Your first dream of the night arrives earlier than most people think. Roughly 60 to 90 minutes after you fall asleep, your brain enters REM sleep, the phase where your eyes flick around under closed lids and most vivid dreaming happens. That first period may last a few minutes. Later ones run much longer, which is why the dream you still have at breakfast is usually the one that was going when the alarm went off.

That part is settled. Very little else is.

Ask a psychoanalyst, a Jungian, a Gestalt therapist, a content-analysis researcher and a sleep scientist why we dream, and you get five answers. The tidy move is to call them five layers of one truth. A previous draft of this piece did exactly that. It was too generous. Two of the five make claims that cannot both be right.

Start with Freud, because everyone else is arguing with him

Sigmund Freud's The Interpretation of Dreams went on sale in November 1899 with 1900 on the title page, which is why the field cites it as 1900. The first run of 600 copies took about eight years to sell.

His claim was that every dream carries two contents: the manifest content, the version you remember, and under it a latent content, a wish the mind will not look at directly. Between them sits a censor, and Freud named two of its tricks. Condensation packs several ideas into one image. Displacement lifts the emotional charge off the thing that matters and parks it on something harmless. A dream is a compromise. The wish surfaces, disguised just enough that you can keep sleeping.

Freud gets a bad rap now and some of it is earned. He wrote that the secret wish must be proved afresh in every case, then found it everywhere he looked. Almost nobody accepts the model as written today. But he was the first to argue at length that dream content is manufactured rather than random, and that working on it tells you something. Every theory below replies to him.

The same staircase drawn twice, with a terracotta circle over its lowest step and a violet circle over its highest
One staircase, read twice. The picture is kind to both men, who stopped speaking in 1913.

Jung's counterweight

Carl Jung is often described as Freud's student. He was not. When the two men first met in Vienna in March 1907, Jung was already a senior physician at the Burghölzli hospital in Zurich, trained under Eugen Bleuler. They collaborated for about six years, Freud treated him as the heir to psychoanalysis, and the friendship broke in 1913.

Jung's dream theory turns on compensation. A dream supplies whatever your waking attitude is short of, pulling a one-sided personality back toward balance. His own example is people who think too well of themselves. They dream of flying, and of falling. The dream fills the gap and, he thought, warns them where they are heading.

He is also the source of the archetypes, figures he believed recur across cultures because they come from a shared inheritance: the collective unconscious. The shadow holds the parts of yourself you have refused to own. The wise old man stands in for the Self. None of that machinery is testable the way a recall rate is, but the instruction still works. Treat an odd dream as a comment on your waking position, then work out what position.

The Gestalt shortcut

Fritz Perls, who co-founded Gestalt therapy with Laura Perls and Paul Goodman, threw interpretation out. In the dream work he published in 1969, you do not analyse the locked door. You speak as the locked door, in the present tense. The assumption is that every person and object in a dream is some part of you that you are not owning. It is the least scientific method here and the fastest to try on a dream that has a mood but no plot.

What people actually dream about

Here the subject stops being philosophy and starts being counting.

In 1966 Calvin Hall and Robert Van de Castle published norms built from 991 dream reports: 100 men, 100 women, five dreams each. The results are unglamorous and they have held up. Forty-five percent of dreams contain at least one act of aggression, 40% a friendly act, 35% some misfortune. Where a dreamer names an emotion at all, it is a negative one 80% of the time, and the commonest single emotion is apprehension.

G. William Domhoff has spent decades on this kind of coding, and his conclusion is the continuity hypothesis: dream content runs continuous with waking concerns and preoccupations. The evidence is long dream series matched against a dreamer's documented activities and relationships, not people reporting what they worry about. The cleanest case belongs to a student the researchers call Kenneth, who recorded 2,022 dreams across his first three years of college. His parents show up in 23.9% of them, a small group of friends in 53.9%. Take eighteen keywords from his waking life and at least one appears in 76.7% of all 2,022 reports.

Does dreaming do a job?

This is where the field splits, and it splits hard.

Antti Revonsuo argued in 2000 that dreaming is a threat simulator, evolved to rehearse spotting danger and getting away from it. He had content to back it up. In 592 dreams collected at home from 52 students, 66.4% contained at least one threatening event, averaging 1.2 threats per dream, and children living through war produce more of them.

Susan Malcolm-Smith and Mark Solms ran the count again in 2004 and got a much thinner result. Physical threats turned up in 21% of reports, realistically life-threatening ones in under 10%, and realistic escapes from a life-threatening event in under 3%. If dreams are practice for escaping danger, the practice almost never includes the escape. Both camps are still publishing.

Antonio Zadra and Robert Stickgold propose a different job. Their model, NEXTUP (Network Exploration to Understand Possibilities), treats dreaming as sleep-dependent memory work: the brain hunts for weak, unexplored links between things you already know and strengthens the useful ones.

Domhoff thinks all of them are chasing something that is not there.

Based on current evidence, it is most likely that dreams are the accidental by-product of two great evolutionary adaptations, sleep and consciousness.

— G. William Domhoff, A New Neurocognitive Theory of Dreams (2001)

That is no friendly amendment to Revonsuo or NEXTUP. It says the hunt for a function is the wrong hunt, and it comes from the man with the largest pile of coded dreams in the field. Anyone who says the science has converged has not read the argument.

Inside the sleeping brain

Eugene Aserinsky and Nathaniel Kleitman identified REM sleep in 1953. It is the state most linked to vivid dreaming, though not exclusively. Pooling 35 studies, Tore Nielsen found that waking people out of REM produced a dream report 81.9% of the time. Waking them out of non-REM sleep, the quieter stages filling most of the night, still produced one 43% of the time. Dreaming is not a REM-only event.

In 1996 Pierre Maquet's group scanned seven sleeping people and mapped what runs hot in REM: both amygdalae, the anterior cingulate cortex, the brainstem. They also found a vast area of the dorsolateral prefrontal cortex switched off. That region handles planning and self-monitoring while you are awake, and its absence is the usual explanation for why a dream can feel urgent and make no sense. The visual system splits the same way: the areas that recognise faces and places light up while the primary visual cortex, the part that would receive actual light, stays dark. You see without eyes.

The pop version of this overshoots. "The rational brain shuts down" is too broad, because the medial prefrontal cortex is highly active in REM. And between 1940 and 1975, hundreds of psychiatric patients had prefrontal pathways surgically cut. Dreaming stopped completely in 70 to 90% of them. Damage the front of the brain and you do not get wilder dreams. You get none.

The famous neural theory here, activation-synthesis, came from J. Allan Hobson and Robert McCarley in 1977: the cortex is handed essentially random signals from the brainstem and improvises a story around them. It is still the version most people quote, and Hobson left it behind. By 2009 he was arguing close to the reverse, that REM sleep is a protoconscious state running a virtual model of a world, and that the model is of real use in building waking consciousness.

Diagram of a sleeping head in profile on a pillow, with a terracotta and a violet region marked inside the skull
Two regions marked, nothing glowing. In REM the amygdala runs hot and a large stretch of dorsolateral prefrontal cortex goes dark.

Where we come down

Two of these five are science and three are reading practices. We lean on the two that measure things. Freud, Jung and Perls measure nothing, which does not make them worthless: they are reliable ways of getting a dreamer to say more about a dream than "it was weird." No dream holds one correct answer waiting to be unlocked, which is why nothing we build will hand you a verdict on last night. On whether dreaming is for anything at all, we find Domhoff's by-product argument the strongest of the three, and we would not bet much on it. None of it is diagnosis, either: a dream tied to trauma, or one wrecking your sleep, belongs with a therapist or a sleep clinic.

Questions We Get Asked

Do dreams predict the future?

No. Between 17 and 38% of people report at least one dream they consider precognitive, which is a fact about memory and coincidence. Ask people to write the prediction down before the event instead of noticing the match afterwards, and the effect collapses. Domhoff rejects precognition outright as incompatible with established physical science.

Which theory has been proven?

None. Neuroscience stands on the firmest ground because it measures what it claims, and even there the measurements describe mechanism, not meaning.

Why do I keep dreaming the same feeling in different places?

Common, and more informative than any one night's imagery. The continuity findings above are why: a feeling that keeps coming back is usually attached to something unfinished while you are awake. Our article on recurring dreams follows the pattern further.

What to do with it tonight

All five perspectives need the same raw material, and it is fragile. Detail goes within minutes of waking. So the useful move is having something on record to apply a framework to. Two sentences and the feeling, caught before you sit up, will serve you better later than a careful account rebuilt at lunchtime. That is our position from doing this daily, not a research finding.

Then let it pile up. One dream read alone is nearly impossible to interpret with confidence. Thirty of them are, in our experience, a pattern you can start to read, and what repeats across them is what Hall, Van de Castle and Domhoff were counting all along. Dreamograph exists to make that pile easy to keep and to search back through, the one part of dream work that improves with time instead of fading.

Sources

  • Freud, S. (1900). The Interpretation of Dreams (Die Traumdeutung). Leipzig & Vienna: Franz Deuticke. Published 4 November 1899 with an 1900 title page. A. A. Brill's English translation is free at Project Gutenberg.
  • Jung, C. G. (1964). Approaching the unconscious. In C. G. Jung, M.-L. von Franz, J. L. Henderson, J. Jacobi & A. Jaffé, Man and His Symbols. London: Aldus Books / New York: Doubleday.
  • Perls, F. S. (1969). Gestalt Therapy Verbatim. Lafayette, CA: Real People Press. Foundational text: Perls, F., Hefferline, R., & Goodman, P. (1951). Gestalt Therapy: Excitement and Growth in the Human Personality.
  • Hall, C. S., & Van de Castle, R. L. (1966). The Content Analysis of Dreams. New York: Appleton-Century-Crofts. Normative tables at dreams.ucsc.edu/Norms.
  • Domhoff, G. W. (2001). A new neurocognitive theory of dreams. Dreaming, 11(1), 13–33. Free full text.
  • Domhoff, G. W., & Schneider, A. (2008). Studying dream content using the archive and search engine on DreamBank.net. Consciousness and Cognition, 17(4), 1238–1247. Free PDF.
  • Domhoff, G. W. (2000). Dreams and Parapsychology. Free full text.
  • Revonsuo, A. (2000). The reinterpretation of dreams: An evolutionary hypothesis of the function of dreaming. Behavioral and Brain Sciences, 23(6), 877–901.
  • Revonsuo, A., & Valli, K. (2000). Dreaming and consciousness: Testing the threat simulation theory of the function of dreaming. Psyche, 6(8).
  • Valli, K., Revonsuo, A., Pälkäs, O., Ismail, K. H., Ali, K. J., & Punamäki, R.-L. (2005). The threat simulation theory of the evolutionary function of dreaming: Evidence from dreams of traumatized children. Consciousness and Cognition, 14(1), 188–218.
  • Valli, K., Revonsuo, A., Pälkäs, O., & Punamäki, R.-L. (2006). The effect of trauma on dream content: A field study of Palestinian children. Dreaming, 16(2), 63–87.
  • Malcolm-Smith, S., & Solms, M. (2004). Incidence of threat in dreams: A response to Revonsuo's threat simulation theory. Dreaming, 14(4), 220–229.
  • Zadra, A., & Stickgold, R. (2021). When Brains Dream: Exploring the Science and Mystery of Sleep. New York: W. W. Norton & Company.
  • Aserinsky, E., & Kleitman, N. (1953). Regularly occurring periods of eye motility, and concomitant phenomena, during sleep. Science, 118(3062), 273–274.
  • Nielsen, T. A. (2000). A review of mentation in REM and NREM sleep: "Covert" REM sleep as a possible reconciliation of two opposing models. Behavioral and Brain Sciences, 23(6), 851–866. Free PDF.
  • Maquet, P., Péters, J.-M., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A., & Franck, G. (1996). Functional neuroanatomy of human rapid-eye-movement sleep and dreaming. Nature, 383(6596), 163–166.
  • Braun, A. R., et al. (1998). Dissociated pattern of activity in visual cortices and their projections during human rapid eye movement sleep. Science, 279(5347), 91–95.
  • Nir, Y., & Tononi, G. (2010). Dreaming and the brain: From phenomenology to neurophysiology. Trends in Cognitive Sciences, 14(2), 88–100. Free full text.
  • Hobson, J. A., & McCarley, R. W. (1977). The brain as a dream-state generator: An activation-synthesis hypothesis of the dream process. American Journal of Psychiatry, 134(12), 1335–1348.
  • Hobson, J. A. (2009). REM sleep and dreaming: Towards a theory of protoconsciousness. Nature Reviews Neuroscience, 10(11), 803–813.