You finish a great non-fiction book on a Sunday afternoon, dog-ear a few pages, and feel sharper for it. A month later a friend asks what it was about. You name the author, gesture at the cover, and find yourself summarizing two anecdotes and a single line from the introduction. The central argument, the thing that changed how you saw the problem, is gone.

This is not a memory defect. It is what reading does by default. When you read passively, encoding is shallow, retrieval never happens, and the small spike of comprehension you felt at the time decays along a curve that Hermann Ebbinghaus mapped in 1885. Without active processing, most of a book is gone within a month. The good news is that the fix is mechanical. Below are seven techniques drawn from cognitive science, ordered so they reinforce each other. Use two or three on your next book and the difference will be obvious by the second week.

1. Active recall

Active recall is the practice of trying to retrieve an idea from memory before you check the source.

Why it works. Retrieval is not a neutral readout; it modifies the memory it touches. In the testing effect, demonstrated repeatedly by Roediger and Karpicke, students who quiz themselves on material outperform students who reread it, even when the rereaders feel more confident. Each successful retrieval strengthens the trace and makes the next retrieval easier.

How to do it.

At the end of every chapter, close the book and write down the three most important ideas in your own words. Do this before you look at your highlights.
Turn headings into questions. A section called "The endowment effect" becomes "What is the endowment effect, and what experiment first demonstrated it?" Answer from memory.
When you feel the urge to reread a passage, pause. Try to reconstruct it first. Reread only to correct what you got wrong.
If nothing comes to mind, that is information. You have located a gap, which is exactly where attention pays the highest interest.

2. Spaced repetition

Spaced repetition is the practice of reviewing material at expanding intervals — a day later, then three days, then a week, then a month — instead of cramming it once.

Why it works. Each review pulls a memory back from the edge of forgetting, and the more times you do this, the slower the next decay. Spacing flattens Ebbinghaus's forgetting curve. Massed practice feels productive because retrieval is easy when the material is fresh, but that ease is a sign you are learning very little.

How to do it.

After finishing a chapter, write 5 to 10 question-and-answer pairs covering the points you most want to keep.
Review them the next day, three days later, a week later, and a month later. Each time you answer correctly, push the next review further out. Each time you miss one, bring it back to the start.
Keep cards short. One idea per card. If a card has a paragraph on the back, split it.
The easiest way to do this without setting up Anki and tweaking decks is Neuraknow, which turns what you read into spaced review questions and schedules them for you.
Five minutes a day is plenty. The point is consistency, not volume.

3. The Feynman pass

Named after the physicist Richard Feynman, this is the practice of explaining what you just read, in plain language, as if to a curious friend who has no background in the subject.

Why it works. Teaching is retrieval with the difficulty turned up. To explain something clearly, you have to rebuild the argument from memory, choose ordinary words for technical ones, and notice the joints where your understanding is thin. Vague phrasing is a flag that you have a phrase but not the idea behind it.

How to do it.

Pick one chapter or one concept. Open a blank page.
Write an explanation a smart twelve-year-old could follow. No jargon. If you must use a technical term, define it.
Mark the sentences that came out fuzzy. Those are your gaps.
Go back to the book, repair the weak spot, and rewrite that section without looking.
If you have someone willing to listen, explain it out loud. Their first puzzled question is usually worth more than the chapter itself.

4. Take notes by hand or in your own words

Whether you use a notebook or a plain text file, the rule is the same: notes are translations, not transcriptions.

Why it works. Craik and Lockhart's framework of depth of processing captures the basic finding: what you remember is determined by what you did with the information, not by how long you spent looking at it. Copying a sentence verbatim engages almost no semantic processing. Rephrasing forces you to extract the meaning, which is what the brain stores. Studies of laptop note-takers find that they tend to transcribe and recall less than handwriters who, constrained by speed, are forced to compress.

How to do it.

Close the book before you write a note. If you cannot remember the idea, you do not understand it well enough to record it yet.
Write notes as full sentences, not fragments. A claim with a verb is something you can later disagree with; a phrase is just decoration.
Keep notes atomic. One idea per note, with a short title. This makes them findable and recombinable later.
Highlight sparingly. Highlights feel like progress and produce almost none. If you must highlight, return at the end of the chapter and convert each highlight into a sentence in your own words.

5. Connect new ideas to what you already know

Memory is not a warehouse where new items sit on empty shelves. It is a network where new nodes only stick if they wire into existing ones.

Why it works. Working memory holds only about four chunks at once, but a chunk can be arbitrarily large if it is well connected. Long-term storage is built when new material is integrated with prior knowledge — the more hooks an idea has, the more retrieval routes lead back to it. Isolated facts are fragile; connected facts are robust.

How to do it.

For every new idea, ask three questions. What does this remind me of? What does it contradict? Where in my life or work could I have used this last year?
Keep a running list of "ideas in conversation." When a new book argues with an old one, write the disagreement down explicitly. Forced contrast is a powerful encoder.
If you keep notes in a system like Obsidian or a plain folder of files, link new notes to at least two existing ones before you save.
When an analogy comes to mind, write it down even if it is imperfect. A bad analogy you can refine beats a clean fact you cannot recall.

6. Sleep on it

The night after a serious reading session is part of the reading session.

Why it works. During slow-wave sleep, the hippocampus replays the day's traces and the cortex gradually integrates them into long-term storage. This is not metaphor; it is observable in neural recordings. People who learn material and then sleep retain substantially more than those who learn the same material and stay awake the same number of hours. Sleep deprivation degrades both the encoding that happens before bed and the consolidation that should happen after.

How to do it.

Do your hardest reading earlier in the day, when attention is sharpest, but plan to revisit the material briefly before sleep.
In the last fifteen minutes of the day, skim your notes from that day's reading. No phone, no new input, just a quick rehearsal.
Protect seven to eight hours. The cost of cutting an hour to read more is almost always negative; you encode less of what you already read.
If a problem you encountered won't resolve, write it down and put it aside. Solutions often arrive the next morning, which is consolidation doing its job.

7. Use the book — apply ideas in real life within a week

If a book is worth reading, at least one idea in it should change something you do. Find that idea and act on it before the week is out.

Why it works. Application is the deepest form of encoding because it forces the brain to retrieve, adapt, and check the idea against reality. The feedback you get — it worked, it didn't, it needed adjusting — fuses the concept to a specific situation, which is far more memorable than any abstract definition. This is the encoder of last resort: even if you skipped every other technique, applying an idea once will preserve it longer than rereading the chapter five times.

How to do it.

At the end of each book, write down one sentence that begins, "Because of this book, I will…" Make it concrete, small, and time-bound. "I will run a five-minute pre-meeting checklist this Wednesday" beats "I will be more prepared."
Schedule the action the same day you finish the book. Calendars remember; intentions don't.
After you try it, write a two-line debrief. What happened? What would you change? File it with your notes on the book.
If nothing in the book is actionable for you right now, that is also a finding. It tells you something about the book, or about your current questions, and saves you from going through the motions.

What to do tomorrow

You don't need all seven of these to start. Pick two. The simplest pairing is active recall plus spaced repetition: at the end of every chapter, write five questions, and review them on a schedule. That alone, done for a month, will change what reading feels like — less like pouring water through a sieve, more like building something you can return to. Add the Feynman pass for the books you most want to keep. The rest will follow because they have to: once you start retrieving, you naturally start connecting, and once you start connecting, you start using. Reading was never the bottleneck. Remembering was, and remembering is a skill.