2—

The metaphor of the camera-eye is constructed of synecdoches. That is to say, the eye and the camera are parts standing for the whole of their respective visual apparatuses. Vision is no more a product of the eye alone than pictures (especially the "moving pictures" of cinema) are made by the camera alone. In each case, what we see is the result of complex processes that only begin in the eye and the camera. No doubt it is because they house the beginnings of their respective ways of seeing that the eye and the camera have acquired their synecdochic weight. They are the outermost extensions of visual systems whose other structures and functions are hidden inside the skull and inside film labs, editing rooms, and projection booths. Even the crucial light-receptors of each system (the retina and the film) are hidden from view. An analysis of the camera-eye metaphor may properly begin with the eye and the camera per se, but if it is to demonstrate the metaphor's relevance to the visual aesthetics of avant-garde film, it must go on to seek other, less apparent correspondences between the two visual systems.

The classic essay on the subject is George Wald's "Eye and Camera," published in Scientific American in 1950. Wald first asserts, "Today every schoolboy knows that the eye is like a camera," and summarizes these likenesses as follows:

In both instruments a lens projects an inverted image of the surroundings upon a light-sensitive surface: the film in the camera and the retina in the eye. In both the opening of the lens is regulated by an iris. In both the inside of the chamber is lined with a coating of black material which absorbs stray light that would otherwise be reflected back and forth and obscure the image.^[26]

Wald goes on to point out similarities in the light-sensitivity of the film and the retina. Just as a fine-grained, "slow" film is designed for high intensities of light and a more coarsely grained, "fast" film for low intensities of light, so the retina has two kinds of receptor cells: the cones, which operate in bright light and provide the more sharply defined details of our visual world, and the rods, which work at lower light levels and are the source of the coarser, less sharply defined details in the peripheries of our visual world.

Moreover, the cones and rods are on the ends of minute stalks that respond to the light's intensity, so that when the light is dim, the rods are pulled forward and the cones pushed back; when the light is bright the cones move forward and the rods draw back. As Wald says, "One could scarcely imagine a closer approach to the change from fast to slow film in a

Comparable structures and functions of the camera and the eye
 (adapted from George Wald, "Eye and Camera," 
Scientific American , July 1950).

camera." In subsequent layers of the retina, according to more recent research by Frank S. Werblin, the bipolar cells emphasize high contrast in the retinal image, while the amacrine and ganglion cells moderate contrasts. "It is as if," Werblin writes, "a camera system could switch automatically from a high-contrast film to a low-contrast film when it encountered a rapidly changing or a very contrasty scene."^[27]

For Wald, the retina and photographic film offer another kind of analogy, because of their chemical response to light. The rods contain a pigment, rhodopsin, that bleaches in the light and is resynthesized in the dark. This led the nineteenth-century physiologist Willy Kühne to devise an experiment in which he was able to take a picture with the living eye of a

rabbit. First, the rabbit's head was covered to allow rhodopsin to accumulate in the rods. Then it was uncovered and held so that it faced a barred window. After a three-minute "exposure," the animal was killed, its eye removed, and the rear half containing the retina "fixed" in an alum solution, so that the bleached rhodopsin could not be resynthesized. "The next day," Wald reports, "Kühne saw, printed upon the retina in bleached and unaltered rhodopsin, a picture of the window with a clear pattern of its bars."

Wald's own variation on this experiment was to extract rhodopsin from cattle retinas, mix it with gelatin on celluloid, expose it to a pattern of black and white stripes, then "develop" it in darkness with hydroxylamine. The result was a "rhodopsin photograph" showing the same black and white pattern. Thus, just as exposure to the light produces a "latent image" in a film's emulsion, so, Wald argues, "light produces an almost invisible result [on the retina], a latent image, and this indeed is probably the process upon which retinal excitation depends. The visible loss of rhodopsin's color, its bleaching, is the result of subsequent dark reactions, of 'development.' "It is now known that the cones also contain rhodopsin-like pigments that make color vision possible, which leads John Frisby to write, "So really the rods and cones are two distinct light-sensitive systems packaged together into a single 'camera'—the eye."^[28]

If the vertical bands of light and dark gray make one think of the barred window that left its lasting impression on the retina of Kühne's rabbit, it is an appropriate—if somewhat ironic—association, so long as one remembers that neither image duplicates actual vision . They are simply chemical traces of rhodopsin's response to the "light flux" that reaches the retina from the outside world; they are images of "the process upon which retinal excitation depends," as Wald put it. Nevertheless, Wald's and Kühne's experiments show the eye to be more like a camera, and seeing more like photography, than is often recognized. They strengthen the metaphor of the camera-eye by grounding it in processes that can be scientifically verified. In Wald's words, "The more we have come to know about the mechanism of vision, the more pointed and fruitful has become its comparison with photography."

As convincing as that may sound, it is not a view all scientists of vision share. In Handbook of Perception R. M. Boynton offers a pointed and thorough rebuttal:

The eye most emphatically does not work just like a camera, and the differences are worth discussing. The eye is a living organ, while the camera is not. In a camera, light passes through the image-forming optics of high refractive index, and then back again into air before striking the film plane. In the eye,

high-index media are encountered as light enters the eye at the outer surface of the cornea, but the light never again returns to air. The control of pupil size begins with the action of light upon the identical photoreceptors that initiate the act of vision, while the camera's photoelectric analog, when there is one, is located so that the light falling upon the photocell is not affected by the size of the opening in the iris diaphragm. The lens surfaces in most cameras are sections of spheres, to which an optical analysis developed for spherical components can properly be applied. There is no spherical surface anywhere in the eye. The camera lens is homogeneous in its refractive index (or at most contains a few such distinct elements, each of which has this property). The lens of the eye is layered like an onion, with the refractive index of each layer differing slightly from the next. Cameras have shutters and utilize discrete exposures, either singly or in succession. The pupil of the eye is continuously open. Cameras must be aimed by someone; the eye is part of a grand scheme which does its own aiming. Images produced by photographic cameras must first be processed and then viewed or otherwise analyzed; the image produced upon the retina is never again restored to optical form, and the mechanisms responsible for its processing are perhaps a billionfold more complex than those used in photography.^[29]

The list of differences "could be expanded," as Boynton says, but it is surely long enough to discourage anyone from turning to literal-minded scientists for validation of the camera-eye metaphor.

The fact that the eye does not work "just like a camera" is indisputable, but it is also irrelevant, since the significant similarities between the two are metaphorical, not literal. Boynton's effort to discredit the camera-eye metaphor is useful, however, for several reasons. First, it specifies the basic difference underlying the likenesses implied by the metaphor. The difference is between a machine and, in Boynton's words, "a living organ"—between Vertov's "mechanical eye" and Brakhage's "flesh window." It is the basis of the dialectical relationship of eye and camera, from which the visual aesthetics of avant-garde film have emerged.

Second, Boynton repeats a common objection to equating the camera and the eye when he emphasizes the difference between the photographic image and the retinal image. It is true that the retinal image is "never again restored to optical form" and is nothing more than a stimulus for retinal cells at one of the earliest stages in the total visual process. What must be stressed, however, is that the production of an optical image in the camera and in the eye, though essential to both visual processes, is not in itself the basis of their most significant resemblances. Light moving in time—not images—is the "essence" they share.

A third point arises from Boynton's critique of the camera-eye metaphor. Like virtually all commentators on the camera and the eye, Boynton

implies that the photographic image is the visible equivalent of the image cast by the lens on the film plane of the camera. In still photography this is more or less true (allowing for the inevitable differences created by the chemistry of processing and printing photographs), but in cinema, it is not. What the film viewer sees are not images on film but images projected on a screen . These images are created by light moving in time, and therefore they much more closely approximate the sources of seeing than do the images fixed in the emulsion of photographic film.

Cinematic images partake of the same "optical flow" described by Gunnar Johansson: "The optical flow of images into the viewfinder of a camera (or into the camera itself when the lens is open) corresponds to the optical flow impinging on the retina during locomotion."^[30] In fact, since the eyes are always in motion, the image falling on the retina is always flowing over the retinal cells. Of course, cinematic images can not reproduce the same "optical flow" that entered the camera. There are too many intervening steps to permit the original "optical flow" to emerge from the projector unchanged (not to mention the fact that cinematic images may be made without the use of a camera at all). They can, however, represent the same kind of "flow" that impinges on the retina, the only difference being that their "flow" is shaped by the filmmaker through the materials and processes of the cinematic apparatus. Thus the camera-eye metaphor continues to be valid, if one takes into account the actual nature of the film image and conditions of film viewing.

A fourth point is suggested by Boynton's sentence "Cameras must be aimed by someone; the eye is part of a grand scheme which does its own aiming." The camera-eye metaphor should remind us that the camera, too, is "part of a grand scheme" that controls the way it is "aimed" at the world. Whether the camera is held in the hands of Stan Brakhage and "aimed" by Brakhage's intuitive response to his feelings and immediate environment, or attached to a motorcycle's handlebars and "aimed" by Vertov's cameraman as he steers around an inclined track, or perched atop Michael Snow's elegant remote-controlled machine and "aimed" at the Québec landscape by electronic impulses scripted by Snow—the camera is integrated in "a grand scheme which does its own aiming." Metaphorically, it is like the eye in its own "grand scheme" of muscles, tissues, nerves, and brain cells. Here, in fact, is another way of comparing the eye and the camera as synecdoches representing a whole—the "grand scheme"—of which each is a particularly conspicuous but totally integrated part.

Despite the objections raised by Boynton, then, the camera-eye meta-

phor not only continues to make sense but gains strength and pertinence as it is given closer scrutiny—so long as (1) it is understood to be a metaphorical juxtaposition, not a literal equivalence, producing a dialectical relationship of mechanical and organic structures and functions; (2) its implied similarities between the retinal image and the photographic image are recognized to be less relevant than its allusion to the flow of light essential to both visual and cinematic perception; (3) it is treated as a comparison of interrelated parts and processes constituting the "grand schemes" of visual and cinematic perception.