The method in one sentence
Capture linearly while the information is arriving, then map it afterwards as a separate pass — because deciding structure and receiving information compete for the same attention, and doing both at once does neither well.
Most mind-map note-taking advice tells you to map live. That advice is why most people try it once, fall behind in the first ten minutes, and conclude the format does not work for them.
Why live mapping fails
Every item you capture demands a structural decision: where does this go? While reading a book you can pause. In a lecture, a meeting, or a conversation you cannot, and the speaker continues while you decide whether point four is a sub-point of two.
You fall behind. You start putting things wherever is nearest. The map ends up structured by what happened to be adjacent rather than by meaning — which is worse than a linear note, because it looks organised and is not.
Split the two jobs and both work.
Pass one: capture, unstructured
Whatever is fastest. Bullet points, a text file, paper. No hierarchy, no formatting, no judgment about relevance.
The only discipline: mark the things you had a reaction to. A star, an arrow, anything. "That contradicts the earlier point." "I don't believe this." "This connects to the other project." Those reactions are the highest-value thing in the note and they evaporate within a day — the facts you can look up again, the reactions you cannot.
Pass two: map, ideally same day
Now build structure. The gap matters: long enough that you have forgotten some of it, short enough that you can still reconstruct it. Same evening is close to ideal.
- Centre it on the question the material answered, not the title of the session. "Why does the deployment keep failing?" beats "Ops meeting 14th."
- Pull out three to six areas from what you captured. Derive them from the material — not from a standard structure you brought with you.
- Place items under them. Some will not fit anywhere, which is informative: either you are missing an area, or that item was not actually part of this subject.
- Discard. Most captured notes contain material that seemed important in the moment and is not. Keeping everything is how maps become unreadable.
- Add what you now realise. Pass two is not transcription — the understanding generated between the two passes belongs in the map, and it is the main reason the method works.
Twenty minutes for an hour of material is typical.
The forgetting is doing work
The uncomfortable part: you should be reconstructing from memory as much as reading your captured note.
What you cannot recall is either unimportant or not yet understood, and noticing which is which is the actual learning. Copying your linear note into a tree shape while reading it word for word produces a prettier note and no understanding — this is the most common way the method gets performed rather than used.
Keep the raw capture attached
Do not throw away pass one. The map holds structure; the raw note holds detail, quotes, numbers, and the things you did not know were important yet.
They belong together. Where this normally breaks is that the map lives in one tool and the note in another, and within weeks they have drifted and you trust neither.
In Minty Path a note attaches to the branch it concerns, so the raw capture sits inside the structure rather than beside it. There is no separate filing decision — the branch is the filing. Quick notes live on the branch itself; longer titled notes sit alongside; the All Notes view reads across every Growth Path when you want the flat list, and each note links back to the branch that gives it meaning.
Where the method genuinely fits
Lectures and courses. Strong. The gap between passes maps onto the spacing effect neatly.
Books. Strong, per chapter or per section rather than per book.
Meetings. Good for the understanding, poor for the record. Keep the linear minutes; map the parts you need to think about.
Research across sources. The best fit, because a map can hold several sources against each other in a way sequential notes cannot.
Reference material you will only search. Skip it. Mapping something you will never reload is wasted effort.
Three worked shapes
A technical talk. Centre: "what problem is this solving?" Branches: the problem, their approach, what it costs, what it assumes, where it would not apply. That last branch is the one you will actually use.
A chapter. Centre: the chapter's claim, phrased as a question. Branches: evidence given, mechanism proposed, objections raised or dodged, what it changes if true.
A messy project meeting. Centre: what has to be decided. Branches: what was agreed, what was assumed without saying so, what is still open, who is blocked on whom. The second and fourth branches are almost never in the minutes.
Doing it once, properly
The method is only convincing once you have felt the difference between transcribing a note and reconstructing one. Take something you took linear notes on in the last week and map it from memory.
Open the workspace without an account — it takes about fifteen minutes, and it settles the question better than any description of the method will.