Skip to main content
GARGLOO

DREAMS

Why Do We Dream?

Why do we dream? Science has several theories, from memory to threat simulation, but no consensus. Here is what the evidence actually supports.

Last updated: 2026-11-15 · 6 min read

First, what dreaming is

Every healthy human brain dreams. Over a normal night, sleep runs in cycles of about 90 minutes, and most adults go through four to six of these cycles. Dreams happen in every stage of sleep, but they are most common and most intense during REM sleep, the stage when the eyes move rapidly, the body is temporarily still, and brain activity rises to levels close to waking.

Add up the dreaming across a night and it comes to roughly two hours. Most of that happens in the second half of the night, because REM periods get longer as the night goes on: the first is brief, the last before waking is the longest and usually the most vivid. Dreams in REM tend to be emotional, strange, and story-like, while dreams in other stages are usually simpler and closer to ordinary thoughts.

During REM sleep, brain scans show strong activity in the areas tied to emotion, memory, and vision, including the amygdala and the hippocampus, while the front of the brain, the part involved in logical thinking and self-control, is quieter. That combination is a good match for what dreams feel like: vivid and emotional, but rarely logical. The question science has not settled is why the brain does this at all.

Some people also experience lucid dreams, in which they realise they are dreaming while the dream is still happening. Lucid dreaming is documented in sleep research, and some people learn to notice it more often, but it is uncommon, it is not a skill everyone can train reliably, and it does not change what dreams are for.

Theory 1: dreams help us remember

The most studied explanation is memory consolidation: the idea that dreaming helps the brain sort the day's experiences and move important ones into long-term storage. Sleep itself is known to matter for memory, and research has linked dreaming, especially during REM sleep, to strengthening memory and recall.

The evidence is suggestive rather than decisive. People who learn a new skill and then sleep tend to perform better than people who stay awake, and dreams sometimes replay fragments of the day, though usually mixed with older memories and invented details. What has never been shown is that the dreaming itself, rather than sleep in general, does the consolidating.

A related idea, sometimes called mental housekeeping, says dreaming is the brain clearing away partial, wrong, or unnecessary information, like tidying a desk at the end of the day. It is an appealing picture, but it is hard to test, and it remains a metaphor rather than a measured result.

Theory 2: dreams help us feel

A second leading explanation is emotional regulation: dreams give the brain a safe place to replay strong feelings, fear, embarrassment, grief, joy, without the body's full stress response. Studies have found that more REM sleep is linked to less reactivity to negative memories the next day, which fits the idea that dreaming helps take the sting out of difficult experiences.

This theory also helps explain why dreams are so emotional, and why stressful periods often bring more intense or more frequent dreams. But the same problem applies as with memory: the link is a correlation. Nobody has proven that the dream, rather than the sleep around it, does the emotional work.

A third explanation, the threat simulation theory proposed by researcher Antti Revonsuo, goes further and says dreaming is an ancient rehearsal system: the brain practices facing dangers, such as being chased or attacked, in a safe simulation, so the body is better prepared in real life. It fits the fact that threatening dreams are common across cultures, but critics note that most dreams are not threatening at all, and that the theory is difficult to test in a way that could prove it wrong.

The sceptical view: maybe dreams are a by-product

Not every scientist thinks dreams have a purpose. The activation-synthesis theory, proposed by Allan Hobson and Robert McCarley in 1977, argues that dreams are the brain's attempt to make sense of random bursts of activity during REM sleep. In this view, the sleeping brain fires off signals, the cortex stitches them into a story as best it can, and the result feels meaningful even though it started as noise.

The continuity hypothesis offers a quieter version of the same scepticism: dreams mostly continue waking life, drawing on the same people, places, and worries, which suggests they reflect what the brain is already doing rather than serving a separate function of their own.

These views are not just contrarian. They explain two things the purpose theories struggle with: why dreams are so often bizarre and incoherent, and why people can lose the ability to dream after certain brain injuries yet still sleep and remember normally. If dreaming were essential to memory, losing it should cause obvious damage, and in the cases studied, it does not.

Why there is still no consensus, and what dreams are not

The honest state of the science is that there is no consensus. Memory, emotion, threat rehearsal, and the by-product view each have evidence behind them and problems in front of them, and sleep researchers continue to disagree. Part of the difficulty is practical: dreams are private, they fade within minutes of waking, partly because levels of the brain chemical norepinephrine are low during REM sleep, and there is no way to watch a dream directly.

That uncertainty is worth saying plainly, because most dream content on the internet sells the opposite: fixed meanings for every symbol, as if a dream about teeth or water always meant the same thing. There is no scientific evidence that dreams predict the future, and no dream dictionary has ever been validated by research. A dream can be meaningful to the person who had it without being a code with one correct answer.

What science can say with confidence is smaller but still remarkable: you dream for about two hours every night, mostly in REM sleep; dreaming is normal and healthy; and the brain is doing something real while you dream, even if researchers cannot yet agree on what that something is for.

Frequently asked questions

Do we dream every night?

Almost certainly yes. Everyone dreams, for roughly two hours a night across four to six sleep cycles, but most dreams are forgotten within minutes of waking. Not remembering a dream does not mean you did not have one.

Why can't I remember my dreams?

Dreams fade fast because the brain chemical norepinephrine, which helps form memories, is at low levels during REM sleep. Waking during or just after a dream, and writing it down at once, makes remembering much more likely.

Do animals dream?

Scientists observe REM-like sleep, with rapid eye movement and twitching, in many mammals, including dogs and cats, and in birds. Whether animals experience anything like a human dream cannot be known directly, since they cannot describe it.

What are nightmares?

Nightmares are vivid, frightening dreams that wake the sleeper, usually during REM sleep. Occasional nightmares are normal, especially during stressful periods. Frequent nightmares that disturb sleep are worth discussing with a doctor, since they can be linked to stress, medicines, or sleep disorders.

Sources

Why Do We Dream? | GARGLOO