Why we don’t remember being a baby

In the first years of life, an explosive learning process unfolds. Yet, paradoxically, we rarely remember any fragment of our existence before the age of three, and complete memories usually don’t form until around six. Some people claim to recall their first steps or being cradled in a crib by their mother, but these are almost certainly false memories — reconstructions shaped by photos or stories from those who were present. As memory researchers have demonstrated time and again, our capacity for recall is less like a recording device capturing reality and more like a narrative we construct to shape our identity and better navigate life.

At the beginning of the 20th century, Sigmund Freud coined the term infantile amnesia to describe this lack of early memories, attributing it to the repression of thoughts related to childhood sexuality or aggression — ideas deemed unacceptable to a civilized mind. Since then, various theories have emerged to explain this phenomenon. Some, like Freud, suggest it results from later reformatting, though not necessarily due to cultural repression. Others argue that the infant brain is simply not capable of forming memories.

This second hypothesis is based on the fact that the different regions of the hippocampus are connected by what is known as the trisynaptic circuit, a neuronal pathway that is still immature in a child’s brain. This immaturity would prevent the hippocampus from encoding episodic information, the ability that later enables us to recall personal experiences in specific places and times. Furthermore, this idea seemed to be reinforced by observations that children’s memory capacity is as limited as that of adults with hippocampal damage-induced amnesia.

However, an article published on Thursday in Science challenges this hypothesis, providing evidence that children do form memories but later, when they grow up, cannot retrieve them.

One common explanation for childhood amnesia is that the hippocampus — the brain region responsible for memory storage — has not fully developed before adolescence, which prevents memories from being encoded. But a team led by researchers at Yale University has found that this is not the case. Using innovative techniques to study memory formation in infants, the researchers showed images of faces, objects, and scenes to children between four months and two years old while monitoring their brain activity with fMRI. Then, while continuing to monitor neuronal activity, they showed them the same objects again along with others they had never seen before.

When infants had previously observed a stimulus, they were expected to look at it more closely when it was presented again. In the experiment, when shown two images — one novel and one familiar — if the infant focused their gaze more on the latter, the researchers concluded that they recognized it as familiar.

With this hypothesis in mind, the team investigated whether hippocampal activity was linked to infants’ memory. The results showed that the greater the activity in the hippocampus when an infant first saw an image, the longer they gazed at it when it was shown to them later, suggesting that the information had been stored in that brain region.

But where do these memories go if they are formed but never retrieved? That’s the question being asked by researchers like Nick Turk-Browne, the study’s lead author. One possibility is that they are stored in ephemeral memory. Another is that they reside in our brains, although they can no longer be accessed.

To find out, Turk-Browne and his colleagues are testing children’s memories with videos taken from their point of view. What the preliminary results show is that memories formed during early childhood are there, but they fade before the age of six.

Santiago Canals, a researcher at the Alicante Institute of Neurosciences, acknowledges the expertise of the scientists who published the study in Science in conducting experiments with children. However, he believes the study does not resolve the questions that remained unanswered before its publication.

“There was already evidence that children store memories,” he says. “There is a conditioned response in babies aged one or two years old, who react in a given context, such as when you put them in a crib and they smile, or when they recognize familiar objects or imitate things their parents do, and imitate them later. There was little doubt that children learn and then forget.”

The fate of these early memories remains an open question: do they gradually fade away, or does something happen that prevents their retrieval, even though they remain in our adult brains?

Nick Turk-Browne acknowledges that the answers to these questions are still hypothetical. “An inability to retrieve memories could suggest that the hippocampus isn’t being sent the correct information to access childhood memories, possibly due to other changes in how the brain processes experiences,” he notes. “For example, as we acquire language, concepts, and skills, the same event, such as going to the grandparents' house, might be processed very differently in an older child or adult than in a baby. This different processing could mean that the hippocampus isn’t receiving the appropriate ‘search terms’ to find the memory as it was stored, based on the child’s experience at the time.”

In the future, he speculates, perhaps we could hope to recover childhood memories by focusing on memories with content processed similarly in infancy and later life — such as smells or faces — while avoiding those that are processed differently, like spoken or written language, which is foreign to infants.

Although the possibility of recovering early memories remains a distant challenge in humans, studies with mice have made significant progress. In 2023, a team from Trinity College Dublin published an experiment in Science Advances in which they used invasive techniques — ethically unfeasible in humans — to label the specific neurons that encoded a maze memory in mice during childhood. The mice initially learned how to escape the maze, but by the time they reached adulthood weeks later, they had forgotten the solution. However, by stimulating the neurons that originally stored the memory using light, the researchers were able to reactivate the lost memory, allowing the mice to recall how to exit the maze.

“Although this type of direct hippocampal stimulation to reactivate memories won’t be possible in humans in the short term, there may be other ways [to achieve it], if those memories still exist,” says Turk-Browne.

The study published on Thursday does not determine how long childhood memories persist beyond the minutes measured by the researchers. However, it is believed that these memories could last until around the age of six or seven. After that, the demands of adulthood and the learning processes that move us beyond childhood — such as written language and abstract thought — make these early memories largely inaccessible.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition