The Discovery of the Occluding Edge and Its Implications for Perception (Gibson, 1979, Chapter 11)

Gibson introduced us to the notion of occlusion in Chapter 5. Occlusion is when one surface is progressively covered up by another surface, and the details of how optical texture is deleted specifies the occluding edge and the fact that occlusion is happening. In this chapter, Gibson details the radical implications of this fact of the optic array for theories of perception. Most importantly, given occlusion, vision cannot be based on images, series of images, or perspective structure, and it is not limited to working when sensations fall on the retina. Vision is a process that extends back and forwards in time, and this is enabled by information. 

Gibson reviews the key occlusion experiments done with Kaplan. They made videos, in which each frame simply contained one textured surface on top of the other, but successive frames progressively deleted optical texture from the rear surface. No edge was visible in any single frame, but when put in motion, the deletion of optical texture revealed the edge and specified the occlusion. Critically, this was fully reversible; when the film was played backwards, the event reversed and the rear surface is de-occluded. 

What was happening was the deletion of optical texture; it was going away. But what is universally perceived is one surface going out of view behind the other (not ceasing to exist). The rear surface is perceived as persisting. Gibson claims that this is indeed perception, and not memory; the time-extended event in the optic array specifies a time extended event in the environment, and the former enables the perception of the latter. 

Gibson notes two things. First, occlusion is not the same as going out of existence. Surfaces can lose their integrity, and running that movie backwards does not produce the event 'coming into existence', or vice versa. So some events are reversible, and so some optical texture events are reversible, and some are not. We can perceive the difference. Second, reversible occlusion supporting the perception of the environment independently of what photons are hitting the retina as a fact of the optic array is a) ubiquitous and b) radically alters what a theory of vision should look like. 

Occlusion can be reversed by the motion of objects in the environment, or by the motion of the observer (locomotion). At any given point of observation, there are visual solid angles from some surfaces and not others. As the observer moves to another point of observation, some surfaces become more occluded and some more revealed (pg 187). This locomotion path is not a dense series of observation points, but a genuinely continuous trajectory through an environment and this path is the 'viewpoint'. That path can then vary in length, over seconds and meters but also minutes and hours and longer distances. Gibson claims that visual experience is not simply of the view at the current point of observation, but the view from the locomotion path. You do not have one view enriched with memory (of the past) and imagination (of the future), but a time-extended view enables by the reversible occlusion in the optic array. 

This is a typically radical hypothesis. Exploring an environment is not building a memory trace, but experiencing the layout of surfaces and how these occlude each other from various locations to support a rich perception of that environment. Gibson suggests that this perceptual mechanism can support orientation to one's environment over wide variation in time and space; that this can replace the notions of cognitive maps that must be assembled and referred to to navigate. Effectively, becoming oriented to an environment and being able to find one's way around it is about becoming a skilled perceiver of the higher order information specifying that environment. Obviously this only works to the extent that the layout remains stable, but this matches experience; "I know where the kitchen sink is, I think I know where the ski boots are stored, but I don't always know where my child is" (pg 189). 

Gibson then discusses how reversible optical transformations and occlusions supports public knowledge; two people inhabiting the same environment can perceive the same environment even though they will have different perspectives on it at any moment to the extent they have locomoted through it in common ways. You can also have secret knowledge; some locations are occluded from certain spots and you can use this to hide objects, events, or yourself (and the seeker can use this perception to seek these things as well). 

The key implication is the rejection of the theory that perception is based in sensations and that experience beyond the 'now' requires mechanisms other than perception.