Why Do We Dream?

Every night, your brain becomes paralyzed from the neck down and starts generating experiences that feel as real as waking life — often more vivid, almost always bizarre — and then erases most of them before you can remember them. If you designed this system from scratch and tried to explain it to someone, they'd probably tell you it sounds like a bug.

We've been trying to understand dreams for as long as we've been keeping records. Freud said they were wish fulfillment and coded messages from the unconscious. Neuroscientists mostly disagree, but they haven't fully converged on what dreams actually *are* — and the more we learn about sleep, the stranger the phenomenon gets.

## What's Actually Happening During Dreams?

Dream sleep has a name: **REM** — rapid eye movement. It's one of four sleep stages your brain cycles through roughly every 90 minutes. During REM, your brain becomes extremely active — in some ways *more* active than when you're awake. Your heart rate and breathing become irregular. And a neural system called **REM atonia** paralyzes your voluntary muscles almost completely. You're physically pinned to your bed while your brain runs what looks, on an EEG, like a waking simulation.

The paralysis isn't incidental — it's probably protective. People with REM sleep behavior disorder, where atonia fails, physically act out their dreams: punching, kicking, shouting. What we call "sleep paralysis" — the terrifying experience of waking up unable to move, sometimes with hallucinations — is REM atonia persisting briefly into wakefulness.

Here's the weird part: the visual and auditory cortices fire during dreams in patterns very similar to real perception. When you dream you're watching something happen, your visual cortex is processing that imagery in a way that overlaps measurably with actually seeing it. Dreams aren't just stories your conscious mind tells — they're apparently using the same hardware as real experience.

## What Are the Theories?

Three main hypotheses compete, and none of them fully wins.

**Memory consolidation**: This is probably the most evidence-backed view. REM sleep is associated with the consolidation of procedural and emotional memories — integrating new experiences into existing memory structures, running simulations that test connections between things you've learned. Studies have found that sleeping after learning a skill improves performance more than staying awake for the same period, and that depriving people of REM specifically disrupts certain types of memory consolidation.

**Threat simulation**: A Finnish researcher named Antti Revonsuo proposed that dreams function as a safe environment for rehearsing responses to threats. Evolutionarily, being able to "practice" evading predators or navigating social confrontations during sleep could be adaptive. Supporting this: dreams disproportionately feature threats, chasing, falling, and danger compared to normal waking experience.

**Emotional processing**: Some evidence suggests REM sleep processes emotional memories differently from non-REM sleep, particularly around fear responses. PTSD is associated with disturbed REM sleep and intrusive nightmares — and some theories suggest normal REM processing strips the emotional charge from memories while preserving their content.

None of these theories fully explains why dreams have *narrative structure*, why they contain bizarre combinations of real and impossible elements, why we sometimes become aware we're dreaming (lucid dreaming), or why the experience is as *vivid* as it is.

> 🔬 **Quick experiment:** Before sleep tonight, spend 15 minutes learning a physical skill — juggling, a guitar chord, a yo-yo trick, anything with motor coordination. Test yourself before bed, then test yourself again the moment you wake up tomorrow (before practicing again). Most people show measurable improvement overnight, especially if they got good REM sleep. The skill consolidation happens during sleep, not during the practice itself. If you can track your sleep stages with a fitness tracker, compare your improvement on nights with longer REM to nights with shorter REM.

## Why It Matters Practically

Even without a complete theory, we know a few things that have real implications.

Alcohol suppresses REM sleep. Even moderate drinking before bed, while it may help you fall asleep faster, reduces the REM sleep you get in the first half of the night. Some common antidepressants (SSRIs, SNRIs) also suppress REM dramatically — which is one reason some researchers have become interested in whether REM disruption has downstream effects on emotional processing.

Chronic REM deprivation produces measurable effects: impaired emotional regulation, reduced ability to read facial expressions, reduced pain tolerance.

The honest answer to "why do we dream" is: probably several things at once, and we don't fully know the proportions yet. The brain during REM is doing something computationally serious — it's not just running screensavers.

// COMMENTS

ON THIS PAGE