Early Verb Morphology in French: Assessing the Contribution of Child-Directed Speech and of Conversation.

Children learning languages like English, French, Italian and Spanish, start using verbs in single forms before they start to produce two morphologically different forms for the same verb type (e.g., Tomasello, 1992; Gathercole et al., 1999). In this paper two main questions are addressed: 1) Why children start producing verb words in a single form? 2) During the single verb form period, why certain verbs are used in a form (e.g., /ka'se/ 'broken/to break') and other verbs in another (e.g., /turn/ 'turn/turns')? This paper investigates these questions by analyzing the videorecorded longitudinal data of two French-acquiring children, collected at home every two weeks when the children were between 1;3 and 2;4, transcribed and linked in CHAT format. The children produce verbs in single forms until 1;9, and 1;11, according to the child. All verbs produced until this period by the children and their partners, as well as the relations between the verb forms in conversation, were analyzed. Results show that most verbs appear in only one phono-morphological form also in Child-Directed Speech (CDS). When two forms of the same verb are used (for example, /kas/ 'you break' and /kase/ 'to break' or 'broken'), one of the two is used most frequently. The reduction of the morphological variation found in CDS provides a partial answer to the single-form verb morphology period in children's early verb production and complements coherently a semantic-pragmatic explanation by which children suppose unique relations between form and meaning (Clark, 1987; Markman, 1992). The particular verb forms produced by the children tend to correspond to those used dominantly in CDS. At the same time, they tend to be also reinforced in conversational contingencies (CC). For about half of the verb forms, CDS and CC provide the same congruent information. Of the remaining verb forms, about three quarters are explained by CC, while less then 15% are explained by CDS, indicating that conversational contingencies are a stronger source of influence than general input. Only 7% and 5% of the of the overall verb types seem to constitute child-constructed patterns whose origins remain to be fully understood. These findings underline the close relationships among patterns of language acquisition, conversational exchanges and child-directed speech and suggest that conversational contingencies, co-constructed by the children, have a stronger impact then general CDS