Summary: New research clarifies how the brain removes information from working memory and shows that intentional forgetting is an effortful process.
Source: Cognitive Neuroscience Society
Forgetting is often harder than it seems.
Anyone who has struggled to stop an earworm, or who repeatedly checks whether they locked the door, knows how distracting an unwanted thought can be. Cognitive neuroscience has long emphasized how we learn and remember, but researchers are increasingly focused on the opposite question: how do we remove information from the focus of our minds?
At the Cognitive Neuroscience Society annual meeting, researchers presented new findings showing that people use distinct strategies to purge information from working memory, and that deliberately forgetting requires substantial cognitive effort.
“Controlling what we keep and what we drop from working memory is essential for switching tasks and refocusing attention,” says Marie Banich of the University of Colorado, Boulder, who chaired a session on this topic. “So it makes sense that we have mechanisms for removing information from the focus of thought—but those mechanisms are not automatic.”
Building on earlier studies, Banich and collaborators have identified three separate ways people remove items from working memory: replacing an unwanted thought with another, actively suppressing the thought, or clearing the mind of thoughts entirely. Their work combines functional MRI, machine learning, and careful behavioral experiments to track these processes in the brain.
Banich’s interest in forgetting grew from a personal experience with loss and a professional concern about intrusive thoughts in conditions like depression, obsessive-compulsive disorder, schizophrenia, anxiety, PTSD, and ADHD. Although the content of intrusive thoughts differs across conditions—paranoia in schizophrenia, contamination concerns in OCD, catastrophic future scenarios in anxiety—the underlying problem is similar: thoughts repeatedly capture attention and resist removal.
To determine whether someone has truly stopped thinking about a stimulus, Banich’s team trained classifiers on fMRI activity while participants thought about distinct categories (faces, places, objects). After instructing participants to forget specific items, researchers monitored whether the category-specific neural pattern persisted. This method let them detect whether a thought remained active, and it revealed distinct neural signatures for replacing, suppressing, or clearing thoughts.
Their results point to four brain networks that play different roles when a memory is maintained or intentionally removed: the somatomotor network, the visual network, the default mode network, and the frontoparietal control network. Suppression and mindful clearing both appear to depend heavily on the frontoparietal control network, suggesting this system is central to effortful removal of unwanted information.
Identifying these networks opens the door to individualized assessments: some people may struggle to control thoughts because their control networks do not distinguish clearly between removal strategies. Banich and colleagues plan to test whether neurofeedback in the scanner can help people learn to engage the most effective forgetting mechanisms.
Directed forgetting reduces interference but demands attention
Research from Sara Festini at the University of Tampa, also presented at the meeting, shows that intentionally removing items from working memory reduces proactive interference and other forms of memory interference. Proactive interference occurs when old information interferes with learning new, similar information—like confusing two people named Emma. Festini’s work suggests that suppression is often more effective than replacement at preventing such interference.
In her experiments, participants receive explicit “forget” cues that instruct them to remove specific items from working memory. Festini notes these cues are not artificial laboratory contrivances: in daily life we provide implicit forget cues when we change our minds—“Never mind, I don’t want that”—or when an instructor tells students to disregard an earlier comment. These instructions can trigger goal-directed removal that is distinct from merely stopping attention to the item.
However, directed forgetting is attentionally demanding. Festini’s lab has shown that additional attention-demanding tasks can disrupt the efficiency of directed forgetting, indicating it requires cognitive resources rather than being a passive process. They also report age-related differences: older adults show weaker directed-forgetting performance than younger adults, though explicit forget cues still help reduce interference for both groups.
Although Festini’s studies have not yet been translated into clinical treatments, the findings imply that people with disorders marked by poor attentional control—such as major depressive disorder or ADHD—may find it harder to intentionally remove intrusive information. Banich highlights potential relevance for PTSD, where overgeneralized memories are common: people with PTSD might struggle to isolate and suppress the specific memory that triggers overgeneralized responses.
A practical paradox emerges: to suppress an intrusive memory you typically must first identify it, which requires briefly bringing it into awareness. That identification step may be difficult or distressing for some individuals, complicating efforts to design interventions based on directed forgetting.
Festini is now investigating when and why people choose to remove items designated as low-value from working memory without explicit forget instructions. She aims to identify the tipping point that motivates effortful removal, since the benefits—reduced interference and clearer focus—must be weighed against the attentional cost of deliberate forgetting.
About this memory research news
Author: Lisa M.P. Munoz
Source: Cognitive Neuroscience Society
Contact: Lisa M.P. Munoz – Cognitive Neuroscience Society
Image: The image is in the public domain
Original Research: Findings presented at the Cognitive Neuroscience Society 30th Annual Meeting