It isn’t fatigue. It’s the door. For nearly twenty years, researchers in cognitive psychology have shown that a simple passage from one room to another is enough to erase an intention that was clear just seconds earlier. The phenomenon has a scientific name, the updating localization effect, more commonly known to the public as the door threshold effect. And it isn’t just an impression: laboratory experiments, some reproduced in entirely virtual environments, show that the brain literally closes a mental file at the moment you cross a doorway.
Key takeaways
- Doors aren’t just passages: they trigger a mental reset that erases intentions
- This phenomenon works even with virtual or imagined doors—the perception of a change of context is what matters
- Recent studies nuance the initial effect: it truly appears only when the brain is already overloaded
The experiment that put researchers on the trail
Everything started with the work of Gabriel Radvansky, a psychologist at the University of Notre Dame in Indiana. His team had participants navigate a virtual environment filled with rooms connected by doors, each containing a table with an object to transport to the next room. In the original setup, the participants played a video game in which they moved freely through a 3D environment made of a sequence of rooms, each hosting a table with an object, and they were required to move every object from one table to the next, either into the neighboring room through the door or within the same room. The result? Underwhelming, yet highly informative: a decline in memory appeared as soon as people passed from one location to another, even though the path traveled was exactly the same in both conditions.
The detail that intrigued the researchers was that physical distance or time elapsed did not account for the drop. The findings show that the drop in memory performance isn’t due to the distance walked or the time spent; it’s the simple act of changing “scenes.” It isn’t the effort of walking that clears the head; it’s the change of scenery itself that acts as a reset signal. The brain seems to treat each room as a distinct chapter of a story, and turning the page erases a portion of the previous chapter, including the intention that had been formed there.
Even without a real door, the brain erases
The most troubling part is that this mechanism doesn’t even require a real door. Radvansky replicated the effect with mere computer windows or by asking participants to imagine a movement. Earlier research suggested that such forgetting occurs at both physical boundaries, such as moving from one room to another through a door, and metaphysical boundaries, like imagining crossing a door, or even moving from one computer window to another. It’s dizzying: if the place doesn’t actually change, but the mind believes it does, forgetting still occurs.
This robustness has been tested from several angles. In a variant using a real physical environment rather than a simulation, to maximize immersion, the updating effects were observed in both conditions, virtual and real. The decisive factor is therefore not the realism of the setting, but the perception of a transition between two distinct mental contexts. A door can be made of cardboard or exist only in imagination, yet the brain responds in the same way: it archives the old episode and opens a new drawer, even if it means letting slip what had just been stored there.
A real effect, but not universal
More recent work adds a crucial nuance. A study published this year in the journal BMC Psychology, led by Oliver Baumann at Bond University in Australia, initially failed to reproduce the threshold effect under standard conditions. The early results surprised the team: at first they found no threshold effect, and they wondered if participants were perhaps too good at remembering everything. It was only after deliberately overloading working memory by asking participants to count backward while moving that forgetting reemerged. The study confirms that the effect does exist, but only when the brain is occupied, and to a lesser extent than earlier studies suggested.
Other teams, notably at the University of Edinburgh, pushed the doubt even further by testing the effect across four protocols that mixed immersive virtual reality with real movements. Across all these experiments, no significant door effect on forgetting was observed under normal conditions; the effect reappeared only under cognitive load. This tempered result doesn’t negate the phenomenon, but it clarifies its boundaries: the threshold effect may be less a universal flaw of human memory than a vulnerability that becomes evident mainly when the mind is juggling several tasks—a scenario most of us experience continually, between notifications, kids calling, and mental grocery lists.
What it changes concretely
A January 2026 study from Radvansky’s lab, published in Memory & Cognition, adds another layer of complexity. Event boundaries can sometimes improve memory for information encoded near them; items learned close to a transition point are remembered better than those learned in the middle of an episode. The threshold effect is therefore not uniformly detrimental—it reflects a broader episodic-organization process that can work in either direction. The threshold isn’t merely an amplifier of forgotten intentions; paradoxically, it can also serve as a marker that reinforces certain memories located right at its edge.
Practically speaking, the takeaway is a single sentence that researchers keep echoing: form a clear and specific intention before crossing a threshold, rather than relying on ambient memory, which reduces the risk of losing it at the boundary. The next time you head to the kitchen to grab something, say it aloud before you pass through the door. It isn’t a superstition; it’s a way to lock the intention into a format the brain won’t archive as quickly as everything else.
Sources: atlantico.fr | envol-descamps.fr
