Why do we dream? Here's what research says
Why do we dream? Research keeps offering better answers — from emotional regulation to creative problem-solving. Here's what we know, and what we still don't.
Contents
It’s one of the most fundamental questions in sleep science — and it’s still not fully answered. Why do we spend a quarter of our lives in a state where the brain creates vivid, narrative, emotionally intense experiences that mostly vanish the moment we wake up?
There isn’t one answer. There are several — and they complement each other.
What we know for certain
Sleep research is reasonably confident about a few things when it comes to dreaming:
Most dreaming happens during REM sleep (Rapid Eye Movement), but dream reports come from every sleep stage. REM dreams are the most narrative, emotional, and vivid. They occur across all cultures, across every age group from small children onward, and in many other mammals.
We dream more than we remember. The average person has three to five dream cycles per night but typically remembers only a fraction — usually just the last dream of the night. That means we live through hours of dreamed experience we have no conscious access to.
Dreams aren’t random noise. They’re thematically coherent, emotionally charged, and show clear patterns tied to the dreamer’s life, worries, and relationships. Someone under stress dreams about stress-related themes. A student facing an exam dreams about the exam.
Theories about the function of dreaming
Emotional processing (Cartwright, Walker)
The most strongly documented theory concerns emotional regulation. During REM sleep, the brain processes emotionally charged material — but in a physiological environment fundamentally different from being awake: noradrenaline, the brain’s primary stress hormone, is nearly absent.
Sleep researcher Matthew Walker calls this “overnight therapy”: the brain can relive and reprocess intense, painful, or frightening experiences without the full biological stress response. Over time, the emotional charge in those memories decreases — they remain available as factual information, but lose their sharp emotional edge.
Rosalind Cartwright documented this in a study of recently divorced adults. Those who dreamed about their ex-partner in the first months after the breakup had markedly better psychological outcomes a year later than those who didn’t recall such dreams. The dreams weren’t a sign of failing to process the breakup. They were the processing itself.
This explanation also makes sense of PTSD, where trauma-related nightmares repeat the event without REM sleep’s normal noradrenaline dampening — sustaining the emotional intensity rather than defusing it.
Memory consolidation (Stickgold, Walker)
REM sleep plays a documented role in consolidating procedural memory — motor skills, patterns, and rules. Deep sleep primarily consolidates declarative memory (facts and episodes). Both contribute to memory consolidation, but in different ways.
Studies consistently show that sleep after learning improves memory more than an equal amount of time spent awake. Students who slept after a study session performed 20–30 percent better than those who didn’t, controlling for total time since learning.
The exact role dreams play in this isn’t fully settled — whether dreams actively drive memory consolidation or are simply a byproduct of the processes that do is an open question. But that REM sleep itself is critical is well documented.
Threat simulation (Revonsuo)
Finnish sleep researcher Antti Revonsuo proposed an interesting evolutionary explanation: dreams are a simulation arena where the brain practices handling threats.
Studies of dream content show that negative, threatening, and stressful scenarios are heavily overrepresented compared to actual everyday experience — we dream far more about being chased, endangered, and losing things than about mundane routine. Revonsuo argues this isn’t random: dreams train the brain’s response patterns for crisis situations in a safe, simulated environment.
This also explains why being chased in a dream is one of the most universal dream experiences across cultures.
Activation-synthesis (Hobson and McCarley)
An older, more materialist explanation from 1977: dreams are a byproduct of random brain stimulation during REM sleep, and their narrative content is the brain’s attempt to make sense of that random activity — a post-hoc synthesis of noise.
This theory has since been modified and refined by Hobson himself, but it points to something important: dreams may not have a single function, and some of their content may genuinely be noise that consciousness tries to interpret.
The unconscious speaking (Jung and Freud)
The two foremost thinkers in dream psychology — Sigmund Freud and Carl Jung — held different but related theories about the function of dreams.
Freud believed dreams were the royal road to the unconscious: an arena where suppressed wishes and impulses found symbolic expression while sleep lowered the mind’s defenses. The manifest content of a dream (what you remember) was a distorted version of its latent content (the actual wishes underneath).
Jung believed dreams communicated something from the unconscious too, but that the communication was compensatory rather than repressive: dreams showed you what conscious life overlooks, what you lack, what’s out of balance. Dreams were helpers, not escapes.
Modern neuroscience doesn’t confirm Freud’s specific theories about sexual symbolism or wish fulfillment, but what it does confirm — that the dreaming brain processes emotional and personal material in ways that are meaningful, not random — sits closer to Jung’s broader framework than to a purely noise-based view.
Want to dive deeper into the psychology of dreams? See our guide on what dreams mean and how they’re interpreted.
What dreams do — in summary
The best explanations aren’t mutually exclusive. Dreams likely do several things at once:
They process emotional material and reduce its emotional charge over time. They contribute to memory consolidation, particularly of skills and patterns. They may simulate threatening scenarios that train the brain’s responses. And for some of their content, they may be partly meaningless noise that consciousness interprets.
What’s certain is that they aren’t random. And they aren’t unimportant.
Can you influence what you dream about?
To a limited degree. Intense engagement with a topic right before sleep increases the likelihood of dreaming about it. A method called Targeted Memory Reactivation — where memories are cued right at sleep onset — has, in experiments, increased the frequency of dreams about specific topics. And lucid dreamers, who are aware they’re dreaming, can to some extent steer the dream’s content.
Read more about lucid dreaming and techniques for achieving it.
What we can’t do — yet — is guarantee dream content. The brain decides for itself what’s important enough to process.
No single answer, several real ones
Research doesn’t give a single answer to why we dream, but it offers several complementary explanations:
- Emotional processing: REM sleep is the night’s therapy session, with reduced noradrenaline
- Memory consolidation: REM sleep strengthens skill and pattern memory
- Threat simulation: dreams train the brain’s crisis response
- The brain’s synthesis: some content may be meaning-making from random activation
- Psychological communication: Jung and Freud both pointed to dreams as carriers of unconscious material
Sources and further reading
- American Journal of Psychiatry (Hobson & McCarley, 1977) — “The brain as a dream state generator: an activation-synthesis hypothesis of the dream process”
- Matthew Walker, Why We Sleep (2017)
- Rosalind Cartwright, The Twenty-four Hour Mind (2010)
- Sleep Foundation — How Sleep Works
