An Anthropologist on Mars (1995) (6 page)

Maxwell himself was acutely aware of the drawback of this additive process: color photography had no way of “discounting the illuminant”, and its colors changed helplessly with changing wavelengths of light.

In 1957, ninety-odd years after Maxwell’s famous demonstration, Edwin Land—not merely the inventor of the instant Land camera and Polaroid, but an experimenter and theorizer of genius—provided a photographic demonstration of color perception even more startling. Unlike Maxwell, he made only two black-and-white images (using a split-beam camera so they could be taken at the same time from the same viewpoint, through the same lens) and superimposed these on a screen with a double-lens projector. He used two filters to make the images: one passing longer wavelengths (a red filter), the other passing shorter wavelengths (a green filter). The first image was then projected through a red filter, the second with ordinary white light, unfiltered. One might expect that this would produce just an overall pale-pink image, but something “impossible” happened instead. The photograph of a young woman appeared instantly in full color—“blonde hair, pale blue eyes, red coat, bluegreen collar, and strikingly natural flesh tones”, as Land later described it. Where did these colors come from, how were they made? They did not seem to be “in” the photographs or the illuminants themselves. These demonstrations, overwhelming in their simplicity and impact, were color “illusions” in Goethe’s sense, but illusions that demonstrated a neurological truth—that colors are not “out there” in the world, nor (as classical theory held) an automatic correlate of wavelength, but, rather, are constructed by the brain.

These experiments hung, at first, like anomalies, conceptless, in midair; they were inexplicable in terms of existing theory, but did not yet point clearly to a new one. It seemed possible, moreover, that the viewer’s knowledge of appropriate colors might influence his perception of such a scene. Land decided, therefore, to replace familiar images of the natural world with entirely abstract, multicolored displays consisting of geometric patches of colored paper, so that expectation could provide no clues as to what colors should be seen. These abstract displays vaguely resembled some of the paintings of Piet Mondrian, and Land therefore terms them “color Mondrians.” Using the Mondrians, which were illuminated by three projectors, using long-wave (red), middle-wave (green), and short-wave (blue) filters, Land was able to prove that, if a surface formed part of a complex multicolored scene, there was no simple relationship between the wavelength of light reflected from a surface and its perceived color.

If, moreover, a single patch of color (for example, one ordinarily seen as green) was isolated from its surrounding colors, it would appear only as white or pale grey, whatever illuminating beam was used. Thus the green patch, Land showed, could not be regarded as inherently green, but was, in part, given its greenness by its relation to the surrounding areas of the Mondrian.

Whereas color for Newton, for classical theory, was something local and absolute, given by the wavelength of light reflected from each point, Land showed that its determination was neither local nor absolute, but depended upon the surveying of a whole scene and a comparison of the wavelength composition of the light reflected from each point with that of the light reflected from its surround. There had to be a continuous relating, a comparison of every part of the visual field with its own surround, to arrive at that global synthesis—Helmholtz’s “act of judgement.” Land felt that this computation or correlation followed fixed, formal rules; and he was able to predict which colors would be perceived by an observer under different conditions. He devised a “color cube”, an algorithm, for this, in effect a model for the brain’s comparison of the brightnesses, at different wavelengths, of all the parts of a complex, multicolored surface. Whereas Maxwell’s color theory and color triangle were based on the concept of color addition, Land’s model was now one of comparison. He proposed that there were, in fact, two comparisons: first of the reflectance of all the surfaces in a scene within a certain group of wavelengths, or waveband (in Land’s term, a “lightness record” for that waveband), and second, a comparison of the three separate lightness records for the three wavebands (corresponding roughly to the red, green, and blue wavelengths). This second comparison generated the color. Land himself was at pains to avoid specifying any particular brain site for these operations and was careful to call his theory of color vision the Retinex theory, implying that there might be multiple sites of interaction between the retina and the cortex.

If Land was approaching the problem of how we see colors at a psychophysical level by asking human subjects to report how they perceived complex, multicolored mosaics in changing illuminations, Semir Zeki, working in London, was approaching the problem at a physiological level, by inserting microelectrodes in the visual cortex of anesthetized monkeys and measuring the neuronal potentials generated when they were given colored stimuli. Early in the 1970
s
, he was able to make a crucial discovery, to delineate a small area of cells on each side of the brain, in the prestriate cortex of monkeys (areas referred to as V4), which seemed to be specialized for responding to color (Zeki called these “color-coding cells”).
17

17. He was also able to find cells, in an adjacent area, that seemed to respond solely to movement. A remarkable account and analysis of a patient with a pure “motion blindness” was given by Zihl, Von Cramon, and Mai in 1983. The patient’s problems are described as follows:

The visual disorder complained of by the patient was a loss of movement vision in all three dimensions. She had difficulty, for example, in pouring tea or coffee into a cup because the fluid appeared to be frozen, like a glacier. In addition, she could not stop pouring at the right time since she was unable to perceive the movement in the cup (or a pot) when the fluid rose. Furthermore the patient complained of difficulties in following a dialogue because she could not see the movement of the face and, especially, the mouth of the speaker. In a room where more than two other people were walking, she felt very insecure and unwell, and usually left the room immediately, because “people were suddenly here or there but I have not seen them moving.” The patient experienced the same problem but to an even more marked extent in crowded streets or places, which she therefore avoided as much as possible. She could not cross the street because of her inability to judge the speed of a car, but she could identify the car itself without difficulty. “When I’m looking at the car first, it seems far away. But then, when I want to cross the road, suddenly the car is very near.” She gradually learned to “estimate” the distance of moving vehicles by means of the sound becoming louder
.

Thus, ninety years after Wilbrand and Verrey had postulated a specific center for color in the brain, Zeki was finally able to prove that such a center existed.

Fifty years earlier, the eminent neurologist Gordon Holmes, reviewing two hundred cases of visual problems caused by gunshot wounds to the visual cortex, had found not a single case of achromatopsia. He went on to deny that an isolated cerebral achromatopsia could occur. The vehemence of this denial, coming from such a great authority, played a major part in bringing all clinical interest in the subject to an end.
18

18. A vivid account of Holmes’s negative influence has been provided by Damasio, who also points out that all of Holmes’s cases involved lesions in the dorsal aspect of the occipital lobe, whereas the center for achromatopsia lies on the ventral aspect.

Zeki’s brilliant and undeniable demonstration startled the neurological world, reawakening attention to a subject it had for many years dismissed. Following his 1973 paper, new cases of human achromatopsia began appearing in the literature once again, and these could now be examined with new brain-imaging techniques (CAT, MRI, PET, SQUID, etc.) not available to neurologists of an earlier era. Now, for the first time, it was possible to visualize, in life, what areas of the brain might be needed for human color perception. Though many of the cases described had other problems, too (cuts in the visual field, visual agnosia, alexia, etc.), the crucial lesions seemed to be in the medial association cortex, in areas homologous to V4 in the monkey.
19

19. The work of Antonio and Hanna Damasio and their colleagues at the University of Iowa was particularly important here, both by virtue of the minuteness of the perceptual testing, and the refinement of the neuroimaging they used.

It had been shown in the 1960
s
that there were cells in the primary visual cortex of monkeys (in the area termed V1) that responded specifically to wavelength, but not to color; Zeki now showed, in the early 1970
s
, that there were other cells in the V4 areas that responded to color but not to wavelength (these V4 cells, however, received impulses from the V1 cells, converging through an intermediate structure, V2). Thus each V4 cell received information regarding a large portion of the visual field. It seemed that the two stages postulated by Land in his theory might now have an anatomical and physiological grounding: lightness records for each waveband being extracted by the wavelength-sensitive cells in V1, but only being compared or correlated to generate color in the color-coding cells of V4. Every one of these, indeed, seemed to act as a Landian correlator, or a Helmholtzian “judge.”

Color vision, it seemed—like the other processes of early vision: motion, depth, and form perception—required no prior knowledge, was not determined by learning or experience, but was, as neurologists say, a “bottom-up” process. Color can indeed be generated, experimentally, by magnetic stimulation of V4, causing the “seeing” of colored rings and halos—so-called chromatophenes.
20

20. Such chromatophenes may occur spontaneously in visual migraines, and Mr. I. himself had experienced these, on occasion, in migraines occurring before his accident. One wonders what would have been experienced if Mr. I.’s V4 areas had been stimulated—but magnetic stimulation of circumscribed brain areas was not technically possible at the time. One wonders, too, now that such stimulation is possible, whether it might be tried in individuals with congenital (retinal) achromatopsia (several such achromatopes have expressed their curiosity about such an experiment). It is possible—I am not aware of any studies on this—that V4 fails to develop in such people, with the absence of any cone input. But if V4 is present as a functional (though never functioning) unit despite the absence of cones, its stimulation might produce an astounding phenomenon—a burst of unprecedented, totally novel sensation, in a brain⁄mind that had never had a chance to experience or categorize such sensation. Hume wonders if a man could imagine, could even perceive, a color he had never seen before—perhaps this Humean question (propounded in 1738) could And an answer now.

But color vision, in real life, is part and parcel of our total experience, is linked with our own categorizations and values, becomes for each of us a part of our life-world, of us. V4 may be an ultimate generator of color, but it signals to, it converses with, a hundred other systems in the mind-brain; and perhaps it can also be modulated by these. It is at higher levels that integration occurs, that color fuses with memories, expectations, associations, and desires to make a world with resonance and meaning for each of us.
21

21. The power of expectation and mental set in the perception of color is clearly shown in those with partial red-green colorblindness. Such people may not, for example, be able to spot scarlet holly berries against the dark green foliage, or the delicate salmon-pink of dawn—until these are pointed out to them. “Our poor impoverished cone cells”, says a dyschromatope of my acquaintance, “need the amplification of intellect, knowledge, expectation, and attention in order to ‘see’ the colors that we are normally ‘blind’ to.”

Mr. I. not only presented a rather “pure” case of cerebral achromatopsia (virtually uncontaminated by additional defects in the perception of form, motion, or depth), but was a highly intelligent and expert witness as well, one who was skilled at drawing and reporting what he saw. Indeed, when we first met, and he described how objects and surfaces “fluctuated” in different lights, he was, so to speak, describing the world in wavelengths, not in colors. The experience was so unlike anything he had ever experienced, so strange, so anomalous, that he could find no parallels, no metaphors, no paints or words to depict it.

When I phoned Professor Zeki to tell him of this exceptional patient, he was greatly intrigued and wondered, in particular, how Mr. I. might do with Mondrian testing, such as he and Land had used with normally sighted people and with animals. He at once arranged to come to New York to join us—Bob Wasserman, my ophthalmologist colleague; Ralph Siegel, a neurophysiologist; and myself—in a comprehensive testing of Jonathan I. No patient with achromatopsia had ever been examined in this way before.

We used a Mondrian of great complexity and brilliance, illuminated either by white light or by light filtered through narrow-band filters allowing only long wavelengths (red), intermediate wavelengths (green), or short wavelengths (blue) to pass. The intensity of the illuminating beam, in every case, was the same.

Mr. I. could distinguish most of the geometric shapes, though only as consisting of differing shades of grey, and he instantly ranked them on a one-to-four grey scale, although he could not distinguish some color boundaries (for example, between red and green, which both appeared to him, in white light, as black). With rapid, random switching of the filters, the grey-scale value of all the shapes dramatically changed—some shades previously indistinguishable now became very different, and all shades (except actual black) changed, either grossly or subtly, with the wavelength of the illuminating beam. (Thus a green area would be seen by him as white in medium-wavelength light, but as black in white or long-wavelength light.)