In three sentences
A Zettelkasten is a networked accumulation system. Each card carries one concept and a unique address, cards point at one another by reference, and decades of accumulation turn the pile into a web of knowledge. The sociologist Luhmann worked this way for life with roughly ninety thousand cards, and in 1981 described the system as his communication partner: it answers you with old cards you had forgotten, rather than merely storing them. The method itself is centuries older than he is. He is its best-known practitioner and the person who theorised it.
Figure 1 · The slip box mechanism: atomic cards, unique addresses, cross-references accumulating into a web, and the emergent layer Luhmann called a communication partner. The figure of roughly ninety thousand cards is an approximation from archival reconstruction, not a census. Redrawn. View full size ▸Share PNG ▸
From the 1950s onward Luhmann accumulated physical cards. Each held one concept, atomically, with a unique positional address whose numbering alternated between digits and letters so that a new card could always be inserted between any two existing ones. Cross-references to other cards were written on the card by hand, unconstrained by any tree, so any branch could point at any other.
The portable lesson is not "use cards". It is three structural choices: atomicity, so that a reference has a precise target; a unique address, so that pointing is possible at all; and cross-reference without copying, so that a concept exists in exactly one version.
In digital tools those map onto atomic notes, stable links and embedded references. Drop any one and the web starts to rot: concepts get smeared through long documents where nothing can cite them, or the same fact is copied into several versions that then drift apart.
It also has a design domain. This is a long-term system for accumulation across time, not a tool for digesting a single input. Work out which one you are building before deciding which parts to borrow.
Verification record
Records how far each claim was checked, not a guarantee that it is true.
Two findings. The 1981 experience report is frequently miscited to its 1992 reprint; this page uses the original venue. And the widely repeated figure of about seventy books is round-number lore. The same reconstruction records a list of more than five hundred titles at his death.
Verification layer · Counter-examples and failure boundaries
Three misuses all mistake the best-known practitioner for the inventor.
Third, this is a long-term networked system for accumulation across time, not a tool for digesting a single input. Comparing it against linear real-time capture methods compares two different design goals.
One concept, one node: why linking instead of copying holds up a whole system
Our knowledge base runs on the same axiom: one concept, one node, nodes point at each other through links, and the same content is never copied a second time. The discipline is not fastidiousness. Copying lets one fact grow two versions, and then which one is true? Linking means an update happens in exactly one place and every citing location follows automatically. The full case carries the criteria for splitting nodes, how large a single concept may be, the rules for cross-citation, and how a violation is cleaned up.
02
Trigpoint PassField record · 2026-07
When a node counts as one card: the splitting criteria, and integrity first
A research report recommended building 5 nodes on one topic. Following it would make the library finer grained, but we have a rule: a complete argument may not be broken apart for the sake of tidy classification. Which of the 5 should be built?
Trigpoint Pass content · includes applicable criteria
How orphan cards are handled: the monthly settlement of nodes with no citations
Three governance documents say nodes with no citations are cleared every month, yet the main library has never kept a single scan result. A rule written but never run; what triggered the few clean-ups of disconnected nodes that actually happened?
How to read the source notes “Bibliographic details checked” means author, year, title and venue were confirmed. It does not mean the source's claims were checked. “Matched against the full text” means the claim was compared with the text of the source itself. Definitions as of 2026-09-20.
[1] Luhmann, N. (1981). Kommunikation mit Zettelkästen. In Öffentliche Meinung und sozialer Wandel. Opladen: Westdeutscher Verlag, 222–228. https://doi.org/10.1007/978-3-322-87749-9_19↩
Tell the editor privately: a correction, a counter-example or where it applies
An editor reviews submissions once a week. Once a note is handled, its summary and response are listed on the discussion and response record; nothing is published before then. No sign-up is needed, and the nickname may be left blank.
