Production Before Comprehension in the Emergence of Transitive Constructions in Dutch Child Language

Although 2-year-old English- or Dutch-speaking children tend to use correct subject-object word order in their own utterances, they appear to make a substantial number of word order errors in their comprehension of other people's utterances. This pattern of adult-like production but poor comprehension is challenging for linguistic theory. While most approaches to language acquisition explain this pattern from extra-linguistic factors such as task demands, the constraint-based approach Optimality Theory predicts this asymmetry between production and comprehension to arise as a result of the linguistic competition between constraints on word order and animacy. This study tests this prediction by investigating how children's comprehension and production of word order in transitive constructions develop, and to what degree their comprehension and production are influenced by animacy. Two- and three-year-old Dutch speaking children (n = 32) and adult controls (n = 41) were tested on their comprehension and production of simple transitive sentences, in which the animacy of the grammatical subject and object were manipulated. Comprehension was tested in a picture selection task and a preferential looking task, and production was tested in a parallel sentence elicitation task. Children's comprehension of transitive sentences in the picture selection task was found to be less accurate than their production of the same sentences in the sentence elicitation task. Their eye gaze in the minimally demanding preferential looking task did not reveal a more advanced understanding of these sentences. In comprehension, children's response accuracy, and to a lesser extent their eye gaze, was influenced by the animacy of subject and object, providing evidence that their poor comprehension is due to the competition between word order and animacy, as predicted by the constraint-based approach. In contrast, animacy may have a facilitating effect on children's production of transitive sentences. These findings suggest that the mature form and meaning of a transitive construction are not acquired together. Rather, the form-meaning pairings of transitive constructions seem to arise gradually as the by-product of acquiring the constraint ranking of the grammar. This leads to the gradual alignment of forms and meanings in child language and hence to the emergence of linguistic constructions.