IELTS 形式のリーディング：睡眠と記憶

Sleep, memory and the learning brain

For much of the twentieth century, sleep was treated by many researchers as a kind of enforced idleness, a daily shutdown during which the brain simply rested. That view has been steadily overturned. Techniques that record the brain's electrical activity have shown that sleep is a highly organised process, and that during it the brain is far from inactive. Some regions are, if anything, more active than they are during quiet wakefulness.

Sleep proceeds in cycles of roughly ninety minutes, each containing lighter stages, a period of deep slow-wave sleep and a phase known as rapid eye movement, or REM, sleep. The proportions change through the night: deep sleep dominates the first cycles, whereas REM sleep becomes longer towards morning. This pattern matters, because researchers increasingly believe that the different stages contribute to memory in different ways, and that cutting a night short may therefore remove some kinds of benefit more than others.

The prevailing account, often called the consolidation hypothesis, holds that memories are fragile when they are first formed and are stabilised afterwards. During deep sleep, patterns of activity that occurred while a person was awake appear to be replayed in the brain, apparently strengthening the connections involved. In one widely cited type of experiment, volunteers learn a list of word pairs in the evening. Those who then sleep recall noticeably more of the pairs than volunteers who spend the same interval awake, even though both groups had an identical amount of time to learn the list in the first place.

REM sleep is linked less to the storing of facts than to making sense of them. Some studies suggest that people who have slept are better at spotting a hidden rule in a set of puzzles, or at linking ideas that seemed unrelated the day before. Researchers are cautious, however. Many such experiments involve small groups of volunteers, and it is difficult to separate the effects of a particular sleep stage from those of sleep as a whole. The claim that REM sleep makes people more creative, repeated often in popular articles, therefore remains a hypothesis rather than an established fact.

What is better established is the cost of too little sleep. Volunteers who are kept awake for a night perform poorly on tests of attention, and they are also less able to absorb new information the following day, as if the brain's capacity to take in new material had been temporarily reduced. A single night of lost sleep can often be recovered, but chronic shortage, in which adults regularly sleep for much less than they need, is associated with lasting difficulties with concentration and mood. Association is not proof of cause, though: people who sleep badly may also differ from others in ways that affect memory independently of their sleep.

Such findings have prompted debate about how sleep should be treated in education and the workplace. Some schools have moved their start times later, hoping to match the natural rhythms of teenagers, whose body clocks tend to shift towards later waking. Early results have been encouraging in some cases, but the evidence is mixed, and it is not yet clear whether the benefits come from the extra sleep itself or from the calmer mornings that follow. Meanwhile, the advice often given to students, to revise late into the night, looks increasingly unwise: if learning is consolidated during sleep, then time sacrificed for extra study may partly cancel out what it was meant to achieve.

Several questions remain open. Researchers do not yet know how much sleep any one individual needs for the purposes of memory, since needs vary considerably, nor can they say with confidence whether short daytime naps offer the same benefits as a full night. Technologies that play gentle sounds during deep sleep in order to boost memory have produced intriguing results in laboratories, but their use outside the laboratory has barely been tested. What seems clear is that sleep, far from being a passive interval between periods of useful activity, is part of the learning process itself.