The current project is part of a larger study investigating the development of infinitival complements in normal-language (NL) and language-impaired (LI) children. Speech-language pathologists need normative information about complex syntactic forms, such as the infinitive, for assessment and goal selection with language impaired children. Although NL children start to produce infinitives by 3 years of age, the meanings they allow for these sentences are not adult-like even for 5, 6 year olds. Infinitives pose a comprehension problem because the subject of the infinitival clause is not explicitly stated. (For example, for the sentence Mickey asks Bert to swim, the child has to determine who does the swimming). The current study will investigate children's interpretations of infinitive sentences with regards to the possible referents for the infinitival clause subject. Experiments #1 and #2 include normal-language children aged 3;0 to 5;6. Experiment #1 elicits judgments for each of 27 infinitive sentences about the interpretations allowed by the child and investigates the relative efficacy of different methodologies for eliciting these judgments. Task types include an open choice task in which each sentence is presented once and the children indicate all possible interpretations about the sentence and a yes/no decision task in which each sentence is presented several times, each time with a different interpretation, and the children indicate if that interpretation is acceptable for the sentence. Presentation modes include a demonstration condition in which the examiner demonstrates possible interpretations and questions the child about whether each interpretation is acceptable and a picture presentation condition in which the child is shown pictures of alternative interpretations and asked to select the pictures that could match the sentence. The analysis will compare responses across task types and presentation conditions for each age group and for individual children. Experiment #2 investigates children's susceptibility to contextual bias in interpreting infinitive sentences and compares these interpretations to performance on a judgment task (as in Experiment #1). Children will act out each of 39 infinitive sentences that have been preceded by either a neutral lead sentence (e.g., It is a hot day. Mickey tells Bert to swim.) or a biased lead sentence that focuses attention on one possible referent (e.g. This is a story about Mickey/Bert/Ernie. Mickey tells Bert to swim.) The analysis will compare responses across lead conditions and across tasks (judgment vs act-out interpretation). Experiment #3 will pilot use of the judgment and interpretation task procedures from Experiments #1 and #2 on normal-language children below 3 years of age and on language-impaired children.
