Main idea

Pillar cards — a few structuring ideas held on cards you revisit and confront — were put in place several months before the discovery of Jirō Kawakita's KJ法. The encounter is therefore real and late, and it authorizes neither of the two easy conclusions.

It does not authorize "I had the same idea": the KJ法 starts from scattered material and brings out a structure through successive groupings, whereas pillar cards keep a few already stabilized ideas alive. The purposes differ.

Nor does it authorize "this is KJ法 without knowing it": a family resemblance — getting units of thought out of your head, making them manipulable, keeping them autonomous, bringing them together to keep thinking — is not an identity of method. A nuclear plant technique is not the ancestor of a memory system for AI because it involves a handoff.

You have to be able to hold both sentences together: this is close enough to be interesting, and this is not the same thing.

Why it matters

This gives the stance that makes an honest prior-art search bearable. Without it, every discovery becomes either a humiliation — it had all been done already — or a claim — I had found the same thing — and in both cases you stop looking.

It also protects against the opposite drift, which is more insidious: through sheer searching for antecedents, you end up manufacturing them, spending ten lines explaining why two things might vaguely resemble each other.

Nuances and limits

The boundary between kinship and identity is a matter of judgment, not of criteria. Two readers in good faith can decide differently on the same pair.

And the position is unstable over time: a method that evolves can move closer to a tradition it was distinct from, without anyone replaying the comparison.

Open questions

• What do you do with a kinship that is recognized but not borrowed when it comes time to publish: cite it as an influence, as a neighbour, or not cite it at all?