Learning has three stages: encoding, consolidation, retrieval. Note-taking advice obsesses over the first and the third. The middle one happens while you're unconscious, and you have almost no control over it except one enormous lever — whether you sleep at all.
What actually happens overnight
Rasch and Born's review of the field describes active system consolidation: during sleep the brain replays newly formed hippocampal traces and gradually redistributes them to the cortex for long-term storage. Not passive rest. A filing operation, running on the material you gave it that day.
The two big sleep states appear to do different jobs:
| Stage | When it dominates | Associated with |
|---|---|---|
| Slow-wave sleep | Early night | Declarative memory — facts, events, what you read |
| REM sleep | Late night, longer cycles toward morning | Procedural skills, emotional memory, recombination and insight |
The practical consequence of that table is underappreciated: cutting sleep short doesn't shave an even slice off both. Sleeping four hours instead of eight doesn't cost you half your consolidation — it disproportionately removes REM, which is concentrated in the hours you deleted. The last two hours of sleep are not the optional ones.
An all-nighter costs you twice
Sleep loss hurts on both sides of the learning event.
After learning, the consolidation window is gone — you studied and then discarded much of the filing pass.
Before learning, the encoding machinery itself is degraded. A tired hippocampus takes in less; the material never lands well enough to be consolidated in the first place. So the night before the exam you're studying at reduced capacity and not filing what you manage to absorb.
This is the single most reliable finding in the entire study-skills literature and the single most ignored. It outranks every note-taking method on this wiki. Spaced repetition and active recall are powerful precisely because they distribute learning across days — which means across nights, and the nights are doing a large share of the work. The spacing effect and sleep are not two separate findings so much as two views of one mechanism.
Sleep, insight, and the note you wrote yesterday
The "sleep on it" folk wisdom holds up better than most folk wisdom. Sleep doesn't only strengthen individual traces; it appears to reorganize them, extracting patterns and connecting items that were encoded separately. That's the plausible mechanism behind waking up with a solution you didn't have at midnight.
It has a direct implication for how you keep notes. If overnight processing works on what you encoded that day, then the notes you write before sleeping are the raw material. A day's worth of scattered captures, reviewed for two minutes at night, gives consolidation something structured to work with. And the connections that surface the next morning are exactly the ones worth writing down — which is why a capture habit that works at 7am matters as much as one that works in a meeting.
What to do with this
- Protect the night after you learn something hard. The study session is only half of the intervention.
- Front-load the difficult material earlier in the day, when encoding is strongest, and use tired hours for mechanical work rather than new concepts.
- Do a two-minute review before bed — a skim of the day's notes, not a study session. Cheap, and it feeds the pass that's about to run.
- Stop treating naps as a failure of discipline. Even short daytime sleep shows consolidation benefits in the literature.
- Never trade sleep for review time the night before a test. You're swapping a reliable mechanism for an unreliable one.
The honest caveats
Popular writing on sleep and memory outruns the evidence in places. Effect sizes vary by task and population, the neat slow-wave/REM division of labour is a simplification of an active debate, and much of the work comes from short lab-based learning tasks rather than semester-long study. Treat "sleep supports consolidation" as solid and any specific number about how much as provisional.
None of that changes the practical advice, because the intervention has no downside. Of everything in memory and focus, this is the one that costs nothing, works while you do nothing, and gets skipped first.