Sleep and Memory: Why the Night Does Half Your Learning

    How Memory Works

    Sleep and Memory: Why the Night Does Half Your Learning

    A century of research says sleep is not time off from learning. It is when the brain replays, sorts and cleans. What the studies show, what it means for when you study, and the one-hour-before-bed habit that helps lines stick.

    By Published Updated

    Say you are an actor learning lines. The familiar habit is to cram in the afternoon, or panic in the morning before the audition. Change one thing: the time of day. Move the main session to the hour before bed, keep it light and playful, then stop and sleep. Come back to the lines the next morning and they are more solid than when you put them down.

    This is not a trick. It is the brain working as designed. While you sleep your brain does two different jobs. It replays the day and strengthens the memories it has chosen to keep, which is called consolidation. And, separately, it clears out the waste products of a day's thinking, which I used to call cleaning mode. Learning is not finished when you close the book. It is finished the next morning, and only if you slept.

    Here is what the studies actually show, and the rules you can use tonight.

    The 1924 experiment that started it

    In 1924 two psychologists at Cornell, John Jenkins and Karl Dallenbach, ran a study that was almost comically simple. Two students learned lists of nonsense syllables, meaningless combinations like "zup" and "kef", and were then tested after one, two, four or eight hours. The twist was that sometimes the gap was spent awake, going about a normal day, and sometimes it was spent asleep.

    The results were clear, and later research points the same way. After eight hours awake the students had forgotten most of the list. After eight hours asleep they remembered far more, and the forgetting curve, the line showing how much of the list was lost over time, flattened out after the first couple of hours of sleep instead of continuing to fall. Jenkins and Dallenbach concluded that forgetting was less about time passing and more about interference: new experiences overwriting old ones. Sleep protected memory partly by switching off the flow of new input.

    That explanation was half right, and it took the rest of the century to find the other half. Sleep does not just protect memories from interference. It works on them.

    What the sleeping brain is doing

    The modern picture, set out most thoroughly in a 2013 review by Björn Rasch and Jan Born in the journal Physiological Reviews, is called active systems consolidation. The short version runs like this.

    During the day, new experiences are held in a fast-learning, limited-capacity store centred on the hippocampus, the structure without which no new facts or events can be laid down. The hippocampus is quick but small, like a notebook. Durable knowledge lives in the cortex, the brain's wrinkled outer layer, which is vast but slow to change, like a library.

    During deep, slow-wave sleep, mostly in the first half of the night, the hippocampus replays the day's experiences. Recordings in animals show the same patterns of brain cells that fired during a task firing again in sleep, in the same order, at high speed. Each replay nudges the cortex, and over many replays the memory becomes less dependent on the hippocampus and more embedded in the cortex. The notebook is being copied into the library.

    Later in the night, in REM (rapid eye movement) sleep, when most vivid dreaming happens, the brain appears to do something different: integrating the new material with what you already know, finding connections and stripping some of the emotional charge from the day's events. People woken from REM are more likely to solve problems that need a creative leap than people woken from deep sleep.

    Not everything gets replayed. The brain is selective, and the selection seems to favour what was tagged as important during the day: material with emotion attached, material you expected to be tested on, material you cared about. Which is a scientific way of saying that the first step in sleeping on something is to make it matter while you are awake.

    The cleaning cycle

    Cleaning mode turns out to be more literal than it sounds, though it explains a different thing from consolidation. In 2013 a team led by Maiken Nedergaard published a study in Science showing that during sleep, in mice, the space between brain cells expands by around sixty per cent, and cerebrospinal fluid, the clear liquid that surrounds the brain and spinal cord, flushes through, clearing metabolic waste including the proteins associated with Alzheimer's disease. They called the system glymphatic, after the glial cells, the brain's support cells, that run it. It is far more active in sleep than in waking.

    The link to memory is indirect but real. A brain full of the previous day's waste is a worse learning machine. Anyone who has tried to study after a bad night knows the feeling: the words go in and nothing happens. Part of what sleep gives you is not consolidation of yesterday but capacity for today.

    What this means for when you learn

    Here are the rules the research supports.

    Learn before you sleep, lightly. A session in the hour before bed gets replayed first, with the least interference afterwards. But it has to be a low-stress session. If learning adds to your anxiety, the stress hormones work against consolidation, and a mind racing with lines or facts is a mind that does not fall asleep. So the pre-bed session is a run-through, a walk through the memory palace (the familiar route in your mind where you have left an image for each thing you are learning), a playful recital, not a cram. Make it enjoyable and let the night finish the job.

    Do not pull the all-nighter. Everything above runs the other way if you skip the night. Material learned the day before an all-nighter consolidates poorly, and the brain you take into the exam has not been cleaned. Students who study until three and sit the paper at nine are throwing away the consolidation they paid for and arriving with a worse brain than the one they started with. If you have to choose between two more hours of revision and sleep, sleep is the higher-scoring option.

    Space it across nights, not hours. A long-standing revision pattern is one hour after learning, then one day, one week, one month. Notice that after the first review every gap contains at least one night's sleep. That is not an accident. Each night of sleep consolidates what you reviewed, and each review the next day finds the material a little more embedded, so the review goes faster. Spacing across nights uses the brain's own schedule.

    A nap counts. Studies by Sara Mednick and others have found that a nap of sixty to ninety minutes, long enough to include deep sleep and some REM, produces measurable gains in memory for material learned that morning, in some tasks comparable to a full night. If you are learning something in the day and have the chance of an afternoon nap, take it after the session, not before.

    Protect the first half of the night. Deep sleep is concentrated early. A late night that cuts your sleep short loses proportionally more of the deep sleep in which replay happens. Going to bed late and getting up late is not the same as going to bed on time, even if the hours add up.

    Can you learn while asleep?

    Not new material. Playing a language recording all night does not teach you the language, and the studies that have tested it are consistently negative. What does work, in the laboratory, is reactivating something you already learned. In a 2009 study in Science, Ken Paller's group at Northwestern had people learn the locations of objects, each paired with a sound. When some of those sounds were replayed quietly during deep sleep, memory for those particular objects improved compared with the rest. The sleeping brain was cued to replay specific memories.

    This is called targeted memory reactivation and it is a laboratory technique rather than a study tip, but it tells you something useful. The sleeping brain is listening for what to replay. The best cue you can give it is to have made the material matter before you lay down.

    The honest limits

    Sleep consolidates what you encoded, that is, what you took in and laid down while awake. It cannot encode for you. A fact that went in weakly, with no picture, no meaning and no attention, will be replayed weakly or not at all. The techniques on this site, turning information into vivid images, placing them somewhere, saying them aloud, are about encoding well during the day. Sleep then does the second half of the work. Neither half is optional.

    So the complete picture of learning something for good looks like this. Prime yourself, by asking before you start what the material means to you, so it is tagged as important. Encode it vividly. Test yourself rather than rereading. Review it on a widening schedule. And between every review, sleep, because the night is where the brain decides what you get to keep.

    For the review schedule in detail, The Forgetting Curve (and How to Beat It) explains why the gaps widen. For the encoding side, start with Memory Techniques: The Complete Guide. For the memory systems that sleep is copying between, see Types of Memory: Working, Short-Term, Long-Term and More and Episodic vs Semantic Memory: The Difference and Why It Matters. And for the actors' version of the pre-bed habit, How Actors Learn Lines Fast: The Seven-Step Recipe (From Learning Your Lines) puts it inside a full seven-step process.

    To use the night on something you have to deliver, How to Memorise a Speech with a Memory Palace (No Notes Needed) builds a talk you can give without notes, and How to Memorise a Script for an Audition Overnight plans the evening and the morning around the sleep in between.

    Memory map

    1. Sleep is not time off from learning; it is when the brain replays, sorts and cleans. Image: a brain in a nightcap sorting papers into a filing cabinet.
    2. Jenkins and Dallenbach, 1924: two students, nonsense syllables, far less forgetting after sleep than after waking. Image: two students at desks, one asleep on his arms, with the clock showing eight hours.
    3. Forgetting is partly interference; sleep switches off new input. Image: a closed door with a queue of new words waiting outside it.
    4. Active systems consolidation: the hippocampus notebook is copied into the cortex library during deep sleep. Image: a small notebook being transcribed onto library shelves by lamplight.
    5. Replay: the same cells fire in the same order, at speed, while you sleep. Image: a film strip of the day running fast-forward inside a sleeping head.
    6. REM integrates and connects; people woken from REM solve creative problems better. Image: scattered puzzle pieces joining themselves in the dark.
    7. The glymphatic cleaning cycle: the brain flushes waste in sleep, far more than when awake. Image: a brain under a gentle shower, grey flecks washing away.
    8. Learn in the hour before bed, lightly and without stress; let the night finish it. Image: a bedside lamp, an open script, a clock at ten and a smiling face.
    9. Never trade sleep for more revision; the all-nighter undoes the consolidation and dirties the brain. Image: a student at 3 a.m. surrounded by books with a greyed-out brain above them.
    10. Space reviews across nights: one hour, one day, one week, one month. Image: four review cards separated by crescent moons.
    11. A sixty to ninety minute nap after learning consolidates like a short night. Image: a figure asleep on a sofa with a textbook closed on their chest and a clock showing ninety minutes.
    12. Sleep consolidates what you encoded; it cannot encode for you. Image: a vivid picture and a faded one going into the same night, only the vivid one coming out the other side.

    Sources

    • Mark Channon, Learning Your Lines (Nick Hern Books): https://www.nickhernbooks.co.uk/learning-your-lines
    • Mark Channon, How to Remember Anything (Hodder Education)
    • Jenkins, J. G. and Dallenbach, K. M. (1924). Obliviscence during sleep and waking. American Journal of Psychology, 35(4), 605 to 612
    • Rasch, B. and Born, J. (2013). About sleep's role in memory. Physiological Reviews, 93(2), 681 to 766. https://doi.org/10.1152/physrev.00032.2012
    • Xie, L., Kang, H., Xu, Q., Nedergaard, M. et al. (2013). Sleep drives metabolite clearance from the adult brain. Science, 342(6156), 373 to 377. https://doi.org/10.1126/science.1241224
    • Rudoy, J. D., Voss, J. L., Westerberg, C. E. and Paller, K. A. (2009). Strengthening individual memories by reactivating them during sleep. Science, 326(5956), 1079. https://doi.org/10.1126/science.1179013
    • Mednick, S., Nakayama, K. and Stickgold, R. (2003). Sleep-dependent learning: a nap is as good as a night. Nature Neuroscience, 6(7), 697 to 698. https://doi.org/10.1038/nn1078

    Loci

    Build the memory palace in this article inside Loci — store your locations, link images, and rehearse with spaced repetition.

    Try it in Loci

    Memory School community

    Learn directly with Mark inside the Memory School community on Skool — monthly lessons, practice, and Q&A.

    Join Memory School

    About the author

    Grand Master of Memory (1995). Author of three books on memory published by Hodder and Nick Hern. Creator of BBC1's Monkhouse's Memory Masters. Memory coach on Channel 4's Can I Improve My Memory?

    Read full bio →