Self-teaching, spaced repetition, and why books don't work
Claims that “books don’t work” for learning spark a broader debate over how people actually acquire durable knowledge. Commenters contrast passive, front-to-back reading with practices like active recall, spaced repetition, problem-solving, and metacognitive strategies, arguing that books are effective only when used within such frameworks. Others highlight domain differences (e.g., math, programming, medicine, languages), the motivational burden of spaced repetition, and emerging tools — from SRS apps to LLM-powered tutors — as attempts to better align learning media with how memory and understanding develop.
Books as Learning Tools
- Many commenters say books “work” extremely well for them, especially for deep, contextual understanding (history, programming, theory).
- Others argue the traditional linear, front‑to‑back textbook format is inefficient and often badly written; the format’s dominance is seen as historical path dependence, not optimal pedagogy.
- Several note that learning from books fails when readers lack metacognitive skills: how to question, summarize, self‑test, and decide what to re‑read.
- Suggestions: skim first, use the index, ask questions in the margins, re‑read with notes, and drop bad books quickly.
Spaced Repetition & SRS
- Many report strong long‑term retention from spaced repetition, especially for vocabulary, terminology, and dense factual domains (languages, medicine, aviation).
- Others find SRS less effective or frustrating for items that don’t naturally “stick” (e.g., birthdays, family tree relations) or see cards becoming “leeches.”
- There’s debate whether SRS mainly helps “know‑what” (facts) versus “know‑how” (skills). Some insist it’s mostly for atomic facts; others argue you can decompose complex concepts into many small prompts.
- Limitations raised: difficulty representing many‑to‑one / compound skills on flashcards; some are building tools for sentence‑level or compound prompts, or leveraging LLMs for adaptive questions.
Practice, Difficulty, and “Clicking”
- Strong consensus that deep understanding in math, physics, and programming requires sustained practice, problem‑solving, and “pushing until it clicks,” not just reading.
- Some emphasize that interest and motivation are critical; without them, even good methods fail. Others resist the idea that some people are simply “unteachable.”
- Analogies to music: reading about an instrument isn’t enough, but practice itself must be skillful, not just repetition.
Format, Tools, and Alternatives
- Incremental reading, interactive “books” with embedded review, and dynamic question generation are praised as promising hybrids.
- Several are building or using SRS products for programming languages, language learning, and other domains; opinions differ on whether programming is better learned by flashcards or by hands‑on coding.
- Critiques of modern textbook economics include frequent new editions, bundled online components, and alleged perverse incentives in academia.