Acquisition of advanced syntax and primary pragmatics: an investigation into children’s referent choice in obligatory and long-distance control

ABSTRACT

The focus of this thesis was on the acquisition of control in typically developing children and the strategies they might employ for referent assignment in control constructions. The goals were to empirically establish the syntactic nature of obligatory control whilst in contrast investigate the development path children take in their acquisition of a pragmatically governed non-obligatory control construct.

Sixty children participated in three picture-selection tasks that tested obligatory object control and non-obligatory long-distance control. The first task established the children’s base-line interpretations, whilst pragmatic topic primes were introduced in the next two tasks to confirm which referent the children preferred and establish which they would permit.

The results of this study confirmed the syntactic nature of obligatory control in a comparison of the children’s results with adult controls as well as by an evaluation of the errors made by some of children in the trials. Further, results showed that children’s development of non-obligatory, long-distance control develops slowly, in stages, as they mature. Like adults, children initially showed a preference for the object as a control referent, despite the potential topic-hood held by the subject of the sentence. As pragmatic discourse was added, however, the children’s developing grammar differed from the adult grammar, and it was shown that the linear locality of the object seemed to hold a precedent for the children that it did not for the adults when presented with strong pragmatic primes as preceding discourse; the children were more resistant than the adults to switching from the object to the subject as the coreferent. Furthermore, the choice of verb had an impact on the referent chosen by the children, with evidence that despite the pragmatic control assumed of non-obligatory control, syntactic properties such as c-command may have an effect on the children’s choice of referent in long distance control. It will be shown, however, that this effect can be broken with the addition of a strong prime, un-like the more persistent impact of linear locality on their interpretations.