WV

It is a different ,The spontaneous speech of three preschool children is analyzed for evidence of the development of the kind of knowledge represented by current transformational treatments of Wh questions. The children are those participating in a longitudinal, naturalistic study of the development of grammar. The first Wh questions seem to be unanalyzed routines or constructions not involving transformations. As Wh questions become more complex and varied, several kinds of evidence of transformational knowledge emerge. All sorts of Wh questions are for a time produced in a form that adults do not model, but which is a hypothetical intermediate in the transformational analysis of adult forms. This child's form seems to involve a preposing operation but to omit a transposing operation. One sort of Wh question, the Why question, seems to be created by one child (Adam) as a direct transform of an antecedent declarative from the mother. In general, there is evidence that children in the preschool years do develop a grammatical structure underlying Wh questions that is much like the structure described in current transformational grammars. Because the abstract underlying structure is not strongly suggested by the surface form of Wh questions, it is difficult to see how it can be learned. However, there are recurrent discourse patterns, involving sentence and constituent exchanges, which are rich in structural information, and these may constitute the basis of a learning process.