Memory capacity limitations on the use of visual context in human language comprehension
Speaker: Emilia Ellsiepen
Institution: Saarland University
Abstract:
Spoken language and visual attention have been shown to rapidly influence each other: While language guides referential and anticipatory gaze in a scene, the acquired information from the scene may conversely help in resolving ambiguities in the utterance. Scenes used in influential studies (e.g. Tanenhaus et al. 95, Altmann&Kamide 04, Knoeferle et al. 05), however, include usually only 4 to 5 scene entities. My aim is to investigate to what extend richer visual scenes can be used in language comprehension and how referential and anticipatory gaze interacts with the decay of information in visual working memory.
I will present the design of a first experiment and possibly preliminary results. In this experiment, I test whether anticipation of referents in the context of a restrictive verb persists, if participants are presented 7 objects prior to listening to the sentence and whether the order in which participants looked at the objects influences anticipation.
