The question of what makes a statement “important” in a story was studied. Causal relations were identified between all pairs of events in six folktales, using context-dependent, logical criteria of necessity, and counterfactual tests of the form: If event A had not occurred, then, in the circumstances of the story, event B would not have occurred. Causal networks were derived from these identifications for each story and two properties of them were found to predict judgments of importance: (1) the number of direct causal connections and (2) whether or not an event was in a causal chain from the opening to the closing of the story. The judged importance of a statement increased with the number of causal connections and causal chain membership. Regression analyses showed that substantial proportions of variance were accounted for jointly by both properties and uniquely by causal connections. The importance of a statement, whether identified by structural analysis or judged by naive subjects, seems to be determined by analogous assessments of the statement's causal and logical relations to other statements in the text.
Read full abstract