Memo on Motion (for Seminar on 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 References to these essays must cite them explicitly as unpublished manuscripts.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 physical motions of bodies and particles have been described and specified with great success, in terms of displacement (and spin) with respect to a chosen “frame of reference,” that is, a coordinate system with three axes using linear (and angular) coordinates. This is the basis of Newtonian mechanics.
Similarly the physical notion of light in space, considered as particles or waves, can also be described and specified by the same system, except that a considerable modification of it was introduced by the theory of relativity. Nevertheless a modified three-dimensional frame of reference is still the basis of physical optics.
But what frame of reference if any, is appropriate for describing the “motion of light” in quite another sense of the term — the “motion” by means of which the motions of bodies in an environment are seen (and the movements of an observer himself)? How are the transformations of pattern in the ambient optic array to be specified? In ecological optics (based on the steady state of multiply reflected illumination between opaque surfaces) what we have to describe is change in the perspective projection of objects, and change in the projection of the whole environment during locomotion. Moreover we have to describe the change from projected to unprojected, and vice versa, in the ambient light, that is, the effect of changing occlusion when the object (or the observer) moves. Is any coordinate system appropriate for these descriptions?
Gibson, Olum, and Rosenblatt (1955) analyzed what they called “motion perspective” for locomotion with respect to the surface of the earth, defining a pair of angles as the coordinates of each point in the array. They were thus able to describe the flow pattern of the array as something distinct from the contrast pattern of the array, but the analysis did not take account of occlusion, that is, of objects being concealed or revealed at an edge. In this case elements of the array disappear and appear (are deleted or accreted). Elements that do not exist have no coordinates. In this case a coordinate system is of no use.
The optical events that seem to constitute the information for the perception of material motions are not themselves motions. If so, they cannot be described by changing coordinates (pairs of numbers). Another kind of analysis is needed than the one which has been so powerful in physics.