Linguistics colloquium: Daphna Heller
University of Toronto
Title: What visual memory can tell us about the context: The case of modification
This work is in collaboration with Danielle Moed
Modifiers, such as adjectives (e.g., “the open cage”) or prepositional phrases (e.g., “the cage with the bunny”), are normally used when the context contains more than one object from one nominal category (e.g., in a situation with two cages). In this paper, we test the idea that modifiers provide information not only about the referent itself (i.e., the object being described), but also about the unmentioned object from the same nominal category (the “contrast”, for short). To test this idea, we developed a new experimental procedure that combines a linguistic and a non-linguistic task. First, participants produced spontaneous instructions for their addressee to select an image out of an array. Then, they had to perform a (surprise) recognition memory task which tested how well they remembered unmentioned images they saw during the production phase. Interestingly, results reveal that visual memory of the unmentioned contrasts is affected by the semantic content of the modifiers used, even though these descriptions were used to describe another image. We discuss the implications for the interplay between visual attention and linguistic representations.
Building 403, room 101