Scott Alexander, curated
← Back to curation

Social Censorship: The First Offender Model

Quality
71
Strong
Claude Shift
54
Moderate
RWI
2
of 10

Summary

A coordination-game model of social taboos. Puzzle: ~40% of Americans support some mild eugenics (Zigerell), yet it's a firestorm taboo — how? Scott's police/mugging analogy (from his anti-Caplan argument): a chief who credibly commits to massive retaliation against the FIRST mugger deters all mugging at zero standing cost, as long as the commitment holds. Eugenics opponents do the same — certainty of punishment for the first to speak out — so nobody is first. The upshot: 'taboo' may mean nothing more than 'one of two equally-sized sides has a tenuous coordination advantage,' not near-universal opposition (the gay-rights 1969->2019 flip shows the equilibrium can switch). This retires his old anti-censorship argument that taboo implies the taboo idea is powerless.

Why this score

Quality 71 · Strong. 71 — Strong. A sharp, generative coordination model of taboos with a memorable mechanism (the police-mugging analogy) and a real update to how censorship is understood. Short and building on his prior metaphor, which places it mid-Strong.

Claude’s paradigm shift 54 · Moderate. 54 — Moderate. The first-offender / 'taboo as tenuous coordination advantage' model is a fresh, non-obvious reframe of how social censorship works.

Real-world impact 2 · Minor. 2 — a within-discourse model in the free-speech/coordination sphere; no material reach.