Why Do We Dream?

The 1953 breakthrough

For most of history, scientists assumed the sleeping brain simply shut down. That changed in September 1953, when University of Chicago graduate student Eugene Aserinsky, tracking eye movements during sleep, noticed his subjects' eyes darting rapidly beneath closed lids several times a night. Woken during these episodes, people reported vivid dreams far more often than when roused from quiet sleep. Aserinsky and his advisor, Nathaniel Kleitman, had found REM sleep -- and with it, the first real evidence that dreaming has a distinct, measurable home in the brain.

Two rival theories

By the late 1970s, Harvard psychiatrists Hobson and McCarley proposed a blunt alternative to Freud: during REM sleep, the brainstem fires off essentially random signals, and the cortex -- the brain's storyteller -- scrambles to weave that noise into something coherent. A dream, on this view, is meaning imposed after the fact, not planted beforehand. Finnish neuroscientist Antti Revonsuo later argued the opposite: dreaming evolved as a threat-rehearsal system, a nightly fire drill for danger. Both theories have real evidence behind them, and neither has fully won.

The memory connection

More recent work points to a third role: emotional processing. Sleep researcher Matthew Walker and colleagues have found that REM is the only brain state where noradrenaline, a stress-linked chemical, drops close to zero while the brain's memory and emotion circuits stay active -- a window where difficult memories might get reprocessed without their original sting attached. It's a compelling idea, though not a settled one: exactly how much of memory processing depends on REM specifically, versus deep non-REM sleep, is still genuinely debated among sleep scientists.

The takeaway

Here's the honest answer: nobody knows for certain, and a 2025 review in the Journal of Sleep Research still calls it "the dreaming conundrum." One popular myth is worth retiring, though -- dreaming isn't exclusive to REM sleep. Wake someone from ordinary non-REM sleep and ask what was on their mind, and they'll often describe something dreamlike too, just fainter and harder to hold onto. Even rats seem to do it: recorded in mazes, then again while asleep, their brains replay the same neural patterns, as if reliving the day.