Do redundancy-driven inferences arise spontaneously during comprehension?

Growing evidence suggests that informationally redundant utterances can elicit pragmatic inferences. This study investigated whether such inferences arise spontaneously during comprehension, using a sentence-completion task (Experiment 1) and a recognition memory task (Experiment 2). Participants read short scenarios involving two U.K. natives and one newcomer to the United Kingdom from China. One U.K. character produced either a redundant or non-redundant utterance. Redundant utterances involved common knowledge about U.K. life and were thus redundant for the U.K. characters but not the Chinese character. Across both experiments, participants, without explicit prompting, were more likely to indicate the Chinese character as the addressee (Experiment 1) or falsely recognize her as present (Experiment 2) in the redundant than in the non-redundant condition. These findings provide novel evidence that redundancy-driven inferences arise spontaneously during comprehension and extend work on informativity-based expectations by showing that comprehenders use utterance content to update assumptions about communicative intent and shared knowledge.