The idea of semantic features has taken some force within psychology and a number of research workers have suggested that semantic acquisition is, in some manner, determined by semantic components. This notion has come to be called the "semantic feature hypothesis". An examination of the semantic feature hypothesis was made by testing 80 New Zealand children (40 Maori and 40 European), four years of age, for comprehension of the semantic components, as specified by Manfred Bierwisch, of a set of spatial adjectives. Results indicated that a componential analysis does not offer a valid model for the acquisition of either words or of word meaning. The problem would seem to be that it is impossible to derive a theory of semantic processing or of semantic acquisition solely from one level of analysis. The acquisition of the words of the spatial set (big - little, high - low, etc.) can best be dealt with by a multi-level model of processing which postulates as its elements verbal forms (signifiers), perceptual preferences, and cognitive operations. (Author)