Verbs and attention to relational roles in English and Tamil

Abstract

English-learning children have been shown to reliably use cues from argument structure in learning verbs. However, languages pair overtly expressed arguments with verbs to varying extents, raising the question of whether children learning all languages expect the same, universal mapping between arguments and relational roles. Three experiments examined this question by asking how strongly early-learned verbs by themselves, without their corresponding explicitly expressed arguments, point to 'conceptual arguments' - the relational roles in a scene. Children aged two to four years and adult speakers of two languages that differ structurally in terms of whether the arguments of a verb are explicitly expressed more (English) or less (Tamil) frequently were compared in their mapping of verbs, presented without any overtly expressed arguments, to a range of scenes. The results suggest different developmental trajectories for language learners, as well as different patterns of adult interpretation, and offer new ways of thinking about the nature of verbs cross-linguistically.