Children are strikingly good at learning the meanings of names for things. There is considerable evidence that they solve this mapping problem through inferring the referential intentions of other people. This mindreading capacity might also underlie children’s assumptions that words do not overlap in reference, and that new words are likely to refer to whole objects. Finally, there is some support for the view that the mindreading that underlies word learning is due to a general capacity to infer mental states, not a submodule that is dedicated to communication.