Now that developmental psycholinguists have been gradually easing themselves out of the syntactic fixation with which they have been afflicted for some time, they have come face to face with the problem of semantics which is, of course, the heart of any account of language and language acquisition. My plan today is to begin by presenting some reasons for rejecting one theory of semantic development -namely, the Semantic Feature Hypothesis which has stimulated a great deal of research on semantic development including some of my own. Second, I will sketch an alternative theoretical model which may be more satisfactory. The latter theory draws from aspects of my own research, the developmental work of Donaldson and Wales (e.g., 1970) Macnamara (1972), Rosch (e.g., 1973) and Nelson (e.g. 1974), and research reported by Bransford, Franks, Posner and others (e.g. Bransford and McCarrell, 1975; Franks and Bransford, 1971; Posner and Keele, 1968).