The fact that wouldn't come
You studied it. You're sure of it. You sat at the same desk, the same lamp throwing the same yellow light, and the term was right there under your pen. Now you're in a bright exam hall that smells of floor polish, and the word has dissolved. It surfaces an hour later, in the corridor, unbidden and useless.
This is not a failure of effort, and it isn't quite a failure of memory either. The information is intact. What's missing is the doorway you built to reach it. Psychologists call this context-dependent memory, and once you understand how it works, you can stop building doorways that only open in one room.
The divers who proved it
In 1975, Duncan Godden and Alan Baddeley ran one of the most quietly elegant experiments in memory research. They asked deep-sea divers to learn lists of words in two settings: on dry land, and fifteen feet underwater. Then they tested recall in both environments.
The pattern was clean. Words learned underwater were remembered best underwater. Words learned on land were remembered best on land. Divers who switched environments between learning and testing recalled noticeably less—around a third fewer words. Nothing about the words changed. Only the surroundings did.
The surroundings, it turns out, get encoded alongside the material. When you learn, your brain doesn't store a clean, isolated fact. It stores the fact wrapped in incidental detail: the light, the sounds, the temperature, the faint background of where you were. Those details become retrieval cues—little hooks that help you pull the memory back up. When the hooks are present, recall is easier. When they're gone, the memory is still there, but you're reaching for it without a handle.
Encoding specificity: the principle underneath
The broader idea comes from Endel Tulving and Donald Thomson, who proposed the encoding specificity principle: a cue helps you remember only to the degree that it was part of the original encoding. Whatever was present when the memory formed becomes part of the address you use to find it later.
This is why a smell can ambush you with a memory from childhood more vividly than any deliberate attempt to recall it. The smell was encoded with the scene, and it remains an unusually direct route back. It's also why staring at a blank wall trying to force a name rarely works, but retracing your steps—returning to where you first heard it—often does. You're not searching the memory; you're rebuilding the context that points to it.
The context doesn't even have to be external. State-dependent memory shows the same effect for your internal condition. Material learned while caffeinated tends to come back more easily when you're caffeinated again. Mood works similarly: a low mood preferentially surfaces memories formed in a low mood. Your physiological and emotional state at encoding becomes part of the retrieval address, just as the room does.
The trap of the perfect study spot
Here's where the practical lesson takes a turn most people don't expect.
The intuitive response to context-dependent memory is to match your contexts—study in conditions identical to the test. Sometimes that helps, and athletes and musicians lean on it deliberately, rehearsing under conditions that mimic the real performance. But for most learning, you don't control the test environment, and you can't predict where you'll need to recall what you know. The word you're learning today might be needed in a meeting, a conversation, an exam hall, a country you haven't visited yet.
If you always study in the same place—the same desk, the same chair, the same playlist—you bind the knowledge tightly to that single context. You build one beautiful doorway. And then life asks you to walk through a wall.
Why varying your context makes memory portable
The counterintuitive fix comes from a 1978 study by Steven Smith, Arthur Glenberg, and Robert Bjork. Students learned a list of words either in one room, or split across two different rooms. When tested later in a neutral third room, the students who had studied in two different places recalled substantially more.
The reason is worth sitting with. When you learn the same material in varied settings, the memory stops depending on any single set of cues. Instead of one fragile doorway, you've built several, and you've quietly told your brain that this knowledge is not about the room—it's about the thing itself. The memory becomes less context-bound and more context-free. It travels.
This is one of the family of "desirable difficulties" Bjork has spent a career mapping: conditions that make studying feel slightly harder or messier in the moment, but produce memory that's sturdier and more flexible later. Varying your context is a small, almost effortless one. You don't have to study harder. You just have to study in more than one place.
What to actually do with this
You don't need a diving suit or a second apartment. A few deliberately loose habits do the work:
Study the same material in genuinely different settings across your sessions—the kitchen table one day, a library another, a park bench, the bus. Each new environment adds another retrieval route to the same knowledge.
Let the incidental conditions vary too. Different times of day, sometimes with background noise and sometimes in silence, standing or sitting. Each variation loosens the grip of any single cue.
Resist the urge to perfect a ritual. The candle, the exact playlist, the one chair—these feel like they're helping you concentrate, and they may be. But they're also weaving themselves into the memory as cues you won't have on the day it counts. A ritual that's identical every time is a context you're quietly depending on.
And notice the internal states too. If you only ever review when wired on coffee, you may find the knowledge harder to reach on a flat, tired afternoon. Spreading your practice across different states of mind and body makes recall less hostage to any one of them.
None of this replaces the heavy lifting of memory—retrieving the answer from scratch, spacing your reviews so they catch you on the edge of forgetting. Context is the quieter layer underneath. It decides not whether the memory exists, but whether you can reach it when the room is wrong.
The point is portability
What you're really after, when you learn something, is knowledge you can carry. Not a fact chained to a desk, but an idea that comes when called, wherever you happen to be standing. Context-dependent memory is the reason that doesn't happen automatically—and the reason it can, if you stop studying in exactly the same way every time.
This is part of why a flashcard app earns its keep precisely by being everywhere. Recall is built to be fast and fully offline, so a deck lives in your pocket and a review fits the two minutes on a train, the wait for coffee, the quiet end of a lunch break—not just the one desk where you sit down to "really study." Its FSRS spaced-repetition scheduling decides when each card should return; the variety of where you answer it is yours to supply, and it costs you nothing but the willingness to review in more than one place. Bring your Anki and Quizlet decks over, then let them follow you around. The knowledge you build in many small, ordinary contexts is the knowledge that shows up when you need it. If that's the kind of memory you're after, you can start at https://recall.lumenlabs.works.