Varvara Kuz (UCL) - 2019-20 Students
Varvara.kuz.19@ucl.ac.uk

The Role of Bilingualism in prediction mechanisms in L1 sentence processing

My project aims to contribute to three fields in linguistics: eye movements in reading, sentence processing and bilingualism. Whilst the importance of prediction in sentence processing is widely acknowledged, little literature exists on the specific cognitive processes involved. Analysing the ‘atypical’ behaviour of bilinguals can inform our understanding of these mechanisms more generally. My proposed eye-tracking methodology will also fill a gap in descriptions of bilingual eye-movement. Prediction is an important psychological strategy used in many daily situations, allowing us to build mental representations based on available information, and to respond accordingly ahead of an event. It also plays a vital role when parsing a sentence to foresee its conclusion. Building predictions about upcoming items in a syntactic structure is what allows us to process sentences quickly and efficiently. Sometimes, however, these predictions are erroneous, resulting in a greater processing cost due to the need to revise the structure of the sentence. Due to their linguistic experience, bilinguals may have cognitive mechanisms which allow them to manage predictions and prediction errors more efficiently than monolinguals, though the mechanisms behind these differences remain unclear. I am interested in uncovering the cognitive mechanisms underpinning prediction mechanisms when parsing a sentence, specifically when these predictions are violated. Overcoming these errors in prediction is costly and may, for example, lead us to slow down when reading. Bilinguals, however, may have cognitive mechanisms which can help them overcome these errors in prediction with lesser costs. My main hypothesis is that bilinguals, due to their unique linguistic experience, are better able to resolve prediction errors during reading because of enhanced cognitive control abilities. I will collect and compare eye movement data of bilinguals and monolinguals when reading sentences that will lead them to make errors in prediction.