What Are the Routes to Face Recognition?
James Craig Bartlett, Jean H. Searcy, Hervé Abdi · Oxford University Press eBooks · 2006
The span of configural information can be small (e.g., specifying the relationship between two adjacent components) or it might be large (e.g., specifying the relationship between nonadjacent components separated by large distances, or specifying the relationship among all of the components in the face). This chapter examines the span of configural information by asking observers to classify upright or inverted face stimuli on the basis of internal or external features. Adapting the Garner interference task, the discussion shows that for many (but not all) observers, internal and external features in upright faces were not coded independently. Configural processing is more disrupted by inversion, though the qualitative way in which features are interpreted depends on the assignment of directions, and, hence, orientation.