Why Do We Dream?

You wake up convinced you were late for a train departing from your childhood kitchen. For a moment, it felt like a place you could visit. Then the scene falls apart. What produced it, and was your brain doing anything useful? Those sound like one question, but they’re two: what causes dreams, and what, if anything, dreams are for.

Let’s start with what we can observe. A dream is private; researchers can’t simply look inside one. If people describe dreams after waking, how do we know what was happening during sleep rather than what they’ve reconstructed afterward?

That’s a real limit. Researchers can record sleep activity and wake people at particular points to ask what they were experiencing. Dream reports are especially common and often vivid after REM sleep, but dreaming also occurs in non-REM sleep. The report is evidence of an experience, not a recording of it. Recall can fade quickly, and putting an experience into words may change it. So the strongest claims concern patterns across sleep stages and reports, not a perfect map of any one dream.

What, then, builds the story? I’ve heard that dreams are just the brain making sense of random signals. That seems plausible when a kitchen becomes a station, but it also seems too easy. Dreams sometimes draw on things we’ve been worried about all day.

Exactly. The sleeping brain remains active, and its activity changes across sleep stages. Sensations, memories, emotions, and recent concerns can all contribute to dream content. The older idea that the brain simply stitches a story around internally generated signals captures part of this: dreams needn’t begin with a deliberate plot. But ‘just random’ misses their links to waking life. Research connects dreaming with activity in systems involved in perception, memory, and emotion. Those connections help explain possible ingredients; they don’t tell us why a particular scene appears.

And ingredients aren’t a purpose. Sleep supports learning and memory, so it’s tempting to say dreams consolidate memories. But couldn’t the brain be doing useful work during sleep while the dream is merely what that work feels like, or even an unrelated by-product?

Yes. There is substantial evidence that sleep contributes to memory and learning. There’s also evidence that experiences can reappear in altered form in dreams. Neither establishes that consciously experiencing a dream is necessary for those benefits. Other proposals say dreams help us process emotions, rehearse threats, or explore social situations. They fit some features of dreams, particularly their emotional intensity, but showing a fit is easier than showing a function. Dreams can also be mundane, repetitive, or distressing without any clear payoff.

That distinction matters for nightmares. If someone has a frightening dream, it would be glib to claim their brain was helpfully practising danger. But it would also be too strong to say dreams never help people think or feel differently afterward. Is that where personal meaning comes in?

It does, with a caution. A dream may reveal what’s on your mind in the ordinary sense that your concerns can shape it. Talking or writing about it might help you reflect. But there’s no reliable universal code in which, say, a train or a kitchen always means one thing. Nor does a dream predict an event because its imagery feels striking. We should separate what a person finds meaningful from a scientific claim about what caused the dream or what dreams evolved to accomplish.

So here’s where we land. We know dreaming occurs across sleep stages, and dream content can incorporate memories, emotions, and recent experience. We don’t know that dreams have one single function—or that the experience itself is required for sleep’s benefits. Worth watching are studies that connect detailed dream reports with brain activity and test whether changing dreams, without simply disrupting sleep, changes what happens when people wake. Until then, a dream can be meaningful to you without being a message your brain has definitively decoded.