We are accustomed to believing that the past has already occurred and therefore remains unchangeable. In this view, memory functions like an archive: the brain stores events and retrieves the relevant record when a decision needs to be made.
Modern memory research, however, paints a far stranger picture. The brain doesn't simply retrieve the past—it partially reconstructs it every time. And sometimes, the version of an event it creates begins to influence the future almost as powerfully as the event that truly happened.
This phenomenon was particularly illuminated by a study from Jianqin Wang, Henry Otgaar, and Mark Howe, published in the Journal of Experimental Psychology: Learning, Memory, and Cognition. The researchers posed an unusual question: can the memory of a reward that never existed alter a person's subsequent choices?
How to Create a Memory of an Event That Never Happened
The researchers utilized a visual variation of the classic Deese-Roediger-McDermott (DRM) paradigm.
Participants were shown groups of related images. Some images were accompanied by a monetary reward, while others were not.
However, within each semantic group, there was a particularly significant item—a “critical lure.” This item was semantically closely related to the other images, yet participants were never shown it.
It was precisely at this point that memory began to construct reality.
Following the training, some individuals were convinced they had previously seen the critical lure. Moreover, they could mistakenly “recall” that this never-presented object was associated with a monetary reward.
A complete false memory of a reward thus emerged.
When a Fictitious Past Begins to Govern Choice
Next, participants were asked to make decisions. And that's when the most interesting part happened.
People more frequently selected critical lures from semantic groups that had previously been associated with a reward. This occurred even though these specific images had never actually yielded money for them—in fact, they had not appeared at all during the learning phase.
Participants also made these decisions more quickly.
In one experiment, an even more striking pattern emerged: the more false memories of a reward an individual formed, the stronger their subsequent tendency was to choose the corresponding lures.
However, simple false recognition alone proved insufficient. It was specifically the erroneous association of the image with a reward that correlated with the subsequent preference. This reveals a surprising chain:
association - memory reconstruction - non-existent episode - real choice.
The event never happened. The memory appeared. Yet the consequences were real.
Why Does the Brain Make Such a Mistake At All?
Memory did not evolve as a perfect video recording of life. Its purpose is to help the organism quickly utilize past experiences to guide behavior in new situations.
If several similar objects proved beneficial, it's advantageous for the brain to form a broader rule: objects of this type might be valuable.
To achieve this, related representations activate one another. The study's authors interpret their findings through the mechanism of spreading activation among related memory elements and propose a constructive association hypothesis.
Most of the time, such generalization is extremely useful. But sometimes, the brain takes an additional step and stops distinguishing between “this is similar to what brought me a reward” and “I remember that this exact thing brought me a reward.”
Thus, a useful mechanism for generalizing experience can generate a memory of an event that never actually occurred.
What Happens in the Brain When a False Memory Is Born
The work by Wang and colleagues was behavioral and did not directly reveal the neural processes that create such memories.
However, a new 2026 study adds an important detail to this picture. Hanyue Liu and her colleagues used electroencephalography and event-related potentials to investigate not the choice itself, but the moment of information encoding from which false recognition later arises. It turned out that false memories can form through different pathways.
When the critical lure was strongly associated with learned objects, a more vivid false memory was linked to schematic integration—the brain seemingly incorporated a new element into an existing semantic structure.
With weaker associations, a different mechanism was observed—more pronounced detailed associative processing during encoding.
In other words, false memory may not arise at a single moment when a person “misremembers.” Its future form begins to be laid down even when the brain first organizes incoming experience.
Memory: A Modeling Machine
Together, these studies change the conventional view of memory. It turns out to be less a repository of the past and more a system for constructing models.
The brain connects events, extracts patterns, fills in gaps, transfers meanings between similar situations, and creates the most plausible picture of what happened. This picture then serves as the basis for the next decision.
Imagine someone walking down an unfamiliar street and suddenly feeling a fondness for one of the cafes. They can't explain why. Perhaps the interior reminds them of a place where a pleasant conversation once took place. Perhaps the music resembles something heard many years ago. Perhaps the combination of smells, light, and details activated an entire network of related memories. Consciousness receives the outcome:
“I like it here for some reason.”
However, the story that gave rise to this feeling might be significantly more complex—and not necessarily align perfectly with the actual past.
Where Does Consciousness Fit In?
Research into false reward memories indicates that a person's decisions can be based on reconstructed memory content, which is subjectively perceived as part of the past, even though the corresponding event never occurred. This calls for a more cautious approach to the simple model:
event - memory - conscious decision.
The actual chain might look different:
events - associations - reconstruction - subjective past - decision.
And consciousness does not necessarily access the original reality, but rather a model of that reality already processed by the brain.
Therefore, these experiments do not, in themselves, disprove the global workspace theory or higher-order theories of consciousness.
However, they raise an important question for these theories: how much information has already been selected, transformed, and reconstructed before becoming the content of conscious experience?
The Past Is Not Just Stored—It Participates in Creating the Future
The brain isn't obliged to remember the past perfectly. It's more important for it to use the past effectively enough to predict what to do next. Memory, therefore, constantly balances between accuracy and generalization. A too-precise memory would be poor at transferring experience to new situations. An overly free reconstruction would turn the past into fantasy. Human memory exists between these extremes.
Sometimes, it errs.
Yet, the very ability to reconstruct experience is likely part of the mechanism that allows humans to imagine situations that haven't yet occurred, evaluate them, and make choices.
This leads to a paradoxical outcome: the brain can create an event that never existed in the past—and then use it to construct a very real future.
Therefore, the question “What happened to me?” is not always entirely separate from another:
“What did my brain construct from what happened to me?”
