Mentally transformed representations in memory are linked to their originals

Our minds are not static storage lockers; they are active workspaces where we hold information and, crucially, where we manipulate it. Imagine trying to drive a car while constantly checking a rearview mirror. To know where a car behind you will be in a few seconds, you must take the image you see in the mirror and mentally shift it forward in time and space. This ability to take a piece of information held in the mind and twist, turn, or update it is a fundamental part of how we navigate a changing world. Scientists call this active workspace "working memory." For decades, a central question has been how this system handles the original information once it has been changed. If you mentally rotate an object, does the brain discard the starting image to make room for the new one, or does it keep the original version safe in the background, perhaps as a backup?

A team of researchers set out to answer this by watching the brain in action while people performed a mental rotation task. They asked volunteers to memorize the angle of a striped pattern, then mentally rotate it by a specific amount, and finally memorize a second, different pattern. Later, a signal told the participants which of these two items—the rotated version or the new pattern—would be tested. By using a technique that reads electrical signals from the scalp, the researchers could see which specific patterns were still being held in the brain's working memory at any given moment. They found something surprising: even though the original pattern was no longer needed for the task, the brain did not throw it away. Instead, the original pattern remained visible in the brain's signals, but only when the rotated version was still being held. The two were linked; if the brain let go of the rotated image, it also let go of the original. If it kept the rotated image, it kept the original too.

The study began with a simple premise. When we perform a mental transformation, such as rotating a shape in our mind, we create a new mental image based on an old one. Previous research suggested that the old image often lingers in the brain even after it becomes irrelevant to the current task. This was puzzling because working memory is known to be efficient; it usually discards information the moment it is no longer useful to save space for what matters. The researchers wanted to know why the brain kept this "useless" original image. They considered three possibilities. First, perhaps the brain simply hadn't finished processing the original image yet, and it was just a lingering echo of perception. Second, maybe the act of rotating the image required such deep attention that the original was locked into memory and could not be removed. Third, perhaps the original and the rotated image were tied together in a single package, so that keeping one meant keeping the other.

To test these ideas, the researchers designed a precise experiment involving forty volunteers. Each person sat in a quiet room and looked at a screen. First, they saw a circular patch with a striped pattern at a specific angle. They had to remember this angle. Then, an instruction appeared telling them to mentally rotate that pattern by a certain amount, such as thirty or ninety degrees. The result of this mental rotation was the "rotation product." The participants were told that this new, rotated angle was the only thing that mattered now; the original angle was no longer needed. Immediately after, a second striped pattern appeared, and they had to memorize this new one as well. Finally, a number appeared on the screen. If the number was one, it meant the rotated pattern would be tested. If it was two, the second, unrotated pattern would be tested. The participants then had to judge whether a test pattern was rotated clockwise or counter-clockwise relative to the cued item.

Throughout the waiting periods between these steps, the researchers flashed a brief, neutral visual impulse on the screen. This impulse acted like a probe, allowing the researchers to take a snapshot of what was currently active in the brain's working memory. By analyzing the electrical activity that followed this flash, they could decode whether the brain was holding the original pattern, the rotated pattern, or the second pattern.

The results were clear. When the participants were holding the rotated pattern in their minds, the brain signals showed that the original pattern was still there, even though it was no longer needed for the task. This happened even after they had memorized a second, completely different pattern. This finding ruled out the idea that the original image was just a lingering echo of perception or that it was stuck because it had been encoded too deeply. If it were just a lingering echo, it should have faded away once the second pattern was added. If it were stuck due to deep encoding, it should have remained visible even when the rotated pattern was discarded.

The key to the mystery came when the researchers looked at what happened when the participants were told to focus on the second pattern instead of the rotated one. In these moments, the brain signals for the rotated pattern disappeared. Crucially, the signals for the original pattern disappeared at the exact same time. The two were not independent; they were linked. The brain did not keep the original image as a separate backup. Instead, it kept the original image only because it was holding the rotated image. When the rotated image was no longer needed, the original image was dropped along with it.

The researchers also looked at how well the participants performed. They found that on trials where the brain showed a strong signal for the rotated pattern, the participants were more accurate. However, on trials where the brain showed a strong signal for the original pattern, the participants were actually less accurate. This suggests that the presence of the original image was not helping the task; in fact, it seemed to be a slight distraction. The brain was holding onto the original image not because it was useful for the current goal, but because of the way the mental transformation was structured. The rotated image was built upon the original, and the brain maintained them as a connected pair.

This discovery changes how we think about the inner world of the mind. It suggests that when we create a new mental image by transforming an old one, we do not simply replace the old with the new. Instead, the new image remains anchored to its source. The brain seems to prefer keeping these representations linked, perhaps to allow for flexibility. In a real-world situation, if a car disappears from view, we might predict its future path based on its last known position. If the situation changes suddenly, having the original information still available might allow us to quickly recalculate the prediction without starting from scratch. While this study showed that keeping the original image offered no advantage in a simple, isolated task, it hints that this linked structure might be a feature of how our brains handle dynamic, changing environments. The mind does not just store facts; it builds a web of connections where new ideas are held in place by the old ones they grew from.