Truth-seeking series · № 19

Preference falsification

Everyone thought everyone else believed it.

People often say what's safe rather than what they think — the economist Timur Kuran called it preference falsification. Each person has a private threshold: "I'll say what I actually believe once at least T others already have." Nobody can see anyone's threshold. Everyone can see the count.

Below are 100 people. Blue means speaking their mind; gray means staying quiet. Run each scenario — the third one is the punchline.

The punchline

"The ladder" has thresholds 0, 1, 2, … 99: one bold soul speaks, which frees the person who needed one, which frees the person who needed two — and the whole society flips.

"One missing rung" is the same society with a single person's threshold nudged from 1 to 2. The instigator speaks… and nothing else ever happens. Two populations, private beliefs identical to a rounding error — one erupts, one is silent forever.

That's the epistemic trap: you cannot read private belief off public behavior. Silence is not agreement; it can be a hundred people each waiting for two more. And an observed "consensus" can be one nudged threshold away from evaporating.

Why this matters for truth seeking

It explains why revolutions blindside the experts — Kuran's example was 1989, when regimes with decades of apparent public support dissolved in weeks. The support was performed, and everyone privately knew it, and nobody knew that everyone knew.

It also cuts the other way: loud online "consensus" can be a threshold artifact, manufactured by a visible early few — the count does the persuading, not the argument. So when you poll a room, remember what you're measuring: stated opinion is public behavior, and public behavior is strategic. If you want the truth, lower the cost of saying it — anonymity, secret ballots, asking people one at a time.