A recent study, published in the Journal of Experimental Psychology: Learning, Memory, and Cognition, challenges previous assumptions about the automatic nature of the relationship between memory and attention. The researchers found that while eye movements are attracted to information retrieved from long-term memory, this influence on attention is not automatic. Instead, actively holding retrieved information in mind is necessary for it to guide attention.
Memory plays a vital role in everyday life, helping us navigate familiar environments, recognize faces, and complete tasks. However, not all memories are useful all the time. Sometimes, memories can be distracting or irrelevant, pulling our attention away from what we need to focus on. Previous research has focused on how controlling memory retrieval affects what we can later recall. However, the researchers in this study wanted to investigate a more immediate effect: how controlling memory retrieval affects where we direct our attention at that very moment.
“Although being able to remember information stored in memory is crucial for daily living, it is not always desirable,” said study author Mrinmayi Kulkarni, a postdoctoral fellow at the Rotman Research Institute in Toronto. “In some cases, it may be beneficial to control memory retrieval. For instance, imagine that you are concentrating on a difficult writing assignment, and one of the words in the text reminds you of an event from your personal life. Information retrieved from long-term memory has the ability to grab your attention and disrupt your ability to complete the task at hand.”
“In such cases, it may be important to push the retrieved memory out of your mind so that you can complete your assignment. This is especially important when the memories are unpleasant or traumatic. In fact, the involuntary retrieval and re-living of traumatic experiences (called ‘flashbacks’) is a common symptom associated with conditions such as post-traumatic stress disorder (PTSD).”
For their study, the researchers recruited 32 right-handed students from the University of Wisconsin Milwaukee, with data from 26 participants being analyzed after excluding those with unreliable eye-tracking results. The experiment involved four phases: an initial encoding phase, a test-and-encode phase, a memory regulation and visual search phase, and a final recognition test.
In the encoding phase, participants memorized a series of scene-object pairs, which included three faces and three tools paired with various scenes. The objects were simple and easily recognizable, and the participants were encouraged to form meaningful associations between the objects and the scenes to aid in memory retention. For example, they might think about how likely it was to find a specific person in a certain place or how a tool might be used in that setting.
To ensure proper learning, participants then underwent a test-and-encode phase where their memory for the scene-object pairs was tested and reinforced. Each scene was shown again, and participants had the chance to confirm the correct object pairing, thus solidifying their memory of the associations.
The core of the experiment was the memory regulation and visual search phase. Here, participants were given specific instructions to either retrieve, suppress, or substitute a memory. In the retrieval condition, participants were asked to recall the object associated with a scene. In the suppression condition, they were instructed to avoid thinking about the associated object. In the substitution condition, participants were told to replace the object with a different one from another category (for example, replacing a face with a tool).
After a delay, participants viewed a display containing all the objects along with black dots and were instructed to look at the display freely until all but one dot disappeared. Eye movements during this phase were tracked to observe where participants directed their attention in relation to the objects.
The study’s findings revealed that attention, as measured through eye movements, was strongly influenced by memory retrieval. In the retrieval condition, participants looked more often at the object they had been asked to recall, suggesting that retrieved memories actively attract attention.
“Our results show that when participants are allowed to retrieve their memories without any control, they are more likely to pay attention to the information they are retrieving, even if that information is irrelevant to the task participants are instructed to perform,” Kulkarni told PsyPost.
In contrast, when participants were instructed to suppress or substitute the memory, they spent less time looking at the associated object. In some cases, especially with tools, participants’ attention to the suppressed objects fell below chance levels, indicating active suppression of the memory. In the substitution condition, participants’ eye movements often shifted toward the object from the other category, showing that they were successfully replacing the memory with the substitute.
“We found that when participants attempt to control their memories—by trying either to push the retrieved memories out of their minds or by replacing the memory with other information—the retrieved information exerted much less influence on their attention,” Kulkarni explained. “These results suggest that the link between memory and attention may not be automatic, as was previously thought. Instead, retrieving information and actively holding it in mind might be necessary for the information to influence attention.”
Interestingly, the study also found differences based on the type of object being retrieved. Memories of tools were more easily suppressed or replaced than memories of faces, which remained more resistant to memory control efforts. In the final recognition test, participants were less likely to accurately recall the scene-tool associations after trying to suppress or replace the memories, while scene-face pairs were less affected. This suggests that faces may be more deeply encoded or more challenging to suppress than other types of objects.
“Finally, we also found that controlling memory retrieval weakens the memory itself,” Kulkarni said. “In a surprise memory test administered at the end of the experiment, participants’ memory for information that had been controlled was worse than information that was retrieved. This indicates that, in addition to reducing the influence of retrieved information on attention in the moment, controlling memory retrieval might have long-term effects on memory.”
The researchers were also surprised to discover that participants’ ability to successfully control their memories appeared to vary depending on the type of memory they were trying to suppress.
“In the experiment, participants were more effective at suppressing memory for tools, relative to memory for famous faces,” Kulkarni said. “This might suggest one of two things: first, the way that the brain processes and stores faces is different from the way it stores other types of information (in this case, tools). Second, faces are a special type of stimulus.”
“Past research shows that faces are biologically relevant. We encounter a lot of faces in our lives, and they can tell us important information, such as the identity, mood, and threat level of a person. Because of the importance of faces in our lives as social animals, face stimuli might be more resistant to being forgotten following memory control. Additionally, the use of famous faces might have enabled participants to build richer narratives to remember them by, making them less likely to be forgotten.”
But the study, like all research, has some caveats to consider. “A major caveat here was that the materials we used were all emotionally neutral (e.g., pictures of buildings, streets, faces, and tools),” Kulkarni noted. “We know from past work that emotional stimuli, particularly negative materials, are treated slightly differently from neutral stimuli by the brain. Hence, before this work can be applied to clinical populations such as individuals with depression or PTSD, we need more research to examine how these effects translate when participants have to control memories for unpleasant information.”
Furthermore, while eye-tracking provided a valuable tool for measuring attention, the study didn’t fully explore how individual differences in memory strength or suppression ability might influence the results. Some participants may have found it easier or harder to control their memory retrieval, and these differences could affect how memory and attention interact. Future research could investigate these individual differences in more detail, potentially using neuroimaging techniques to measure brain activity.
“In the next steps for this project, we hope to probe how the brain accomplishes memory control,” Kulkarni said. “We plan to examine which parts of the brain are involved in this process, and how they interact to enable memory suppression.”
The study, “Control of memory retrieval alters memory-based eye movements,” was authored by Mrinmayi Kulkarni, Allison E. Nickel, Greta N. Minor, and Deborah E. Hannula.
