Summary: New research shows that conscious control over memory—deciding to remember or forget—has a stronger influence on what we retain long-term than the emotional charge of an item. Participants reliably recalled words they were instructed to remember more often than emotionally charged words, even though negative emotion sometimes improved recall or led to false memories.
Surprisingly, simply sleeping did not produce a consistent benefit for memory in this study. However, specific sleep-related brain rhythms, especially sleep spindles, were linked to improved recall for emotionally negative words. Together, the results point to intentional, top-down control as a primary force in determining which memories survive.
Key Facts:
- Intentional remembering is strongest: Directing participants to remember particular items had a larger effect on later recall than emotional content.
- Sleep physiology plays a role: Sleep spindles correlated with better recall of negative, remember-cued words, while other sleep features related differently to memory outcomes.
- Emotion can mislead memory: Negative words increased the risk of false recall—participants sometimes “remembered” negative words they had not actually seen.
Source: Frontiers
Background: It is well known that sleep supports memory consolidation, but the mechanisms that decide which memories are kept or discarded remain unclear. Emotional salience and conscious goals both shape memory, but they operate in different ways: emotion often acts automatically, while intentional remembering depends on deliberate attention and control.
To compare these influences, researchers asked participants to either remember or forget words presented on a screen; some words were neutral, and others carried negative emotional meaning. The team compared immediate recognition and delayed recall after a 12-hour interval that either included nocturnal sleep or wakefulness.
“What we intend to remember and to forget can be powerful,” said Dr. Laura Kurdziel of Merrimack College, lead author of the study published in Frontiers in Behavioral Neuroscience. “We have more control over our memories than we often think we do.”
Study design and sleep monitoring
Two matched studies were conducted. Study 1 was online with 45 participants; Study 2 took place in the lab with 53 participants. In each study half the participants learned words in the morning and were tested that evening (wake condition), while the other half learned in the evening and were tested the following morning (sleep condition).
Lab participants in the evening group also wore EEG headbands overnight so researchers could link specific sleep physiology to memory outcomes. Each participant completed two sessions: during the first, they viewed 100 words, each followed by a cue to remember or forget; half of the words were negatively valenced and half neutral. Immediately after encoding, they completed a recognition test with 100 words (50 old, 50 new foils). Twelve hours later they attempted free recall of the remember-cued items, and EEG data were analyzed for correlations with performance.
Findings: Intent beats emotion, but emotion still matters
Across both studies, explicit instructions to remember significantly improved both recognition and delayed recall. Emotional valence alone did not consistently boost overall memory performance, yet it interacted with instruction: negative words that were also cued to be remembered were especially likely to be recalled. At the same time, negative foils were more likely to be falsely recognized, indicating that negative emotion can increase memory distortion.
Kurdziel explains, “During encoding, we devote more attentional resources to items we are told to remember. Top-down control can ‘tag’ those items as relevant, biasing hippocampal processes and increasing the chance that they are reactivated and consolidated.” She adds that instructions can also suppress irrelevant information, reducing interference and making remember-cued items easier to retrieve.
Sleep physiology, not sleep duration, predicted specific effects
Contrary to expectations, whether participants slept across the interval did not produce a general memory advantage. Still, EEG measures showed meaningful relationships with memory content. Greater sleep spindle activity predicted better recall of negative, remember-cued words. REM theta power was associated with an increased tendency to falsely recall negative foils, consistent with emotional generalization. Unexpectedly, deeper slow-wave sleep measures (SWS and delta power) correlated negatively with total recall, which may reflect selective downscaling or active forgetting of information deemed irrelevant.
These sleep–memory links were based on a smaller subset of participants with usable EEG, so the authors caution that the physiological associations need replication in larger and more diverse samples.
Key Questions Answered:
A: Emotional content can slightly enhance recall for some items but also raises the likelihood of false memories, especially for negative material.
A: No overall memory boost from sleep was observed, although specific sleep rhythms (like sleep spindles and REM theta) correlated with different memory outcomes.
A: Intentionally deciding which items to remember produced the clearest and most consistent improvement in long-term recall.
About this memory and neuroscience research news
Author: Angharad Brewer Gillham
Source: Frontiers
Contact: Angharad Brewer Gillham – Frontiers
Image: The image is credited to Neuroscience News
Original Research: Open access. “Top-Down Instruction Outweighs Emotional Salience: Nocturnal Sleep Physiology Indicates Selective Memory Consolidation” by Laura Kurdziel et al., Frontiers in Behavioral Neuroscience. DOI: 10.3389/fnbeh.2025.1643449
Abstract
Top-Down Instruction Outweighs Emotional Salience: Nocturnal Sleep Physiology Indicates Selective Memory Consolidation
Introduction: Sleep supports memory consolidation but may also contribute to forgetting. How sleep prioritizes information when emotional salience and intentional goals conflict is not fully understood.
Methods: Two studies tested interactions between emotional content and directed memory instructions across a 12-hour interval that either included sleep or wake. Participants completed a directed-forgetting paradigm with neutral and negatively valenced words, followed by immediate recognition and delayed free recall.
Results: Instruction to remember consistently enhanced recognition and recall in both online and in-lab studies, while emotion alone did not show uniform benefits. The sleep condition did not produce a general memory benefit. In the lab study, overnight EEG revealed that sleep spindle activity predicted recall for negative remember-cued words; SWS and delta power were negatively correlated with total recall; and REM theta power related to increased false recall of negative foils.
Discussion: These results extend prior nap-based findings by showing full-night sleep physiology reflects selective consolidation mechanisms even when behavioral sleep effects are absent. Overall, top-down instruction appears more influential than emotional salience in guiding which memories persist, and sleep physiology can help explain selective consolidation processes.