AbstractA narrative comprises of singular causal links connecting a sequence of actions/events (i.e. a chronology). A technique which inserts these links using Bayesian inferences from items of evidence is outlined. The so constructed Bayesian narratives underpin both the analysis of historically based case studies and a possible theory of action which complements the standard theories of normative and rational actions.
Keywords: Bayesian narratives, causal inference, action theory, small-n, learning.