Why do we dream? The honest answer is that nobody knows for certain, and it is worth being upfront about that before surveying the theories — because this is a topic where confident explanations vastly outrun the evidence. What we have are several plausible, partially-supported accounts that are not mutually exclusive.
What We Know For Sure
- Dreaming occurs mainly during REM sleep, though it also happens in non-REM stages
- Most people dream several times a night but forget most of it
- REM periods get longer toward morning, which is why you remember dreams from just before waking
- During REM, the brain is highly active while the body’s voluntary muscles are temporarily paralysed
- We dream roughly every 90 minutes after falling asleep
That muscle paralysis during REM is worth noting — it is what stops you physically acting out dreams, and when the mechanism fails, people develop REM sleep behaviour disorder.
The Leading Theories
| Theory | Core claim | Support | Weakness |
|---|---|---|---|
| Memory consolidation | Dreams process and file the day’s experiences | Sleep deprivation impairs memory; REM increases after learning | Does not explain why dreams are so bizarre |
| Threat simulation | Dreams rehearse responses to danger | Negative content is common; children dream of threats often | Many dreams are not threatening at all |
| Emotional regulation | REM processes emotional memory while stress chemistry is quiet | REM deprivation worsens mood regulation | Mechanism is not fully mapped |
| Activation-synthesis | Dreams are the cortex making sense of random brainstem signals | Explains the incoherence well | Does not explain apparent meaning or structure |
| Continuity hypothesis | Dreams reflect waking concerns | Dream content correlates with daily life | Descriptive more than explanatory |
These overlap considerably. Emotional regulation and memory consolidation in particular are probably describing related processes from different angles.
The Emotional Processing Account
This is currently among the better-supported ideas, and it has a striking detail behind it. During REM sleep, activity in the brain’s noradrenergic system — the circuitry underlying stress and the fight-or-flight response — drops to near zero.
So the brain appears to revisit emotionally charged memories while the chemistry of alarm is switched off. The proposal is that this lets you process difficult experiences without re-triggering the full stress response. It is a neat explanation for why a good night’s sleep makes problems feel more manageable, and it is consistent with why sleep disruption hits mood so hard.
What Dreams Are Not
Worth separating from the evidence:
- Not prophecy. No controlled evidence supports precognitive dreaming. The apparent hits are selection bias — you remember the one that matched and forget thousands that did not.
- Not a symbolic code. Dream symbol dictionaries have no empirical basis. Freud’s ideas were historically influential but are not supported by modern evidence, and most of the “universal symbols” are cultural rather than innate.
- Not necessary in every form. Some people report never dreaming and are perfectly healthy — they almost certainly dream and simply do not recall it.
Why You Forget Almost All of It
Dream recall is poor mainly because the neurochemistry of REM does not favour memory formation. Levels of acetylcholine are high while norepinephrine is very low, and that combination is not conducive to encoding durable memories.
Practically, this means recall depends almost entirely on waking up during or immediately after a dream. If you wake later, it is gone. That is why dreams remembered from the middle of the night feel so vivid and vanish within minutes.
Keeping a notebook beside the bed genuinely improves recall — not because it captures more dreams, but because the act of attending to them, and writing immediately on waking, catches them before they degrade.
Nightmares and Recurring Dreams
None of this settles why do we dream, but it narrows the field considerably.
Nightmares are common in children and become less frequent with age. Frequent, distressing nightmares in adults are worth taking seriously — they are associated with stress and poor sleep quality, and there is legitimate treatment (imagery rehearsal therapy) that works well.
Recurring dreams usually track unresolved waking concerns. In the continuity framing, they persist because the underlying concern persists.
The Honest Summary
Dreaming almost certainly does something useful, because it is metabolically expensive and evolutionarily ancient. The most defensible statement is that REM sleep contributes to memory and emotional processing, and that dreams are the conscious experience of that work happening.
Beyond that, the specific functions remain genuinely open. Anyone telling you they know why we dream is ahead of the evidence.
Related Reading
- How does memory work? — the consolidation story
- Do we only use 10% of our brain? — the myth that will not die
- What is hibernation? — sleep taken to the extreme
2 thoughts on “Why Do We Dream? What the Evidence Actually Supports”