An alternative assessment of children's comprehension of relative clauses

Methodological difficulties are identified in the assessment of children's comprehension of relative clauses (RCs) by means of the standard acting-out procedure. An alternative experimental task is proposed which attempts to minimize these difficulties while exploring the advantages of the acting-out technique. An experiment is reported in which Portuguese-speaking 3- to 6-year olds were tested by means of this alternative task. The results show children being affected by both the position and the focus of the RC in the same way adults have been reported to be and an error analysis suggests that most of children's difficulties can be ascribed to memory overload while RCs are processed as noun modifier sentences. These results contradict the view that there are developmental differences with regard to the way sentences with RCs are parsed and that processing load induces errors particularly manifested in right-branching subject focus sentences. Developmental changes taking place between the ages of 3 and 6 are considered and questions are raised concerning development from the age of 5.