Reflexives and pronouns in sentence processing : an experimental study of children and adults

• Main Author: Clackson, Kaili
• Published: University of Essex 2012
• Subjects: 415.55
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.573732

Research investigating how adults make use of structural and non-structural information during online processing of reflexives and pronouns has yielded conflicting results. Some have concluded that structural constraints are privileged in being applied early in processing (Nicol & Swinney, 1989; Sturt, 2003), while others maintain that non-structural factors such as discourse prominence can also influence early processing (Badecker & Straub, 2002). While results from the few studies on children seem to support the former view (McKee, Nicol, & McDaniel, 1993; Love, Walenski, & Swinney, 2009), children's processing of reflexives and pronouns has not been studied using a continuous time- sensitive measure. Similarly, a number of studies have examined adults' on line processing of cataphoric pronouns (Cowart& Cairns, 1987; van Gompel & Liversedge, 2003; Kazanina, Lau, Lieberman, Yoshida, & Phillips, 2007), but this has not yet been investigated in children. The aim of this thesis was to fill this gap in the literature by systematically comparing adults' and children's performance on reflexives, forwards pronouns and cataphoric pronouns, using the time-sensitive method of eye-tracking-while-listening. It ,was found that while children's final interpretation of reflexives was adult-like, during processing children were temporarily more distracted than adults when multiple cues supported a prominent competitor antecedent, with performance becoming more adult-like over development. In the processing of forwards pronouns children's performance both offline and online differed from adults and developed with age. In processing cataphoric pronouns children applied structural constraints in the same way as adults, but when structural constraints allowed alternative interpretations, their on line processing was not fully adult- like. These results are interpreted as being consistent with parallel processing models, and as showing that children apply syntactic constraints in the same way as adults, although their more limited ability to dynamically control multiple sources of information leads to adult/child processing differences.