Anderson, Joseph and Barbara Fisher. "The Myth of Persistence of Vision." Journal of the University Film Association XXX:4 (Fall 1978): 3-8.
The Myth of Persistence of Vision
JOSEPH ANDERSON and BARBARA FISHER
University of Wisconsin-Madison
The notion of 'persistence of vision' is ubiquitous in film literature. Almost every writer, whether one of film's great poets or a minor one, is to some degree responsible for its general pervasiveness. André Bazin marveled at how it took so long for the motion picture to come into being "since all the prerequisites had been assembled and the persistence of the image on the retina had been known for a long time."1 Raymond Spottiswood explained that:
... the movie camera is essentially a machine for taking pictures intermittently, the separate, spaced-out pictures afterward being fused together in the observer's brain. Persistence of Vision, a sort of mental hangover, prolongs the image of what the eye is seeing. In this way a rapid succession of slightly different still pictures deceives it into thinking that it has seen real continuity of movement. If the eye were entirely sober, there would be no movies!2
One could cite many more examples to attest to the frequency with which the notion of persistence of vision is mentioned, but frequency is hardly the issue; it is more important to realize that persistence of vision has become embedded in the foundations of film scholarship and has taken on the status of myth.
Myths develop, Levi-Strauss tells us, as a way of dealing with otherwise irresolvable conflicts inherent in the foundations of a culture. For two irresolvable cultural imperatives to be allowed to exist side by side in open opposition is dangerous. At any moment the assumptions about reality upon which the culture is based may come apart along this fissure, and the entire society may be split in two. Myth embodies the opposing imperatives, and the unity of the myth obscures the underlying schism such that it is possible to get on about the business of everyday living.
If one were to ask a student of film to identify the basic schism in film scholarship, he might answer the realist/formalist split, thinking of the basic divergence concerning the nature of film voiced by Bazin and Eisenstein. But there is an even more fundamental schism, one that is deeper and more dangerous to the field, the one that Munsterberg faced squarely when he suggested that we must first ask, 'What psychological factors are involved when we watch the happenings on the screen?' Secondly, he said. 'We must ask what characterizes the independence of an art, what constitutes the conditions under which the works of a special art stand.' 'The first inquiry,' he rightly observed, 'is psychological, the second esthetic."3
It is the schism between psychology and aesthetics that has been covered over by the myth of persistence of vision. It has served us well. On the one hand the myth of persistence of vision accounted for how it was possible for a viewer to perceive a motion picture at all, and on the other hand it prescribed the inherent nature of cinema itself.
Proponents of both realist and formalist schools of film aesthetics have invoked persistence of vision to lend a kind of organic or evolutionary inevitability to their positions. Bazin asserted that the recording process in film is so automatic--'For the first time, between the originating object and its reproduction there intervenes only the instrumentality of a non-living agent"-- that we perceive not an interpretation of reality but reality itself through the cinema, whose function then becomes to reveal not so much the physical reality as the phenomenal reality of the world. Inherent in that position are the two basic assumptions which characterize realist theories: (1) that the camera records reality objectively, and (2) that we perceive that record of reality directly. The myth of persistence of vision has provided the basis for the second, more fundamental assumption, by postulating a succession of veridical imprints upon the retina, and has allowed the first to go unchallenged. The explosive potential of the divergence pointed out by Munsterberg is thus contained by the myth.
Eisenstein years earlier had argued similarly that it is reasonable to seek a definition of "the principal style and spirit of cinema from its technical (optical) basis." But he said that the popular understanding of a 'blending' of successive images in film was misleading and suggested that it was more a matter of overlaying images. "For the idea (or sensation) of movement arises from the process of superimposing on the retained impression of the object's first position, a newly visable further position of the object.'5 Thus for Eisenstein, a collision occurs between every two frames of a film resulting in the impression of motion. This notion of collision, rooted in his formulation of the psychology of motion perception, provided the basis for Eisenstein's theory of montage.
The wedding of the structure of human perception to the structure of the work of art itself via the myth of persistence of vision has allowed film scholars through the years to paste over that crack in their discipline and get on with the business at hand, which with few exceptions has been the business of developing aesthetic theories. For many years the myth of persistence of vision made it unnecessary to question the psychological and physiological systems of the spectator which interact with the images upon the screen. When aesthetic theories fell short, film scholars found themselves unprepared to deal with the empiricism inherent in the development of psychological theory. In recent years it has become increasingly apparent that while films may be viewed as aesthetic texts, meriting study in their own right, nevertheless they are texts perceived by human beings in a social context. But when film scholars needed a theory of mind. they turned not to psychology but to Freud, a circumstance which prompted Richard Griffith to observe that they evidenced more interest in therapy than in theory.6
Two outstanding architects of film history, Terry Ramsaye and Arthur Knight, attribute the concept of persistence of vision to Peter Mark Roget. One suspects, however. that they did a bit of myth building themselves in the process. Ramsaye, in his inimitable style (that of a snake-oil peddler), provides this account:
The star of motion picture destiny, always traveling westward in the course of the race, reached England with Roget. One day while he was engaged in his inquiry into the affairs of the external senses he chanced to glance from his study window to note the approach of a vehicle. It was only a baker's cart and Roget hurriedly turned back to his papers. But, as he turned away, the line of his vision swept past the interferences of the slats of a Venetian blind. Through the slitted apertures the scientist caught the impression that the cart was proceeding by jerks. He saw it, despite its rapid motion, momentarily at rest in each slit, and, through each successive opening, he saw it in a different phase of motion. 7
Ramsaye goes on to say that Roget presented his finding in a paper before the 'Royal Society' entitled "Persistence of Vision with Regard to Moving Objects."
Arthur Knight provides the identical citation and goes on to recount the spread of Roget's theory throughout Europe. He lists a number of parlor toys that served to establish "the basic truth of Roget's contention that through some peculiarity of the eye an image is retained for a fraction of a second longer than it actually appears." Knight assures us that "upon this peculiarity rests the fortune of the entire motion-picture industry."8
It is questionable whether either Ramsaye or Knight bothered to check the text of Roget's original paper. If one consults the Philosophical Transactions of the Royal Society of London for the Year MDCCCXXV, one discovers that on 9 December 1824, P.M. Roget read a paper entitled "Explanation of an optical deception in the appearance of the spokes of a wheel seen through vertical apertures."9 Roget reports that if one views a revolving carriage wheel (he makes no mention of a baker's cart) through a series of vertical openings as in venetian blinds, 'the spokes of the wheel, instead of appearing straight, as they would naturally do if no bars intervened, seem to have a considerable degree of curvature" (Fig. 1). Roget makes no mention of the cart "proceeding by jerks" that Ramsaye reports. Nor does Roget ever use the term "persistence of vision." He does describe in great detail, with the aid of drawings, the optical illusion mentioned above.
Roget warns that one should not leap to the conjecture that the curved appearance of the spokes of the wheel is the result of visually combining short straight pairs of several spokes into a curved image of one spoke. The illusion of curvature, he points out, still persists when only one spoke is rotated behind a moving grating. In fact, the geometric explanation of the curvature is fairly obvious and is explained quite adequately by Roget. He even provides a mathematical formula for determining the curvature under the described conditions. In its simplest form we can say that if a given line rotates about a center (such as a spoke of a wheel) and if that line is intersected by a moving line of consistent orientation (such as the edge of a venetian blind), the plot of all possible points of their intersection will describe a curve (Fig. 2). It is apparent that if one were to view a rotating line through a narrow moving slit, he would never see the entire line at once but would see only a point moving in a curved path. He would thus have no way of knowing that the line generating the moving point is in reality a straight one. His perception of a point moving along a curved path is veridical; it is not an illusion.
The curved spoke effect is, however, a bit more complicated. What is reported is not simply the observation of a point moving in a curved path but instead, an array of complete curved spokes that appear not to rotate at all but remain stationary. The more complex image is no doubt made possible by the more complex viewing situation, i.e., a series of spokes are viewed through a series of slits. But how is the effect produced? It is reasonable to assume (if spacings remain equal and velocities constant) that the curved path traced by the intersection of the first slit and the first spoke will be retraced by the second slit and the second spoke, and so on. The remaining question is how the eye combines the sequential tracings of moving points into static forms, i.e., the curved spokes of a wheel. It is at this point, where Roget attempts to provide an explanation for the phenomenon he had so adroitly described, that he departs from the discipline of scientific observation and lapses into speculation.
Along with the precise description of the phenomenon itself he provides us with an explanation of the fact that the spokes of a wheel when passing behind a grating "leave in the eye the trace of a continuous curve (sic) line and the spokes will appear to be curved." He likens the phenomenon to the
. . .illusion that occurs when a bright object is wheeled rapidly round in a circle, giving rise to the appearance of a line of light throughout the whole circumference- namely, that an impression made by a pencil of rays on the retina, if sufficiently vivid. will remain for a certain time after the cause has ceased. 10
It cannot be, as Roget suggests, so simple a process as the fusing of slowly decaying tracings upon the retina, for the curved spokes persist even when the eye is moved about over the display. To put the matter another way, when Roget speaks of tracings remaining upon the retina from a sufficiently bright object, he is apparently referring to after-images. Yet it is well known that after-images, since they are in fact tracings of stimulation left upon the retina, yield stabilized images (an image that is on the retina itself and therefore moves as the eye is moved). If after-images were involved in the spoked wheel illusion, the result would be a plethora of images resulting from the tracings scattered about the retina according to each separate fixation of the eye. This is, of course, not what one sees; what is seen is a single unified, symmetrical pattern of curved spokes, no matter how one moves his eyes about.
Thus we are forced to conclude that while Roget explained very well the mathematical origin of the curved paths of the stimulus itself, he was unable to account properly for the phenomenal experience as a whole. It is at the level of accounting for human processing of the stimulus array that he falls short. No psychologist today would attempt to explain the phenomenon solely in terms of processing occurring at the level of the retina. Unfortunately, modern psychologists have not attempted to explain this phenomenon at all. Direct references to Roget's observation are nonexistent in contemporary literature on perception.
They are prevalent, however, in film literature. Unfortunately, film scholars were not equipped to recognize Roget's error in accounting for the curved spoke effect. They therefore used his incorrect explanation to account for the perception of the successive frames of a motion picture as a continuously moving image. In doing so, they erred on two counts: assuming (1) that the process in question occurs at the retinal level and involves the fusion of successive "tracings," and (2) that such a process could account for the perception of motion. Clearly, a simple fusion of a succession of images would not result in the perception of motion. It would result in the perception of one composite, still image. In short, Roget's 'illusion' is not an illusion of motion, nor does he claim that it is.
The phenomenon which he describes has little, if anything, to do with filmic illusion. Roget has described a case in which a series of moving points results in the perception of a static image. In cinema a series of static images results in the illusion of motion. Nevertheles, Roget's explanation of 'an optical deception in the appearance of the spokes of a wheel seen through vertical apertures" has been accepted by a generation of film scholars as the valid explanation of the perceptual combination of successive frames of a motion picture.
Other more recent psychological inquiries would to be far more relevant. In 1963, Roget's modern professional counterparts, David H. Hubel and Toren N. Weisel published a paper which passed unnoticed by film scholars, which not only reopened the field of visual perception, but also threatened the tranquility of any field which has embedded in its foundations assumptions about how see. Their paper, entitled 'Receptive Fields, Binocular Interaction and Functional Architecture in the Cat's Visual Cortex," demonstrated that processing in mammals proceeds from the eye through a systematized hierarchy of specialized cells in the brain.11 And while in certain lower animals there is a great deal of processing in the retina, Hubel and Weisel demonstrated that in mammals there is a rather direct path from the receptors in the eye to the brain, and that for the most part, processing is deferred until the signal reaches the brain.
In the brain, although processing proceeds in an orderly manner according to the constraints of the hierarchical system, that processing becomes exceedingly complex. The complexity staggers our imagination, for physically the system consists of vast networks of microscopic cells connected to each other with gossamer-like fibers. A single cell may have as many as a thousand connections with other cells in the network. And there are millions of such cells all in continual communication with each other in a yet to be deciphered neural code. Somewhere in this system reside the mechanisms for processing a series of rapidly presented static images as continuous motion. Hubel and Weisel probed only the first few levels of the visual system, but they revealed enough of its vast and intricate structure to change forever our concept of what it means to see.
In the light of contemporary discoveries about visual processing, the simple myth of persistence of vision appears as naive as the story of Adam and Eve in the light of contemporary anthropology. But just as anthropology is hard-pressed to replace the myth of creation with an equally satisfying story, so contemporary psychology is in the embarrassing position of having destroyed our faith in the unity of psychology and aesthetics as bound up in the myth of persistence of vision, and yet remaining unable to supply a surrogate explanation to bridge the basic schism in our film theory.
The simple truth is that even though we have been looking at motion pictures for three quarters of a century, we still do not understand their most basic perceptual principles. Our predicament has been summed up quite succinctly by Irvin Rock:
Everyone, it would seem, knows that moving pictures are made by projecting a series of stationary frames on a screen in rapid succession. Yet few people seem to be curious about the basis of this effect and those who are, seem to be satisfied with an incorrect explanation. . . . The fact of the matter is that we do not know why movement is perceived.12
Film scholars, like cuckold husbands, are apparently the last to know that the myth of persistence of vision is no longer true.
The process of demythologizing reopens some basic questions of classical film theory, and in particular necessitates a re-evaluation of optically-based concepts of the film medium. It calls into question certain claims made by the young Eisenstein in his attempt to support his theory of montage by appeal to isomorphic perceptual processes in the viewer (more so than in Eisenstein's later work, where persistence of vision is no longer the primary basis for his model of visual processing). There are more serious and far-reaching implications, however, for realist theories of film. Roy Armes stated the realist position when he argued that the cinema
. . . is indissolubly linked to the camera as a means of obtaining basic images. The cinema has a potential for realism because, though film projection is a process of illusion, relying on a defect of the eye (the inability to differentiate images which follow one another at a rate of sixteen or twenty-four frames a second), the camera itself does not cheat. The images it gives are those which record the successive stages of movement as they occurred in real life.13
These arguments about the nature of the film medium, like the historical claims for evolutionary inevitability, are much less persuasive when denied the support of the persistence of vision myth. Other aspects of the medium, too, such as the "invisible style" of classical Hollywood cinema, which is founded upon and in turn reinforces these ontological assertions, require further investigation.
Without the myth of persistence of vision to obscure the breach in our discipline. we are able to face directly the opposing imperatives of psychology and aesthetics. Rather than simply reciting a familiar story about Roget and "Persistence of Vision" and then proceeding with an aesthetic inquirv, film scholars are now in a position to avail themselves of the literature and methodology of perceptual psvchology. As we come to greater understanding of the psychological factors involved in viewing a motion picture. the issues alluded to above may well appear in quite a different light. It is now possible to pursue psychological inquiry into the cinema with the same intensity we have in the past devoted to aesthetic inquiry. As Munsterberg observed, "they belong intimately together."14
1 André Bazin, What is Cinema?, Vol. I, trans. Hugh Gray (Berkeley: University of California Press, 1967), p. 19,
2 Raymond Spottiswood, Film and its Techniques (Berkeley: University of California Press, 1963), p. 47.
3 Hugo Munsterberg, The Photopay: A Psychological Study (1916: rpt. as The Film: : A Psychological Study, New York: Dover Publications, Inc., 1970), p. 17.
4 Bazin, p. 13.
5 Sergei M. Eisenstein. Film Form, and ed. and trans. Jay Leyda (New York: Harcourt, Brace and World. Inc., 1949), p. 49.
6 Richard Griffith, "Introduction" to Dover Edition of Munsterberg, p. xii.
7 Terry Ramsaye. A Million and One Nights (New York: Simon and Schuster. 1926), p. 10.
8 Arthur Knight, The Liveliest Art (New York: New American Library, 1957). p. 14.
9 Peter Mark Roget, "Explanation of an optical deception in the appearance of the spokes of a wheel when seen through vertical apertures," in Roval Society of London. Philosophical Transactions, MDCCCXXV, Pt. I (London: Nicol, 1825). pp. 131-140.
10 Roget, p. 135.
11 David H. Hubel and Toren N. Weisel, "Receptive Fields, Binocular Integration, and Functional Architecture in the Cat's Visual Cortex," Journal of Physiology 160 (1962): 106-154.
12 Irvin Rock. An Introduction to Perception (New York: Macmillan Publishing Company, Inc., 1975), pp. 193-194.
13 Roy Armes. Film and -Reality (Middlesex, England: Penguin Books, Ltd., 1974), p. 17.
14 Munsterberg, p. 17.