Why We Dream: REM Since 1953

Explore the discovery of REM sleep and the mysterious dreams—like talking dogs and Alps-opening kitchens—that reveal how our sleeping brain works each night.

Tonight, while you are doing the most ordinary thing in the world—sleeping—your brain may stage a private film festival. It can cast an old teacher, a missing suitcase, a kitchen that opens onto the Alps, and a dog that speaks with perfect authority. By breakfast, most of it is gone, not because the mind has switched off, but because sleep has structure.

Sleep is not one smooth country. It is more like a rail route with repeated stops. Over about ninety minutes, most adults pass through light non-REM sleep, deeper slow-wave sleep, and REM sleep, named for rapid eye movement. That cycle repeats about four or five times a night, and the REM stretches usually grow longer toward morning, when dreams often become most elaborate.

REM sleep was identified in 1953 at the University of Chicago by Eugene Aserinsky and Nathaniel Kleitman, with important follow-up work by William Dement. Aserinsky noticed that sleeping infants’ eyes sometimes darted beneath their lids. When adults were awakened during similar bursts, they often reported vivid dreams. The discovery turned the night from a blank pause into an active landscape that could be measured with electrodes and careful awakenings.

Dreams can happen outside REM too. In non-REM sleep, they are often less cinematic—more like thoughts, fragments, or simple scenes—though exceptions are common. REM dreams, especially late in the night, are more likely to be visual, emotional, and odd in the classic dream way: your apartment is also an airport, and no one there thinks this needs explaining.

How many dreams do we have in one night? There is no perfect counter; a dream has no barcode. But when researchers wake sleepers repeatedly in labs, many report several dreams, especially after REM awakenings. A reasonable estimate is four to six dream episodes, sometimes more. Remembering them depends heavily on timing. Wake during a dream or just after it, and you have a chance. Sleep through the handoff, and the memory tends to dissolve.

Forgetting also has chemistry behind it. During REM, norepinephrine, a brain chemical linked with alertness and memory tagging, is low. The hippocampus and cortex are busy, but they may not file experience the way they do when you are awake and paying attention. A dream can be a brilliant voice memo recorded in an app you forgot to save.

So why dream at all? The evidence is strongest around sleep itself, and more cautious around the dream story. Sleep helps stabilize some memories and connect new learning with older knowledge; studies show benefits for skills, vocabulary, and pattern finding. One hypothesis says dreaming is part of that sorting. Researchers still debate whether the plot is essential or a side effect.

Another hypothesis is threat simulation, associated with the Finnish neuroscientist Antti Revonsuo. Dreams may let the brain rehearse danger in a low-risk virtual world. That does not mean every nightmare about a late train is useful training. Still, threatening themes appear often across cultures, and the idea remains plausible: your sleeping brain may run emergency drills with a very eccentric scriptwriter.

A third idea is emotional housekeeping. REM sleep is linked with strong activity in emotion-related brain regions, while some areas used for cool, logical control are less dominant. Many studies connect sleep with how people process fear, stress, and painful memories. The cautious version is simple: sleep appears to help recalibrate feeling, and dreams may revisit charged material without the full force of waking life.

Lucid dreams are the moments when the dreamer realizes, “I’m dreaming.” This is not just folklore. In laboratories, lucid dreamers have sent prearranged signals with their eyes while still asleep. Most voluntary muscles are inhibited during REM, so a hand signal will not work, but the eyes can move. In the late 1970s and early 1980s, researchers including Keith Hearne and Stephen LaBerge used left-right eye movements to verify lucidity during REM.

The same REM muscle inhibition explains sleep paralysis. The brain normally blocks signals to major muscles so you do not act out dreams. Sometimes awareness returns before the block lifts. You can breathe, and your eyes may open, but you cannot move for seconds or minutes. If dream imagery leaks into waking perception, a shadow in the room or pressure on the chest can feel intensely real.

That sensed “presence” is not proof of an intruder; it is the half-awake brain explaining a body map that feels wrong. Text and clocks are another clue: in dreams, they often change because the world is generated moment by moment, not received from stable light. Dreams are not a universal code, but not mere nonsense either. By morning, the talking dog is gone, and the kitchen no longer opens onto the Alps.