How a rumor keeps its shape

A rumor is not just a false fact. It is a claim passed along while the evidence is still thin, because the people passing it want a story before they can get a confirmation.

The classic laboratory picture comes from Gordon Allport and Leo Postman in the 1940s. They had people repeat a story down a chain, the way a message moves through a party, and described three habits. Leveling: the story got shorter. Details that did not serve the point fell away. Sharpening: a few details grew louder than they had been, until a minor object in the original became the center of the retelling. Assimilation: what survived bent toward what the listener already expected about the people involved. A rumor does not only lose information. It is rebuilt to fit the mind it lands in.

Allport also offered a blunt formula that later researchers have argued with and never quite thrown away. Rumor travels when the topic is important to the people hearing it and the situation is ambiguous. Neither factor alone is enough. A trivial mystery does not get repeated. A grave event that has already been clearly explained has less room for a rival version. Importance times ambiguity is the gap. Into that gap, people would rather place a story than sit with nothing. The formula is not a law of physics. It is a good first question: why this claim, in this room, now?

Those habits did not retire when the chain became a group chat. A forwarded message drops context, which is leveling. It turns up the outrageous clause, which is sharpening. It flatters the politics of the room it lands in, which is assimilation. The speed is new. A chain that once took an evening now takes a minute, and the same text can arrive in a thousand rooms before anyone in a position to know has spoken. The edits are old. What changed is that the edited version can outrun the correction by a distance that used to be impossible.

Platforms add a few pressures the laboratory did not have. A message that produces a strong reaction is more likely to be shown again, so sharpening is rewarded by the machinery, not only by human memory. The audience is invisible and large, so people write for the room they imagine, and the imagined room is usually more like themselves than the real one. Context that lived in a face — a shrug, a “someone told me” — is stripped by the forward button. What remains looks like a finding. The format of a screenshot is part of the persuasion. It looks like evidence because it looks like a document.