Discourse models, pronoun resolution, and the implicit causality of verbs

Some interpersonal verbs, such as admire and amaze, describe an action or property of one person (the reactor) that is necessarily a response to an action or property of another (the initiator). We hypothesized that these verbs make the initiator relatively more accessible in a comprehender's discourse model and that this change in relative accessibility aids identification of the referent of a pronoun in a subsequent because clause. We predicted that, as a result, subjects would be faster to recognize a character's name after a because clause that uses a pronoun to refer to that character than after one that refers to some other character. Four experiments confirmed this prediction. Three further experiments demonstrated the importance of the verb's causal structure and of the presence of the connective because to this result.