Australasian Science: Australia's authority on science since 1938

To Sleep, Perchance to Learn

By Tim Hannan

New research finds that we can learn while we’re asleep under certain conditions.

We spend one-third of our life asleep, but those who don’t seem to appreciate the joys of oblivion at the end of a day bemoan that this is a waste. It has often been speculated that the time spent asleep could be used productively to learn verbal information, such as a new language, although evidence for this has been so limited that some neuroscientists have declared it impossible. Now, a new study has found that if complex auditory information is presented at exactly the right time in the sleep cycle, learning of verbal material may indeed occur.

It has long been known that during sleep the brain engages in processes that help to strengthen or consolidate information acquired during the day. Through the continued activation of relevant networks, memory traces are strengthened and integrated with other information. This observation that the brain is active in aspects of learning and memory during sleep has given rise to attempts to market products that purport to facilitate sleep learning, promising that the sleeper will learn a language in 30 days without any daytime effort.

In contrast, neuroscientists have argued that sleep learning is not only unproven but implausible, as the brain is not in a state in which learning could occur. It is known that learning requires long-term potentiation of neurons in the hippocampus, and all available imaging techniques show that these are relatively inactive during sleep. Learning of complex verbal material, such as the meanings of words and phrases in a foreign language, has been thought to depend on memory systems that are understood to require conscious awareness.

However, a Swiss research team has drawn on prior neuro­psychological studies suggesting that learning does not always require conscious awareness, but can involve the implicit learning of associations between stimuli. They speculated that it may be possible for the learning of verbal information in certain specific stages of sleep when greater excitation of the cortex is evident, along with other neural properties similar to the waking state.

In the study, reported in Current Biology (, the researchers presented word pairs to sleeping participants during slow-wave sleep. In each pair, one word was the name of a well-known object (e.g. house, elephant – in German) and the other was a made-up word purported to be from a foreign language (e.g. tofer, guga). Each pair of words was presented four times, with presentation timed to align with slow-wave peaks. On waking, participants were presented with the non-words, and their learning was assessed by a test of implicit memory: they were simply asked to state whether the object to which each referred was larger or smaller than a shoebox.

The results indicated that the participants classification of the “foreign words”, which they did not recall hearing during sleep, was significantly better than could be expected on the basis of chance. Recall of non-words presented during slow-wave peaks was higher than recall of those heard during other periods of brain activity.

Functional MRI during the testing of the classification of words indicated that the hippocampus was active during recall, suggesting that hippocampal-based learning had indeed occurred during sleep. In summary, though asleep, participants had learned something of the meaning of the non-words through their pairing with known words, without conscious awareness of the learning itself, and this learning appeared to involve the usual learning mechanisms of the awake brain.

If replicated, the study’s finding that learning of verbal associations can occur during sleep obviously falsifies the conventional assumption that it is impossible, and challenges some popular theories of the nature of sleep and its stages. Although the amount of information learned in this study was modest, the possibility that extended and repeated exposure to verbal information during a specific stage of sleep may result in learning will likely stimulate a wave of further research in this process, and a flood of new sleep learning products.

A/Prof Tim Hannan is Head of the School of Psychology at Charles Sturt University, and the Past President of the Australian Psychological Society.