Lewis Wolpert: 'Once, I dreamt that I had changed into an umbrella and had difficulty opening it'

I have weird dreams but do not remember them well. I have in the past dreamt that I was hanging from the tail of an elephant and once, when feverish, that I had changed into an umbrella and had difficulty opening it; more recently, I have dreamt that I am a metaphor. So, what really is known about the nature and function of dreams and sleep itself? No one could believe that such universal human processes have no important function.

Yet the function of sleep is largely unknown. Possible explanations are that it is for energy conservation, or a time for recovery of tissue, including the brain. It is also possible that sleep is a means of keeping animals inactive when it is dark. A rather different explanation is that it is a time when the brain is plastic and memories can be consolidated. Memory traces may be in a fragile state until the first exposure to sleep.

Experiments have shown that sleep deprivation does affect learning, but the non-specific effects of lack of sleep make this rather less reliable evidence. It is hardly surprising that one does less well when one is sleepy and cannot concentrate.

The discovery of rapid-eye-movement – REM – sleep has led to many studies of its role and its relation to dreaming. REM is accompanied by rapid brain rhythms and relaxed muscle tone. REM sleep increases when a new task has been learnt, in both humans and animals. There is also support from the observation that in rats that have learnt a new environment, the nerve cells in the hippocampus involved in the learning process fire again during sleep.

This idea – that dreaming may help to consolidate memories – goes back more than 200 years to David Hartley. Freud's theories a hundred years later do not explain their function, though he believed dreams were about wish fulfilment and provided the true road to the unconscious; there is no evidence for that whatsoever. A current view is that dreams may be a manifestation of the process of memory consolidation. Deprivation of REM sleep can result in less retention of learnt tasks and emotional memories. Brain-imaging studies show that most regions are less active during sleep, but in REM sleep certain regions, such as the amygdala, which is at the core of emotional responses, are more active than in the waking state. But the significance of these patterns is not understood. Moreover, monoamine oxidase inhibitors are a major component of some antidepressants, and this compound completely suppresses REM sleep. Yet patients on such drugs for years have not shown any impairment in their memory.

Nevertheless, there is evidence that in REM sleep the brain is more capable of fluid thinking – waking after REM sleep seems to facilitate solving anagrams. The relationship between dreams and memories of related events is also far from clear, and whole events are rarely included in dreams. By contrast, hypnagogic dreams, which occur around the onset of sleep, can incorporate recent waking activities.

It is possible to go for 40 hours without sleep and still recall all sorts of information acquired shortly after waking. Animals that are born relatively mature, such as the guinea pig and the dolphin, have low levels of REM sleep – both less than one hour out of 10 – whereas those born immature, such as the platypus, have high amounts: eight out of 14. Humans are somewhere in between – two hours out of eight. REM sleep may thus play a key role in the development of new neuronal connections as the animal matures.

There is a long way to go in the understanding of sleep and dreaming. I would love to know if my need for at least eight hours' sleep, plus sleeping in lectures, is genetic – I envy those who manage on much less. How are some of my colleagues able to come home from a party and then work till 2am? Or, like the late Roy Porter, start writing at 5am each morning? Is it in the genes, or could I blame my parents for putting me to bed so early as a child?

Lewis Wolpert is professor of biology as applied to medicine at University College London