UNDERSTANDING THE NEURAL REPRESENTATIONS OF ABSTRACT CONCEPTS: CONVERGING EVIDENCE FROM FUNCTIONAL NEUROIMAGING AND APHASIA

Abstract

While the neural underpinnings of concrete semantic knowledge have been studied extensively, abstract conceptual knowledge remains enigmatic. In the first experiment, participants underwent a functional MRI scan while thinking deeply about abstract and concrete words. A functional connectivity analysis revealed a cortical network, including portions of the left temporal parietal cortex (TPC), that showed coordinated activity specific to abstract word processing. Alternatively, concrete words led to cooperation of a network in the inferior, middle and polar temporal lobes. In a second experiment, participants with focal lesions in the left TPC, as well as matched control participants, were tested on a spoken-to-written word matching task, in which they were asked to select either an abstract or concrete word, from an array of words that were related or unrelated to the target. The results revealed an interaction between concreteness and relatedness. Participants with lesions did not have an overall deficit for abstract words, relative to concrete words, in this task. However, their accuracy was significantly lower for abstract words in related arrays, compared to words in unrelated arrays. These results confirm that the TPC plays an important role in abstract concept representation, and that it is part of a larger network of functionally cooperative regions needed for abstract word processing. These results also provide converging evidence that abstract concepts rely on neural networks that are independent from those involved in concrete concepts, and have important implications for existing accounts of the neural representation of semantic memory.