Processing Load and Children's Comprehension of Relative Clause Sentences.

Two experiments investigated the role of capacity in children's comprehension of relative clause sentences. Sentences varied in number of participant roles (2; 3), focus (object, subject) and embeddedness (centre-embedded, right-branching). For example: 2-role, object-focused, right branching "... the pig that the cow bumped"; 2-role, subject-focused right-branching "... the cow that bumped the pig"; 3-role, objectfocused, centre-embedded "The cow that the pig bumped ate"; 3-role, subject-focused, right-branching "The cow bumped the pig that ate". A fifth type, 3-role, subjectfocused, centre-embedded "The cow that bumped the pig ate" was included in Experiment 2. Centre-embeddedness and object-focus were expected to constrain individuals toward assigning more nouns to their roles in the same decision. Estimates of sentences' processing loads were based on number of roles assigned in parallel, and quantified in terms of a relational complexity metric (binary, ternary relations). In Experiments 1 and 2, 135 children (4 to 8 years) and 48 children (4 to 7 years) respectively, responded to comprehension probes (e.g., "Who bumped?"). As predicted, the 3-role sentences were more difficult than 2-role sentences if sentences were centreembedded, object-focused or both, but not if sentences were right-branching. Comprehension improved with age. Children's working memory capacity (listening span), and their capacity to process complex relations (hierarchical classification, transitivity) were also assessed. Age-related improvements in performance were observed on all tasks. Regression analyses showed that Relational Complexity tasks accounted for variance in comprehension independently of age and listening span. The processing load involved in comprehension of relative clause sentences seems to due, in part, to the complexity of the relational information entailed in the role assignment process.