Neural substrates of object identification: Functional magnetic resonance imaging evidence that category and visual attribute contribute to semantic knowledge

Abstract

Recent findings suggest that neural representations of semantic knowledge contain information about category, modality, and attributes. Although an object's category is defined according to shared attributes that uniquely distinguish it from other category members, a clear dissociation between visual attribute and category representation has not yet been reported. We investigated the contribution of category (living and nonliving) and visual attribute (global form and local details) to semantic representation in the fusiform gyrus. During functional magnetic resonance imaging (fMRI), 40 adults named pictures of animals, tools, and vehicles. In a preliminary study, identification of objects in these categories was differentially dependent on global versus local visual feature processing. fMRI findings indicate that activation in the lateral and medial regions of the fusiform gyrus distinguished stimuli according to category, that is, living versus nonliving, respectively. In contrast, visual attributes of global form (animals) were associated with higher activity in the right fusiform gyrus, while local details (tools) were associated with higher activity in the left fusiform gyrus. When both global and local attributes were relevant to processing (vehicles), cortex in both left and right medial fusiform gyri was more active than for other categories. Taken together, results support distinctions in the role of visual attributes and category in semantic representation.