January 1971
Current Problems in Ecological Optics
(Projects for Psychol. 512. Seminar on Perception, Spring 1971)
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
Critical reading of Ch. 10-12 in The Senses Considered as Perceptual Systems is presupposed, and also Ch. 1-4. Other references and memoranda will be supplied.
A. Questions concerning the validity of ecological optics.
1. The distinction between seeing light and seeing things by means of light is basic, but it can be subdivided into several distinctions. Consider the following questions: How do we see light? How do we see luminous things (stars, flames)? How do we see non-luminous things (surfaces) by means of light (illumination)? How do we see pictures of non-luminous things by means of illumination? Finally, what are the possible kinds of visual displays (modes of displaying optical information)?
2. Contrast the theory that ambient light is composed of a bundle of light-rays (each ray being a stimulus) with the notion of ambient light as having structure. Consider the theory of a projective correspondence between elements of the physical world and rays of light coming to an eye. Survey existing statements of the correspondence between elements of the world and rays of light (examples can be found in Hull, 1943, p. 33; Gibson, 1950, p. 45-47; and Rock, 1966, p. 2). Wherein is the force of this ray-theory of projection?
3. Contrast the above theory with the new conception of a correspondence between geometrical features of the world and invariants in the structure of the ambient light. Is this latter kind of correspondence projective? In what sense?
B. Consider what is known about the environmental “structuring” of ambient light, i.e., the natural causes of its stricture in a terrestrial world. The following topics fall under this heading.
1. Read critically the last half of Ch. 10 in Perceptual Systems. What is the value of the classification into layout structuring, reflectance-structuring, and shadow-structuring if the classification is not exhaustive?
2. Examine the theory of the structuring of ambient light by variations in surface layout. What are the main geometrical features of layout? What are the corresponding optical invariants? How does occlusion enter into the problem of perceiving layout?
3. What is known about reflectance-structuring and shadow-structuring? (Reflectance of a surface and shadowing of a surface are non-geometrical features of the environment ñ agreed?) How does the old problem of whiteness-constancy in the presence or absence of shadow enter in? What is blocking progress in understanding this kind of perception? How do we reformulate the problem of the perception of surface-color?
4. What are some other natural causes of an optic array, and are they important for everyday vision? (Cf. memo on the perception of “intangible things.”)
5. The structuring of light by artifice (e.g., pictures) is considered to be of a special kind — a technology of displaying optical information. (Ch. 11 and additional references, esp. J. Kennedy.) What is the difference between the way a picture structures light and the way an environment structures light?
C. Consider the general problem of defining the invariants over time of a transforming ambient array (p. 160 ff, 195 ff, 206 ff, and 264 f.).
1. Is it established that an ambient array changing in time specifies both a unique path of locomotion and a unique layout of environmental surfaces?
2. The geometrical features of objects, and of the opaque environment, consist of corners, edges, convexities, concavities, occluding edges and occluding convexities, according to the memo of August 1970 and the definitions given therein. Is this list adequate? Does it provide a sort of “feature geometry” for describing the environment as contrasted with the abstract three-dimensional, empty-space geometry of Euclid and Descartes?
3. The optical information for the perception of these surface-features is said to lie in the formless invariants of a persisting optic array, not in the “forms” or “figures” of the array (i.e., the array considered as a patchwork). This hypothesis needs to be elaborated. What is the evidence for it?
D. Consider the general problem of environmental events, and event perception, as distinguished from the general problem of environmental layout and layout perception (“objects” and “space”).
1. Read critically the paper entitled The Problem of Event Perception (unpublished) and the paper What gives rise to the perception of motion? (Gibson, 1968).
2. Consider the new hypothesis that the optical information to specify an ecological event is a “disturbance” of the structure of an optic array, and the corollary of the hypothesis that there is no such thing as the “motion” of an element in an optic array, either absolute or relative. This is an upsetting idea. Is it intelligible?
3. A tentative classification of ecological events is offered. What kinds of optical disturbances go with what kinds of events?
E. Other problems
The pickup of the information in the simultaneous disparity of two arrays by the binocular system.
The perception of non-rigid surface layout as distinguished from rigid layout.
The perception of the horizon.
In general, how much of ecological optics has now been established and what remains to be done? What should be the program for the future?
References to published material:
C. Hull. Principles of behavior, 1943.
I. Rock. The nature of perceptual adaptation, 1966.
J. Gibson. The Perception of the Visual World, 1966.
J. Gibson. The Senses Considered as Perceptual Systems, 1950.
J. Gibson. What give rise to the perception of motion? Psyhcol. Rev., 1968, 75, 335-346. (86)