Highlighting a book produces a list of sentences the author wrote. It does not produce a record of what the book argues, which is why re-reading your highlights six months later is such a strange experience: every line is familiar and none of them add up to anything.
A map fixes that by forcing the one thing highlighting never asks for: where each idea sits relative to the others. The method here takes about twenty minutes after finishing a non-fiction book, and produces something you can actually revisit.
Map after, not during
Read the book first. Mapping while reading turns you into a stenographer, following the author's chapter order and reproducing their structure, which is the structure of an argument being delivered, not the structure of the idea.
So: read normally, mark the passages that matter (a highlight, a page corner, a line in a notebook), and map afterwards from those marks plus memory. What you can't remember mostly wasn't load-bearing, and the map is better for losing it.
For a dense textbook you'll want a per-chapter pass instead. That's a different job; see using mind maps to study.
The four-branch skeleton
Put the book's central claim in the middle, in one sentence, in your words. If you can't write it, that's the finding, and the map is already worth the twenty minutes.
Then four branches, always the same four:
1. Claims. The three to six things the book asserts. Not chapters; chapters are packaging. Under each claim, the evidence the author gives for it.
2. Evidence and examples. The studies, stories and numbers you'll want to cite. Keep the page number on each. This is the branch that turns a book you read into a book you can use in an argument.
3. Objections. Where you disagree, where the evidence is thin, what the author didn't address. Most people skip this branch and it's the one that makes the map yours rather than a summary.
4. Actions and applications. What changes because you read this. Three things at most. Anything that survives here should leave the map and go into whatever system actually runs your life.
Everything in the book that fits none of these four is trivia. Let it go.
Keeping quotes honest
One rule, and it matters more than any layout decision: mark the author's words as the author's.
Six months on, you will not remember whether a striking line was theirs or yours. Put quotes in quotation marks with a page number; put paraphrases in plain text; put your own thinking in a note or a differently-styled topic. It costs nothing during capture and saves you from quoting yourself as an authority, or worse, publishing someone else's sentence as your own.
Notes on any topic for the long quotes, icons or styling to separate the author's words from yours, and crosslinks to tie an objection to the claim it undercuts. Folders keep a year of books findable.
What it doesLinking books to each other
The value of book maps compounds only if they connect. Two low-effort habits do most of that work:
Crosslink contradictions inside a map. When the author's chapter 7 undercuts their chapter 2, draw the line. That tension is the most interesting thing you will take from the book.
Keep one running map across books. A single map per theme ("attention", "pricing", "how habits form") with a branch per book underneath. When two authors disagree, you can see it, which is the only reliable way to notice you have read six books that all say the same thing.
Name maps consistently (Author — Title) and put them in a folder. Search
across your library is what turns forty maps into a reference; inconsistent
names are what stops it working.
Fiction, briefly
The four-branch skeleton is for non-fiction. For a novel, the useful branches are different: characters and their relationships, timeline, motifs, the questions the book leaves open. The relationships between characters matter more than the hierarchy, so expect to lean on crosslinks or a free-form layout rather than a neat tree.
Most people don't need this. Map a novel only if you're studying it or writing about it.
The revisit, which is the whole point
A map you never reopen is a slightly more elegant version of highlights you never reread. Two occasions to open it, and both are cheap:
- When you're about to write or argue about the topic. Search your library for the term, and the evidence branch does the work.
- Once a quarter, collapse every branch to the first level and try to reconstruct the claims from memory. Anything you can't recall, expand. Five minutes per book, and it's the difference between having read something and knowing it.
That second one is retrieval practice, and it is the only part of any note system with real evidence behind it. Everything else, the layouts, the themes, the perfect folder structure, is arrangement.