Readwyn 0.1.0

How to take reading notes with a Zettelkasten

Niklas Luhmann, the sociologist behind the best-known Zettelkasten, didn't write in it while he read. Between 1952 and early 1997 he filled about 90,000 slips, and the Niklas Luhmann Archive in Bielefeld says he made them in a second step, from reading notes he had taken first. Taking notes for a Zettelkasten can follow the same two stages. While you read, note what the text says and where it says it; afterwards, write the cards: one thought each, in your own words, filed beside a card it answers and linked to others.

The first box, 1952 to 1960

Luhmann started the box as a trainee lawyer in Lüneburg and kept it going as an official at the education ministry of Lower Saxony in Hanover. His first collection, which the archive calls Zettelkasten I, holds about 22,000 slips from roughly 1952 to 1960, though notes on world society kept going into it until about 1973. He wrote on thin DIN A6 paper, often cut from used A4 sheets, and mostly on one side, so a slip could be read without lifting it out of the drawer.

These early notes stay close to the books. The archive describes them as running text that follows the source, and topic sections often open with a list of things to read. The bibliography of this box was itself a list: just under 900 titles, sorted by author, on about 120 slips. The box had 108 numbered sections, among them organisation and law, authority, and causality.

The second box, from 1961

During a stay at Harvard, where he sat in on Talcott Parsons's sociology courses, Luhmann started again. The first box had stayed in Germany. Zettelkasten II grew to about 67,000 slips by early 1997, and its notes got shorter and more like theses.

His reading notes changed with it. From about 1964 he gave every reference a slip of its own (around 11,000 of them), and on the back of many he wrote a brief excerpt in keywords. From that excerpt he wrote the real cards. If an excerpt ran long, it went on extra slips marked in red with the author and the year.

Every card got a fixed number that never changed. A card that took up one point of card 1/1 became 1/1a and was filed between 1/1 and 1/2. A point raised on 1/1a became 1/1a1, and so on, until some numbers ran to 13 characters. He also wrote the numbers of related cards in other parts of the box onto a card, often with a link back. The archive's sample count puts these references at 25,000 to 28,000 in the second box alone. The way in was a keyword register of about 3,200 entries, and as a rule it named no more than four cards for a keyword: from there, the links led on.

The method after Luhmann

Luhmann wrote about the box himself, in an essay the archive dates to the early 1980s, “Kommunikation mit Zettelkästen”. In Manfred Kuehn's English translation, bibliographic notes taken from a text go into the box, while the works he has actually read get slips in a separate box, so he can tell what he read from what he only meant to read. The slip for a book he'd read could also carry the numbers of the cards it led to. He thought excerpts were only useful for especially remarkable formulations.

The labels most people search for came much later. Sönke Ahrens's How to Take Smart Notes, first published in 2017, uses the names literature notes and permanent notes for two kinds of note. In the appendix to the 2022 edition, he says one app can hold both, provided the types stay visibly apart and only the permanent notes form the linked network. He also advises writing each new note, where you can, as a direct reply to a note that's already there. Set against Luhmann's box, the literature notes are his keyword excerpts on the back of a reference slip, and the permanent notes are his numbered cards.

Ebook highlights belong to the first stage. A list of highlights with page numbers gives you the source's words and where to find them again, which is the job Luhmann's excerpts did. In Readwyn, the Notebook exports the highlights and notes of one book as a Markdown or plain-text file, with the title, the authors, each chapter title, and every highlight with its page and its note. That file is the raw material for a literature note. Ahrens says literature notes are short (as Andy Matuschak quotes him), so you'd still cut the list down to what you might use. A highlight can't go into the box as it stands, because it's someone else's sentence: the permanent card restates one idea in your words and carries the number, or the link, of the card it replies to.

Reading Luhmann's own cards

You can look at the original slips. In April 2019 the archive put the first part of the box online, and Heise reported that about 4,000 cards of the first box were fully transcribed by then, with about 35,000 more up as images. The work is a project of Bielefeld University and the Cologne Center for eHumanities, planned to run for 15 years.