Ask someone where they were when they heard a piece of huge news, and most people answer instantly, in detail, with total confidence — the room, who else was there, what they were holding. Ask the same person what they had for lunch two Tuesdays ago and you'll get nothing. Both are memories. Only one felt worth keeping at full resolution, and the part of your brain that made that call is a small, curled structure called the hippocampus.
A librarian, not a filing cabinet
The hippocampus doesn't store memories the way a filing cabinet stores paper — it doesn't hold the memory itself for the long term. Its job is closer to a librarian's: when an experience is new and worth keeping, the hippocampus binds together the scattered pieces of it — the sound, the place, the emotion, the faces — into one retrievable record, and gradually hands that record off to the cortex for long-term storage. Damage the hippocampus and a person can still recall childhood in full; what breaks is the ability to lay down new long-term memories at all, because the librarian that files them is gone, even though the shelves it used to file them to are intact.
Emotion is the hippocampus's fast-track stamp. A memory tagged with a strong emotional charge — fear, surprise, grief — gets consolidated more aggressively than a neutral one, courtesy of a neighbouring structure, the amygdala, leaning on the hippocampus while the memory is still forming. That's the actual mechanism behind "I'll never forget that day." It isn't that the day was recorded more completely. It's that the recording got flagged, at the moment it was made, as worth the extra effort.
Vivid and accurate are not the same claim
This is where the story gets uncomfortable: the flag that makes a memory feel unshakeably clear says nothing about whether the memory is correct. Flashbulb memories — the vivid, confident kind tied to shocking news — are studied precisely because people trust them more than ordinary memories, and yet they drift and distort with time at roughly the same rate ordinary memories do. Confidence and accuracy are graded on separate scales inside the same brain, and the gap between them is exactly where false memories and eyewitness misidentification live.
What we're still unsure about
How reconsolidation works — the process by which recalling an old memory can quietly rewrite it, each time it's retrieved — is still an active area of research, and popular science tends to overstate how settled the mechanism is. What's well established is that memory is reconstructive rather than a plain a playback; what's still being worked out is exactly how much a given act of remembering changes what gets remembered next time.
This sits inside The Neuroscience of Memory (Hippocampus), one of seven topics in Neuroscience, one of four domains in Psychology, one of seventeen subjects the app can quiz you on.