readings> why we sleep and dream

Does sleep have a purpose? Do you really know what its like to dream? The answers may surprise you.

Every night we shut our eyes, slip the moorings of consciousness and enter the confused limbo of sleep and dreams. But why? Couldn't evolution have allowed us to take a nightly rest without this apparently dangerous loss of awareness? Or do we need to sleep because the brain has to have its time in dreams?

Despite nearly 2,000 sleep clinics around the world investigating this vital subject, scientists still seem, ahem, a little in the dark about the answer. But first a few of the facts that are certain. Sleep is a universal among animals. Even cockroaches have moments of quiescence. Their antennae will droop and they will become insensitive to gentle stimuli. Fish and reptiles also have their drowsy periods while birds and mammals have proper sleep cycles.

And don't be fooled into thinking that humans are the great slugabeds. Humans average seven and a half hours sleep a night. True, large ungulates make do with less. Giraffes sleep as little as two hours a night while elephants and cattle sleep about four. But for the rest of the mammalian kingdom, sleep can seem to be what they do best. Cats and lions will happily nap for 18 hours a day; sloths, armadillos, owl monkeys and opossums will snooze away just about as many; while bats can go 20 hours - they might come out at night, but not for very long it seems.

And once scientists began tapping electrodes to the heads of sleeping animals and humans, they found that all higher mammals and birds share a similar sleep cycle. Recordings of the brain's EEG rhythms - the crackle of neural activity - show that sleep has a distinct pattern with phases that in humans follows a 90 minute period.

First we descend through four increasingly deep stages of slow wave sleep - so named because the EEG rhythm changes from a busy jagged buzz to a slow relaxed throb. At its deepest, it becomes very difficult to rouse a person from slow wave sleep. The sleeper may take fully a minute to become orientated. This is also the time when people are liable to sleep walk (see box) or, in children, wet the bed. Then after an hour and a half of this slow wave slumber, the sleeper's brain will rise back through the levels before eventually convulsing with its first period of REM - a sleep stage as dramatic as it is misunderstood.

In REM (which stands for rapid eye movement) the eyes leap about in their sockets as if chasing some phantom vision. Electrodes show that the brain is highly aroused as if back in a waking state. The body also seems a bit out of control with the heart pumping, temperature control gone awry, and blood pressure all over the place. But despite this metabolic furore, outwardly the sleeper appears at his or her most inert as REM also brings on complete motor paralysis. A neurochemical block is put on messages entering and leaving the brainstem, preventing any movement apart from a few twitches.

When sleep researchers first discovered REM sleep in the 1950s, they found that subjects awoken during this stage invariably reported having vivid dreams. Putting two and two together, REM sleep became dreaming sleep - an idea that has stuck in the public mind. But in fact it has since been shown that dreaming of a sort goes on all night long. Subjects roused from slow wave sleep can usually tell of vague ruminations or mundane fantasies.

David Foulkes, a psychologist at Emory University in Atlanta, Georgia, says these slow wave dreams seem identical to the kind of drifty daydream thinking we have during idle moments of waking. The only difference is that in our sleep, our mental experiences create no permanent record. Each thought is forgotten almost as soon as it occurs. Foulkes says the dreams of REM sleep are certainly brighter, more emotional, and more bizarre. But these differences are mostly ones of internal arousal level than of mechanism. So whatever the reason for our having dreams, it cannot explain the different stages of the sleep cycle. REM sleep simply increases the intensity and detectability of something we were already doing.

Our first REM period of the night actually lasts only a few minutes. But as the night goes on, the REM phases grow longer until much of our time in the last few hours before waking are spent in REM. With such a strikingly set pattern of sleep in both higher animals and humans, it is natural to feel that it must serve some equally defined purpose. And yet the best current explanations for the sleep cycle are surprisingly prosaic.

Certainly, there is much support for the idea that sleep is restorative. In the depths of stage four slow wave sleep, when brain and body activity is turned down about as low as it can go, there are surges in various growth hormones. Some researchers have claimed that the adolescent body can grow as much as a centimetre at this time of night, as if the body is taking advantage of a time when it can devote itself entirely to building and repair. And feeling good in the morning is linked to the quality of our slow wave sleep - if we toss and turn without properly hitting stage four, then we wake feeling rough.

But sleep is not completely mandatory because humans have lasted a couple of weeks awake and despite being dog tired, could still rouse themselves to beat sleep researchers in games of pinball. And while laboratory rats will die after some weeks of sleep deprivation, it seems to be a collapse of the immune system that actually kills them.

Another favoured theory is that sleep is a good way of saving energy. This certainly seems true in small animals - bats and hummingbirds go into a hibernation-like torpor when they sleep - but the metabolic savings are not actually that great in larger animals. Even deep stage four sleep sees a drop of just 20 percent while the rest of sleep hardly differs from just lying still all night and resting.

Then there is the ever-popular idea that sleep - especially the wild activity of REM - might play a special role in the fixing of memories accumulated during the day. Dreams often seem like a confused riffling through the memory banks or - shades of Freud - the surfacing of repressed thoughts. However careful work by Avi Karni of the Weizmann Institute in Israel seems finally to have put paid to this theory.

 It is already known that it does indeed take time for the brain to consolidate a memory trace. Growth changes are needed to strengthen the connections between brain cells and so turn a jangling pattern of neural activity into some more permanent record of an event. By testing subjects at intervals after learning a task, Karni showed that this hardening process goes on in the background all during the day but seems to be interrupted during the metabolic drop of slow wave sleep. So slow wave sleep actually appears to suspend the fixing of memories. Then in REM sleep, the fixing resumed - but Karni argues this is only because REM is near enough a waking mental state in terms of arousal levels. There is nothing special about REM itself.

Summing up the various findings, Jim Horne of Loughborough University's sleep lab says slow wave sleep seems the key. It is a necessary downtime for certain kinds of renewal. But this job only takes the first few hours of the night and the question is why we then stay asleep in dreams? His answer is that it is simply safer. Leaving an animal exposed all night with a fully alert and worrying mind would only lead to trouble. Much better to keep it in a state of unremembering, dimly sensing, dream rumination. In this view, REM sleep would merely act as a way of regularly bringing the mind back close to the surface without actually awakening it.

Horne says that while it may not always feel like it, we can tumble out of the REM sleep phase and go straight into action. So it turns out then that our slumbers are mostly just nature's way of keeping us all out of crepusculan mischief. Now there's a thought to sleep on.

why we dream

Yes, yes, I know other people's dreams are boring. But let me tell you about one of mine - the one where I discovered I even dream of smells and tastes!

It was one of those usual Technicolor, jump-cut, jobs. There I was racing along in a Flintstones's cartoon car, when the next instant I was soaring high like a bird over the ocean. Passing a cruise ship, I swept down to the bridge where I saw a bearded figure standing alone eating a bag of greasy chips. I immediately knew this person was me, that I was a spy and that I was in command of an empty ship.

However there was no time to linger. Next I was sweeping down the length of the vessel - yet now it was a grey painted frigate. The intricate ironwork of a radar mast caught my eye and I had a momentary close-up of a few struts. Then when I pulled back, I found the scene had changed completely and I was now seated at a desk with the parts of a model ship strewn about. My final thought as I awoke was how difficult it was going to be to glue all the fiddly bits of the mast together.

It was not the contents of the dream that got me. Instead, I felt that for the first time - after some months of trying - I had glimpsed the mechanics of the dreaming process. I had seen the kind of features reported by famous Victorian documenters of dreams like Hervey de Saint-Denys, and also what a modern neuroscientific understanding of the brain might lead us to expect.

The first point was that my dream seemed to consist of a series of supernaturally sharp, yet oddly frozen, pictures. We tend to take our dreams at face value, remembering them as if they were a smoothly projected slice of life. But looking closely I could see that each frame seemed quite disconnected from the next like a succession of tableaux. And while there was always a powerful sense of movement in these images, it came from the way my mind's eye seemed to sweep across and around. Despite the feeling that there were tossing waves or billowing smoke from a funnel, nothing really moved.

The second big feature of my dream was that the scene might change utterly from one frame to the next, yet the thinking, commenting, part of my mind did not appear to notice. Clearly, there was a vague theme about ships connecting all the images, but each scene was of a completely different type of ship. Nor was I fussed about the way in one frame I might be the soaring viewer of the scene, then the following I was the bearded figure or a modeller at a desk.

Thinking back on the dream, I realised that I had not even been right about that. The bearded figure had actually looked nothing like me, but was the German actor from a U-boat film I had recently seen. And while I had thought the modeller was me in my bedroom as a teenager, it was really nothing like my old bedroom - or, indeed, me. As turns out to be the rule in dreams, everything was cheerfully misidentified or assumed.

Finally there was the puzzle of why it had taken so much effort to notice these things. For example, only the day before I had been wondering why my dreams had most kinds of sensations but not tastes or smells. Then I realised that when I had seen the soggy bag of chips in the bearded figure's hand, there had been a sudden intense experience of their savour. I had to think hard to catch such details.

So what does science have to say about dreams? The vividness of our dreams means it is difficult not to believe they must serve some purpose. Why put on a nightly display of weirdness and leaping emotion unless it signified something - perhaps a Freudian battle with the unconscious or a ritual cleansing of the memory banks? Yet sleep researchers like John Antrobus of the City University of New York have mostly come to just the opposite conclusion. They say the sleeping brain dreams simply because it has no proper way of switching itself off.

Unlike a computer, the brain cannot pull out the plug and shut down its circuits for the night. Living cells have complex metabolisms to support and all 100 billion neurons of the brain must discharge at least once or twice each second just to maintain their electro-chemical tone. And as Antrobus says: "If you can't switch the brain off, then you can't really stop it thinking or trying to make sense of the state it then finds itself in." So even in the depths of slow wave sleep, there is some meandering mental activity. And in the bright pseudo-arousal of REM, this activity can take on a waking intensity.

But why do dreams take the form that they do? For Antrobus, the key is that when we fall asleep, we enter a state of deep sensory deprivation. The brainstem - the first thickening of the spinal cord where it enters the brain - puts a neurochemical block on all but the most urgent traffic entering or leaving the brain, so cutting us off from our senses and also preventing movement (a motor damping that turns into total paralysis during REM). In fact, sometimes we can actually catch this sensory curtain falling. Many a drowsy commuter has probably had the experience of nuchal atonia - a sudden slump of the head and eerie silence as they are cut off from the external world.

This crude brainstem gating serves the purpose of holding us in a state of physical quiescence. But it does not mean our brains are then empty of thoughts or even perceptions. Antrobus says it surprises people that a sensation-starved brain will immediately begin manufacturing substitute experiences. It would seem that producing mental imagery should involve some sort of selective effort. But instead the brain is built to generate a constant flow of images.

During waking life, we need to go into each moment armed with a set of perceptual expectations. As we reach for a door knob, we must already be making mental predictions about what the handle will feel like in our hands. Or if we hear a siren coming down the street, we need to be prepared soon to see the flashing blue lights of a police car or ambulance. There is so much information to deal with during every instant of life that we have anticipate as much of it as possible to be able to assimilate the world smoothly - and also to recognise when some event is actually surprising, such as finding a door handle is made of squishy jelly or the siren is mounted on a milk cart.

This built-in ability to anticipate sensations is believed to be why we have mental imagery in the first place. As Ulric Neisser, a psychologist at Cornell University in New York state, points out, a mental image is an expectation without the answering sensation. We ask ourselves to think what it would be like to see something - an elephant in a balloon perhaps - without actually going on to see it.

Furthermore, experiments show that it takes about half a second to generate a full blown mental image - and it fades just as fast. Also, while most people can conjure up a momentarily convincing image of some object or situation, it tends to be rather static. The picture is like a sudden view - or succession of views - rather than an animated film. In other words, our waking mental images are remarkably similar to those we have in dreams, except that in sleep the visions appear to erupt unbidden rather than coming as part of an organised train of thought.

So shut off the brain from real sensation and it cannot help but throw up a series of fleeting anticipatory states in a restless perceptual search. Exactly the same happens in real sensory deprivation experiments where subjects are left a few hours floating in a muffled, darkened tank of salty water. It is not long before they report disorientated thoughts and weird visions. But a hallucinatory parade of images is only part of the story of the dreaming state. There is also the question of the dazed misidentifications and almost immediate forgetting of our dream experiences.

Antrobus says the latest evidence suggests that as well as the brain putting a block on outside traffic, it also appears to become internally decoupled. The brain is made up of many processing areas - perhaps hundreds - each responsible for some small aspect of mental experience. There are speech centres, planning centres, memory-fixing centres, and even separate centres for a sharp sense of colour or sense of direction. Normally, all these areas of activity are tightly bound into an integrated whole - neuroscientists believe that some form of synchronised firing may tie the many bits of the brain into a common rhythm. It appears that in sleep, this co-ordination unravels, leaving each module of processing only weakly in touch with the activity of all the others.

The result is confused thinking about the relentless parade of images. The frontal areas of the brain that do the planning and the commenting on experience will be trying to make sense of what they see, but because they are only loosely in touch with the sensory evidence, they will make mistaken assumptions about identity and fail to notice the crazy shift in scenery every half second or so.

This lack of integration would also mean there was no coherent story feeding into the parts of the brain responsible for capturing a record of mental life. So to remember the content of a dream, we would have to wake up and relive its last few moments. Then unless we had learnt how our dreams should appear, we would tend to gloss over their disjointedness, smoothing our memory into something that looked more like the steady logical flow of normal waking consciousness.

So that is what dreams are made of. What's that? You don't believe me? Well put a pen and notebook by your bedside tonight, set your alarm clock for a little earlier than usual, and we can compare notes in the morning. We will see if you too can taste those chips.

sleep walking and other oddities

We all know about sleep-walking. But can you rape and even murder in your sleep? In 1994, a Sunderland prison officer was cleared of breaking into the home of his neighbour at 5am and attempting to rape her because experts said he suffered from a serious sleep disorder. In 1992, a London policeman was similarly found not guilty after nearly strangling his uncle with a rope in the course of a nightmare.

Researchers can tell you hundreds of other equally extraordinary tales of what people do in their sleep such as the woman who packed her dogs into the car and drove 20 miles before waking up, the dieter who raided the fridge at night to gorge on raw meat and vegetables, and the holidaymaker who walked straight off the balcony of her hotel, fell 15 feet, smashed her hip, but did not wake up until she reached the hospital.

That the sleeping mind is capable of confused and largely automatic behaviour is no surprise if the brain never really shuts down, just becomes dazed and unremembering. However researchers have found there is a clear difference between ordinary sleep walking and a more specific REM sleep syndrome.

The less alarming activity takes place in the depths of slow wave sleep. We come awake enough to act on automatic pilot - to get up and do something routine like go to the toilet, or perhaps take the car for a midnight spin - but not awake enough to recall doing so. However murders and other violent acts must occur in the high arousal of REM sleep. The explanation seems to be that sufferers lose their usual REM paralysis that prevents them acting out their dreaming behaviour. In such cases, doctors often find evidence of a degenerative disorder like multiple sclerosis affecting the brainstem.

six wacky reasons given for dreams

We dream to rinse our eyeballs: David Maurice of Columbia University in New York suggests the violent eye-movements of REM sleep are to shake-up the aqueous humor inside, so allowing fresh oxygen to circulate.

We dream to practice our response to danger: Antti Revonsuo of the University of Turku, Finland, says dreams evolved in our cavemen ancestors as a safe virtual reality playground in which to get used to handling threatening situations.

We dream to get in touch with our internal organs: Steven Maier of the University of Colorado in Denver believes that dreaming sleep is essential to give the brain a sensitive time to listen to what our immune system has to report about the state of our body. Some of what is learnt may be symbolised as dream imagery.

We dream to exercise rarely used brain circuits: Jim Krueger of the University of Tennessee in Memphis thinks that the wild activity of REM might tone critical, but rarely employed, brain pathways - for example, the brainstem feedback mechanism which causes us to breathe more deeply when exposed to high levels of carbon dioxide.

We dream to stockpile key neurotransmitters: J Allan Hobson of the Harvard Medical School in Boston wonders whether REM sleep builds up levels of important brain transmitters just prior to waking. He notes that the consumption of both noradrenaline and serotonin, two of the most important messenger molecules, drops to almost zero in REM.

We dream to reinforce our genetic individuality: Michel Jouvet of the University of Lyon, France, argues that REM sleep allows time for our brain to run through its genetic program and so prevent the day's experiences from causing us to deviate too far from our meant to be, DNA-encoded, personality profile.

home> back to readings