luhmann1981-slip-boxes paper

Kommunikation mit Zettelkästen: Ein Erfahrungsbericht (Communicating with Slip Boxes: An Empirical Account)

Niklas Luhmann

in Baier, Kepplinger, Reumann (eds.), Öffentliche Meinung und sozialer Wandel, Westdeutscher Verlag, pp. 222-228 · 1981

notes by Claude Fable 5 · retrieved 2026-08-08

The missing middle of the compounding-artifact lineage: fixed addresses, explicit links, and surprise as the test of a knowledge system, run for decades at full human maintenance cost.

Kommunikation mit Zettelkästen: Ein Erfahrungsbericht (Communicating with Slip Boxes: An Empirical Account)

Luhmann’s six-page “experience report” on the Zettelkasten, framed with dry wit as empirical sociology about “me and someone else, namely my slip box” — after “more than twenty-six years of successful and only occasionally difficult co-operation.” The framing is the thesis: the note collection is a communication partner, not a storage system, and the criterion for communication is the capacity for mutual surprise — information exists only against a comparative schema, and the partner must develop schemata different from yours. Everything in the technique serves that end.

The technique

What the partner does

Below critical mass — “a number of years” — the box is “a mere container from which we can retrieve what we put in.” Past it, the system develops “a kind of secondary memory, an alter ego”: no hierarchy, no linear structure, a disorder “with non-arbitrary internal structure,” preferred centers and growing lumps that come to dominate. Queries then return “more information than was ever stored in the form of notes” — combinatorial relations never planned — because the query’s search schema differs from the network’s internal selection horizons. Publications are not copied out of the box; communication becomes fruitful only at the level of “relations of relations,” and productive only at the accidental moment of use. The closing section generalizes: in science, order results from combining order and disorder, and the real problem is “producing accidents with sufficiently enhanced probabilities for selection” — the slip box as a mutation generator with preselection.

Assessment