Matuschak on evergreen notes
📚SOURCE
💎PERMANENT NOTES
🧵SYNTHESIS
Paraphrase, not quotation: the cluster of pages around 📰 Evergreen notes in his public working notes.
Matuschak separates notes that are transient (jotted, used once, forgotten) from notes that are evergreen: written to be developed over years and to add up. His complaint is that almost everything people call note-taking is the first kind, and so nothing accumulates. The aim is not better notes. The aim is better thinking, and notes are where it is kept.
He gives the properties as short rules. A note should be atomic: one idea, so that it can be linked from anywhere. It should be concept-oriented, organised around an idea and not around the book or meeting it came from. Notes should be densely linked, and he prefers associations to a filing tree. They are written for yourself, without the smoothing an audience demands.
His sharpest rule: give a note a title that states a claim.
✨KEY IDEAS
Most note-taking is transient: it captures and never compounds.
A note should hold one idea, so that it can be reused wherever that idea is needed.
Organise by concept, not by source, author or event.
Prefer associative links to a hierarchy of categories.
Titles should be complete claims, so that a title can be used in a sentence and tested for truth.
Write for yourself by default; an audience makes you tidy up thoughts that are not finished.
🕸️CONNECTIONS
🔗LINKED FROM
📝READING NOTES
How I read it
His notes are themselves written in the form they describe, so each rule is a page of perhaps a hundred words linking to the others. I clipped the main page on Friday, read outward from it for an hour this morning at Lantern Coffee, and then closed the laptop and wrote this. It arrived through the clipper and was processed within a day. That is the rule working.
The rules, and what each one does
| His rule | What it prevents |
|---|---|
| Atomic | A note that can only be used in the place it was written |
| Concept-oriented | A library of book reports |
| Densely linked | Ideas that can be found only by remembering where they were put |
| Associative, not hierarchical | The August night I deleted 600 notes |
| Claims as titles | Notes called Thoughts on estimation |
| Written for yourself | Polishing in place of thinking |
The rule I use differently
He says to write for yourself and ignore the reader. I agree about performance and disagree about context. The self who opens a note in eighteen months has forgotten the meeting, the book and the mood. If I write for present me I leave out exactly what future me needs. So my version is: write for someone with my vocabulary and none of my memory.
That reading turned into its own note today.
What it does for the newsletter
If notes are atomic and their titles are claims, then an issue is not written from nothing. It is three notes that already agree, an opening story, and a question. I have sketched that as 🧭 Newsletter issue from three notes and will try it on issue 32.
What I am not taking
- His stack of tools and his spaced-repetition practice. Not this year.
- The implied scale. He has thousands of notes. I have, as of this morning, two.
Filed on the map 04. 🗃️ Notes & Writing. Ahrens says much of this at book length in 📖 Ahrens on notes that outlive projects; Matuschak is the version I would hand to a colleague.
What it produced
- 💎 A note is finished when it can be read without its source (shared with Ahrens)
- 💎 Notes should be written for the person who has forgotten everything
- 💎 An idea is found by what it connects to and not by where it is filed (finished after Bush, two weeks later)
- 💎 An issue is three notes that already agree with each other
Added 13 March: the 400-word issue after launch night was assembled from three notes in forty minutes with almost no sleep. It works.
Related notes
🗣️IN SHORT
- In : Evergreen notes and the linked pages of his working notes
- Matuschak argues that : notes are worth keeping only if they are written to develop and accumulate, and that this depends on a few properties of each note
- by stating that : notes should be atomic, organised by concept, densely and associatively linked, titled as claims and written for oneself, because those properties let an idea be reused and built on