The environment is continuous, but sensory input is not. Eye movements deliver a succession of views over time: as do hand movements (in the absence of vision). How does a coherent representation of space emerge based upon these discrete inputs? I will argue that the spatiotemporal character of perceiving is reflected in the mental representation of natural scenes. The representation frequently includes the unseen but highly expected layout not present in the studied view, but expected to exist just outside its borders (boundary extension; Intraub & Richardson, 1989). Research demonstrating the generality of this effect in memory for: a) 2D photographs, b) imagined scenes, c) regions of 3D scenes explored visually or haptically, and d) regions of 3D scenes explored by a deaf and blind "haptic expert" will be presented. The special status of view-boundaries and the processes they apparently elicit will be considered in the context of research on the representation of occluded objects, and research on the impact of planned eye fixations on memory. Although in one sense anticipatory representation of unexplored layout is a memory "error," it may actually play an important role in scene perception by facilitating integration of successive views and supporting a coherent representation of a continuous world that can be perceived only a part at a time.

Department Faculty Hosts: Dr. Landau & Dr. McCloskey