Two studies addressed the relative strengths of object kind bias and syntactic knowledge in 2-year-olds' inductions of word meaning. The study looked at children's interpretations of novel proper names for familiar and unfamiliar objects. In each study, 10 children were assigned to each of 2 conditions (familiar and unfamiliar) and shown 2 cats (familiar) or 2 monsters (unfamiliar), similar within each pair except for their dress. One was named, and one was not. The children were told to do something specific with the named object. As predicted, they chose the named object much more often in the "familiar" condition than in the "unfamiliar" condition. Spontaneous comments made by the children were of two types: one referring to the second object as having the same name as the first, indicating an extension of the given name as a common noun; and the other asking the name of the second object, suggesting an extension of the given name as a proper noun. The former was more common in the "unfamiliar" condition, and the latter more common in the "familiar" condition. It is concluded that the first words used for objects are interpreted as referring to object kind, and that object-kind bias is a bias to acquire a lower-level term. (MSE)