A yellow flannel blanket, a sofa, a pair of hands, soft daylight from a window. That is all the neurobiologist Olga Ivashkina has of her grandfather, who died when she was about one year old. She believes he is holding her in the scene. She has tried to check. She questioned her mother and grandmother, and they could not settle it, because they are not even sure the blanket was in that room at that moment.
So is it a memory, or a story she heard so often that it turned into a picture? She cannot tell. By her own account she should not have it at all: memories from before the age of one are almost always wiped. If you carry a scene from your first years, it may be partly assembled from what the grown-ups around you said afterwards. Ivashkina can say how that happens, and where the real fragment might sit.
The usual name for the wipe is infantile amnesia. A baby is born with an unfinished brain. The frontal lobes keep developing until about 23 to 25, and the other networks are rebuilt through childhood, the hippocampal ones included. New neurons are made there fastest in those early years. Mouse pups go through the same thing: their hippocampal connections change, and they forget what happened to them young.
Ivashkina's bet is that several processes run at once, and one of them is pruning. Connections get cut. Some of a memory is lost outright. Some of it is still there, but the road out is gone.
To see how, picture an engram: the physical trace of a memory, a group of neurons that together encode it. Yesterday's breakfast is an active engram, and you can pull it up on demand. A silent engram is different. Some of its neurons have lost their links to each other. Others have lost their links to the outside. In animals, researchers can switch such an engram on artificially, and the memory comes out. In people that is impossible.
Studies on young animals give this picture some support. In January 2026, a paper in PLOS Biology by Erika Stewart's group reported that a drug damping the activity of microglia, the brain's resident immune cells, during a particular window after birth prevented infantile amnesia for a fear memory. Forgetting, in that experiment, was something the brain did actively. A June 2026 preprint on bioRxiv, led by Joo Hee Yang, reports that cells which carried an infant memory in one brain region, the IL, were recruited again into adult memory traces.
What is left of childhood, then, looks like this:
What stays of childhood memories, it seems, is such scraps, fragments of silent engrams. Sometimes something catches them in a half-random way.
И, судя по всему, от детских воспоминаний остаются такие куски, обрывки молчащих энграмм. Иногда что-то может их зацепить каким-то полуслучайным образом.
— Olga Ivashkina, Osnova, 06:08
Then the family gets involved. A relative tells you something. You replay it in your head, then in conversation, and each time the trace is written again, and again. Now you have a memory. Is it real? Ivashkina's answer is no, not cleanly and not exactly, because we "record nothing" like a camera. Something has certainly been hung on it, perhaps your mother's view of that day, perhaps another relative's.
How much of the scene was ever there, the blanket, the light, the hands, she leaves as a separate question.
