Perception, ontology, and word meaning.

In their commentary on our paper (Soja, Carey, & Spelke, 1991), Landau, Jones and Smith (1992) (hereafter LJS) clarify the position taken by Landau, Smith, and Jones (1988), and briefly describe results from several subsequent papers. As they describe their research program, they are concerned with the relations between the syntactic category of a newly heard word (count noun, adjective, preposition), the ontological category of the word’s referent, and the perceptual basis of projecting the meaning from the referent to other entities. Their experiments provide evidence that, by age 3, a count noun referring to an object is projected on the basis of shape similarity to the referent, an adjective is projected on the basis of color or texture similarity to the referent, and prepositions are projected on the basis of spatial relations between the referent and other objects. On some points, we are in agreement with LJS. Like them, we are concerned with the role of syntax in lexical development. Of most relevance to our concerns is LJS’s restriction of the projection of word meaning on the basis of shape to cases of count nouns referring to rigid objects. They state that “implicit in our