ABSTRACT The research explores the availability and usefulness of linguistic cues to verb meaning that might help the child learner to choose among the multitude of hypotheses consistent with observation of world events. This approach, known as syntactic bootstrapping, involves the use of syntactic information about verbs to constrain inferences about their meaning. One line of research looks at the judgments of adult speakers of English to ask what inferences from syntax to semantics are supported by the language. Subjects will make judgments of semantic similarity between verbs, building up a picture of semantic relatedness within a set of verbs. Other subjects will judge sentences for acceptability, providing information about the structural properties of the same verbs. Statistical comparisons of these two data sets will determine whether various syntactic distinctions among verbs match semantic differences. A second line of research explores what children know about syntax/meaning relations, and whether they will make form-to-meaning inferences in interpreting novel verbs. Preschoolers are shown video-taped scenes, and asked to interpret a nonsense verb used to describe them. The question of interest here is whether their interpretations of the same scene change depending on the sentence structure in which the nonsense verb is presented.
