No One Remembers Their First Two Years: Science Explains Why Those Early Memories Persist

No photos, no smell, no faces. Your first two years of life disappear from your memory, as if they never happened. Yet a team of researchers at Yale University has just demonstrated that these memories do exist somewhere in your brain: it is access that is missing, not the trace itself.

Freud had a name for it, science took a century to understand why

This phenomenon, dubbed infantile amnesia by Sigmund Freud, has intrigued scientists for more than a century. Two competing hypotheses had stood against each other until now: either the baby’s brain is too immature to form memories, or it does form them but later loses them. This seemingly simple question proved difficult to settle given the technical challenges of studying a infant brain. How do you ask a six-month-old if they remember an image seen the day before? They do not yet speak.

It was precisely this technical roadblock that Nick Turk-Browne’s team, a psychology professor at Yale and director of the Wu Tsai Institute, managed to overcome. “The hallmark of these kinds of memories, what we call episodic memories, is that you can describe them to others, but that is impossible with infants who can’t speak yet,” he explains. His lab developed a method to keep babies awake and still inside a functional MRI scanner (fMRI), using pacifiers, blankets, and stuffed animals to comfort them. In itself, an achievement: until now, most neuroimaging studies on infants were conducted during their sleep, for lack of a better option.

Twelve-month-old babies filmed as they memorized

Published in March 2025 in the journal Science, the experiment involved showing infants aged 4 to 25 months a series of novel photographs while recording activity in their hippocampus, the brain region considered the memory center. Later, some of these images were shown again, paired with new images, and researchers measured how long each baby looked at the already seen image: a classic index of recognition in nonverbal toddlers.

By comparing brain activity during the successful formation of memories versus the forgotten images, the researchers confirmed that the hippocampus is active in memory encoding from the earliest ages, and this was true for 11 of 13 infants over one year old, but not for those under one year—a milestone crossed around the first birthday. The researchers also found that babies who performed best on memory tests showed greater hippocampal activity, a direct link between the strength of brain activation and the quality of the memory.

The result overturns a dominant hypothesis that had persisted for decades. Researchers had thought until now that we did not retain these experiences because the hippocampus, still developing into adolescence, was simply incapable of encoding memories during our early years. But this new research shows that is not the case. As Turk-Browne summarizes, “What we can conclude precisely from our study is that infants have the capacity to encode episodic memories in the hippocampus from about age one.” 

The trace exists, but the key to accessing it has disappeared

This shifts the entire problem. These findings suggest that infantile amnesia, the inability to recall early childhood, could be explained more by retrieval failures than by an inability to form memories, with those early memories persisting yet remaining inaccessible. The baby brain records. It’s the key to reopen the missing drawer.

Hints in this direction had already appeared in animals. Recent studies on rodents, monitoring brain activity, showed that engrams, those patterns of cells that store memories, do form in the infant hippocampus but become inaccessible over time, though they can be artificially reactivated with a technique using light to stimulate neurons. In mice, one can literally reawaken a childhood memory using implanted optical fibers in the brain. In humans, we are not there yet, but the suspected mechanism is the same: dormant, locked traces rather than erased data.

One question remains for which the research does not yet have a firm answer: why does this door close? The dominant theory held that, with the hippocampus remaining immature until adolescence, infantile amnesia arose from the hippocampus’s inability to encode experiences in early life. But we now know encoding works from around one year old. The real culprit would thus lie with the circuits connecting the hippocampus to the rest of the brain, still developing, and lacking the maturity needed to organize long-term retrieval of these memories. The hippocampus itself doubles in size during the first two years of life, before gradually weaving the connections that enable durable episodic memory storage.

What this changes for understanding the brains of very young children

There is something almost vertigo-inducing in this finding: your baby brain, unable to speak or walk properly, was already able to memorize a face, a room, a specific toy. By around the age of one, children become extraordinary learners, acquiring language, walking, object recognition, understanding social connections, and yet we remember none of these experiences—a striking mismatch between this incredible plasticity and this learning capacity, summarizes Turk-Browne. This paradox, long accepted as a biological fate, is now becoming a real research project: the Yale team is now trying to understand precisely what happens to these childhood memories in the months and years after their formation, before they fade from our personal story.

Source : news.yale.edu