Tacit verification of determinate and indeterminate text ideas.

Singer presented evidence that reading time reflects, in part, processes of continuously verifying current clauses with reference to their text antecedents. This study extended that analysis to text ideas of indeterminate truth in their discourse context. For example, having read simply that Norm's house was destroyed, one might later read that Norm's cousin believed that Norm's house was destroyed by a fire. Research concerning people's reports of not knowing something suggested that the reading of clauses of indeterminate truth would be faster than for false or inconsistent clauses. In two experiments, target sentences varied in their truth, verb entailments, and use of negation with reference to their antecedents. Reading time for indeterminate targets was sometimes distinctly shorter and was never longer than for corresponding false targets. The effects were modulated by the entailments of factive and nonfactive main verbs and by negative expression. All of the conditions that were identical to ones of Singer (2006) replicated the former findings. It was concluded that encountering text ideas of indeterminate truth does not initiate protracted memory searches for comparable concepts.