Dream on ...

The one third of our lives that we spend asleep is an enigma to scientists. They still don't even clearly understand why we have to sleep. But perhaps the greatest puzzle of all is the role of dreaming, which great thinkers have puzzled over for millennia. Aristotle held that in dreams the sleeper saw things that he wished or intended to do, while Thomas Nashe, the 16th-century pamphleteer, wrote them off as "a bubbly scum or froth of the fancy which the day hath left undigested".

Increasingly, though, research suggests that dreams aren't just neural rubbish. They have something to do with the processing of memories. They may aid creativity — the tune to Yesterday reportedly came to Paul McCartney in a dream — but may also help us play out threatening situations, preparing us for danger.

The meaning of dreams remains tantalisingly inconclusive. Research on the subject is thriving, as underlined by the work of Robert Stickgold, associate professor of psychiatry at Harvard Medical School's Centre for Sleep and Cognition.

Scientists once equated going to sleep with the flicking off a light. Then, in the 1950s, a team in Chicago found a sleep state in which the brain is every bit as active as when it's awake — the heart beats faster, blood pressure and blood flow to the brain increases, while the eyes move rapidly beneath their lids. This state, which takes up one fifth of our slumbers, is called Rapid Eye Movement (REM) sleep.

This discovery seemed to confirm the ancient Indian philosophy that there are three modes of existence — sleep, dreams and wakefulness, and the finding seemed to promise a way to solve the mystery of dreams. But, half a century later, we are still struggling to understand what our nocturnal imaginings really mean.

The late Francis Crick, co-discoverer of the double helix structure of DNA, and the mathematician Graeme Mitchison proposed that REM sleep was the brain's way of discarding unwanted memory traces to allow new ones to take hold. In the last decade it has emerged that dreaming occurs outside periods of REM, notably deep sleep, the kind of sleep characterised by slow brain wave activity.

Many of those who study dreams look for some accommodation between writing off dreams as a nonsensical byproduct of memory consolidation and the irresistible urge to get our dreams to tell a story, and reveal a deeper meaning.

Dr Stickgold is fascinated by the question of whether dreaming enables us to build associations between different memories, extract the meaning of events and learn from them.

Such research may help explain why children, teenagers and especially, infants need more sleep than adults, given that their brains are developing rapidly as they learn new things.

In one experiment, Stickgold's team managed to get 17 people to have the same dreams. They trained 27 subjects — 12 novices, 10 experts and five amnesiacs — to play the computer game Tetris for seven hours over three days, and monitored their dreams as they were drifting to sleep. More than 60 per cent reported dreaming in the hour after they fell asleep, and all reported the same images of Tetris blocks falling and rotating, as they do on the computer screen.

The biggest surprise was that the amnesiacs — who had no short-term memory owing to damage to a brain region called the hippocampus — had the same dreams as the normal subjects, even though they could not remember playing the game. Compared with later stages of dreaming, these initial (hypnagogic) dreams that occur as we slip into slumber were thought to be more tightly linked to recent memories. But the fact that some amnesiacs saw the Tetris pieces suggests a powerful role played by the unconscious, too.

Dr Stickgold believes the amnesiacs' unconscious Tetris memories may have influenced not only their dreams but their waking behaviour. Most of this group had to be taught the game afresh each day. But then a colleague noticed that at the start of a session, one of the amnesiacs placed her fingers on the exact three keys used in playing Tetris. Dr Stickgold said: "This is Freud's unconscious — things activated in our brain, which are, in fact, memories that guide our behaviour but are not conscious."

Weird associations

The amnesiacs were probably drawing on memories in their neocortex, the rind of the brain, which are more abstract and are linked by associations. This may show why dreams often contain weird associations of events and places.

In subsequent work, the researchers tried to establish how real events influence dreams by focusing on the terrorist attacks of September 11, 2001.

A group of 14 students had recorded their dreams before and after attacks. The day after the attacks, they reported changes in dream features — to include fire, planes, terrorists and police — and there was a strong relation between these changes and the exposure to the events on television. The study suggests that the media's endless replay of horrific images may have boosted stress levels.