People can remember several things at once, but when an interruption forces them off course, they may rely on only one memory item to get back on track. A new study suggests that the brain doesn’t simply return to its original plan after a distraction—it quickly builds a new one based on where a person ends up, although that flexibility weakens as memory demands grow.
Working memory allows people to hold information in mind for a short time while carrying out tasks. It supports everything from remembering a shopping list to navigating an unfamiliar building. But daily life rarely unfolds without interruptions. A phone notification, a conversation, or an unexpected detour can force someone to pause what they were doing before trying to continue.
Most laboratory studies of working memory examine whether memories survive distractions, often using brief visual interruptions while the person’s situation remains unchanged. The new research instead focused on a different question: what happens when a distraction changes both the task and the person’s position within it, making the original plan less useful?
To investigate that problem, researchers designed an experiment inspired by the classic Snake arcade game, creating a dynamic environment where participants constantly had to combine memory with decision-making.
A memory game became a test of real-world planning
The study involved 50 participants across two experiments.
At the start of each trial, participants memorized the locations of 1, 2, or 4 hidden apples scattered among bushes. Once the apples disappeared, they controlled a snake across the playing field to find them from memory.
Half of the trials introduced an interruption. Instead of immediately searching for the hidden apples, participants first had to collect newly appearing grapes worth more points. This forced them to abandon their original route before returning to the memory task.
The experiment also included a “reminder” option. At any point, participants could briefly reveal the hidden apple locations again, but doing so came with both a time delay and a point penalty. This allowed the researchers to measure when people relied on memory alone and when they decided it was safer to refresh their information.
Interruptions became much more costly as memory demands increased
The results showed a clear pattern.
When participants had only one location to remember, distractions had relatively modest effects. As memory load increased to 2 and especially 4 locations, interruptions caused substantially more errors.
Participants became more likely to search incorrect locations before finding a target, and they increasingly relied on reminders after being interrupted. Even the first target they searched for—the one expected to have the strongest memory representation—became harder to retrieve without assistance when several locations had to be remembered.
These findings indicate that interruptions become progressively more disruptive as working memory has to support larger amounts of information.
People did not return to their original plan
One of the study’s most striking findings involved how participants resumed the task after an interruption.
Without distractions, participants naturally chose the nearest hidden apple first, minimizing travel and creating an efficient search route.
After collecting the distracting grapes, however, participants did not simply continue following the route they had likely planned before the interruption.
Instead, they reorganized their search based on the snake’s new position. Apples that had become closer after the distraction were more likely to be collected first, showing that participants updated their plans to match their current situation rather than trying to resume an outdated sequence.
That flexible replanning remained remarkably effective when only 2 locations had to be remembered.
Working memory dropped to about one usable item after a distraction
Although participants could use multiple remembered locations during uninterrupted trials, interruptions dramatically changed how much information they actually relied on.
Without distractions, people increasingly used working memory as memory load grew, averaging nearly all of the remembered targets before requesting reminders. At the highest memory load, participants used memory to find an average of 3.13 items before checking the reminder.
After a distraction, that pattern almost disappeared.
Regardless of whether participants had memorized 1, 2, or 4 locations, they relied on working memory for only about one item before turning to reminders. Average memory-based performance remained close to a single remembered target across all memory loads.
The researchers suggest this does not necessarily mean the remaining memories vanished. Instead, people may have retained additional information but lacked enough confidence to depend on it, choosing instead to refresh their memory before continuing.
Longer interruptions made replanning harder under heavier memory loads
The second experiment tested whether participants simply predicted where they would be after the distraction ended.
To rule out that possibility, researchers made the interruption last an unpredictable 4, 7, or 10 seconds, preventing participants from knowing exactly when the search would resume.
Even under these uncertain conditions, participants still reorganized their search around their updated position, indicating that replanning occurred during or immediately after the distraction rather than only before it.
However, longer interruptions gradually reduced this flexibility when participants had to remember 4 locations. The tendency to prioritize the nearest target weakened as distraction duration increased, suggesting that maintaining and updating several possible goals becomes increasingly difficult when interruptions last longer.
In contrast, participants remembering only 2 locations maintained nearly the same efficient strategy regardless of how long the distraction lasted.
Memory and the outside world appear to shape each other
The findings challenge a simple view of working memory in which stored information merely waits unchanged until it is needed again.
Instead, the study suggests that remembered information and the outside environment continually interact. As a person’s position changes during an interruption, the memory that receives priority appears to change as well.
Rather than preserving a fixed action plan, working memory seems to support ongoing adjustments that reflect the person’s current circumstances.
The researchers also argue that participants balanced the costs of relying on memory against the costs of checking the reminder. Since reminders imposed both a time penalty and a point loss, participants appeared willing to trust memory for the first target before deciding whether refreshing the remaining information would be worthwhile.
The study leaves important questions for future research
The researchers note that their interruptions always occurred before participants began searching for hidden targets. Future studies will need to determine whether similar replanning occurs when an interruption happens in the middle of an already unfolding action.
They also suggest exploring more complex situations in which memory-guided behavior and new sensory demands overlap. In everyday life, people often face interruptions that require responding to new events while continuing to pursue older goals at the same time.
The new game-based framework provides a controlled way to examine those situations while preserving many of the challenges found outside the laboratory. By showing that interruptions reduce practical reliance on working memory to roughly a single immediately useful goal—even when more information has been remembered—the study offers new insight into how people adapt their plans as the world around them changes.
Publication details
Ziyao Zhang et al, Working memory constraints on replanning following distraction, Communications Psychology (2026). DOI: 10.1038/s44271-026-00497-6
