Yee, E. 1, 2 , Chrysikou, E. 2 , Hoffman, E. 2 & Thompson-Schill, S. 2
1 Basque Center on Cognition Brain & Language
2 University of Pennsylvania
How do we know the meaning of words? Sensorimotor-based theories of semantic memory claim that semantic information about an object is distributed over the neural substrates that are invoked when we perceive and interact with it. Hence, occupying a neural substrate that is an important part of an object’s representation (e.g., with a concurrent secondary task) should interfere with accessing that representation. In the current work, participants made concreteness judgments about (heard) names of objects while either simultaneously performing a patty-cake -like task on a table, mentally rotating objects, or performing no concurrent task. Objects varied in the extent to which one interacts with them manually (e.g., tiger=low manual interaction, pencil=high manual interaction). We found that performing a concurrent task increased errors for all objects. Critically, however, during the patty-cake task, errors were greatest for objects rated as high in manual interaction. (In contrast, the concurrent mental rotation task did not disproportionately increase errors for manual objects.) These findings suggest that engaging brain regions underlying manual interaction (with an incompatible manual task) interferes with comprehending the names of that are manually experienced. Hence, these regions appear to be part of (rather than peripheral to) the representation of frequently manipulated objects.