"Optical illusion is optical truth." (Goethe) 

"Seeing is believing."

"The Metamorphosis of things themselves into the sight of them." -- Merleau-Ponty, The Primacy of Perception, p. 171

Any theory of vision must describe some relation between the eye and the brain. Humberto Maturana studied the visual cortex of the frog and summarized his research in an article entitled, "what the frog's eye tells the frog's brain." Maturana and his co-authors demonstrated that the frog's sensory receptors speak to the brain in a language that is highly processed and species specific. If every species constructs for itself a different world, which is the world? Thus Maturana's credo: There is no observation without an observer.(K. Hayles, "Simulated Nature and Natural Simulations," in Uncommon Ground.) Further research led Maturana to conclude that perception is not fundamentally representational, that the perceiver encounters the world through his own self-organizing processes, through autopoesis.

Monocular perspective was the model of both painting and of human vision until the mid-nineteenth century. The possibility of stereo vision escaped Euclid, Archimedes, and Newton, and even Leonardo did not fully appreciate it, although he did notice that the two eyes have different views of a sphere. In the "classical era" of representation, to use Michel Foucault's epistème, the camera obscura provided a model of vision in which perception was a relatively passive process of reception, structured by geometry. In the nineteenth century, with the emergence of physiological optics, it became increasingly clear that the makeup and capacities of an observer contributed to the making of a perception. 

In Techniques of the Observer, Jonathan Crary traces the transformations of vision into something both more subjective and more available to (objective) measurement. Suspensions of Perception continues this study and focusses on attention. For Crary, the problematic of attention in the late nineteenth century requires (1.) an embodied observer, (2.) an understanding of perception as fundamentally durational, and (3.) an active processing of an aggregate of information. (p. 154) 

Stereo vision was not discovered until 1838, by Charles Wheatsone, a physicist and inventor. Wheatstone wrote:"It will now be obvious why it is impossible for the artist to give a faithful representation of any near solid object, that is, to produce a painting which shall not be distinguished in the mind from the object itself. When the painting and the object are seen with both eyes, in the case of the painting two similar pictures are projected on the retinae, in the case of the solid object the two pictures are dissimilar , there is therefore an essential difference between the impressions on the organs of sensation in the two cases, and consequently between the perceptions formed in the mind; the painting therefore cannot be confounded with the solid object." (quoted in Steven Pinker, How the Mind Works, p. 219) 

For David Marr, vision is the process of discovering from images what is present in the world, and where it is. Marr was the first to describe vision as solving the "ill-posed problems" (problems that have no single solution) of reverse optics -- the fact that the retinal image has countless possible interpretations -- by adding assumptions about the world into the process of vision. Marr was a forceful defender of the computational theory of mind. He defined vision as "a process that produces from images of the external world a description that is useful for the viewer and not cluttered with irrelevant information." The study of vision must also include an inquiry into the nature of the internal representations by which we capture this information and make it available for decisions. But the quintessential purpose of human vision is to tell about shape and space and spatial arrangement by building descriptions of the shapes and positions of things out of images. 

For Marr, vision passes through two early retinocentric stages: In the first, representations are obtained of the changes and structure of the image: the local variations of geometry, reflectance, and illumination result in the primal sketch. A number of processes operating on the primal sketch result in the 2 1/2 dimensional sketch, which yields the geometry of the visible surfaces. Others have called this a "visible surface representation." Depth is downgraded to half a dimension because it does not define the medium in which the visual information is held. It is just a piece of information held in that medium. According to Steven Pinker's analysis of How The Mind Works, the 2 1/2 D sketch is incarnated in neurons in a topographically orgainzed cortical map. For Pinker, the mind is a collection of modules, a system of organs, or a society of experts. (p.255) Modules for vision include shape analyzers, light- and reflectance analysts, that evolved in the primate brain as adaptations to their visual environments, with distinct advantages for survival. 

Information in the 2 1/2-D array is specified in a retinal frame of reference, a coordinate system centered on the viewer. According to Marr, the last phase of vision is object-oriented, yielding a 3-D model representation, in which the information from the retinal frame, or "visual field" is correlated to a world-aligned frame, or "visual world." But rather than resulting in a single world-frame, research into shape recognition has lead to the theorem that objects have their own reference frames. 

In certain respects, Marr's account of vision ressembles the visually motivated account of art by Adolf Hildebrand of 1893. In "The Problem of Form in the Fine Arts," Hildebrand described the visual processes by which we develop concepts of form out of changing appearance. Hildebrand used the way an artist steps back from his work to take it all in at once as the model of a complete image. (Fernbild ) Hildebrand described two different modes of seeing, the kinesthetic "close vision", in which the eye constantly scans from a close vantage point, transforming perception into a temporal sequence of images; and the calm coherence of the distant view, which alone yields the image of the "three-dimensional complex," with its pure surface and implied depth. For Hildebrand, the main task of the artist, whether he was a painter or a sculptor, or even an architect, was to unify surface images with movement into depth. He saw relief , a kind of 2 1/2 dimensional art, based on the idea of the object as a planar stratum of uniform dept, as the artistic mediation between surface and depth. Again, Hildebrand linked possibility of any synthetic unity of of vision with the distant image what the "calmly observing eye is able to take in without any kinesthetic activity." For Hildebrand, the establishment of a coherent impression of depth in relation to surface is the central problem of form in art.