cSCAN Rounds

Error and expectation in child language

Our expertise at using language -- whether holding conversations, understanding sentences, or recognising words -- is to a large degree driven by prediction: By anticipating what others are likely to say, we can ease the burdens of speaking and listening. But is prediction only a characteristic of expert systems, or is it also a means for becoming expert in the first place? I will present a series of studies in which we used eyetracking and iPad apps to test whether children's language development can be explained by predictive, error-driven learning.