Our brains can distort long-term memories in adaptive ways, helping us distinguish between similar events by amplifying the differences that we recall between them, neuroscientists from the University of Oregon and New York University discovered in a new study.

The research challenges the conventional view of memory, which posits that new memories are first encoded separately and distinctly from old memories. Yufei Zhao, a doctoral student in the Department of Psychology at the University of Oregon and lead author of the paper, published Feb. 22 in The Journal of Neuroscience, said that other robust studies have found that similar memories are represented like unrelated memories within the brain. This study counters those assumptions, suggesting that memory may be far more fluid from its inception.

In popular psychology, concepts like Sigmund Freud’s “narcissism of small differences” have long held appeal, as they seem to explain why people exaggerate minor distinctions to better conceptualize the world. Because we tend to forget or blend similar memories, the scientists theorized that people would exaggerate small differences, such as the weather, to counteract the interference between nearly identical memories. They likened this exaggeration, which they called the “repulsion effect,” to the way that two magnets of the same polarity repel each other.

“Our memory system is dynamic. On the one hand, memories could fade away or lose accuracy. On the other hand, adaptive changes could happen to the memory,” said Zhao. “In fact, memory distortion could be helpful in solving our daily problems.”

Over two days of testing, 29 participants learned a series of associations between faces and object images. Participants were separated into “competitive” and “noncompetitive” conditions.

In the competitive version of the test, each object image was paired with another that was identical to it in every way except its color and the associated face. In the noncompetitive version, on the other hand, the paired images were also different colors but belonged to different object categories — for example, umbrellas and balloons.

The noncompetitive paired images differed in color to the same degree as the competitive paired images. This meant researchers could measure just how much participants exaggerated color differences between otherwise identical objects.

On the second day, participants were scanned by a functional magnetic resonance imaging machine while carrying out perception and cued recall tests. After leaving the scanner, participants completed a color memory test. They had to recollect an object’s color by pointing to one object along a continuous color wheel.

Participants recalled far starker color differences between similar objects than dissimilar objects, reflecting a significant degree of bias. In addition, participants whose color memories exhibited more bias had better associative memory performance. The researchers concluded that this repulsion effect, which exaggerates the differences between very similar memories, acted against memory interference, the process by which very similar memories are forgotten or blended together.

“It is intuitive to think that inaccurate memory is harmful and thus, it is fascinating to see that memory distortions actually could have positive effects on memory performance,” Zhao said.

The team’s fMRI data shed light on how and when the brain recalls color information. Zhao and her colleagues scanned key regions in the visual and parietal regions of the brain, and their images suggested that the brain recalls color information more strongly when distinguishing between similar objects.

While the hippocampus has long been tied to competition between memories, the researchers did not observe any major differences in hippocampal activity between the competitive and noncompetitive conditions. “It is very hard to probe the hippocampal activities at feature level based on the current techniques,” said Zhao.

“In our study, we found the critical difference between the two similar objects, instead of being ‘unrelated,’ was directionally exaggerated,” Zhao said. She also suspects that the adaptive exaggeration does not occur in short-term memory, as the study participants initially had difficulty telling objects apart.

This research signals that seemingly harmful or unwanted features of cognition — in this case, the distortion of recalled color differences — can ultimately aid the accuracy of long-term memory. The study may help in the development of brain-simulated neural networks, according to Zhao.

Zhao and her colleagues are currently delving deeper into the repulsion effect to identify the prerequisites for it to occur.

“We think it is necessary for people to concretely bring the old and similar memory event to mind when they are learning a new memory event before [the] repulsion effect could happen,” Zhao commented. “Also, a strong subjective awareness of the difference might also be critical for [the] repulsion effect to happen.”

The study, “Adaptive memory distortions are predicted by feature representations in parietal cortex,” was published in a Feb. 22 issue of The Journal of Neuroscience. The authors of the study were Yufei Zhao and Brice A. Kuhl, of the University of Oregon, and Avi J.H. Chanales, of New York University.