Published online by Cambridge University Press: 26 September 2008
Third- and fifth-grade children, who had passed pretests of comprehension for words and sentence types used in the experiment, were tested on their ability to assign given/new roles in active, passive and cleft sentences controlled for stress. The experimental task involved a best-fit judgement that required the child to decide which one of two context sentences was the first part of a story continued by the target sentence. There was a significant difference in response to passive and cleft sentences, as opposed to actives. Despite a strong statistical tendency to perform correctly on passives and clefts, there were a substantial number of subjects who performed near chance in all sentence conditions. Five individual patterns of response to the task are identified. Strategies used by subjects in responding to the task, and the implications of the results for functionalist theories of language acquisition are discussed.