Seeing in Black and White
Seeing in Black and White
Abstract and Keywords
Do people who see in black and white see black and white? Most people assume that totally color‐blind people see the whiteness of their shirts, the grayness of their trousers, and the blackness of their shoes. They feel the same way about animals that see only in black and white. This chapter's thesis is that only those who see in color see black and white. Its premise is that some positively equivalent representations have unequal amounts of negative information.
Do people who see in black and white see black and white? Most people assume that totally “color‐blind” people see the whiteness of their shirts, the grayness of their trousers, and the blackness of their shoes. I argue that only those who see in color see black and white.
Color vision without colors
You are viewing a single black‐and‐white chessboard by means of two televisions (fig. 12.1). The television images look the same to the unaided eye. (The light coming from each television has different properties, but these can only be differentiated with devices such as a prism.) Despite the indistinguishability, the color television is more informative because it is sensitive to hues. Unlike the black‐and‐white television, the color television gives you an opportunity to discern any hue that an object might have. When the object lacks a hue, the color television accurately conveys that absence of hue.
The shadow is primarily a change in luminance (bright to dark) whereas the grass/pavement border is a change in both colour (green to grey) and luminance (dark to light). Colour is therefore a potential cue for helping disambiguate shadows from reflectance changes via the following rule: luminance variations that are accompanied by colour variations are variations in illumination. (Kingdom et al. 2004, 907)
Computers have particular difficulty drawing the line between objects and their shadows. Developers of machine vision have made progress by exploiting color television.
Seeing black is a matter of seeing the absence of light by seeing the consequent absence of color. One can see an absence of light in other ways. Seeing darkness does not entail seeing blackness. Monochromats have no color vision at all. They can see an absence of light without seeing blackness.
In the headless woman illusion, a magician puts a black hood over a woman sitting in front of a black background (Armstrong 1968b). (p.222) The lady looks like she has no head! Gullible members of the audience mistake the absence of a representation of a head as a representation of the absence of a head. Similarly, when someone looks at the black‐and‐white television and infers that the chessboard is black and white, he mistakes the absence of a representation of hues with a representation of an absence of hue.
In the chessboard scenario, you are not given the clue that one of the televisions is a color television. When I ask whether you are seeing in color, I mean nonepistemic seeing. This sense of “see” does not entail that any particular belief is acquired by seeing. It is the kind of seeing easily attributed to the animals upon which color scientists experiment.
Eye doctors assume the nonepistemic sense of “see” when testing people for color deficiency. A motorist who cannot see red can see that a traffic light is red, for he can rely on a principle he learned from normal observers: the bottom light of a traffic box is a red light.
One cannot cure color deficiency by teaching the sufferers how to infer the colors of object. Hue differences positively correlate with differences in brightness and texture. Color‐deficient people can improve their perception of these correlated differences by varying illumination conditions and by use of colored filters. For instance, James Clerk Maxwell (1855) proposed the construction of a pair of spectacles with one red glass and one green glass. Once a red‐green blind person is informed which eyepiece is red and which is green, she can improve her color discriminations. But she still does not see red or green. (Such spectacles have been manufactured commercially, but they reduce the amount of light available to the viewer.)
When we describe a color‐deficient motorist as seeing that the traffic light is red, we are using “see” in the epistemic sense. This epistemic sense entails characteristic beliefs. The man sees that the light is red only if he believes that it is red.
John Dalton's (1798) own difficulties in discerning red from blue led to his discovery of “color blindness.” In recognition, Dalton was to be presented to the king in 1832. However, as a Quaker, Dalton could not wear the mandatory formal dress. One solution was to have (p.223) Dalton wear doctoral robes. These were acceptable to Quakers and were formal enough for court presentation. The hitch was that the doctoral robes were scarlet. Quakers prohibit scarlet clothing. Dalton's optical research came to the rescue. Organizers realized that Dalton would be oblivious to the infraction, so they proceeded with the plan, and Dalton received regal recognition for his research.
Often this anecdote is embellished with the claim that robe looked gray to Dalton. That exaggerates his color deficiency. Like most color‐deficient people, Dalton could see several hues.
Dalton was fascinated by the fact that his inner experience differed from the vast majority of people. But his concern about the phenomenal aspect of experience never played any role in the subsequent testing for color deficiency. Nor does it play any essential role in the nature of color deficiency. Scientists who think of animals as robots still attribute color vision to them. What matters is the perceiver's responsiveness to hues. This functionalism extends to human beings.
Since sensitivity to hues does not require triggering, you are getting more information from a color television than from a black‐and‐white television even if the color television registers no hues. The advantage of having a chromatic receiver for achromatic stimuli is that one gets the information that the stimuli are achromatic. There is information about what is not there.
Two representations are informationally equivalent if and only if they are positively and negatively equivalent. They are positively equivalent if and only if they provide the same information about what is the case. They are negatively equivalent if and only if they provide the same information about what is not the case. The two television images of the chessboard show that positive equivalence does not entail negative equivalence.
The presence–absence search asymmetry
People spot the presence of a feature more quickly than the absence of a feature (Treisman and Souther 1985). For instance, experimental (p.224) subjects spot a Q in a field of Os quicker than they can discover an O in a field of Qs (fig. 12.2). People more easily find a moving figure in a field of stationary figures than a stationary figure in a field of moving figures. (A figure is stationary if it has an absence of movement.) Silhouettes of bumps are easier to spot in a field of black ellipses than is an ellipse in a field of bumps. (Bumps have three‐dimensional curvature; black ellipses have an absence of three dimensionality.) People search for an orange figure in a field of red figures more efficiently than they search for a red figure in a field of orange figures. (Relative to red, orange involves the presence of yellow. Relative to orange, red involves the absence of yellow.) (Wolfe 2001)
One explanation of these asymmetries is that the presence of a feature allows us to do a parallel search instead of examining the items one by one. A Q is an O with a line added. We just look for the presence of that line in the scene as a whole. The target just pops out of a homogeneous field. Therefore, adding extra Os has little influence on the search. In contrast, a search for an absence is always a search for something relative to a figure. There are no autonomous absences. We are forced to engage in a serial search. Consequently, adding extra Qs increases search time.
There is an asymmetry in checking whether a patch is dark blue rather than black. Finding some blueness is enough to show the patch
Confirming achromaticity is harder than confirming chromaticity. There is no extra difficulty in disconfirming achromaticity; verifying the presence of any hue refutes the chromatic hypothesis. The asymmetry of confirmation and disconfirmation also affects absences of light. So blackness requires confirmation of a double absence.
Human beings have excellent color vision. They look for an ab‐ sence of light by searching for hues. Hence, blackness is detected as much by our cones as by our rods.
Proofs of absences are subject to vagueness despite the aura of absoluteness suggested by the language of privation. As the day progresses, the sky gradually changes from blue to gray and gradually changes from gray to black. There is no definite first moment at which the sky becomes achromatic. There is no definite first moment at which the sky becomes lightless.
Attributions of blackness are also affected by the relativity of “black.” The moon's sky is blacker than the Earth's night sky because Earth has an atmosphere. Air in the upper atmosphere scatters starlight, creating a faint airglow. This permanent low‐grade aurora annoys some astronomers. They will deny the night sky is really black. Even if the atmosphere were removed, perfectionists would complain about the sunlight reflected off of interplanetary dust (zodiacal light). If the dust were removed, they would complain about the background light from faint, unresolved stars and nebula.
The only as/as only fallacy
Christopher Boorse (1994) warns against the syntactic slide from “only as” to “as only.”
Golf Illustrated treats Annika Sorenstam only as a golfer. But Golf Illustrated does not treat her as only a golfer—that would be demeaning. Exclusively focusing on the golfing aspects of Sorenstam does not (p.226) carry the message that this is an exhaustive treatment of her—that Annika Sorenstam is nothing more than a golfer.
Boorse's distinction helps us distinguish only representing in black and white from representing as only in black and white. In the 1950s television series Father Knows Best, the characters are portrayed only in black and white. They are not portrayed as being only black and white. This contrasts with the 1998 movie Pleasantville. The main character is David, a 1990s teenager addicted to the 1950s situation comedy Pleasantville. With the help of a mysterious television repairman, David and his unaddicted twin sister, Jennifer, enter Pleasantville. Everybody in this 1950s utopia is white, employed, and clean‐cut. The weather is always nice. When a basketball player makes a shot, it always goes in. There are no toilets in the restrooms. And everything and everyone is in black and white. Things begin to change as Jennifer and David begin to make people act out of character. In particular, some people begin to acquire hues—to the consternation of the community.
The director/screenwriter Gary Ross depicts the town of Pleasantville by shooting scenes with black‐and‐white film. (More precisely, the scenes were filmed in color and then the hues were subtracted so that Ross could selectively reintroduce hues as the plot developed.) We would miss the point of Ross's depiction if the movie were interpreted as merely involving a switch to a less sensitive medium. We are supposed to pretend that we are seeing in color and infer that there are just no colors to see.
Irreducibility of information about absences
Negative information is trickier than positive information. If it could be reduced to positive information, philosophical difficulties would be bypassed.
The history of logical atomism suggests that this cannot be done. Bertrand Russell initially assumed he could paraphrase negative true (p.227) statements as being indirectly about positive facts. But eventually he gave up and admitted that there are negative facts.
The basic problem is that negative statements are more powerful than positive statements. Knowing how things are not gives you knowledge of exhaustiveness. If there is any reduction to be achieved, it runs from positive statements to negative facts. For instance, one reductive strategy is to exploit a kind of double negation; to say that the cat is on the mat is to say that there is no negative fact of the cat not being on the mat.
Absences of absences are part of the political ontology of negative campaigning. In 1959 the Democratic Party candidate for president, John F. Kennedy, alleged that the Republican Eisenhower administration had permitted the growth of a missile gap between the United States and the Soviet Union. After his narrow electoral victory over Vice President Richard Nixon, Kennedy acquired overwhelming evidence from aerial surveillance and a spy that his allegation was false. Although Kennedy conceded the absence of a missile gap to his secretary of defense, he continued to rely on the fear of a missile gap to support an American buildup of missiles—thereby fueling the arms race.
Negative facts are repugnant to human philosophers. People have a strong intuition that reality is fundamentally positive. But from a logical point of view, negative facts provide a more powerful reductive base than do positive facts. Negative metaphysics may be as effective as it is repugnant.
Seeing and conceiving colors
Black‐and‐white televisions do not enable the viewer to distinguish between black and dark red or between white and light yellow. A red‐and‐yellow chessboard is indistinguishable from a black‐and‐white chessboard. When a black‐and‐white chessboard is seen through a black‐and‐white television, the viewer cannot see the blackness of the black squares. To see the blackness, he would have to be sensitive to (p.228) the absence of hue. Only the color television gives him this sensitivity. The viewer of the color television of the chessboard sees the blackness of the black squares even if he does not realize that he is watching color television. He does not see that the squares are black but he will see the blackness of the squares.
In dim lighting, I cannot tell whether I am holding my gray tie or my green tie. The tie in my hands looks gray. But my green tie would also look gray in such low lighting. I gradually increase the illumination. The tie continues to look gray. Eventually I see that the tie in my hands is indeed gray. But I only saw the grayness of the tie after the light was increased.
Seeing the grayness of the tie is more than a matter of it looking gray. Nor is it good enough that the grayness of the tie causes me to experience the tie as gray. To see the grayness of the tie, I must see an absence of hue. “Gray” is a privational concept like “colorless” and “sober.” Monochromats are insensitive to hues. Therefore, they cannot see the grayness of my tie. Nor can they see any other achromatic color such as black and white. Consequently, people who see in black and white cannot see black and white.
Some monochromats make remarks that I construe as acknowledging this inability to see gray. In “The Case of the Colorblind Painter,” Oliver Sachs describes an artist, Jonathan I., who became completely monochromatic at age sixty‐five. Jonathan probably had a stroke that disabled the V‐4 area of his visual cortex. Although his cones continued to signal, Jonathan could no longer synthesize the information into hues. This meant that he could not dream in color or hallucinate in color or have colored afterimages. Nor could he visualize colors in imagination. Jonathan initially compared what he experienced with what we see when watching a black‐and‐white television:
Subsequently, he said neither “grey” nor “leaden” could convey what his world was actually like. It was not “grey” that he experienced, he said, but perceptual qualities for which ordinary experience, ordinary language, had no equivalent. (Sachs 1995, 11)
Loss of color vision plunged Jonathan into months of depression. Frances Futterman, who was born monochromatic, wrote Sachs after reading his account of Jonathan:
I was struck by how different that kind of experience must be, compared to my own experience of never having seen color before, thus never having lost it—and also never having been depressed about my colorless world. . . . The way I see in and of itself is not depressing. In fact, I am frequently overwhelmed by the beauty of the natural world. . . . People say I must see in shades of gray or in “black and white,” but I don't think so. The word gray has no more meaning for me than the word red or blue—in fact, even less meaning, because I have developed inner concepts of color words like red and blue; but, for the life of me, I can't conceive of gray. (1995, 33 fn. 24)
Talk of what we can and cannot conceive is tricky. If Futterman had no clue at all as to what “gray” means, she would not be able to comment on gray. Even if she cannot conceive of grayness, she can conceive others conceiving the grayness.
Meta‐conception is enough for linguistic competence when the speaker can rely on linguistic division of labor. Futterman became a member of a linguistic community dominated by color‐normal people, some of whom taught her “gray.” She in turn could teach “gray” to children. Indeed, Futterman could impart the grammar of color terms to children along with her ample encyclopedic knowledge about colors: stop signs are red, violets are blue, pink resembles red more than green, and so on.
One of the standard tests for color deficiency is the ability to identify colors. But unless the tests are administered carefully, color‐deficient people can pass by employing their background knowledge and by exploiting indirect visual clues. For instance, monochromats (p.230) can tell whether a photograph is in color by looking at it edgewise, thereby spotting the layers of pigment. They also exploit the fact that color photographs lack the sharp contours of black‐and‐white film.
Color‐deficient pedestrians can see that the traffic light is green by its location. They can see that a sign is red by its hexagonal shape and the word “stop.” People with normal color vision do not need to rely on these clues. Since nonepistemic seeing does not involve inferences, the deficit constituted by color deficiency is clearer when we stick to nonepistemic seeing.
Futterman can also conceive of grayness in the way she can conceive of the fourth dimension. Topologists can imagine the fourth dimension discursively with the help of algebra. From high school, Futterman knows Cartesian geometry in which shapes are described in terms of two variables along the x‐ and y‐axes. From there she can proceed to projective geometry by adding a third variable. Algebraically, it is now a trivial step from the third dimension to the fourth dimension. Color mixing is also governed by an algebra. Once one learns the rules for adding light and subtracting light by means of pigments, one's conception of color is deepened. Even if one cannot visualize hues, one can appreciate how they constitute an extra dimension to vision.
Both of these alternative ways of understanding grayness require a language community containing people with color vision. In the past, there may have been isolated populations that have been entirely monochromatic (Sachs 1997). On the island of Pinglap in the South Pacific, more than 5% of the population are genetic achromats. Heavy inbreeding caused this after one of the island's periodic floods. In 1775, typhoon Lengkieki reduced the population of this remote island from roughly a thousand to about twenty. One of the survivors must have carried the recessive gene for monochromaticity. In four generations, achromatic children were born. Since genetic achromats are often segregated, another flood could leave a population composed solely of achromats. These survivors would breed true. Memories of hues would die out over the generations. In this purely achromatic population, no one could conceive of grayness by using linguistic division of labor.
(p.231) In recent history, educational institutions enable monochromats to acquire a detailed understanding of color. One of the foremost experts on color deficiency, Knut Norby, was born monochromatic:
Although I have acquired a thorough theoretical knowledge of the physics of colours and the physiology of the colour receptor mechanisms, nothing of this can help me to understand the true nature of colours. From the history of art I have also learned about the meanings often attributed to colours and how colours have been used at different times, but this too does not give me an understanding of the essential character or quality of colours. (1990, 305)
But Norby competed for his post at the University of Oslo on the strength of subtle insights into the nature of color and color deficiency. Why would Norby ask color‐normal people about how azure relates to blue if he could not understand the answer?
Sampling and simulating color deficiency
When Thomas Nagel asked, “What is it like to be a bat?” he chose a creature that is very alien to us. But color‐deficient people are our neighbors and colleagues (and some of my readers).
The “best” way to learn what it is like to be color deficient is to become color deficient. Many drugs and diseases disable elements of color vision. There are also ways of becoming temporarily color deficient. Staring at a bright light through a colored filter can induce a type of partial color deficiency resembling anomalous trichromatism. The V‐4 section of your visual cortex can be disabled by magnetic stimulation. You can sample the perspective of Jonathan I.
Some people have one color‐deficient eye and one normal eye (Graham and Hsia 1958). These people can experience color deficiency by simply closing the normal eye. One such unilaterally color‐deficient woman helped researchers construct a room that gives normal occupants a good idea how things look to a person (p.232) who is insensitive to red and green (Kalmus 1965, 12–13). She was normal in her right eye but dichromatic in her left eye. Specifically, she was deuteranope. When using both eyes, her vision was normal. But when she used only her left eye, she was limited to three kinds of color sensations in her dichromatic eye: gray, yellow, and blue. She lacked any green or red sensations. The experimenters then furnished a room exclusively from gray, yellow, and blue materials. The woman verified that the room looked the same to each eye.
Those who cannot visit the dichromatic room can see color photographs of it. This is not as effective as an environmental simulation. The photograph of the deuteropic room is surrounded by a normally colored background. A better approximation is a slide presented in darkness. Better yet is a movie. Dean Farnsworth made a documentary movie, Color Vision Deficiencies, that shows, among other things, the color world of this woman.
Environmental and photographic simulations are helpful in showing the limitations of what a dichromat sees. Designers of traffic lights make important safety improvements by examining dichromatically doctored videotapes of traffic lights. The 8% of Western men who are color deficient now face fresh difficulties navigating the World Wide Web. As becomes evident to those who employ a monochrome monitor, Web pages are heavily color coded. There are now Web sites (based on Brettel et al. 1997; Viénot et al. 1999) that help by allowing color‐normal people to see what their Web pages look like to color‐deficient people. You insert the HTML code and enter the type of color deficiency you wish to simulate. Out comes your Web site in not so glorious color. Other Web sites let you enter digital photographs to see how they look like to color‐deficient viewers.
These simulations are highly instructive. However, they tempt us to overestimate how much color‐deficient people see. Although the monocularly dichromatic consultant helping in Dean Farnsworth's film could not distinguish what she was seeing from her dichromatic left eye and normal right eye, I say she was seeing more through her normal eye, for only her normal eye was enabling her to see absences of green and red. There is positive equivalence without negative equivalence.
When we try to imagine what it is like to see in black and white, we have trouble subtracting our color vision. A photographer who wants to convey a dog's perspective will focus on something of interest to a dog from a low angle and use black‐and‐white film. (The effectiveness of this technique survives the debunking of the myth that dogs see only in black and white.) We take ourselves to be seeing through the dog's eyes because we see blacks, whites, and shades of gray. True, the surface of the photograph has achromatic color patches, and we are seeing their blackness, whiteness, and grayness. But we ought not to infer from this perception of the photograph's surface that a monochromatic dog is also seeing the blackness, whiteness, and shades of gray. Features of the surface of the photograph need not be features of what the photograph depicts. A black‐and‐white photograph depicts only differences in brightness.
When I try to “get into the head” of a monochromat, I tend to picture his eyes as delivering black‐and‐white images. It is as if I were a little man lodged in his head watching black‐and‐white television. From my internal vantage point, I see the blackness and whiteness and grayness of portions of the image. This tempts me to infer that the color‐deficient person also sees the blackness and whiteness and grayness of his internal image. I have been warned of similar fallacies. Yet I keep committing these “Cartesian theater” mistakes. I note the source of temptation for the edification of my fellow recidivists.
The closest approximation we color‐normal people have to monochromaticity is night vision. At night, trichromats see in black and white. But this environmental color deficiency is not as restrictive as monochromatism. Our cones are not turned off in dim conditions. Just the reverse: dark‐adapted cones are at their peak sensitivity. When Sachs (1997, 55) accompanied night fishermen, he could see the yellowish light of bioluminescent protozoa Noctiluca. They light up when the water is disturbed. Their flash of light marks the lift‐off of a flying fish and then its splashdown.
(p.234) The ability to see colors varies with external conditions. At night, the environment offers too little light for us to see the colors of objects. The objects are still colored at night. Our physiological ability to discriminate colors is highest at night. But it is not enough to compensate for the light shortage. We are blind to the colors of objects in dim conditions. We see only in black and white. This is environmental color deficiency.
The phenomenal view
According to the phenomenal analysis of seeing in color, one sees in color if and only if one has colored experiences (“qualia”). If a sufferer of jaundice is watching the black‐and‐white television broadcasting a black‐and‐white chessboard, then he sees the white squares as light yellows. On a simple phenomenal view, sufferers of jaundice see in color in virtue of this yellowish illusion. Under this view, seeing in color does not entail accurately seeing in color. Nor does it entail that anything is colored or that anything exists besides the perceiver. If color experiences suffice for seeing in color, then a brain in a vat can see in color.
Of course, one could combine the sensitivity analysis with the phenomenal view by demanding that both conditions be satisfied for color vision. Given this double requirement, you see in color when your colored experiences reliably track chromatic stimuli.
The combined view has trouble with mixed viewing conditions. Do you see the whole scene in color only when every bit of it is colored? If so, I can end your color vision by showing you this in black and white.
Is it good enough for there to be some color—like a red cherry in a field of snow? If so, I can take away your color vision by snatching away the cherry. Or suppose the scene alternates between having a green dot and gray dot. Is your color vision going in and out with the dot?
In Frank Jackson's (1986) thought experiment, Mary grows up in a black‐and‐white room, reading black‐and‐white books and watching black‐and‐white television. Although there are no hues in Mary's environment, she has access to physics textbooks and becomes an expert on color. She knows all there is to know about the physics of color. Even so, Mary learns something when she is released from her room and sees a ripe tomato. Now she knows what red looks like. Since the facts about color cannot be reduced to physics, Jackson concludes that physicalism is false.
Since Jackson is assuming that Mary has normal sensitivity to hues, Mary sees more than does a totally color‐deficient person. Mary is seeing absences of hues (whether she realizes it or not). Unlike the monochromat, Mary sees the blackness of her telephone, the whiteness of her chalk, and the grayness of her couch.
Since Mary is seeing an absence of red, she must have the concept of red. Mary should be understood as a counterexample to concept empiricism. According to concept empiricism, we acquire all of our concepts through experience of their positive instances. We start life as blank slates and acquire basic concepts by perceiving things that correspond to those concepts. From this stock of basic concepts, we derive complex concepts. For instance, we have the concept of a unicorn by conjoining the concept of a horse with the concept of a horn. Unlike basic concepts, derivative concepts need not have positive instances.
“Red” is generally taken to be a simple concept. So Mary should first acquire the concept of red only upon first contact with something red. But I say Mary had the concept of red before she saw any red thing. She has it independently of her expertise in physics. If Mary had been untrained in physics, she would still have the concept of red in virtue of her color vision. The same is true if the people on the outside of her room were monochromats with no concept of hues.
(p.236) Concept empiricism antecedently faces David Hume's famous counterexample of the missing shade of blue. Although Hume is a concept empiricist, he concedes that if one has never experienced a certain shade of blue, one could still have the concept of that shade of blue by virtue of experiencing the shades leading up to that shade of blue. We acquire some concepts by experiencing near misses of instances.
Mary is a different kind of counterexample. She has the concept of red even though she is entirely bereft of positive colored experience. So we cannot say Mary learns a new fact in virtue of having newly acquired the concept of red. If concept empiricism were true, then depriving the subject of instances can limit her conceptual range. But concept empiricism is false. Mary has the concept of red by virtue her disposition to recognize red things when they are presented to her.
Lastly, consider the case of “Colored Colleen.” Her environment is censored so that she has never seen achromatic colors (black, white, shades of gray). When she is presented for the first time with a black telephone, she is excited: “That is what black looks like!”
Is Mary any more of a counterexample to physicalism than Colleen? Mary is stepping up to a new dimension of experience. Colleen is stepping down. The step‐down does not threaten physicalism because Jackson is complaining that physicalism leaves something positive out.
Step‐downs generally count as novel experiences. People are aesthetically stimulated by restrictive color schemes. Consider Dartmouth College alumni who decorate their kitchens in the school colors of green and white. The absence of nongreen hues leaps out at you. (And also simulates the rare color deficiency of a tritanope—they lack the yellow/blue sense.)
Chapter 13 focuses on the limit case of a step‐down: total darkness.