Source — Good Sleep, Good Learning, Good Life
A comprehensive synthesis of sleep science by Dr Piotr Wozniak (SuperMemo), originally published May 2012 and updated through October 2017. The article (~50,000 words) synthesises neurophysiology, circadian biology, and memory research into a practical framework for optimising sleep to maximise learning and creative output. Wozniak draws heavily on his decades of work with spaced-repetition software (SuperMemo) and its built-in SleepChart tool, which tracks the relationship between sleep patterns and recall performance.
Raw file: raw/good-sleep-good-learning-good-life-full.md.
Structure of the source
- Importance of sleep — why we sleep, the “disk and RAM” metaphor, what happens when sleep is lost (garbage collection, anabolic function, immune function, death from deprivation).
- two-component-sleep-model — Borbély’s two-process model: circadian-rhythm (the clock) and homeostatic sleep drive (the hourglass). The “fundamental theorem of good sleep”: optimal sleep requires both strong homeostatic pressure and ascending circadian sleepiness.
- Formula for good sleep — free-running-sleep as the gold standard; the free-running algorithm; optimising brainwork timing; why alarm clocks destroy sleep quality; sleep-inertia; health effects of shift-work and jet lag.
- Sleep habits — body clock mechanics; lark-owl misconceptions; DSPS; ASPS; segmented sleep; baby sleep; insomnia; hypersomnia; sleep apnea.
- napping — evolutionary basis; cognitive benefits; the napping “rulebook”; myth-busting; optimal circadian timing for naps; why polyphasic-sleep doesn’t work.
- sleep-and-learning — the central thesis. Neural optimisation during sleep; NREM for declarative memory consolidation (hippocampal-to-cortical transfer); REM for procedural learning and creativity; sleep deprivation’s devastating effect on recall and encoding; SuperMemo data showing sleep quality predicts learning performance.
- Sleep and society — school start times vs. adolescent circadian rhythms; sleep and the economy; sleep in the military; drowsy driving.
- Science of sleep — Borbély model in detail; three-process model; NREM stages and functions; REM functions; thermoregulation; sleep evolution; sleep across species.
Key takeaways
- The fundamental theorem of good sleep: high-quality sleep requires two conditions simultaneously — strong homeostatic sleepiness (15–19 hours of wakefulness) AND ascending circadian sleepiness. Neither alone suffices. This is the two-component-sleep-model in one sentence.
- free-running-sleep is the gold standard. Sleeping without alarm clocks, letting the body’s own circadian + homeostatic signals determine sleep timing, produces the highest quality sleep. Most people’s natural period is slightly >24 hours.
- Sleep is a neural optimiser, not mere rest. The brain’s primary use of sleep is to consolidate memories: NREM sleep transfers declarative memories from hippocampus to cortex; REM sleep consolidates procedural skills and enables creative recombination. See sleep-and-learning.
- Alarm clocks are the enemy of learning. Waking during deep NREM interrupts consolidation and causes severe sleep-inertia. “Do not wake up kids for school” — early school hours conflict with adolescent circadian-rhythm and produce chronic sleep-deprivation.
- napping restores encoding capacity. Even a short nap can “reset” the hippocampus, restoring the ability to encode new memories that declines across the waking day. The optimal nap window is the early-afternoon circadian dip (~7 hours after waking).
- polyphasic-sleep is a trap. Uberman-style schedules produce chronic sleep deprivation. REM deprivation impairs complex learning in proportion to task complexity. “Less is more” applies to sleep only in the sense that quality matters — not that you can compress it.
- Sleep deprivation compounds. Cognitive costs of lost sleep accumulate and are not fully recovered by a single good night. This is compound-interest in reverse: small daily deficits produce catastrophic long-term damage to learning, health, and creativity.
- DSPS is not laziness. It is a genuine circadian misalignment, often genetic, where the internal clock runs later than society demands. The solution is respecting the shifted phase, not fighting it with alarm clocks.
- Baby sleep should be on-demand. Imposing adult schedules on infants disrupts developing circadian rhythms. Co-sleeping supports synchronisation.
How it connects to the rest of the wiki
- sleep-and-learning — the article’s central thesis; master concept page for memory consolidation, NREM/REM roles, and the SuperMemo data.
- circadian-rhythm — the biological clock that governs sleep timing, alertness, and hormone release.
- two-component-sleep-model — Borbély’s framework that underpins the entire article.
- free-running-sleep — the practical recommendation: let the body decide when to sleep.
- napping — the science and practice of strategic daytime sleep.
- polyphasic-sleep — debunked as a learning-optimisation strategy.
- sleep-inertia — the cost of waking at the wrong time.
- sleep-deprivation — effects on learning, health, and mortality.
- delayed-sleep-phase-syndrome — circadian misalignment, especially in teenagers.
- compound-interest — sleep debt as negative compounding; sleep quality as positive compounding on learning.
- first-principles-thinking — Wozniak’s approach: decompose sleep into circadian + homeostatic components, then reason from the physiology rather than folk wisdom.
- inversion — “How would I guarantee poor learning?” → disrupt sleep, use alarm clocks, ignore circadian timing, pull all-nighters.
- nudge-theory — school start times and workplace napping policies are choice-architecture problems.
- mental-models — the two-component model, the disk/RAM metaphor, and the garbage-collection analogy are all mental models for understanding sleep.
- high-agency — choosing free-running sleep despite social pressure is a high-agency move.
Confidence notes
- The two-component model (Borbély) is well-established and replicated across decades of research. High confidence.
- Neural optimisation claims (NREM declarative, REM procedural) reflect mainstream consensus as of 2017, though the field is still refining the details. High confidence on the broad claims, medium on the mechanism specifics.
- The SuperMemo data is proprietary and self-reported (Wozniak’s own sleep logs and user data). The patterns are consistent with published research, but the specific quantitative claims should be treated as illustrative rather than definitive.
- Polyphasic sleep debunking aligns with mainstream sleep medicine. High confidence.
- Some specific numerical thresholds (e.g., “15–19 hours of wakefulness”) are Wozniak’s heuristics, not precise physiological constants.
Sources
raw/good-sleep-good-learning-good-life-full.md— the ingested document.