Down the hotel stairs, round the corner, and there it is: a convenience-store fridge with rice balls lined up in rows. Olga Ivashkina can still see which ones sat where. She ate that breakfast on a trip to Japan, and she says she won't forget it.
Ivashkina is a neurobiologist who studies memory. Ask her what she ate two days ago and the answer comes out of a different drawer. "Probably a sandwich and an egg," she says. Roughly.
Roughly is the point. You can't pull a specific ordinary morning out of your head, and the reason is not that you were careless. The brain stops keeping the morning and keeps a type of morning instead. Nothing in there is a recording; what you hold of last month is partly built.
If you're not a super-mnemonist, you won't tell me what you ate for breakfast a month ago on a Saturday with a new moon. But you can tell me what you usually eat for breakfast.
А если вы не супер-мнемонник, то вы не скажете, что вы ели на завтрак месяц назад в субботу с новолунием. Но вы можете сказать, что вы вообще обычно едите на завтрак.
— Olga Ivashkina, Osnova, 15:09
The machinery sits in the hippocampus, a structure shaped like a seahorse. It is a fold of cortex: the skull can't be blown up like a balloon, so over evolution the brain kept tucking itself in to fit more tissue. The hippocampus is a hub, not a vault. A memory is a network of neurons spread across the brain, with the emotional parts here and the sounds of a melody there. Some memories form without the hippocampus at all.
The physical trace of a memory, the engram, is also not a still photograph. Picture a crowd seen from above where everyone in a red shirt takes hold of a neighbour's hand. That chain is the memory. Next day somebody changes shirts, somebody else puts one on, a stranger passes a ribbon across a gap. The chain keeps going and keeps changing.
Over months, the hippocampus takes part in re-indexing these traces, and the memory leans less on it and more on the cortex. The memories get less exact and more systematic. How far this goes, Ivashkina says plainly, nobody knows.
What comes out is a type with edges. She will not have borscht for breakfast. She won't travel to the other end of Moscow for a meal. Another person might eat every breakfast in a new café, and each café will wash out, but "I breakfast in new places because it's fun" stays. The individual mornings go. The habit of the person stays.
Some mornings survive. A new place, a new feeling, a meal that hits you "right in the heart" (Ivashkina's words) is kept as itself. That is why the Japanese rice balls are still lined up in her head.
Then there is Solomon Shereshevsky, the Soviet mnemonist whom the psychologist Alexander Luria wrote about in a short book on a vast memory. He remembered strings of digits and invented words, and his own life stayed vivid as well. For him the ordinary episodes did not blur. Yet his brain was the same as everyone's, Ivashkina points out; he had no giant head. For the rest of us the default is blurring, and only the very bright or very significant episodes stay sharp.
Researchers are looking at the replay itself. A paper by Nicholas Hindy in the Journal of Cognitive Neuroscience in August 2026 reports that when the hippocampus reactivates a memory, the replay builds an abstract, general version while weakening the details of the individual episode.
So what happens when you do try to recall last Tuesday's breakfast? Ivashkina's answer is that you may be recalling something that did not happen on Tuesday at all. You can't retrieve what was on the plate, but you can retrieve what could have been there.
