Un-Bayesian Minds

This monograph develops a unified theory of  Un-Bayesian updating, in which inference is nolonger understood as the revision of probability distributions, but as the transformationof epistemic fields of meaning. Departing from the classical Bayesian paradigm - where evidence isassumed to be given and updating proceeds via conditionalization - the theory establishes thatevidence is constructed through interpretation, and that updating operates on structured objectsdefined by sensitivity, curvature, and memory.