February 1974
A Neglected Set of Facts about Vision that can only be Comprehended by Ecological Optics
J. J. Gibson, Cornell University
The World Wide Web distribution of James Gibson’s “Purple Perils” is for scholarly use with the understanding that Gibson did not intend them for publication. References to these essays must cite them explicitly as unpublished manuscripts. Copies may be circulated if this statement is included on each copy.
The acceptance of traditional optics by psychologists as the basis for the study of vision has led to the neglect of many important facts. The neglect of these facts has sometimes been justified by supposing that they are connected with “perceiving” instead of “sensing”, and that we have to understand the latter before we can study the former, but this is dubious. One set of such facts has to do with surfaces and objects that are said to be “out of sight” or more particularly, are occluded. What are these facts?
1. The Opaqueness of surfaces. In contrast to the transparency of planes and forms in geometrical space, the opacity of most surfaces and objects in the environment has not been recognized. Opacity explains the fact that the earth is “cluttered”, i.e. that the “furniture” of the earth covers the earth. The principle involved is very general. I have called it optical occlusion.
Students of vision have long known about optical diminution with distance, or minification, but have not studied occlusion. They have talked about the so-called “vanishing” of objects at the horizon but not about the vanishing of one object behind another. This is a quite different way of going out of sight.
There are several ways in which an object or a surface can go out of sight besides the two mentioned. For example, we say that a thing goes out of sight “in the dark” or “under cover of darkness”. But actually this is a case where the whole optic array fails for lack of illumination ñ where optical information has ceased because stimulation has ceased. An optic array depends on an illuminated medium.
2. Occluded surfaces. The rule is that, for any point of observation, some of the opaque surfaces of the world are projected and the remainder are unprojected. The surfaces of the layout are thus divided into two portions. The obverse of this rule is that, for any piece of surface, it is projected at some points of observation in the air and unprojected in the remainder. All surfaces are projected to some point of observation and there is some point of observation at which any surface is projected.
3. Occluding edges or convexities. Between a projected and an unprojected surface (or between the projected and the unprojected part of a surface) there is always an occluding edge or an occluding convexity. This is what separates occluding from occluded surfaces. It is of two sorts (a) the edge or convexity separating the surface of a “foreground” from the surface of a “background” or (b) the edge or convexity separating the “near” surface of an object from the “far” surface. This says that an object hides both the background and the backside of the object itself, i.e. the near surface conceals both kinds of far surface.
4. The projections of projected surfaces. Corresponding to edges and convexities there are only margins between adjacent areas in the ambient optic array. This is what should be meant by saying that there is no “depth” in the array. It is a question-begging to speak of “superposed forms” and there is no such thing as a “cue” of “superposition”. But there is optical information for occlusion.
5. Reversible occlusion-disocclusion. Optical occlusion occurs in time, and is perfectly reversible. Points of observation usually move, and so do objects. Change of occlusion is caused by either (a) displacement of the point of observation relative to the permanent layout or (b) displacement or rotation of an object relative to the permanent layout and a point of observation.
In general, any surface that goes out of sight with a given movement of the point of observation (locomotion) comes into sight with the opposite motion. Hence concealed and revealed surfaces are continually interchanging. Along with this exchange goes a continual shifting of the occluding edges and occluding convexities.
6. Reversible changes in the optic array. Corresponding to any changing occlusion of surfaces is a unique change at the margins between areas in the optic array. This change of optical structure or texture is also reversible. An example is the deletion or accretion of structure on one side of the margin (cf. film on occlusion, and paper by Gibson, Kaplan, Reynolds, and Wheeler, 1969)
All the facts so far described are facts of ecological optics, not of physical or geometrical or physiological optics. The facts of reversible occlusion, for example, do not fall into either the category ofobjective facts or that of subjective facts. They are neither facts of physics as now taught nor of experience as now conceived. Occlusion expresses the complementary of the point of observation and the world, the reciprocity of the observer and the environment. It thus escapes the dilemma of dualism.
Consider next some of the facts of the perception of surfaces:
7. Experiments on occlusion. There is evidence to show that becoming-concealed (or revealed) is perceived as such, by children and adults. The studies by Michotte of the “tunnel phenomenon” and the “rabbit hole phenomenon” suggest this, and so does Kaplan’s recent experiment. Note that the perceiving of an object becoming concealed entails the perceiving of an object (or at least a surface) that isconcealed. So does the perceiving of an occluding edge. This is a paradox only if the assumptions of retinal image optics are applied to perception, the assumption that a field of color-patches is basic to visual perception.
8. Vanishing in the distance. In films with optical minification or magnification, children perceive Mickey Mouse going out of sight in the distance and coming into sight out of the distance. Such vanishing into a point is not to be confused with vanishing into thin air. Just as going and coming are opposite motions, so minification and magnification are opposite transformations in an array.
9. Change of state of a substance. It is at least a hypothesis that the going out of existence of a surface is distinguishable from its going out of sight (cf. film with Kaplan et al). In a later film on “reversible events” Gibson and Kaushall have illustrated the fact that the process of going out of existence in never the exact reverse of the process of coming into existence, whereas going out of sight is always the reverse of coming into sight.
The art of conjuring consists in large part of inducing the perception that something has gone out of existence whereas in fact it has only gone out of sight, or that it has come into existence when it has only come into sight. The optical information to specify occlusion is suppressed by the magician and the optical information to specify explosion or evaporation for example is substituted.
10. Orientation to hidden places. There is no doubt that animals and men can orient to hidden places like home. This means that one can perceive a hidden place or object behind the opaque surfaces that conceal it if one has perceived it going out of sight. The place goes out of sight as the observer departs and the vista closes in; an object may be put out of sight by hiding it in a drawer, or by burying it in the ground. The ability to visualize and orient to a hidden object is not adequately explained by the theory that the observer has a memory image of the object somewhere within him.
Finally consider some of the facts of social occlusion (hiding, concealing, screening, covering).
11. Hiding. To “hide” is to make one’s body unprojected at the point of observation of another person (or points likely to be occupied by other persons). Among animals, the prey hides from the predator and sometimes the predator hides in ambush from the prey. Concealment is practiced by children in the game of “hide and seek”. The rules of reversible occlusion are learned by babies in the game of “peek-a-boo”. None of these activities (behaviors and perception combined) has ever been studied by developmental psychologists. Motivation to achieve the state of seeing but not being seen is universal, but the understanding of it depends on what might be called social optics (peepholes, one-way vision screens, spyglasses and the like).
12. Privacy. The seeking of “privacy” seems to derive from the act of hiding one’s body. Shelters are also usually opaque enclosures or semi-enclosures having the same functions are screens, curtains, andblinds. In ecological optics we distinguish between the transmission of light through a screen or window and the transmission of optical information through it.
The human motivation for covering parts of the body with opaque clothing can be analyzed in this connection.