February 1958
Note on the Responses of the Eye to Focusable Light
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
Woodworth says that the function of eye-movements is to “serve the requirements of clear vision” (Exp. Psych., Ch. 23, or revised edition, Ch. 17). This is a subjective statement. Can the functions of the ocular responses be stated objectively and yet meaningfully? What does the eye (or a pair of eyes) do in response to light? Assume a spherical array of focusable light at a given station point. A human eye responds to this stimulation in the following objective ways:
1. The eye admits a sector of the array and focuses it on the retina. Accommodation. More exactly, the definition of the image is maximized in the foveal region and is sacrificed in the periphery if necessary.
2. The eye maintains a posture (over an interval of time) such as to admit the same sector of the array. Fixation. This stabilizes the image on the retina, and therewith maintains the same bit of detail on the fovea. To speak accurately the retina is stabilized relative to the image, not the reverse. The stability is not perfect, since fixation involves tremor (physiological nystagmus). This activity seems to be different from the wandering posture of the eye occurring with a homogeneous array, or with a dark field. The implication is that a certain vibratory stabilization of the fovea is the best way for the receptor mosaic to register detail. So defined, fixation may require several types of compensatory movement of the eyes in the head:
a. With voluntary turning of the head, there is a coordinate compensation for active head movement. This response has zero latency; its onset is not a reflex.
b. With rotation of the body and head in a Barany chair there is compensation for passive head movement. This response has a considerable latency, being a reflex to labyrinthine acceleration when the eyes are closed. When the eyes are open and the Barany chair is not rotated but the surrounding environment (cylinder) is rotated instead, the response is purely to the retinal motion stimulus, and it still has a longer latency. This last is the same as a pursuit movement.
3. In the case of an optic array which contains motion, the eye so moves as to stabilize that part of the retinal image that falls on the fovea. A continually changing sector of the optic array is admitted to the eye. Pursuit fixation, or pursuit movement. There are two cases (only the first being described by Woodworth).
a. Pursuit of a moving object. With a stationary observer and movement of an object from place to place, one detail of the array moves relative to the remainder of the array. The fovea will be fixed on this detail. (It is interesting to note that whether the eye maintains such a moving fixation or a stable fixation on some detail of the background, the movement of the object relative to the background as equally perceived.
b. Pursuit of a part of the environment during locomotion. When the observer himself moves from place to place, there is a transformation of the total array insofar as this is projected from nearby objects and surfaces. This is termed motion perspective. The fovea may be fixed on any detail, and all other details of the total image will then move in graded angular velocities with reference to it. (Or the fovea may be fixed on a detail such as the horizon, which results in a stable fixation instead of a moving fixation). In any case, what seems to be perceived is locomotion of the observer himself.
4. The eye changes its posture so as to admit different overlapping sectors of the optic array during successive intervals of time. Exploratory fixation or saccadic movement. This brings the fovea to different details of the array in succession. Stability of the image at the fovea is generally maintained since the intervals are long relative to the eye movements, which are rapid. There are two cases:
a. Exploration of an environment. The outcome of saccadic movements (with which head movements are often combined) is to register both the overall pattern of the optic array and the subordinate patterns of its parts. There seems to be a search for the information contained in these variables of texture and form. Studies of exploratory fixation do not show any tendency for “scanning”, in the sense of fixating adjacent details of the array in successive order (e.g., Buswell).
b. Reading. In the special case of the artificial optical array reflected from a page , i.e. conventionalized marks on a surface, the exploratory fixations of the eye must be ordered instead of jumping about in a random fashion. The array must be scanned, since it has meaning only if it is taken part by part in sequence (e.g., from left to right and top to bottom). The outcome of this ocular behavior is quite different — to register not a “scene” but a sequence equivalent to speech .
5. Actually, a pair of eyes responds to a pair of slightly different optic arrays. (All the above ocular responses are conjugate; there is binocular accommodation, fixation, compensation, pursuit, and exploration). The two eyes admit two sectors of the disparate arrays so selected so as to center the fovea of each eye as corresponding details of the pattern of each sector. This is achieved by varying amounts of convergent fixation or convergence . The disparity of the two arrays (one being a perspective transformation of the other) cannot be registered by the binocular system unless each fovea is centered on the same detail. (The traditional but misleading way of putting this is to say that the two images cannot be “compared” for disparity without “superimposing” the “pictures”, with coincidence of one detail and non-coincidence of others. This leads to insoluble paradoxes about “fusion” of “mental images”, or the like).
Conclusion . The ocular system seems to be adapted to pick up information from the luminous flux of an environment by responding to the arrays projected to two station points, the difference between there arrays, and the changes in each of them over time, taking them one sector at a time in conjugate sectors. The fundamental responses of the system, to which eye “movements” are subordinate, are accommodation, fixation (including compensation, pursuit, and convergence), and exploration . It should be noted, then, that the “cues” or “clues” to the outer world must be in the light, primarily, not in the responses of the eyes to light. Accommodation and convergence of the eyes cannot yield more information about depth than is already present in the arrays.
Eye-movements cannot yield any more information about form than is already present in a static array. Pursuit movements cannot yield any more information about motion than is already present in a moving array. And fixation cannot yield any more information about the direction of an object in the world than is present in the array of light-as projected to a point in space.