How to Learn

A Manual System for Durable Knowledge — No Software Required

Based on cognitive science research  ·  2026

· 01

The Core Problem

Retention failure is not a character flaw. It is a predictable outcome of how human memory actually works — and how most learning environments are designed against it.

Hermann Ebbinghaus mapped the decay in 1885, and a 2015 PLOS One replication confirmed it holds across modern subjects and materials.[1] The curve is steep.

50% forgotten within 20 minutes[1]
90% gone within a week without reinforcement[1]
The Forgetting Curve
100% 75% 50% 25% 0% Day 0 Day 1 Day 7 Day 14 Day 30 with spaced repetition 100% ~50% ~20% ~10%

Storage Strength vs. Retrieval Strength

Robert Bjork (UCLA) distinguishes two separate memory properties.[2] Storage strength is how deeply a memory is embedded in long-term memory — it builds through deliberate practice and remains stable. Retrieval strength is how easily you can access it right now — it rises with recent exposure and drops when unused.

Skills used daily have permanently high retrieval strength. Knowledge consumed once and never revisited never builds storage strength — retrieval drops within days.

Transfer-Appropriate Processing

Memory retrieval is strongest when the context of retrieval matches the context of encoding.[3] Information encoded in one context — relaxed podcast listening — is harder to retrieve in a different context: a conversation, a work problem, a piece of writing. This is why passive consumption alone is structurally insufficient.

· 02

The Big Three Methods

Three interventions — identified through decades of cognitive science research — stack multiplicatively. Spaced repetition handles when to review. Active recall handles how to review. The generation effect handles how to encode in the first place.

Spaced Repetition
Review just before forgetting — the optimal moment to reset the curve
80%+ retention at months

Each retrieval just before a memory would fade resets and extends the forgetting curve. A 2024 medical study confirmed spaced repetition outperforms cramming for long-term retention across materials and timescales.[5]

1d → 3d → 1w → 1mo → ?
Active Recall
Test yourself — retrieval strengthens the memory trace, not just the readout
+23% vs re-reading

Re-reading produces almost no durable memory. A 2024 study found active recall students scored 23% higher than re-readers on final exams.[4] The act of retrieval itself modifies and strengthens the memory trace — it is not passive.

The Generation Effect
Information you create yourself is encoded more deeply than information you receive
+40% vs passive reading

Memory for information you generate is ~40% stronger than for passively read information, confirmed by meta-analytic review.[7] Paraphrasing, summarizing, and explaining to others all qualify — the act of production is the mechanism.

· 03

Passive Audio While Living Your Life

Podcast learners scored higher on immediate knowledge tests than textbook readers in a randomized trial.[8] Driving while listening does not significantly reduce knowledge acquisition.[9] This is encouraging — but there is a ceiling.

Mode 1-week retention
Passive listening only 10–20%
Passive + generation sprint 40–60%
+ Spaced repetition 80%+
Physical Activity Bonus

Walk while you listen. Physical activity while listening increases long-term memory encoding by 22% compared to passive seated listening, with 18% higher engagement throughout.[12]

The 5-Minute Fix

Without active follow-up, passive listening retains only 10–20% after one week.[10][11] "Active production" after passive listening — summarizing, generating questions, writing in your own words — dramatically improves this.

Practical rule: within 30 minutes of listening, spend 5 minutes writing the 3 things you want to keep, in your own words.

· 04

Knowledge Architecture

Facts with no connections to other knowledge have no retrieval handles. The brain stores and recalls through association networks — isolated facts are structurally disadvantaged from the moment they are encoded.

Isolated Facts vs. Connected Notes
Isolated Facts
fade within a week
Connected Notes
retrievable for years

The Zettelkasten Principle

Niklas Luhmann published 50 books and 600+ articles using a 90,000-card linked note system.[13] Every note written in his own words (generation effect). Every note linked to at least one existing note — building a retrieval network. The system compounds over time because every new note has more potential connections than the last.

The Linking Rule

For every new thing you learn, add one line: This connects to ___ because...[14]

Any notes tool works: Obsidian, Notion, Apple Notes. Each new item: 2–3 sentences in your own words plus one explicit connection. The connection is the retrieval handle. Roman supply chain logistics connects to modern SaaS distribution. That link is what makes it findable later.

· 05

Sleep — The Consolidation Layer

Sleep is not rest from learning. It is when learning is completed. New information is encoded into the hippocampus during the day — fast, but fragile. During slow-wave sleep, the hippocampus reactivates memory traces and transfers them to the neocortex for long-term storage.[15][16]

Sleep following learning produces superior memory performance compared to equivalent wakefulness.[17] Sleep deprivation is a learning tax with no compensating technique.

Daily Learning Timeline
CONSUME
WORK
GENERATE
SLEEP
Morningcommute & exercise
Afternoonapply, work
Eveningnotes + Anki cards
Nightconsolidation
Timing Insight

Review before sleep — the consolidation happens during the sleep that follows. Brief eyes-closed rest (10 minutes) after learning also shows memory benefit: reactivation begins even during short wakeful rest.[18]

· 06

The Complete System

The three-phase daily loop integrates everything above into approximately 15 minutes of added time per day, layered onto existing habits with no dedicated study blocks required.

01
Consume
During dead time — 0 added minutes
  • Podcasts, audiobooks, or video audio during commute, exercise, cooking, chores
  • Walk or move when possible for the 22% encoding boost
  • No time penalty: this replaces idle listening, not productive work
02
Generate
5 minutes immediately after — before switching to anything else
  • Write or voice-memo 3 things in your own words
  • Add one connection: "This relates to X because..."
  • Do this before the short-term trace decays (within 30 minutes)
03
Review
5–10 minutes Anki daily[6]
  • Convert Phase 2 notes into 3–5 Anki cards each session
  • Anki schedules all reviews automatically with the FSRS algorithm
  • Mature deck: under 5 min/day once intervals grow long
~15 min/day

Total added time — the consume phase costs nothing, it replaces idle time

· 07

Trade-offs at a Glance

Different methods serve different goals. This table maps the honest costs, effort levels, and appropriate use cases.

Method Time cost Effort Retention Best for
Passive listening only 0 Low 10–20% at 1 week Broad exposure only
Passive + generation sprint +5 min Low ~40–60% Daily habit, any content
Spaced repetition (Anki) +10 min/day Medium 80%+ at months Facts, quotes, frameworks
Linked notes (Zettelkasten) +2 min/note Medium Compounds over time Building a knowledge network
Application same day 0 added None Very high Directly applicable skills
· 08

Edge Cases & Limitations

Every system has conditions where it performs poorly. Knowing these protects against misapplication.

  • Anki card quality matters more than quantity. One good card beats ten bad ones. Cards should test recall of one specific thing, not recognition of a long passage.

  • The generation effect requires prior knowledge to connect to. It works poorly on completely foreign domains with no existing knowledge hooks. Build the base first through broader reading or listening.

  • Sleep deprivation fully negates the system. No technique compensates for chronic poor sleep — consolidation requires slow-wave sleep, which deprivation eliminates.

  • Passive audio at high speed (1.5–2x) may reduce comprehension. There is no strong research on retention at high playback speeds. Comprehension is a prerequisite to encoding — you cannot retain what you did not understand.

  • Spaced repetition works for factual and declarative knowledge. It does not substitute for practice-based skill acquisition — sales calls, writing, building. For procedural skills, doing is the review.

· 09

References

  1. [1]Murre, J.M.J. & Dros, J. (2015). Replication and Analysis of Ebbinghaus' Forgetting Curve. PLOS ONE. journals.plos.org
  2. [2]Bjork, E.L. & Bjork, R.A. (2011). Making things hard on yourself, but in a good way. Psychology and the Real World. bjorklab.psych.ucla.edu
  3. [3]Transfer-Appropriate Processing. Wikipedia. en.wikipedia.org
  4. [4]Spaced repetition and active recall improves academic performance among pharmacy students. ScienceDirect (2024). sciencedirect.com
  5. [5]Implementation of a spaced-repetition approach to enhance undergraduate learning in paediatrics. PMC. pmc.ncbi.nlm.nih.gov
  6. [6]A Cohort Study Assessing the Impact of Anki as a Spaced Repetition Tool on Academic Performance in Medical School. PMC (2023). pmc.ncbi.nlm.nih.gov
  7. [7]Bertsch, S., et al. (2007). The generation effect: A meta-analytic review. Memory & Cognition. link.springer.com
  8. [8]A Multimodal Evaluation of Podcast Learning, Retention, and Attention in Medical Trainees. PMC (2022). pmc.ncbi.nlm.nih.gov
  9. [9]The impact of driving versus undistracted listening on podcast knowledge acquisition. PMC (2025). pmc.ncbi.nlm.nih.gov
  10. [10]Influence of Active Production Versus Passive Consumption of Podcasts on Learning Outcomes. PMC (2025). pmc.ncbi.nlm.nih.gov
  11. [11]Podcasts and Informal Learning: Exploring Knowledge Acquisition and Retention. MDPI Education (2024). mdpi.com
  12. [12]Why Your Brain Loves Listening — The Research Behind the Science. uStudio. ustudio.com
  13. [13]Introduction to the Zettelkasten Method. zettelkasten.de. zettelkasten.de
  14. [14]A Comparison of Self-Explanation and Elaborative Interrogation. PubMed. pubmed.ncbi.nlm.nih.gov
  15. [15]About Sleep's Role in Memory. Physiological Reviews (2012). journals.physiology.org
  16. [16]Sleep smart — optimizing sleep for declarative learning and memory. Frontiers in Psychology (2015). frontiersin.org
  17. [17]Memory consolidation during sleep: a facilitator of new learning? ScienceDirect (2025). sciencedirect.com
  18. [18]'Sleep-dependent' memory consolidation? Brief periods of post-training rest. PMC (2021). pmc.ncbi.nlm.nih.gov