AbstractThis paper presents an experimental study of the acquisition of rootwh-questions in French, in whichwh-movement is optional. Thirty-twowh-questions were elicited from 33 French-speaking children and 22 adults to allow for a quantitative and qualitative comparison with respect to the various word orders. A comparison reveals the following findings: 1° Children produce morewh-in-situ structures; 2° Children produce no subject-verb inversion structures while adults use them often; 3° The choice of question-structure is related to thewh-word that is used, for all subjects; 4° Children are aware of adults’ preferences but do not imitate them; 5° They acquire the various structures for eachwh-element separately; 6° The acquisition of optional rules reflects an interaction of external factors (the input) and internal factors (economy).

No abstract

We present a new method for investigating children's language comprehension and argue that it has the potential to mitigate known task-related biases and expose children's grammatical and lexical knowledge in a more natural and ecologically valid manner. The new method consists of filling in a digital coloring page, according to sentence stimuli (e.g., The green monkey is being scratched by the blue monkey; The rabbit is red.). Through the playful act of coloring in the page, children reveal their interpretations of grammatical constructions and lexical items. We argue that this method gives more accurate results than existing methods, in which children are asked to choose among several pictures representing a number of possible interpretations. We present two experimental studies: one with Dutch-speaking children, tested on four types of grammatical constructions, and a second study with children learning Dutch as a second language, tested on their vocabulary knowledge. In both studies, the new method was compared with one of the most widely used methods: the picture selection task. In the first study where children's performance is said to be underestimated, the new method reveals better performance whereas in the second study where children's performance is assumed to be overestimated, the new method reveals lower performance. The results suggest therefore that the new task indeed decreases external task-related effects and offers a more reliable measurement of children's linguistic knowledge.

No abstract