The Six Years You Spend Hallucinating

Tonight, if the night goes as most nights do, you will hallucinate for roughly two hours. Your eyes will jerk beneath closed lids as though tracking something across a room. The muscles in your limbs will slacken into a near-total paralysis, a merciful safeguard against acting out whatever your mind invents. And your brain, far from resting, will blaze with an activity almost indistinguishable from waking. It will build a world, populate it, run you through it, and then, in the seconds after you open your eyes, erase nearly all of it.

We do this every night. Across an average lifespan, a person spends something on the order of six years dreaming. That is six years of vivid, self-authored theater staged in the dark, watched by no one, remembered by almost no one. And after more than a century of serious inquiry, backed by electrodes, brain scans, and controlled laboratory sleep, science still cannot say with confidence why any of it happens.

This is not a failure of effort. Dreaming is one of the oldest questions in the study of the mind, and some of the sharpest researchers of the twentieth century built careers trying to answer it. The strange thing is how the answer keeps receding. Every time the tools improved, the mystery seemed to deepen rather than resolve. We learned exactly when we dream and what the sleeping brain looks like while it happens. We simply never learned why.

From prophecy to physiology

For most of recorded history, a dream was not a biological event. It was a message. The ancient Egyptians kept dream books that catalogued omens. The Greeks slept in the temples of Asclepius hoping the god would arrive in the night with a cure. In the Hebrew and Christian traditions, dreams carried divine instruction, warnings, prophecies of famine and rescue. The dream was something that came from outside the sleeper, delivered by a force with intentions.

That framing survived, in one form or another, until the beginning of the twentieth century, when Sigmund Freud turned the dream inward. In The Interpretation of Dreams, published in 1900, Freud argued that dreams were not messages from gods but disguised expressions of the dreamer’s own forbidden wishes.¹ The mind, he proposed, contained desires too threatening to face directly, and a kind of internal censor kept them submerged during waking hours. At night, the censor relaxed but did not vanish, so the wishes surfaced in coded form. A dream’s obvious content, what Freud called the manifest content, was a scrambled cover story for the latent meaning underneath. Interpretation meant working backward through the disguise. “The interpretation of dreams,” he wrote, “is the royal road to a knowledge of the unconscious activities of the mind.”

It was a bold and elegant system, and it dominated the way educated people talked about dreams for half a century. It was also, from the standpoint of experimental science, close to useless. Freud’s theory could explain any dream after the fact, which is another way of saying it could never be proven wrong. If a dream seemed to contradict the theory, that only showed how effective the disguise had been. There was no measurement to take, no prediction to test. Dreams remained, in practice, a matter of storytelling.

Then, in 1953, a graduate student at the University of Chicago made a discovery that gave dreams a physical signature for the first time.

The eyes that gave it away

Eugene Aserinsky was not a promising figure by the usual measures. He was broke, older than most students, and working under Nathaniel Kleitman, a demanding physiologist who had founded the world’s first laboratory devoted entirely to sleep. Aserinsky’s early task was tedious: monitor the eye movements of sleeping subjects, on the assumption that the eyes would slowly roll and then go still as sleep deepened. To watch closely, he wired subjects to a machine that recorded electrical activity, including the movement of the eyes, on a scrolling paper trace.

One of his earliest subjects was his own young son, Armond. Watching the trace one night, Aserinsky saw something that should not have been there. Periodically, long after the boy had fallen asleep, his eyes began to dart rapidly back and forth beneath his closed lids, and the pens tracking his brain activity jumped into a pattern that looked almost like wakefulness.² This was not the slow drift of a settling eye. It was quick, coordinated, active. Aserinsky, worried the equipment was broken, checked it repeatedly. It was not broken. The brain was doing something during sleep that nobody had catalogued.

Aserinsky and Kleitman named the phenomenon rapid eye movement sleep, or REM, and published their findings in the journal Science in 1953.² The crucial follow-up came quickly. When they woke subjects during REM periods, those subjects reported vivid dreams the overwhelming majority of the time. Woken during other phases of sleep, they usually reported nothing, or only vague fragments. For the first time in the history of the question, dreaming had a fingerprint. It was no longer a purely private, unmeasurable event. It happened at particular moments, and those moments could be seen on a machine.

The discovery cracked the field open. Kleitman, already regarded as the father of modern sleep research, now had a tool for catching dreams in the act. Researchers mapped the architecture of the night and found that sleep moved in cycles of roughly ninety minutes, each one ending in a burst of REM that grew longer toward morning. They confirmed that during REM the brain consumed nearly as much energy as it did while awake, sometimes more, even as the body lay paralyzed and unresponsive. The sleeping mind, it turned out, was not a mind switched off. It was a mind running hot on a different track.

But a fingerprint is not a motive. Knowing precisely when dreams occurred, and what the dreaming brain looked like, brought scientists no closer to the question that had haunted the subject since antiquity. Why does any of this happen at all? With the machinery finally in place to study dreaming properly, the theories began to multiply, and they pulled in wildly different directions.

Four theories, no winner

The first major challenge to Freud arrived in 1977, and it was less a revision than a demolition. J. Allan Hobson and Robert McCarley, both at Harvard Medical School, proposed what they called the activation-synthesis hypothesis.³ Their argument began in the brainstem, the primitive structure at the base of the brain. During REM sleep, they said, the brainstem fires off bursts of essentially random electrical signals. These signals have no message, no hidden wish, no meaning. They are noise. The higher brain, the forebrain, receives this chaos and does what it always does: it tries to make sense of it. It weaves the random inputs into a narrative, because narrative is what the waking mind produces reflexively. The bizarre logic of dreams, the sudden scene changes, the impossible geographies, all of it reflected the forebrain scrambling to impose story onto static.

In Hobson’s view, the dream was not a coded message from the depths. It was the brain interpreting itself, an internal act of pattern-finding forced onto meaningless data. This directly inverted Freud. Where Freud saw a censored text full of secret significance, Hobson saw improvisation over noise. The meaning we find in dreams, by this account, is meaning we manufacture on waking, not meaning that was ever encoded.

Hobson’s theory was clean and provocative, but it did not settle anything, because it treated the vivid emotional weight of dreams as an accident. Other researchers found that hard to accept, and they went looking for function instead of noise.

One of the most influential functional accounts came from the sleep scientist Matthew Walker, who described REM sleep as a kind of overnight therapy for the emotional brain.⁴ Walker’s argument rests on a chemical detail. During REM, the brain’s levels of noradrenaline, a stress-related neurotransmitter, drop to their lowest point of the entire sleep-wake cycle. In this uniquely calm chemical bath, the brain replays emotionally charged memories from the day. The idea is that by reactivating a painful or frightening memory in the absence of the stress chemistry that originally accompanied it, the brain slowly strips the emotional charge from the memory while leaving the information intact. You wake up remembering what happened, but the raw sting has been softened. Dreaming, in this reading, is how the mind digests difficult experience, processing emotion overnight so it does not accumulate into damage.

A very different function was proposed by the Finnish cognitive scientist Antti Revonsuo, who looked at what dreams are actually about and noticed something dark.⁵ Dreams, he pointed out, are disproportionately full of threat. We are chased, attacked, falling, trapped, pursued by figures we cannot escape. Surveys of dream content consistently find that negative and threatening scenarios outnumber pleasant ones. Revonsuo argued this was no coincidence. In his threat-simulation theory, dreaming is an evolved rehearsal system. For our ancestors, life was genuinely dangerous, and a brain that could practice escaping predators and enemies in the safety of sleep would produce a body better prepared to escape them in daylight. The tiger you flee tonight trains the reflexes that might save you tomorrow. Dreams, by this logic, are not therapy or noise. They are a training ground built by natural selection.

And then there is the theory that has arguably the strongest experimental support: memory. During sleep, the brain appears to consolidate what it learned while awake, transferring and strengthening the connections that store new skills and facts. A striking demonstration came from the laboratory of Matthew Wilson at MIT. Researchers recorded the brain activity of rats as they ran through mazes, capturing the specific sequences of neurons that fired as the animals learned the route. Then they recorded the same rats while they slept. During sleep, the rats’ brains replayed those exact neural sequences, running the maze again internally, sometimes fast-forwarded, as though rehearsing the day’s lesson in the dark.⁶ The finding suggested that sleep is not a passive break from learning but an active phase of it, a period when the brain locks in what it acquired. We do not just rest in order to remember. In some sense, we dream in order to remember.

Four theories, then, each pointing to a different function: interpret the noise, process the emotion, rehearse the threat, consolidate the memory. Each has evidence. None has defeated the others. And just when the field seemed to be circling toward some pluralist truce, a deeper problem surfaced that undermined the assumption everyone had been building on since 1953.

The crack in the foundation

The entire modern science of dreaming had been erected on Aserinsky’s discovery: dreams live in REM sleep. Wake someone in REM and they dream; that was the fingerprint that made the whole enterprise possible. But over the following decades, a set of stubborn observations began to chip at that equation.

People whose REM sleep is suppressed, whether by certain antidepressant medications or by specific brain injuries, do not stop dreaming. In some cases they report dreams that are as vivid and as narratively rich as ever. Meanwhile, people woken from non-REM sleep, the phases where dreaming was supposed to be sparse or absent, frequently do report dream experiences, sometimes elaborate ones. The neat correspondence between one brain state and one mental event turned out to be leakier than the founding studies suggested.

The neuropsychologist Mark Solms compiled some of the most influential evidence here, studying patients with brain damage. He found that damage to the brainstem regions responsible for generating REM did not necessarily abolish dreaming, while damage to certain forebrain regions, areas involved in motivation and the generation of imagery, could abolish dreaming even when REM sleep itself remained intact.⁷ The implication was unsettling for the whole field. REM and dreaming might not be the same thing. They might be produced by different machinery, correlated most of the time but genuinely separable. If so, then decades of research that treated REM as a proxy for dreaming may have been measuring the wrong thing, or at least an incomplete thing.

This is why the honest answer to why we dream remains, after all this, that nobody really knows. It is not that we have no theories. It is that we have too many, each supported, none conclusive, and the very phenomenon they try to explain resists being pinned to a single mechanism.

The theater that plays to no one

Perhaps the mistake is the search for a single purpose. Biology rarely builds tidy, single-function systems. The mouth eats, speaks, and breathes. The same organ can serve several masters at once, and dreaming may be exactly this kind of overloaded process, doing emotional regulation and memory consolidation and perhaps a little threat rehearsal all in the same overnight window, with the vivid imagery as a byproduct the waking mind insists on treating as a message. It may be that asking what dreams are for is like asking what a river is for. It does many things at once and was designed for none of them.

What remains, underneath the competing theories, is a genuinely strange fact about the shape of a human life. Every night, without instruction or effort, the mind assembles a complete world, with places and people and events and a version of the self moving through them. It commits enormous energy to this construction. And then it discards nearly all of it. Studies suggest we forget something like ninety-five percent of our dreams within minutes of waking, the whole production dissolving before we can carry it into the day.

So the most elaborate theater any of us will ever stage runs, night after night, to an empty house. There is no audience. There is barely a record. The dreamer is both the author and the only witness, and even the witness usually forgets. We spend six years of a single lifetime inside these vanishing worlds, and we cannot say, with any of the confidence science normally demands, what we are doing there or why. Tonight the lights will go out, and the story will start again. It will be vivid, urgent, and entirely convincing while it lasts. And in the morning, as it always has, it will refuse to tell you what it meant.

Sources

1. Freud, Sigmund, The Interpretation of Dreams, Franz Deuticke, 1900. — https://www.gutenberg.org/ebooks/66048
2. Aserinsky, E. and Kleitman, N., Regularly Occurring Periods of Eye Motility, and Concomitant Phenomena, During Sleep, Science, 1953. — https://www.science.org/doi/10.1126/science.118.3062.273
3. Hobson, J. A. and McCarley, R. W., The Brain as a Dream State Generator: An Activation-Synthesis Hypothesis of the Dream Process, American Journal of Psychiatry, 1977. — https://pubmed.ncbi.nlm.nih.gov/21570
4. Walker, Matthew, Why We Sleep: Unlocking the Power of Sleep and Dreams, Scribner, 2017. — https://www.simonandschuster.com/books/Why-We-Sleep/Matthew-Walker/9781501144325
5. Revonsuo, Antti, The Reinterpretation of Dreams: An Evolutionary Hypothesis of the Function of Dreaming, Behavioral and Brain Sciences, 2000. — https://pubmed.ncbi.nlm.nih.gov/11515147
6. Louie, K. and Wilson, M. A., Temporally Structured Replay of Awake Hippocampal Ensemble Activity During Rapid Eye Movement Sleep, Neuron, 2001. — https://www.cell.com/neuron/fulltext/S0896-6273(01)00186-6
7. Solms, Mark, Dreaming and REM Sleep Are Controlled by Different Brain Mechanisms, Behavioral and Brain Sciences, 2000. — https://pubmed.ncbi.nlm.nih.gov/11515143

Related reading

• The Longest Anyone Has Ever Stayed Awake
• The Shiver That Music Leaves Behind

More from the Body edition →