Why Do We Dream?

By Neuroskeptic | April 7, 2010 12:40 pm

A few months ago, I asked Why Do We Sleep?

That post was about sleep researcher Jerry Siegel, who argues that sleep evolved as a state of “adaptive inactivity”. According to this idea, animals sleep because otherwise we’d always be active, and constant activity is a waste of energy. Sleeping for a proportion of the time conserves calories, and also keeps us safe from nocturnal predators etc.

Siegel’s theory in what we might call minimalist. That’s in contrast to other hypotheses which claim that sleep serves some kind of vital restorative biological function, or that it’s important for memory formation, or whatever. It’s a hotly debated topic.

But Siegel wasn’t the first sleep minimalist. J. Allan Hobson and Robert McCarley created a storm in 1977 with The Brain As A Dream State Generator; I read somewhere that it provoked more letters to the Editor in the American Journal of Psychiatry than any other paper in that journal.

Hobson and McCarley’s article was so controversial because they argued that dreams are essentially side-effects of brain activation. This was a direct attack on the Freudian view that we dream as a result of our subconscious desires, and that dreams have hidden meanings. Freudian psychoanalysis was incredibly influential in American psychiatry in the 1970s.

Freud believed that dreams exist to fulfil our fantasies, often though not always sexual ones. We dream about what we’d like to do – except we don’t dream about it directly, because we find much of our desires shameful, so our minds disguise the wishes behind layers of metaphor etc. “Steep inclines, ladders and stairs, and going up or down them, are symbolic representations of the sexual act…” Interpreting the symbolism of dreams can therefore shed light on the depths of the mind.

Hobson and McCarley argued that during REM sleep, our brains are active in a similar way to when we are awake; many of the systems responsible for alertness are switched on, unlike during deep, dreamless, non-REM sleep. But of course during REM there is no sensory input (our eyes are closed), and also, we are paralysed: an inhibitory pathway blocks the spinal cord, preventing us from moving, except for our eyes – hence why it’s Rapid Eye Movement sleep.

Dreams are simply a result of the “awake-like” forebrain – the “higher” perceptual, cognitive and emotional areas – trying to make sense of the input that it’s receiving as a result of waves of activation arising from the brainstem. A dream is the forebrain’s “best guess” at making a meaningful story out of the assortment of sensations (mostly visual) and concepts activated by these periodic waves. There’s no attempt to disguise the shameful parts; the bizarreness of dreams simply reflects the fact that the input is pretty much random.

Hobson and McCarley proposed a complex physiological model in which the activation is driven by the giant cells of the pontine tegmentum. These cells fire in bursts according to a genetically hard-wired rhythm of excitation and inhibition.

The details of this model are rather less important than the fact that it reduces dreaming to a neurological side effect. This doesn’t mean that the REM state has no function; maybe it does, but whatever it is, the subjective experience of dreams serves no purpose.

A lot has changed since 1977, but Hobson seems to have stuck by the basic tenets of this theory. A good recent review came out in Nature Neuroscience last year, REM sleep and dreaming. In this paper Hobson proposes that the function of REM sleep is to act as a kind of training system for the developing brain.

The internally-generated signals that arise from the brainstem (now called PGO waves) during REM help the forebrain to learn how to process information. This explains why we spend more time in REM early in life; newborns have much more REM than adults; in the womb, we are in REM almost all the time. However, these are not dreams per se because children don’t start reporting experiencing dreams until about the age of 5.

Protoconscious REM sleep could therefore provide a virtual world model, complete with an emergent imaginary agent (the protoself) that moves (via fixed action patterns) through a fictive space (the internally engendered environment) and experiences strong emotion as it does so.

This is a fascinating hypothesis, although very difficult to test, and it begs the question of how useful “training” based on random, meaningless input is.

While Hobson’s theory is minimalist in that it reduces dreams, at any rate in adulthood, to the status of a by-product, it doesn’t leave them uninteresting. Freudian dream re-interpretation is probably ruled out (“That train represents your penis and that cat was your mother”, etc.), but if dreams are our brains processing random noise, then they still provide an insight into how our brains process information. Dreams are our brains working away on their own, with the real world temporarily removed.

Of course most dreams are not going to give up life-changing insights. A few months back I had a dream which was essentially a scene-for-scene replay of the horror movie Cloverfield. It was a good dream, scarier than the movie itself, because I didn’t know it was a movie. But I think all it tells me is that I was paying attention when I watched Cloverfield.

On the other hand, I have had several dreams that have made me realize important things about myself and my situation at the time. By paying attention to your dreams, you can work out how you really think, and feel, about things, what your preconceptions and preoccupations are. Sometimes.

ResearchBlogging.orgHobson JA, & McCarley RW (1977). The brain as a dream state generator: an activation-synthesis hypothesis of the dream process. The American journal of psychiatry, 134 (12), 1335-48 PMID: 21570

Hobson, J. (2009). REM sleep and dreaming: towards a theory of protoconsciousness Nature Reviews Neuroscience, 10 (11), 803-813 DOI: 10.1038/nrn2716

CATEGORIZED UNDER: freud, papers
  • http://www.blogger.com/profile/05398721312156704738 passionlessDrone

    Hi Neuroskeptic –

    Thanks for the post. I saw a Nova Science Now a year or so ago about sleep.

    There was a very interesting piece where they used a maze with specific signposts, recorded a rodents brainwave pattern as it ran the maze, and then found identical brainwave patterns when the rodent slept. They took this to mean it was 'learning' during sleep.


    – pD

  • Anonymous

    The notion that dreams are epiphenomena or the products of random neuronal firings is complete bullshit. Hobson ignores a lot of research that tends to refute his random hypothesis. For example, there is a very high correlation between age and the experience of the “assassin dream”– you know, the one where you are being chased by someone who wishes to kill you. Clearly, the assassin is a metaphor of death, or at the very least, physical infirmity– concerns that increase with age. This finding has been replicated in cross cultural samples. There are countless other examples from the developmental psychology research base– for example, as men grow older, they tend to depict women as more active and agentic in dreams while as women age, they tend to depict men as more passive– findings which correlate with documented developmental changes in personality across adulthood, especially in the post parental phase of life. Sorry Hobson– to say this is all random is just plain weird. Dream on.

  • http://www.blogger.com/profile/04803674886685831282 Mozglubov

    Do you have a source for the claim that children don't report dreams before the age of five? I am just curious, because I certainly had dreams before that. However, I also have a mild form of narcolepsy which messes with my sleep stages, so perhaps I was just weird.

  • http://www.blogger.com/profile/04182376640295763312 Lee Charles Kelley,

    Sometimes a cigar is just a cigar.

    Sometimes a dream is just the product of random neural firings. Sometimes it has relevance to deep emotional issues.


  • http://www.blogger.com/profile/13491763060736064288 zule

    This comment has been removed by the author.

  • http://www.blogger.com/profile/13491763060736064288 zule

    Interesting point of view, although I do not think that the subject of dreams be so simple.

    However, it could explain the fact that depresive people have more nighmares. Simply, their brain would not work well either when sleeping.

  • Anonymous

    When I quit smoking I had wicked anxiety dreams about smoking– I would dream that I smoked a pack of my beloved Camels and then awake in a panic fearing that I had gone back to smoking. There was nothing random about these dreams! Pure old Freudian wish fulfillment/anxiety dreams!

  • http://www.blogger.com/profile/05660407099521700995 petrossa

    I always believed in the dreams to be random noise theory.
    Still there is this one vivid dream:

    A few months after the death of my wife she appeared to me, surrounded in a withish light and asked me: Why don't you visit my tomb?

    I answered that i thought visiting tombs was a bit forced and that i rather remembered her as i went along through life. (which is my deepest conviction)

    She seemed to consider this, nodded and disappeared never to bother me again.

    Anecdotal i know, but still the vivid dream i can still visualize after all these years.

    Dreams are still far from being understood i guess.

  • http://www.blogger.com/profile/13491763060736064288 zule

    Anonimous: I gave up smocking almost seven years ago and still today, I dream I go back to smoke. I was a stron scmoker (two to three pacvkets a day).

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    Mozglubov: I took it from Hobson although I have read it elsewhere as well. Hobson says that “Although the exact timing of the emergence of adult-type dream consciousness is debated, it is not before age 5 years and may be as late as age 8 years” and gives two references:

    * Foulkes, W. D. Children’s Dreams: Longitudinal Studies
    Wiley, New York,1982).
    * Resnick, J., Stickgold, R., Pace-Schott, E., Williams, J. & Hobson, J. A. Self-representation and bizarreness in children’s dreams. Conscious Cogn. 3, 30–45 (1994).

  • http://www.blogger.com/profile/13491763060736064288 zule

    I remember dreaming when I was 3. I even remember a dream I had when I was that age.

  • Anonymous

    Hi Neuroskeptic,
    A very interesting piece, although a lot of Hobson's Theory has largely been refuted by Mark Solms & Oliver Turnbull who demonstrated with neurological patients, that even with damage to the area of the brain responsible for REM sleep generation (PGO) that REM type dreams were still reported. Jak Panksepp the father of affective neuroscience identified a dopaminergic circuit which is activated during REM dreaming, thought to be involved in “Seeking behaviour” and rewards, Solms & Turnbull suggest this supports Freud's hypothesis that dreams are a form of wish fulfillment…

  • http://www.blogger.com/profile/14330808659093289808 Paul

    Have you read 'Dreaming Reality' by Joe Griffin and Ivan Tyrell. I really find their 'expectation fulfilment theory' interesting.

  • http://www.blogger.com/profile/02219505720807067694 NeuroKüz

    As a guitar player, I have found that it is much easier to play a newly learned song after sleeping. This is also the case for learning new movements (e.g. for dance or sports). I think the study passionlessDrone referred to applies to humans too – there is some kind of motor learning happening during sleep.

    As for dreams, they might forever be a mystery.

  • http://www.blogger.com/profile/17486178248700016699 dr. bob

    maybe we sleep to allow defragmentation of our memory and dreams are an artifact of verification


  • Anonymous

    I thought dreaming was small release of DMT? Or is that still theoretical? I read two books on DMT this year and thought I found that in there.

    To the guitar player – I recently read a study that young songbirds practice in their sleep. I found that very touching.

  • Anonymous

    Almost 80 years after Freud's death, a lot of neuroscience research actually supports many of his hypotheses regarding mental function, including dreams.

  • Anonymous

    How would sleep keep us safe from nocturnal predators? May be doors and locks but during sleep most people move make noises and cannot see if anything nasty is around. I would think that our ancestors took it in turns to sleep just like soldiers today when there are enemies around

  • Anonymous

    I want to know how come every time I dream of mounting this hot chick and riding her hard, a priest enters the scene and I stop. Is that all the product of random neuronal firings?

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    Anonymous 7th April: Thanks – I'd vaguely heard of the dopamine system lesions impairing dreaming data, but I didn't know who was responsible for it – I'll read into that.

    Another interesting thing is that sleep and dreaming abnormalities are common in Parkinson's disease, including REM behaviour disorder in which people act out their dreams.

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    NeuroKuz: I've noticed the same thing re: guitar learning as well… hard to know whether that's related to dreaming or other stages of sleep however. I have never dreamed about playing guitar, although last night I dreamed I had invented a new genre of music fusing jazz and death metal: I was wondering whether to call it death jazz or jath metal. I love dreams.

  • http://www.blogger.com/profile/03928365682331811860 wichitarick

    Thank you
    I have been blessed and cursed with VIVID dream and nightmares all my life. I most certainty reported dreams back to my parents before age 5 and my daughter did this also around age 3.
    I do not seem to understand what connection that medications have on our dream state ?
    The rem I guess I understand but what connection do certain medications have on that dream state or even a serious nightmare state of mind ?
    While I have not spent a lot of time on this I have looked for the connections from my youth and adulthood and the constant vivid dreams and the nightmares later on.
    Then my vivid dreams and nightmares and my electrical misfiring?
    And then the connection with medications and my nightmares?
    This many years later ,I now sleep walk and my dreams are low key and near as colorful or vivid? fascinating .
    The mind what a terrible thing to waste. Rick

  • Anonymous

    le hasard des rêves relevé par Hobson est tout de même lié à l'expérience personnelle du sujet: le cerveau ne peut pas concevoir d'image mentale sur ce qu'il ne connaît pas, c'est évident!

  • http://www.blogger.com/profile/06438220914490074792 Marcia

    Hobson's theory doesn't explain the story-like content of dreams.

    Random firings of neurons wouldn't create the complex plots of dreams. They would create brief sensations like those experienced during migraine auras.

    There are summaries of various theories about why we dream at http://www.meaningofdreams.org/dream_theory/

  • http://www.blogger.com/profile/17665315588139184543 robot archie

    Firstly I remember a vivid nightmare that I experienced when I was 2-3. There is a theory that consciousness (memory (in this instance) occurs after trauma). Oddly the dream I had was of SS herding jews into a circus tent (I cannot account for this type of dream at that age but it may be I had seen references to it somewhere).

    Sleeping habits of primates are indicative of how we might have slept in our earlier evolutionary stages “I would think that our ancestors took it in turns to sleep just like soldiers today when there are enemies around”. I seriously doubt that we would have slept in staggered guarded stages all year round. Amazonian tribes (“keep the river on the right”) provide interesting anthropological descriptions of sleeping habits of “primitive peoples” for example.

    Like Freudian psychology it may be a case of one size does not fit all. Freudian analysis works well on a particular type of personality, jungian works equally well on another personality type, gestalt, NLP etc etc

    Even if it is random firing, the material that it is firing with are substantively constructed by the conscious and waking mind (Burroughs cut up method provides a possible insight into the use of random social artifacts and consequent interpretation to provide an insight of depth and immediate relevence).

    The ghost dreamer above might actually be another type of dream experience altogether as to the DMT dreamer query. DMT has a particular and universally consistent effect on the mind and may sit slightly outside of the normal REM sleeper dream vocabulary.

    Just some random thoughts from reading the comments and article above , which I would love to research and provide proper data for. All interesting points which deserve attention I think.

  • http://blog.cas-group.net jofr

    I wrote this in 2004: “Dreaming has the function of stimulation and preparation. A baby before it is born is paralyzed and isolated from its environment, the same conditions as during REM sleep. Yet the brain must be active to develop itself. This paradox is resolved by dreaming. It seems to be a mechanism to ensure brain development during the time before birth. REM sleep and dreaming help in preparing the child for the unlimited stimulations it will soon have to face. It is necessary for development and facilitates wake-up.” see http://bit.ly/9uzLh0

  • Anonymous

    I'm 12 years old and before I went to bed this night, I was thinking… Why do I sleep? More importantly… Why do we dream? It is honestly a perfect question for past generations and future generations. Dreams is a beautiful state of neither awake or asleep. Kind of like a limbo state. Neither heaven nor hell. I agree with many of these theories, though some I don't completely understand. I've heard people say most of these theories state in the post and on these comments. But some people have said that the brain is so full and compact of ideas, thoughts, emotions and many other scycological things, that your brain releases all of those things and mushes them all together. When you are going to sleep, that's when most of your ideas are active. But why is only the last few hours of sleep, REM sleep?
    I also remember a dream from when I was about 4. Persay a nightmare.
    I honestly never understand my dreams. Just a heads up. But me and my family went on a vacation to someplace andnit was all fun. But then this bear came and attacked my whole family but they escaped and hid with me behind some chairs. I vaguely remember us running away though free that. Now me being about 4, that was scary as heck. You might not be thinking I'm 12, but trust me.. I am.

    Thank you,



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar