Acceso usuarios
To acquire a language, a child must keep track of input from the linguistic environment and infer what grammar generated it. However, not all of the information in the input is available to the child. What the child can encode from the input changes across development, as a function of what information the child can process, and what he knows and expects about the system being acquired. In this paper I present a novel word learning experiment where children generalize differently depending on whether they are learning word meanings or word classes. While the linguistic input could lead to multiple generalizations, children use it in systematically different ways depending on the problem they are trying to solve. This result reflects both content of the input and constraints on learning stemming from hypotheses the learner brings to the task.
