Note on the Terminology of Distortion in the Experiment on Adaptation to Prismatic Spectacles
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
One of the main concerns of the Innsbruck experiments (Kohler, 1951, 1964) was the adaptation of visual perception to the wearing of wedge prisms in front of the eyes as spectacles. This is an old experiment but it has recently been taken up by new investigators. The alterations of optical stimulation thus produced are of several sorts, and it is important but difficult to specify them exactly so that the adaptation to such alterations can be measured. What are they?
It is easy to assert that the field of view (the sheaf of rays entering the eye ) has been displaced toward the base of the prism (by refraction) but this description is incomplete. It is not hard to note thespectral fringes at otherwise abrupt intensity-borders in the array (because of differential refraction) but they are not simple. One can observe the apparent curvature of straight borders or lines in the prismatic field of view but the alteration is much more complicated than this. There are many dimensions of distortion in this field, not just curvature. The prismatic field is also “tapered” relative to the normal field, and its “density” is graded in a very different way. And finally, most important, the observer with spectacles notices what he describes as rubbery stretching of phenomenal objects and surfaces whenever he moves his head. The world seems to have elasticity instead of its usual rigidity. The cause of this movement-produced distortion over time, as distinguished from the causes of the spectacle-produced distortions need to be clarified. I propose to call the former “deformation” instead of distortion, and I wish thus to emphasize the difference between a sequential transformation, a change in time, and a geometrical transformation, a simultaneous relation between two forms (Gibson, 1957).
A number of terms, with somewhat different meanings, are being used by various investigators to describe how a prism alters the sheaf of light rays available to an eye with respect to pattern or, more usually, how it alters the retinal image in an eye. The terms are distortion, rearrangement, transformation, and deformation. To avoid confusion, I suggest the following definitions.
1. Distortion (of form or pattern). This is the pattern entering the eye from a stable environment with a spectacle relative to the pattern entering the eye from that environment without a spectacle. It can be termed a “transformation” in the geometrical sense of the term, a relation between two forms that are in point-to-point correspondence. It could also be called a “rearrangement”, which is the term used by Held (1962), but the idea of a permutation of order implied by that word must be excluded. A prism does not permute the elements of optical texture. Note that the word distortion suggests departure from a norm or standard. It is the alteration of form introduced by the prism.
2. Gaze-contingent Distortion. This is a new term employed by Hay and Pick (in preparation). It is the alteration of the form from an object in the environment as this depends on different amounts of refraction in the overall prismatic field, that is, the difference in distortion that corresponds to the difference of right-left or up-down in the field of view (See illustration, from Hay). Hay and Pick have recently confirmed the fact of the “conditioned” or “conditional” aftereffects discovered by Ivo Kohler as a consequence of prismatic adaptation. They are “gaze-contingent” in that they are measurably different (without spectacles ) when the eye looks up, down, right, or left. Just so, but in an opposite way, the distortion was different (with spectacles) when the eyes looked up, down, right, or left. Note that a changing distortion of form is obtained at the eye when either (a) an object is moved across any meridian of the total field with the head-and-spectacles stationary, or (b) a stationary object is fixed by the eye while the head-and-spectacles are turned on any axis. The amount of distortion depends on the position of the foveal gaze-line of the eye relative to the head or the position of the head relative to the foveal gaze-line. But the changing distortion depends on movement, and this leads to the third definition.
3. Sequential Deformation. This is a change of form with time. It is a change of the same form, not a geometrical relation between two forms, nor an abnormality of one form relative to a standard. It is a sequential transformation as distinguished from a geometrical transformation.
Note that a form that undergoes deformation by this definition will already have undergone distortion by previous definitions and that the two are not at all the same (See illustration). Evidence suggests that a sequential deformation is immediately noticeable, whereas a distortion, unless it is an abnormality, may not be noticeable. A “running” deformation can be detected as a visual event, whereas a distortion (as defined) can only be perceived by comparing or judging and, unless the perceiver has absolutes like vertical, horizontal, rectilinear and rectangular, this perception may be indirect. It is important to remember that a sequential deformation, as defined, is a visual accompaniment of eye or head turning, a feedback or reafferent, whereas of course a distortion is not. Deformation results from either eye-pursuit with stationary head or head movement of any sort with fixed eyes. The experience is that of rubbery motion or elastic change of phenomenal objects. [Presumably it is called elasticbecause the information corresponds in some special way to the information that, in ordinary vision, comes from elastic physical motions in the ordinary environment (Fieandt and Gibson, 1959) rather than from rigid physical motions of the environment relative to the observer or of the observer relative to the environment.]
DiscussionInvestigators of the aftereffects of distorting spectacles, including the writer, have tended to think of the adaptation as simply an adjustment of the visual sensory system to the abnormal pattern of stimulation, considered as a constant thing. That is, it is an adaptation to altered pattern vision. It could be conceived as a sort of phenomenal normalization of pattern, and the aftereffects as a normalization. And, indeed, in the center of the field of view there can occur adaptation to the curvature and tilt of lines, with prolonged looking, the curves tending to become straight and the tilts tending to become vertical or horizontal (Gibson, 1933, 1937, 1937). There are subsequent negative aftereffects analogous to afterimages.
On this line of theorizing, Kohler’s discovery of aftereffects that were conditional upon eye-posture came as a great surprise. Since the gaze-contingent distortions differ, it would seem that the subforms in the outer areas of the field of view must be normalized differently than the subform in the center of the field (See illustration), a requirement that violates our ideas of how the retina works. Evidently, however, the visual system as a whole, the exploratory head-eye-retina system, can somehow become habituated to the array as a whole, despite the seeming paradox that adaptation in the foveal region must be different for different eye-head positions. And the phenomenal trend that goes with this rehabituation is not so much a normalization of sensations of form as a tendency for visual perceptions of objects to become constant in their proportions and for visual perception of the environment to become constant in scale. But perhaps the eye responds not simply to pattern but also to sequential deformation. Perhaps the most important adjustment of the perceptual system to wearing of spectacles is one for which there is, as yet, no clear theory. This is the adjustment to abnormal optical motions, the sequential deformations obtained with prisms. It is not to a biased pattern, or even to a biased pattern of patterns, but to a biased change of pattern. The spectacles alter the visual propriospecific feedback from movements of the body, head, and eyes. This is not muscular, or articular, or vestibular proprioception, but strictly visual proprioception (and proprioception in general is not a sense but a multiple input system that cuts across the senses). Some investigators have been appealing to classical proprioception as an explanation of visual adaptation to distortion (Harris, Add) but this is quite unnecessary. The so-called “rubbery motions” that appear with spectacles (and reappear in reverse when the spectacles are removed) might be understood as the products of abnormal visual kinesthesis from movements of the body, head, and eyes. The normal stimulus motions are taken as kinesthesis with a phenomenally rigid world; the abnormal motions are taken as kinesthesis in a phenomenally elastic world but the new component is gradually converted into kinesthesis; the normal stimulus motions then must be taken as kinesthesis in a phenomenal world of opposite elasticity. This sort of adaptation is quite rapid. The reason may be that we learn to separate promptly the propriospecific component of perceptual input from the exterospecific component even when the relation between these components is altered. At least we do so when a perceptual system is allowed to be active and exploratory. In this situation the variations in the stimulus-flux are self-produced and probably can thus be distinguished from the variations that might be environment-produced. The rubbery motions are in fact propriospecific and they come to be experienced as propriospecific instead of as visible motion. If this is possible, the present-day concern with the contribution of proprioception to visual perception is on the right problem but off on the wrong track (Held, 1962, Harris, 1964). We should isolate and study propriosensitivity as we find it, unhindered by dogma of separable sense modalities and the fallacy, as I believe, of specific nerve energies. Visual proprioception is a fact (e.g., Gibson, 1958) and the apparently “rubbery” character of visual feedback with prismatic distortion deserves investigation in its own right.
References (in order of mention)
Ivo Kohler, Formation and transformation of the perceptual world.(Original in 1951. Translated in 1964 as a monograph in Psychological Issues.)
J. Gibson, Optical motions and transformations, Psychol. Rev., 1957, and Continuous perspective transformations, JEP, 1957.
R. Held (any of the series of reports).
J. Hay and H. Pick (mimeo paper on “Contingent aftereffects”).
K. Fieandt and J. Gibson, Two kinds of transformations, JEP, 1959.
J. Gibson, Curvature adaptation, in JEP, 1933. Tilt adaptation (two papers), in JEP, 1937. Adaptation with negative aftereffects, Psych. Rev., 1937.
R. Harris, Prismatic adaptation as a “proprioceptive shift,” 1964.
J. Gibson, Visually controlled locomotion, Brit. J. Psychol., 1958, p. 49, 182-194.