Investigating the alternative-sensitivity of 'know'

Several semantic analyses of know have been proposed in recent years to account for the so-called Gettier Problem. These analyses make distinct predictions regarding the sensitivity of know to the alternative possibilities expressed by its complement, as induced by expressions such as disjunction. These predictions were tested in two experiments. Results show that knowledge sentences with a disjunctive complement (e.g., John knows that Mary has a son or a daughter) are more likely to be judged as false than classically-equivalent sentences with non- disjunctive complements (e.g., John knows that Mary has a child) under Gettier-like scenarios. We discuss how these findings provide evidence for the alternative- sensitive approach to know.