Sleeping on It

Sleeping on It

How the resting brain turns fresh experience into lasting memory

A For much of the twentieth century, sleep was regarded as a period of pure inactivity — a nightly shutdown during which little of interest happened in the brain. Learning, it was assumed, took place only while a person was awake and attentive. This view began to crumble in the 1990s, when researchers using controlled laboratory tasks noticed something odd: volunteers who had slept after practising a new skill performed markedly better the next day than those who had stayed awake for the same number of hours. The improvement could not be explained by additional practice, because the sleepers had done nothing further. Slowly, a once-fringe idea gained respectability — that sleep is an active state in which the brain continues to work on what it has recently learned. As the evidence accumulated across dozens of studies, sleep came to be seen not as time lost to learning but as an essential part of it.

B To understand how sleep aids learning, psychologists first divide memory into two broad categories. Declarative memory covers facts and events that can be consciously recalled and stated in words, such as a historical date or the name of a colleague. Procedural memory, by contrast, concerns skills and habits — riding a bicycle, typing quickly, or playing a musical phrase — which are difficult to describe but easy to demonstrate. The distinction matters because the two types appear to depend on different stages of sleep. What benefits a chess player memorising openings is not necessarily what benefits a pianist rehearsing a difficult passage.

C The consolidation of declarative memory is closely tied to the deepest phase of sleep, known as slow-wave sleep, which dominates the first few hours of the night. During this phase the brain produces large, slow electrical waves, and it is thought that newly acquired facts are repeatedly reactivated and gradually reinforced. In one experiment, students who learned lists of paired words and then slept remembered far more pairs than those tested after an equivalent period awake. Crucially, the benefit was greatest when the early, slow-wave-rich part of the night was preserved, suggesting that the timing of sleep, not merely its presence, shapes what is retained.

D Skills follow a different schedule. Procedural learning appears to benefit most from rapid-eye-movement sleep, or REM, the stage associated with vivid dreaming, which becomes more plentiful towards morning. When people learn a demanding motor sequence — tapping a pattern of keys as fast as possible, for instance — their speed and accuracy improve after a night's rest even without further training, and the size of this overnight gain tends to track the amount of REM sleep they obtained. Cutting REM short by waking sleepers repeatedly reduces the improvement, while leaving the earlier stages of the night largely undisturbed.

E Underlying these effects is a process of reorganisation. During waking hours, new experiences are captured quickly by the hippocampus, a small structure that acts as a temporary store. This arrangement allows rapid learning but is fragile, and the memories can easily be overwritten. As a person sleeps, patterns of neural activity recorded during the day are 'replayed', and the information is gradually transferred to the cortex, the brain's long-term archive. There it becomes more stable and is woven into existing knowledge, which may explain why a problem left overnight is sometimes solved on waking: the sleeping brain has quietly rearranged the pieces. In this sense, forgetting the exact circumstances of learning can be useful, since the underlying facts survive in a more general and durable form.

F The consequences of missing sleep are equally revealing. Volunteers kept awake for a single night after learning are consistently worse at recalling the material days later, even once they have caught up on their rest — the window for consolidation, it seems, cannot simply be reopened. Sleep loss also blunts the ability to form new memories in the first place: a tired hippocampus absorbs information poorly, rather like a wet sponge that can hold no more water. Both findings underline that sleep is not an optional extra but a biological requirement for stable learning.

G These findings carry obvious practical lessons. Cramming through the night before an examination is doubly counterproductive, since it sacrifices the very sleep that would have secured the day's studying. A short daytime nap can recover some of the benefit, and spacing revision across several nights allows each portion to be consolidated in turn. For workers mastering complex procedures, employers who protect regular sleep may gain more than those who demand longer hours. The emerging message from the laboratory is simple: to remember something tomorrow, sleep on it tonight.