How we know

Method follows the phenomenon. The event trace can locate an adaptation in time and sometimes catch its speed — a bonus, never the load-bearing evidence — because much of the redundancy (relay software, infrastructure choices, client web-of-trust internals, interop specs) never appears in public events at all. The interviews carry the claim ({{ref:claim-interviews-carry-claim|interviews carry the claim}}). What is being established is a community disposition: people orient to an open problem with no global solution and keep adapting anyway. The signal question — why are people who have seen the worst of Nostr still here, rather than retreating permanently to a safe single-relay community? — has its answer in the staying itself. Refusing the global solution is the finding, not a gap. Two rules keep this honest: "everyone adapts" never stands alone without a named instance (a spec after a specific wave, an operator who tightened then loosened), and the trace's limits are stated, not hidden — the elicitation methods of the open-content literature are replicable here minus their retrospective memory bias, since the mechanisms leave typed traces ({{ref:src-open-content-governance|Schroeder & Wagner}}).