In this talk I shall describe recent work that capitalises on the relationship between language and visual attention: specifically, we can use language-modulated shifts in visual attention to study, as a sentence unfolds in time, the nature and content of the mental representations that are constructed during this time; the manner in which these representations are evaluated against the (visual) context; and the manner in which different kinds of information (e.g. lexical, syntactic, semantic) interact during this interpretive process. A series of studies in languages with grammatical structures as diverse as English (typically subject-verb-object) and Japanese (typically subject-object-verb) suggest that sentence processing proceeds through the prediction at each moment in time of what linguistic material will follow. These predictions are driven by a range of information, from knowledge of syntactic conventions to knowledge of real-world plausibility.
If you are interested in meeting with the speaker, please contact Nina
Kazanina .
To go back to the Colloquium schedule, click
here.