Superman Takes a Taxi: Testing Theories of Validation with Inconsistencies in Fantastic Narratives. Issue 3 (16th March 2022)

• Record Type: Journal Article
• Authors: Saerys-Foy, Jeffrey E.; LoCasto, Paul C.; Burn, David; Ferranti, Daniella
• Is Part Of: Discourse processes. Volume 59:Issue 3(2022)
• Page Start: 137
• Page End: 158
• Publication Date: 2022-03-16
• Subjects: Discourse analysis -- Periodicals; 401.41
• DOI: 10.1080/0163853X.2021.1994298
• Languages: English
• ISSNs: 0163-853X

Abstract

ABSTRACT: According to theories of validation, people routinely check incoming information against prior knowledge during comprehension. On these theories, information is validated if it fits with prior knowledge. Some researchers argue that information needs to be successfully validated before being incorporated into the situation model. We report five experiments that tested these claims. In each experiment, participants read fantastic narratives with inconsistencies in the first half of the text. We measured the effect of these inconsistencies on subsequent reading times. In Experiments 1, 2a, and 2b, narratives were about familiar fantastic characters (e.g., Superman), and in Experiments 3a and 3b they were about unfamiliar characters. General linear models with fixed and random effects revealed that inconsistencies in the first half of the narrative (e.g., Superman taking a taxi) facilitated reading times on subsequent sentences that repeated the inconsistency (e.g., Experiments 1 and 3a) but novel inconsistencies did not (e.g., Superman running away from danger; Experiments 2a, 2b, and 3a). This facilitation disappeared when participants intentionally judged the plausibility of the inconsistency (Experiment 3b). Our findings suggest that during passive reading people incorporated information into their situation model even when it did not fit prior knowledge. Only when people made intentional evaluations did we find evidence that they did not incorporate inconsistencies into their situation models. We argue that these findings raise questions about theories of validation and provide suggestions for future research.