The role of the agent in sentence comprehension by preschool children

Lead Research Organisation: University of Kent
Department Name: Sch of Psychology

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
 
Description This project assessed how we develop the parsing strategies that are required to interpret English sentences successfully. This is an important question because the answer tells us whether we learn these parsing strategies via experience or whether they are built into the architecture of the brain's language learning mechanism. Ours was the first project to assess this question in children as young as 2 years of age using eye gaze data, which allows us to follow children's developing interpretation of a sentence in real time and shows us how their parsing strategies change as the sentence unfolds. We assessed whether 2- and 3-year-olds prefer to interpret the first noun in a sentence as the agent of an action, even when such a strategy is misleading (as in the case of passive sentences such as 'the baby is being saved by the fireman'). Our findings support the view that both 2- and 3-year-olds process sentences incrementally, having different biases at different structural positions. Neither age group has a whole-sentence first-noun-as-agent bias, but they do show this bias whilst processing the first noun phrase. Another finding is that 3-year-olds are able to override this bias in order to correctly map both active and passive sentences onto appropriate meanings. This contrasts with previous findings that even 5-year-olds are unable to override initial sentence misinterpretations. In sum, this research has led to a deeper understanding of how first language learning mechanisms are intertwined with the sentence processing mechanisms seen in adults.
Exploitation Route Directly: Preferential-looking research into the first language acquisition of active transitive sentences is a relatively large field which has attracted much attention from academics in the USA, Canada, Germany, France, Australia and UK. The primary beneficiaries will be these scholars and researchers in the international and national language development community, as well as researchers who model this development and researchers in sentence processing in adults looking to inform models of sentence processing with models of language acquisition.
Indirectly: Although the impact of our grant will primarily be scientific, there are two findings which could have impact on early years' professionals and speech and language therapists (SLTS). Firstly, these groups (particularly SLTS) are often taught that passive sentences are a complex area of grammar which typically developing children do not learn until they reach school age. Our results indicate that this is not the case. 3½-year-olds were clearly able to demonstrate correct understanding of full passive sentences, even with novel verbs, with both measures. This is relevant in terms of what SLTs should expect children with language impairments to be able to do.
Sectors Education