Development and adaptation of active dependency completion mechanisms

点击翻译按钮获取中文摘要

Understanding how children learn to comprehend language as efficiently as adults do is a central goal for theories of both normal and abnormal language acquisition. Developmental research has shown that, while much linguistic knowledge develops in the first few years of life, children are not necessarily adept at quickly using this knowledge during real-time language comprehension. For example, as sentences unfold, adult listeners and readers are able to continuously predict how the sentences are likely to continue, based on their knowledge of language or the distribution of sentence structures. However, children's prediction mechanisms sometimes diverge from those of adults, and they must somehow develop adult-like predictive mechanisms during the course of language acquisition. Understanding this developmental process could lead to advances in the pedagogical techniques that aim to enhance language comprehension in children. This may also improve clinical intervention methodologies for adult patients with language disorders, whose comprehension mechanisms show similar characteristics to those of children's.This project investigates the development and adaptation of filler-gap dependency processing mechanisms in children and adults. In comprehending questions like "What was Emily eating the cake with -- ?," adults incrementally complete the dependency by associating the dislocated wh-phrase "what" with the verb, but children wait to complete the dependency until after the verb and its object is encountered. The main hypothesis of this project is that language experience and distributional regularities in the input play a key role in the development and adaptation of predictive processes in filler-gap dependency completion. The motivation for this hypothesis is two-fold. First, it is well established in developmental psychology that distributional regularities can bootstrap children's cognitive and linguistic development. Second, much work in adult psycholinguistics research has argued that syntactic priming, a processing facilitation that results from repetitions of abstract syntactic structures, reflects an implicit learning mechanism that adapts the comprehension procedures in accordance with the recent language experience. Taken together, children's comprehension mechanisms may adapt to an adult-like mechanism based on a) a long-term accumulation of filler-gap dependencies that complete at the verb position, as well as b) syntactic priming of such filler-gap dependencies within an experimental session. Conversely, adults' comprehension mechanisms should also adapt to the child-like, conservative mechanism after exposure to filler-gap dependencies that are completed at a post-verbal position. To test these predictions, Dr. Omaki will conduct i) a corpus study that investigates the distributional patterns of English filler-gap dependencies in adults' conversations and child-directed speech, and ii) five eye-tracking experiments with children and adults that explore the time course of filler-gap dependency processing, as well as its interaction with priming sentences that are designed to create a bias towards the alternative dependency completion mechanism. As such, this project links questions and methodologies in two traditionally separate fields of adult psycholinguistics and language development.