How To Get From Words to Sentences
To a large degree, the nativist position formulated by Chomsky and others has held the role of an antagonist position. Subsequent analysts have spent much effort to show that the structure of language contained considerable roots in nonlinguistic cognition (e.g., Schlesinger, 1971; Sinclair, 1971). The use of cognitively based semantic structures in the description of what children formulate has grown steadily. Analysts have generally continued to attribute to the child considerable powers of general linguistic category formation, however, though these categories may be semantically based categories such as agent and action rather than grammatical ones such as subject and verb. This emphasis on broad semantic categories has recently met, however, with analyses suggesting a more specific, distributionalist bent to the child's early analyses of constructions (Braine, 1976; Clark, 1974). The result is a descriptive flux without apparent satisfactory resolution in favor of one kind of description over another. The question is what can be done about this irresolution.