Representing 3D space in working memory: Spatial images from vision, hearing, touch, and language

Abstract

The chapter deals with a form of transient spatial representation referred to as a spatial image. Like a percept, it is externalized, scaled to the environment, and can appear in any direction about the observer. It transcends the concept of modality, as it can be based on inputs from the three spatial senses, from language, and from long-term memory. Evidence is presented that supports each of the claimed properties of the spatial image, showing that it is quite different from a visual image. Much of the evidence presented is based on spatial updating. A major concern is whether spatial images from different input modalities are functionally equivalent-that once instantiated in working memory, the spatial images from different modalities have the same functional characteristics with respect to subsequent processing, such as that involved in spatial updating. Going further, the research provides some evidence that spatial images are amodal (i.e., do not retain modality-specific features).