predict the future. They also sought to do this without explicitly prompting participants to think about déjà vu, thereby reducing the influence of social expectations.
“Previous research has shown an association between déjà vu and illusions of prediction. In other words, when people experience déjà vu, they often feel like they know exactly what will happen next,” the researchers told PsyPost. “This feeling of prediction seems to be illusory and not linked to any actual predictive ability. A proposed reason is that intense sensations of familiarity seem to drive this feeling of prediction. We aimed to find direct experimental evidence in support of this theory.”
For their study, the researchers employed a virtual reality paradigm across three separate experiments. They used a series of video clips that provided first-person tours of various virtual environments. These environments included common locations such as bowling alleys, cafeterias, and courtyards. The study utilized a concept known as the Gestalt Familiarity Hypothesis. This theory posits that déjà vu occurs when the spatial arrangement of a new scene matches a previously stored memory, even if the specific details are different.
Experiment 1 involved 66 undergraduate students. The procedure began with a study phase where participants watched a series of these virtual tours. To vary the strength of the memory traces, half of the scenes were presented once, while the other half were presented three times. Following this, participants entered a test phase where they viewed completely novel scenes. Although these test scenes were new, half of them shared the exact spatial layout and movement path as a scene from the study phase. The other half were entirely unstudied layouts.
During the test phase, the video tour would pause just before the camera made a final turn. Participants were asked to identify the studied scene that resembled the current layout. Following this recall attempt, they indicated whether the current scene felt familiar. Finally, they reported whether they had a sense of knowing which way the tour would turn next and made a specific prediction of left or right.
The researchers focused their analysis on trials where participants failed to recall the specific studied scene. This was to ensure that any feeling of prediction was based on a general sense of familiarity rather than an explicit memory.
The results indicated that test scenes sharing a layout with videos viewed three times elicited the highest rates of familiarity. These were followed by scenes resembling those viewed once, with unstudied scenes generating the lowest familiarity. Importantly, the reported feelings of prediction followed this same pattern. Higher levels of experimentally induced familiarity led to higher rates of feeling able to predict the turn.
“We found that manipulating familiarity directly increased feelings of prediction,” the researchers explained. “Basically, we have people watch these virtual tours of various environments. Then, later on, they watch tours of new scenes, some of which resemble studied scenes in terms of spatial layout. This can create feelings of déjà vu.”
However, in Experiment 1, participants demonstrated actual predictive accuracy that was above chance levels. The authors hypothesized that the order of the questions influenced this outcome. By asking participants to try to recall the source memory at the beginning of the trial, the researchers may have inadvertently helped participants access the memory before making their prediction. If the participant successfully retrieved the memory of the original path after the initial prompt, they could use that information to predict the turn correctly.
To address this, the researchers conducted Experiment 2 with 94 participants. This experiment mirrored the first but changed the order of the prompts. The question asking participants to recall the source memory was moved to the very end of the sequence, after they had reported their feelings of prediction and made their directional choice. This change aimed to isolate the feeling of prediction from the successful retrieval of memory.
The findings from Experiment 2 confirmed the illusory nature of the sensation. As in the first experiment, increased exposure to the spatial layouts led to increased feelings of familiarity and stronger feelings of prediction.
However, unlike the first experiment, participants’ actual accuracy in predicting the turn was no better than random guessing. This suggests that while familiarity creates a compelling subjective sense of knowing what comes next, it does not provide actual predictive ability when the specific memory remains unrecalled.
“We found that manipulating the spatial familiarity of the otherwise novel test scenes increased people’s feelings like they knew where to turn next,” the researchers said. “Thus, the feeling of prediction seems to be a product of environmental familiarity, rather than because of a general belief that déjà vu means an ability to predict the future or something paranormal as some may believe.”
Experiment 3 included 124 participants and was designed to rule out the possibility that asking about familiarity biases the participants. The researchers wanted to ensure that the act of evaluating familiarity was not what primed the feeling of prediction. Participants were divided into two groups. One group answered questions about both familiarity and prediction, similar to the previous experiments. The second group was only asked about their feelings of prediction and the direction of the turn, with no mention of familiarity.
The results showed that the pattern remained consistent across both groups. Even when participants were not explicitly asked to evaluate how familiar a scene felt, the spatial overlap with studied scenes still drove higher reports of feeling able to predict the outcome. This implies that the process is relatively automatic. The detection of familiarity influences the subjective feeling of prediction regardless of whether the person is consciously focusing on that familiarity.
A specific pattern emerged regarding the “dose” of familiarity. Across the studies, the scenes associated with layouts viewed three times consistently generated stronger effects than those viewed once. This dose-response relationship is suggestive of a causal link. It indicates that the intensity of the underlying memory trace directly calibrates the intensity of the illusory prediction.
“In all of our experiments, we actually never asked participants about déjà vu,” the researchers told PsyPost. “It was a little bit surprising that people did not need to be prompted about déjà vu and still experienced these feelings of prediction. The feeling of familiarity all by itself can likely contribute to an illusory sense of predicting what is going to happen next. The take home is that we have found causal evidence showing that the sense of familiarity itself can cause illusory feelings of prediction.”
The experiments relied exclusively on spatial layout to induce familiarity. It remains to be seen whether other forms of familiarity, such as recognizing a face or a voice without context, would trigger similar predictive illusions. This is sometimes referred to as the “butcher on the bus” phenomenon, where a person recognizes a face but cannot place it. Future work could determine if such non-spatial familiarity also generates feelings of knowing what the person will do or say next.
Another direction for future inquiry involves the timing of these sensations. Some evidence suggests that feelings of prediction might fade quickly, similar to how the biases associated with “tip of the tongue” states diminish over seconds. The current study noted that participants in Experiment 3 who were asked about familiarity first had slightly lower overall rates of predicted feelings than those who were not. This might suggest that the extra time taken to answer the familiarity question allowed the illusory feeling to dissipate slightly.
“There are a number of interesting future directions,” the researchers said. “One is whether familiarity can create feelings of prediction outside the realm of déjà vu. Like I mentioned earlier, though the focus was on déjà vu, we never specifically asked about déjà vu in this study (so that we could directly examine the role of familiarity in driving feelings of prediction). This made us wonder if familiarity can create these feelings under more normal circumstances.
“We are also interested in seeing if we can experimentally show that sensations of familiarity can create other related feelings. For example, we published some work a while ago on how déjà vu is also associated with a hindsight bias. That is, when people experience déjà vu, they also seem to think they predicted a random event after it happened. We would love to apply our methodology to investigate this issue as well.”
“Déjà vu has been such a seemingly puzzling sensation,” the researchers added. “People have many non-scientific explanations for déjà vu and the associated feelings. Only over the last decade and a half or so have we really started to be able to investigate it experimentally. We are very happy that our study can contribute to a better understanding of this seemingly puzzling sensation.”
The study, “Experimental evidence that illusory feelings of prediction can be caused by familiarity detection,” was authored by Andrew M. Huebert, Sarah J. Myers, and Anne M. Cleary.
