Context-dependent memory deserves a closer look.

Change the room and you change what you can remember.

Divers learned lists of words in two places: on a sunny beach, and twenty feet underwater. Later they were tested in both places. The result was beautifully symmetrical. Words learned on the beach were recalled best on the beach. Words learned underwater were recalled best underwater. Learn and test in mismatched places and recall dropped sharply. The knowledge was identical. Only the surroundings changed, and the surroundings turned out to be part of the memory.

This is context-dependent memory, demonstrated in the famous 1975 experiment by Duncan Godden and Alan Baddeley, and it reveals something fundamental about how memories are stored. A memory is not a file saved in isolation. It is a pattern that includes the background: the room, the sounds, the smells, the lighting, the bodily state. At recall, the current context acts as a set of retrieval cues, silently pointing at the stored pattern. Match the context and the cues line up. Mismatch it and some of the pointers go missing.

The effect is everywhere once you look. Students who study in the same quiet room where they will take the exam outperform students who study in varied noisy places, on average, because the exam room reinstates the study cues. A song from years ago can unlock a flood of associated memories, because the music was part of the original pattern. Walk into a room and forget why you came: the doorway effect, where crossing a boundary flushes the context and the intention with it. Police know that witnesses remember more at the scene than at the station, which is why cognitive interviewing tries to mentally reinstate the original context.

The practical lessons cut both ways. If you want to remember something in a specific place, study it there, or vividly imagine the place while studying. But if you want knowledge that travels, that works anywhere, study it in varied contexts, because each new context adds independent retrieval routes and frees the memory from any single one. Musicians rehearse in different rooms. Athletes train under different conditions. The goal is memories that do not need their birthplace.

The principle behind the divers has a grander name: encoding specificity, proposed by Endel Tulving and Donald Thomson in 1973. A retrieval cue works to the extent that it was present at encoding and reinstated at retrieval. The beach was a cue. The water was a cue. The theory predicts, correctly, that even cues you never noticed consciously, background music, ambient smells, the particular chair, become part of the memory’s address. You cannot choose to encode cleanly. The world always comes along.

Police interviewers turned the principle into a technique. The cognitive interview, developed by Fisher and Geiselman, asks witnesses to mentally reinstate the context of the event, the weather, the sounds, their feelings, before describing what happened. Reinstated context unlocks details that direct questioning misses, because the details were filed under that context. It is Godden and Baddeley’s divers, run in reverse: instead of returning to the water, you bring the water back in your mind.

There is a poetic edge to the finding. We think of memory as being about the thing remembered, the words, the facts. Godden and Baddeley’s divers showed that memory is also about the beach and the water, the light and the pressure, the world the thing happened in. You never remember a thing alone. You remember it with its surroundings attached, and the surroundings hold part of the key.