Abstract
Prominent "embodiment" theories argue that semantics is represented in distributed cortical networks that overlap with sensorimotor regions, and that sensory and motor activation is automatic and necessary during written word comprehension. It is unclear, however, how words become "embodied" as children learn to read. We reasoned that if sensory and motor activation is both automatic and necessary when processing word meaning, automatic picture-like neural responses for printed words should be activated as soon as printed word meaning is acquired. In contrast, if sensorimotor activation depends on post-lexical processing, automatic sensory motor activation for words may only gradually emerge as a secondary consequence of reading experience. To test these predictions we asked 21 7- to 10-year-old children and adults to perform a one-back basic level name-matching task with animal and tool pictures and their associated written words in an MRI scanner. Performance was consistent across age and conditions. All age groups showed clear differential cortical specialisation for tool and animal pictures. In adults, regions with a preference for tool or animal pictures, most notably the tool-specific left medial temporal gyrus and inferior frontal gyrus, showed corresponding activation patterns for tool and animal words. Thus, consistent with previous studies, object category representations were elicited merely upon presentation of the corresponding written word in adults. In contrast, even older children, who were proficient readers, showed no picture-like sensorimotor responses on the presentation of written tool or animal words, suggesting that automatic activation of semantic tool and animal networks for words emerges years after initial acquisition of word comprehension. In summary, our findings clearly show that picture-like sensorimotor activation during single word comprehension is absent in the brains of skilled reading children. This forms a challenge for embodiment theories claiming that this activation is both necessary and automatic for comprehension to occur.
Meeting abstract presented at VSS 2013