(Un)predictable false memories: the role of sensorimotor language prediction in episodic memory
This study demonstrates that while semantically predictable words increase false memory regardless of motor similarity, encoding motor-similar but unpredictable action-object sentences through action observation enhances both accurate recall and metacognitive monitoring, highlighting the distinct roles of semantic and sensorimotor prediction in episodic memory.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Human brains are not passive recorders of the world; they are active forecasters. When we listen to a story or read a sentence, our minds do not simply wait for the next word to arrive. Instead, we constantly guess what is coming next, using the context of what we have just heard to predict the future. If someone says, "The chef chopped the vegetables with a sharp," our brain instantly fills in the missing word "knife" before the speaker even finishes the sentence. This ability to anticipate is a fundamental part of how we understand language, but it comes with a hidden cost. Because our brains are so eager to complete the pattern, they sometimes remember words that were never actually spoken. We might recall hearing "knife" even if the speaker said something entirely different, or if they stopped before finishing the thought. This phenomenon, known as false memory, suggests that our predictions can overwrite reality in our minds.
While scientists have long known that these linguistic predictions can create false memories, a new study from researchers at Sapienza University of Rome asks a deeper question: does the physical body play a role in this process? The researchers wanted to know if the part of our brain that controls movement and senses physical sensations—the sensorimotor system—is involved when we predict the next word in a sentence. They suspected that when we hear an action described, our brain might automatically simulate the physical movement required to perform it. If this simulation is strong enough, it could either help us remember what actually happened or, conversely, trick us into remembering a movement that never took place. Understanding this link is crucial because it reveals how our physical experiences shape our memories, potentially explaining why we sometimes remember events with vivid detail that never occurred.
To investigate this, the team designed an experiment that separated the meaning of a sentence from the physical movement it described. They created three types of sentences involving everyday actions. The first type was highly predictable, such as "stirring coffee with a spoon." In this case, both the meaning and the physical movement are obvious. The second type was totally unpredictable, like "stirring coffee with pliers." Here, the object makes no sense, and the movement is awkward and unusual. The third type was the most innovative: "motor-similar unpredictable" sentences, such as "stirring coffee with a thermometer." In this scenario, the object is strange for the action (you don't usually stir coffee with a thermometer), but the physical way you would hold and move the thermometer is very similar to how you hold a spoon. This allowed the researchers to test whether the similarity of the hand movement alone was enough to trigger a false memory, or if the meaning of the object was the dominant factor.
One hundred and eighteen young adults participated in the study, divided into three groups to test how different ways of learning affect memory. The first group watched videos of a person performing the actions described in the sentences. The second group was asked to imagine performing the actions themselves, feeling the movement in their minds without actually moving. The third group simply repeated the sentences silently in their heads, a method known as mental rehearsal. After listening to the sentences and engaging in their assigned activity, the participants waited twenty-four hours. They then returned for a memory test where they heard a mix of the sentences they had seen before and new ones they had never heard. They had to decide if they remembered the sentence and how confident they were in that answer.
The results confirmed that the brain's habit of predicting the future is a powerful force in memory. Participants were significantly more likely to falsely remember the predictable object, such as the spoon, even when they had actually seen or heard a different object. This happened across all three groups, whether they watched the action, imagined it, or just rehearsed the words. However, the study found that the physical similarity of the movement did not create the same level of false memory. When participants saw or imagined stirring coffee with a thermometer, they did not falsely remember the spoon as often as they did when the sentence was fully predictable. This suggests that while the body's movement system is active during language comprehension, the meaning of the words—the semantic content—is the primary driver of these memory errors. The brain is more likely to be tricked by what makes logical sense than by what feels physically similar.
Interestingly, the way the information was encoded made a difference in how well people remembered the actual events. Those who watched the videos of the actions performed were better at correctly identifying the sentences they had seen, particularly when the sentences involved the strange, motor-similar objects like the thermometer. This group also showed higher confidence in their correct memories. In contrast, those who simply rehearsed the words in their heads made more false alarms and were less confident in their answers. The study suggests that watching an action creates a rich, detailed memory trace that helps the brain distinguish between what happened and what it expected to happen. Imagining the action also helped, but not as much as actually seeing it.
The researchers also looked at how confident people felt about their memories. Participants were generally more confident when they remembered unpredictable events correctly, perhaps because the unusual nature of the event made it stand out in their minds. However, when they made a mistake and falsely remembered a predictable object, their confidence did not necessarily drop, indicating that the brain can be very sure of a memory that is actually wrong. This disconnect between confidence and accuracy highlights the complexity of human memory. The study concludes that while our bodies and movements are involved in how we process language, the meaning of the words is the strongest force in creating false memories. The brain prioritizes what makes sense in the world over what feels physically familiar, leading us to remember a spoon even when we saw a thermometer, simply because a spoon is the only thing that makes sense for stirring coffee.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.