February 1966
Rationale of a Current Series of Experiments on the Visual Perception of Superimposition
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
Two very old assumptions about vision should be reexamined. The first asserts as self-evident the impossibility of perceiving two things both lying on the same line of sight (Berkeley). The second states that basic visual experience is a two-dimensional patchwork of color sensations analogous to a painting. The assumptions are contradicted by certain facts.
1. A transparent surface is often seen as such, and this entails the seeing of one surface behind another(surface of water, or glass).
2. The back surface (rear side) of a transparent object can be perceived, both the front and back being visible.
3. A surface can be perceived as lying behind an opaque surface. The latter is said to occlude the former or to be superimposed on it.
4. A continuous surface can be perceived behind a continuous congruent transparent surface if two optical textures undergo relative motion (Gibson, Gibson, Smith, and Flock, JEP 1959, 58, p.40).
5. An object can be perceived to go behind another or surface (Michotte). This is “kinetic occlusion”. Reynold’s experiments.
6. The back surface (rear side) of an opaque object can be perceived as continuous with the front surface, as in the textured cylindrical drum experiment now being carried out.
7. The momentarily occluded faces or surfaces of an opaque rotating object can be perceived vividly. This fact deserves further study. All parts of the solid object seem to coexist simultaneously. Berkeley said, “a distance of itself and immediately cannot be seen. For distance, being a line directed endwise to the eye, it projects only one point in the fund of the eye, which remains the same whether the distance be longer or shorter” (1709). Koffka (Principles, p.115) was aware that this conception is invalid since we cannot “investigate space by examining its individual points separately one by one”. But he did not quite go far enough. He might have replied that, while the argument justly holds for point sensations it does not hold for surface perceptions. The concept of stimulus information for the perception of surface-layout escaped him except for his occasional references to “external” forces of organization. Surface-layout includes slant, curvature, dihedral edge (corner), occlusion-edge, and superposition, of bothopaque and transparent surfaces.
Note that perceived superpositions is of two types, transparent superpositions and opaque superposition. Only the latter is the type of perception that Michotte called amodal. What is the difference between the information for the two kinds of superposition?
The idea that one can see a point or spot behind another point or spot in the same direction is a contradiction. Berkeley was right here. The idea that one can see a color behind another color sounds like a contradiction but is not in fact. One sees a patch of color behind another patch of color when the two patches are not optically congruent (e.g. Wallach).
What about one surface behind another? It is true that one cannot “see” (sense) a surface behind an opaque surface, but one can “see” (perceive) it as behind one. However one can both “see” (sense) a surface behind a transparent surface and also “see” (perceive) it as behind one.
The cases of perceiving one surface behind another opaque surface are (1) an object occluding a larger background, (2) an object occluding only a part of another object, (3) the front of an object occluding the rear surface of the same object. There may or may not be relative optical motion (occlusion transformation) in these three cases. And there may (or may not) be binocular disparity (due to parallax) in these three cases. So there are many possible combinations of information for perceiving occlusion.