In the future, scientists may learn to predict exactly what you will remember.

Do you remember your first kiss? And the death of your grandmother? There is a possibility that you remember - and this is because emotionally colored memories are the core of the story of your personal life. Some rare moments are very tense and stand out against the background of a routine existence, consisting of sleep, food and work. But everyday life is filled with experience that can have personal emotional significance - for example, a dispute with someone or a compliment received.



Some people are able to describe emotional memories to a certain limit, even after a long time, while memories of more ordinary events disappear. But why exactly this is happening, and how we keep our memories, is still unclear. In a new paper published in the journal Psychological Review, we describe a new computer model that can help explain this fact.



To study the effect of emotions on memory in a laboratory setting, scientists usually show subjects films, stories, and images that evoke an emotional response. They can then ask volunteers to describe what they remember. However, in emotional reactions, people are very different from each other. Therefore, researchers are trying to use materials that have a more or less permanent effect on people, positive or negative. For example, the image of conducting a medical procedure on a child seems unpleasant for most of us.



Such studies have given us good evidence in favor of the fact that memory really better remembers materials that have an emotional response.



Over the years, many ideas have been accumulated explaining why this is so. Someone claims that people simply pay more attention to experience that is not indifferent to them - that is, they assign high priority to one memory and discard the rest. According to this theory, it is the attention shown during the initial coding of information that helps people to recall these memories later.



But this is not a complete explanation. It is clear that the events that happened immediately before and immediately after the incident also matter. It is quite easy to recall a more or less interesting experience, if it was followed by a calm period, after which a particularly exciting event occurred. Similarly, the situation in which a memory check occurs affects what memories come to mind. It is easier to remember winning a school competition if you return to this school for a graduate meeting.

Math memory

In our recent work, we gathered all these ideas in an attempt to give a more structured explanation of emotional memories. We started by studying the steps of processing information in the human brain, on which we encode, store and retrieve neutral information. We relied on an existing and proven theory of memory retrieval, especially clear and precise, as it expresses all its statements with mathematical equations.



According to this theory, each of your memories is associated with a mental state that you had at that moment - that is, with a mental context. For example, if you were in a hurry one morning, then your memory of what you ate for breakfast will be influenced by this more general mental context. The memory of breakfast will also be associated with the memory of what you read in the newspaper. Such mental states change with each subsequent experience, but they can be used later to extract memories from memory. For example, if someone asks you what you ate for breakfast in the morning, it will be easier for you if you recall the feeling of haste or the process of reading an article in a newspaper.







Then we asked how emotions can correct each other at the stages of memory formation, using the discoveries made in experiments regarding emotional memory, and recorded the potential effect in a mathematical form. Specifically, we hypothesized that the connection between experience and its mental context is stronger when the experience was emotional. Finally, we fed the equations into a computer program that simulated how a person learns and remembers certain materials.



If our ideas about memory were true, the program should have more accurately “remembered” those things that people remembered better. We found that it was. However, our model reflected not only situations in which emotions reinforce memories, but also those situations in which this does not occur.



For example, my previous experience has shown that although people's memory works better with emotional material in the form of a mixture of emotional and neutral images, this does not work if people show several emotional images in a row or several non-emotional images in a row, for example, the painting process the door. In each such experiment, people's memory abilities are similar. This is still a mystery. However, the model also produced this counterintuitive result, which gave us confidence that our mathematical code could be on the right track.



Many amazing conclusions can be drawn from our work. Apparently, the mechanism underlying well-preserved emotional memories is not as unique as previously thought - both emotional and neutral experience are processed in approximately the same way. However, emotions affect certain processing steps, and differ in such things as the degree of connection of different things and the connection of things and the context of their coding.



These small changes lead to important global changes in the whole process of memorization. Perhaps, since it is so important for us to remember emotional experience, evolution has adjusted many aspects of the process of remembering so that it is sensitive to it - for example, to the threat posed by a predator, or to the possibility of finding food.



Since we describe the influence of emotions in mathematical equations, our work may allow scientists to someday predict what kind of experience a person will have in his memory. The ultimate goal will be attempts to understand this at the level of individuals. So far, there is a lot of uncertainty in the assumptions about what is happening in the head of a particular person, especially about how strongly the various experiences experienced are related, and what attention people pay to them.



But, when we collect more data about these intermediate steps, the predictions of our model may be able to more accurately reproduce the sequence of extracting memories from specific people. Of course, we may be wrong, which may make us reconsider our model. After all, science goes forward through the creation of hypotheses and their verification on empirically obtained data.