The history of visual perception includes innumerable theories . I'd like to quote you two, from the studies of Molineux and Clerambault . Molineux asks: "If a man blind from birth suddenly recovers his sight and sees a sphere and a cube of which he has previous tactile knowledge, will he be able to tell them apart by sight alone?" This is a question which has provoked many contradictory replies . But whether we decide (as the nativists do) that like any other human being the blind man is equipped from birth with archetypal images of both shapes, or that the interconnection between tactile and visual experiences allows immediate recognition (the empiricist belief), or that a period of transition is required, or that visual objects appear as continuous surfaces (such that a joint operation of touch, sight, and movement is necessary in order to understandthem), still the underlying principle of each of these responses will be the same, namely that reality can be articulated and reproduced. The outside world possesses a grammar which we can describe and use to invent an entirely artificial world, to which absolutely fresh experiences can be added, even if they are experienced only in that controlled reality which we call a utopian image. But the problem is not really to decide whether or not we are capable of inventing a world which can replace the entire world of our senses, but to discover what other mechanical worlds are accessible through this utopian vision.

​Here are two very simple examples which belong to the audiovisual world that prefigures utopian images . In his memoirs describing a cataract operation, Gaetan de Clerambault says of the moment in which vision suddenly returns: "Naturally, at first there was a general impression of visual flux, as though underwater. Then, an imprecise notion of distance, bringing things into closer range: if I wanted to pick something up, I knew from experience I had to reach some ten centimeters further than where I saw the object . . . . Every source of light caused an imperfectly geometrical figure of constant form . My right eye saw something like a treble-clef, leaning backwards with the lower element obliquely elongated . At night, the brilliant light of the street lamps and display windows appeared like so many treble-clefs . . . . For my left eye, less affected, the false image was smaller : it was like a somewhat scalene raspberry, I mean with an oblique base, sketched out in glowing filaments . . . . When the light sources are numerous and close together, for instance watching sunlight in the leaves of a tree, the whole forms a most curiously disciplined ensemble. All the figures seem to be resting on a singular kind of grid more intuited than perceived . For the right eye (the one seeing the treble-clefs) this grid is lozenge-shaped. . . . For the left eye (the one seeing the flaming raspberries) the links of the grid are square. . . . The eye from which the cataract was removed tends to modify all colors by the addition of a bit of blue . . . . Strong, dark colors are not changed ; light colors change slightly in dominant tone, sometimes agreeably so: pink takes on a violet hue, a violet-pink turns a rarer color still ; stark tones tend to disappear." A painter who had recently undergone a cataract operation described how he saw cylinders everywhere, and had lost the notion of right-angles : everything he saw was trapezoidal . It seems to me that the visual phenomena described by Clerambault are two kinds. The first, arbitrary, compensatory images remind me of Florenski's canonical signs . The other could be called aquatic images, or flux forms, which invade areas left empty by defective vision . This process of compensation is what preoccupies the architects of utopian images, which are better known as virtual reality or computer graphics . There is a superstition-or belief, or scientific truth supported by experiment-which says that cinema is the art of stimulating a part of the brain that normally functions during sleep, by bombarding it with static images juxtaposed so as to create the illusion of movement . Video, on the other hand, in which the image is liquid, is said to stimulate another part of the brain which functions only when the body is awake. Whether the distinction is scientifically valid or not is irrelevant here. What is interesting is the suggestion that we can intervene to provoke virtual images by using the brain's compensatory mechanisms. A group of people who are involved in manufacturing special effects for the Lucas company in Hollywood discussed with me the possibility of making "personalized" animated films exclusively out of such images. The principal obstacle is that the brain needs twenty to thirty seconds to process the first image, but once the first image is reconstituted the others can run off in an animated series using the same basic pattern . We went further, though, and from these flux-images we imagined film sequences in which abstract animated images would provoke different responses in each one of us. Each spectator would be watching a different threedimensional film than his neighbor, for each would have visual uncertainties (fluxes) of his own.​​