Child and adult construal of restrictive relative clauses: Knowledge of grammar and differential effects of syntactic contextThis research was supported by an FCAR (Fonds pour la Formation de Chercheurs et lʼAide de la Recherche) postdoctoral fellowship to Cathy Fragman, and by Social Sciences and Humanities Research Council of Canada grant # 410-98-0511 to Helen Goodluck and Eithne Guilfoyle. Helen Goodluck was at the University of Ottawa when the research was done. Nicolas Kessous drew Figure 1 and Sheila Scott made the recordings for Experiments 3 and 4. We are grateful to Danijela Stojanović, Paul Hirschbuhler, Gerard Van Herk, our anonymous reviewers, the action editor of JCL and participants at the 1999 European Cognitive Science Meeting (Siena) and the 2000 Boston University Conference on Language Development for advice and comments; and of course we thank the children and adults who took part in the experiments for their time and patience.
FRAGMAN, CATHY; GOODLUCK, HELEN; HEGGIE, LINDSAY; FRAGMAN CATHY; Montreal; GOODLUCK HELEN; University of York; University of York; HEGGIE LINDSAY; Queens University
Журнал:
Journal of Child Language
Дата:
2007
Аннотация:
We report four act-out experiments testing the sensitivity of adults and three- to five-year-old children to the distinction between restrictive and non-restrictive relative clauses in English. Specifically, we test knowledge of the fact that restrictive relative clauses cannot modify a proper name head, and of the fact that relatives introduced by that (as opposed to a wh-pronoun) are obligatorily restrictive. Both children and adults show knowledge of these properties. No support was found for the hypothesis that children extend the block on proper name heads to wh-relatives. Both children and adults are sensitive to the syntactic context (double object vs. existential) in which the relative clause is embedded. However, adults differ from children in four respects. First, in the double object context, adults are more likely than children to commit the error of construing a that relative as referring to a proper name head. Second, the effect of syntactic context on selection of a head is larger for adults than for children. Third, for adults, but not for children, the effect of syntactic context interacts with the type of relative clause. Fourth, adults, but not children, are influenced by whether they hear the existential context before the double object context. We propose that by three to four years of age children have acquired an adult-like grammar of relative clauses, and that the differences we see in child and adult performance can be attributed to that grammar in combination with a mature (adult) or immature (child) sentence processing capacity.
189.3Кб