Key Insight
A holiday post examining how ideological labeling substitutes for argument in public discourse, with a discussion of bans and decibans as units of evidential weight in Bayesian reasoning.
The title refers to a pattern that recurs across ideological divides: the substitution of identity labeling for argument. When a participant responds to evidence not by engaging with it but by attributing the source's position to ideological contamination, the conversation has shifted from science communication to cultural conflict.
A ban is a unit of evidential weight in Bayesian reasoning — the base-10 logarithm of the Bayes factor. A deciban is one-tenth of a ban. These units provide a precise vocabulary for discussing how much any piece of evidence should shift our beliefs.
Identity-protective cognition operates by subverting Bayesian updating — causing individuals to weight evidence not according to its actual probative value but according to its consistency with identity-defining group positions. Understanding evidence in terms of bans and decibans forces the question of evidential weight into explicit view, where it can be evaluated on its merits.