On Zettelkasten Productivity

Edmund

Let's start with some data from reconstructions of Luhmann's second, sociological Zettelkasten:

• Permanent Notes: about 67,000 cards.​
• Literature Notes: about 15,000 entries (mostly books and articles).
• Books he authored: a bit over 70.

This means as an estimation, not a directly documented figure:

• For each book or article he read, he only creates on average ≈ 67,000/15,000 ≈ 4.5 Permanent Notes per book or article. ​
• He processed on the order of 15,000 books and articles in depth, with perhaps roughly half of those being books, so about 7,500 books.
• So the ratio is approximately: books read per book written ≈ 7,500/70 ≈ 100.
• Productivity expressed as output per unit of input: ≈ 70/7,500 ≈ 1 %

The “books read per book written” ratio is a quiet reminder that visible output sits on top of a large, invisible foundation. A writer who publishes one substantial book for every hundred they read is not inefficient; they are distilling, selecting, and discarding far more than they keep.

This flips a common obsession with productivity: instead of asking “How many books have I written?”, one might ask “How many have I truly digested deeply enough that a book could grow out of them?” Measured this way, high productivity is less about speed and more about the density of thought each finished work contains.

zettelsan

The I/O ratio of 100 books read per a book published is an interesting heuristic.

There was this very prolific author of science/nonfiction whom I greatly admired (and who passed away relatively recently). His advice to budding authors was similar: you have to read at least 100 books to write one book of some quality.

He was a modern-day polymath, pathologically collecting on the order of 100 thousand books in his personal library. (He bought a 6-story building in Tokyo to keep a faction of his collection.)

Maybe one book written per 100 books read is intuitively what it takes for those people to condense knowledge worthy of output.